-$Id: AUTHORS,v 1.22 2008/09/23 23:55:29 jharg Exp $
+$Id: AUTHORS 3096 2010-04-30 14:32:49Z chrfranke $
This code was originally developed as a Senior Thesis by Michael
Cornwell at the Concurrent Systems Laboratory (now part of the Storage
Casper Dik <casper@holland.sun.com>
Christian Franke <franke@computer.org>
Guilhem Frézou <guilhem.frezou@catii.fr>
-Douglas Gilbert <dougg@torque.net>
+Douglas Gilbert <dgilbert@interlog.com>
Guido Guenther <agx@sigxcpu.org>
Geoff Keating <geoffk@geoffk.org>
Dr. David Kirkby <drkirkby@ntlworld.com>
CHANGELOG for smartmontools
-$Id: CHANGELOG 3077 2010-03-16 20:48:06Z chrfranke $
+$Id: CHANGELOG 3124 2010-07-12 19:21:00Z chrfranke $
The most recent version of this file is:
http://smartmontools.svn.sourceforge.net/viewvc/smartmontools/trunk/smartmontools/CHANGELOG?view=markup
<DEVELOPERS: ADDITIONS TO THE CHANGE LOG GO JUST BELOW HERE, PLEASE>
+ [CF] drivedb.h USB updates:
+ - Iomega LDHD-UP (ticket #83)
+ - WD Elements Desktop 2TB
+ - Maxtor OneTouch (0x0d49:0x7300)
+
+ [MS] drivedb.h updates:
+ - Intel X25-M SSD first Generation
+ - ExcelStor J8160
+ - OCZ Agility2
+
+ [CF] drivedb.h updates:
+ - Transcend Solid State Drives (ticket #80)
+
+ [CF] drivedb.h USB update:
+ - LaCie Rugged Hard Drive
+
+ [CF] smartctl: Add options '--scan, --scan-open'.
+
+ [CF] Windows: Use also VendorId from IOCTL_STORAGE_QUERY_PROPERTY.
+
+ [CF] smartd: Change defaults of '-C' and '-U' directives to 0 (disabled)
+ if attribute name is changed by '-v 19[78],...' directive.
+
+ [CF] configure.in: Fix include path for MinGW.
+
+ [CF] Move 'posix/reg*' to 'regex/reg*'.
+ Add configure check for regex.
+
+ [MS] cciss.cpp: avoid redefining be32toh
+ megaraid.h: replace use of undefined preprocessor macro BITS_PER_LONG
+ by union construct (thanks to [DL]).
+ Add assert for sizeof(ptr_t) == 8 (thanks to [CF]).
+
+ [CF] Makefile.am: Add os_qnxnto.* to EXTRA_smart*_SOURCES.
+
+ [MS] drivedb.h update:
+ - WD My Passport Essential SE 1TB variant (USB interface)
+
+ [CF] Use getopt_long() from getopt/getopt* if necessary.
+ Add missing cast to os_qnxnto.cpp.
+ This fixes build on QNX (ticket #1).
+ Thanks to Stefan (stevestereo) for testing.
+
+ [CF] drivedb.h update:
+ - WD Caviar Green (Adv. Format) family
+
+ [CF] drivedb.h USB update:
+ - Verbatim External Hard Drive 47519
+
+ [DL] Fix regression in smartctl option '-t select,M-N' which prevents
+ that more than one test span can be specified (ticket #75).
+
+ [CF] drivedb.h updates:
+ - Add raw64 attributes 1, 210-213 to all SSD drives with
+ 64-bit attribute format.
+
+ [CF] Support smartd '-l xerror' also for disks which use reserved
+ byte as log index.
+
+ [CF] Fix initialization of values missing in smartd '.state' files.
+
+ [CF] Add smartd directive '-l xerror' to check error count from
+ the Extended Comprehensive SMART Error Log (ticket #34).
+
+ [CF] Fix max number of cciss devices, 128 devices are supported
+ again (ticket #49). Regression was introduced during migration
+ to new interface.
+
+ [CF] Update man pages (include Debian patch
+ 60_remove-redhatism.diff and Debian Bug 570892).
+
+ [CF] Add SVN revision number to man pages.
+
+ [CF] Windows: Read default drivedb.h and smartd.conf from exe
+ directory instead of current directory.
+
+ [CF] drivedb.h update:
+ - SAMSUNG SpinPoint M series
+
+ [CF] Replace runtime check of byte ordering by compile time check.
+
+ [CF] drivedb.h USB updates:
+ - ALi M5621 (unsupported)
+ - LaCie with JMicron (ticket #69)
+ - JMicron (0x2352)
+ - Enable 48-bit commands for Hitachi drive
+
+ [CF] Read USB ID info from drivedb.h (ticket #44).
+
+ [CF] Create branch RELEASE_5_39_DRIVEDB with last drivedb.h file
+ compatible with smartmontools 5.39[.1].
+
+ [MS] drivedb.h updates:
+ - WD Raptor 80GB variant
+ - correct Regex for some WD AV-GP variants
+ - Hitachi Ultrastar A7K2000
+ - Hitachi Travelstar 5K500.B
+ - Hitachi Deskstar 7K1000.C
+ - adjust naming of Hitachi Travelstar and Deskstar drives
+
+ [CF] Move 'posix/getopt*' to 'getopt/getopt*'. Can be used for
+ platforms with regex() but without getopt_long() (QNX, ticket #1).
+
+ [CF] smartd '-l selftest' directive: Print info if error count
+ decreased. Avoid misleading warning if error count decreased
+ to zero (ticket #67).
+
+ [CF] smartctl: Rework ataPrintMain(). Issue ATA SMART commands only if
+ necessary. Improve handling of SMART STATUS command failure when
+ ATA output registers are missing (ticket #27).
+
+ [CF] USB ID updates:
+ - A-DATA SH93
+ - Hitachi/SimpleTech 1TB
+
+ [CF] configure.in: Print configuration summary.
+
+ [CF] smartctl -l xselftest,selftest: Print old log if extended self-test
+ log index is out of range. Workaround for bad log data from Intel
+ X25-M G2 (ticket #66).
+
[CF] USB ID updates:
- LaCie Desktop Hard Drive
- Prolific PL2507 (unsupported)
## Process this file with automake to produce Makefile.in
#
-# $Id: Makefile.am 3074 2010-03-05 23:00:30Z chrfranke $
+# $Id: Makefile.am 3115 2010-06-02 17:23:05Z chrfranke $
#
@SET_MAKE@
AM_CPPFLAGS += -DSMARTMONTOOLS_ATTRIBUTELOG='"$(attributelog)"'
endif
+if NEED_GETOPT_LONG
+AM_CPPFLAGS += -I$(srcdir)/getopt -DHAVE_GETOPT_LONG -D__GNU_LIBRARY__
+endif
+if NEED_REGEX
+AM_CPPFLAGS += -I$(srcdir)/regex
+endif
+
sbin_PROGRAMS = smartd \
smartctl
os_netbsd.h \
os_openbsd.cpp \
os_openbsd.h \
+ os_qnxnto.cpp \
+ os_qnxnto.h \
os_solaris.cpp \
os_solaris.h \
os_solaris_ata.s \
if OS_WIN32_MINGW
-smartd_SOURCES += \
- posix/regex.h \
- posix/regex.c \
- os_win32/daemon_win32.h \
- os_win32/daemon_win32.cpp \
- os_win32/hostname_win32.h \
- os_win32/hostname_win32.cpp \
- os_win32/syslog.h \
- os_win32/syslog_win32.cpp
-
-# Included by regex.c:
-EXTRA_smartd_SOURCES += \
- posix/regcomp.c \
- posix/regexec.c \
- posix/regex_internal.c \
- posix/regex_internal.h
+smartd_SOURCES += \
+ os_win32/daemon_win32.cpp \
+ os_win32/daemon_win32.h \
+ os_win32/hostname_win32.cpp \
+ os_win32/hostname_win32.h \
+ os_win32/syslog_win32.cpp \
+ os_win32/syslog.h
endif
os_netbsd.h \
os_openbsd.cpp \
os_openbsd.h \
+ os_qnxnto.cpp \
+ os_qnxnto.h \
os_solaris.cpp \
os_solaris.h \
os_win32.cpp \
dev_legacy.cpp \
megaraid.h
-if OS_WIN32_MINGW
+if NEED_GETOPT_LONG
+
+smartctl_SOURCES += \
+ getopt/getopt.c \
+ getopt/getopt.h \
+ getopt/getopt1.c
-smartctl_SOURCES += \
- posix/regex.h \
- posix/regex.c \
- os_win32/syslog.h
+smartd_SOURCES += \
+ getopt/getopt.c \
+ getopt/getopt.h \
+ getopt/getopt1.c
+
+endif
+
+if NEED_REGEX
+
+smartctl_SOURCES += \
+ regex/regex.c \
+ regex/regex.h \
+ regex/regex_internal.h
+
+smartd_SOURCES += \
+ regex/regex.c \
+ regex/regex.h \
+ regex/regex_internal.h
# Included by regex.c:
-EXTRA_smartctl_SOURCES += \
- posix/regcomp.c \
- posix/regexec.c \
- posix/regex_internal.c \
- posix/regex_internal.h
+EXTRA_smartctl_SOURCES += \
+ regex/regcomp.c \
+ regex/regexec.c \
+ regex/regex_internal.c
+
+EXTRA_smartd_SOURCES += \
+ regex/regcomp.c \
+ regex/regexec.c \
+ regex/regex_internal.c
endif
endif
MAN_FILTER = \
- sed "s|CURRENT_CVS_VERSION|$(releaseversion)|g; \
- s|CURRENT_CVS_DATE|`sed -n 's,^.*DATE[^"]*"\([^"]*\)".*$$,\1,p' svnversion.h`|g; \
- s|CURRENT_CVS_TIME|`sed -n 's,^.*TIME[^"]*"\([^"]*\)".*$$,\1,p' svnversion.h`|g; \
+ sed "s|CURRENT_SVN_VERSION|$(releaseversion)|g; \
+ s|CURRENT_SVN_DATE|`sed -n 's,^.*DATE[^"]*"\([^"]*\)".*$$,\1,p' svnversion.h`|g; \
+ s|CURRENT_SVN_REV|`sed -n 's,^.*REV[^"]*"\([^"]*\)".*$$,r\1,p' svnversion.h`|g; \
s|/usr/local/share/man/|$(mandir)/|g; \
s|/usr/local/sbin/|$(sbindir)/|g; \
s|/usr/local/etc/rc\\.d/init.d/|$(initddir)/|g; \
- s|/usr/local/share/doc/smartmontools-5.1/|$(docsdir)/|g; \
+ s|/usr/local/share/doc/smartmontools/|$(docsdir)/|g; \
s|/usr/local/etc/smartd\\.conf|$(sysconfdir)/smartd.conf|g; \
s|/usr/local/etc/smart_drivedb\\.h|$(sysconfdir)/smart_drivedb\\.h|g" | \
$(MAN_CAPABILITIES) | \
smartmontools NEWS
------------------
-$Id: NEWS 3076 2010-03-12 22:23:08Z chrfranke $
+$Id: NEWS 3119 2010-06-11 16:21:25Z chrfranke $
The most up-to-date version of this file is:
http://smartmontools.svn.sourceforge.net/viewvc/smartmontools/trunk/smartmontools/NEWS?view=markup
- configure: New default value for '--with-docdir'.
- Drive database is in a separate source file 'drivedb.h'
which can be downloaded from SVN.
+- USB ID info is now included in 'drivedb.h'.
- New script 'update-smart-drivedb'.
- smartd libcap-ng support, option '-C, --capabilities'.
+- smartd directive '-l xerror' to check Extended Comprehensive
+ SMART Error Log.
- smartctl option '-l scterc[,...]' to get/set the
SCT Error Recovery Control time limit.
+- smartctl options '--scan, --scan-open'.
- Linux: Add '/dev/sd[a-c][a-z]' to smartd DEVICESCAN.
+- Windows: Read 'drivedb.h' and 'smartd.conf' from exe directory.
- Windows: Support for 64-bit executables.
- Windows: Support for cross compilation on Linux.
+- Fix regression in smartctl option '-t select,M-N'.
- Fix SCT temperature table commands on big endian CPUs.
Date 2010-01-28
#include "utility.h"
#include "dev_ata_cmd_set.h" // for parsed_ata_device
-const char * atacmds_cpp_cvsid = "$Id: atacmds.cpp 3065 2010-02-10 22:16:50Z chrfranke $"
+const char * atacmds_cpp_cvsid = "$Id: atacmds.cpp 3117 2010-06-08 15:41:04Z chrfranke $"
ATACMDS_H_CVSID;
// for passing global control variables
bool & increase)
{
unsigned char id = (!offline ? 197 : 198);
- increase = !!(defs[id].flags & ATTRFLAG_INCREASING);
+ const ata_vendor_attr_defs::entry & def = defs[id];
+ if (def.flags & ATTRFLAG_INCREASING)
+ increase = true; // '-v 19[78],increasing' option
+ else if (def.name.empty() || (id == 198 && def.name == "Offline_Scan_UNC_SectCt"))
+ increase = false; // no or '-v 198,offlinescanuncsectorct' option
+ else
+ id = 0; // other '-v 19[78],...' option
return id;
}
{"194,10xCelsius" , "194,temp10x,Temperature_Celsius_x10"},
{"194,unknown" , "194,raw48,Unknown_Attribute"},
{"197,increasing" , "197,raw48+,Total_Pending_Sectors"}, // '+' sets flag
- {"198,offlinescanuncsectorct" , "198,raw48,Offline_Scan_UNC_SectCt"},
+ {"198,offlinescanuncsectorct" , "198,raw48,Offline_Scan_UNC_SectCt"}, // see also get_unc_attr_id() above
{"198,increasing" , "198,raw48+,Total_Offl_Uncorrectabl"}, // '+' sets flag
{"200,writeerrorcount" , "200,raw48,Write_Error_Count"},
{"201,detectedtacount" , "201,raw48,Detected_TA_Count"},
#include "utility.h"
#include "knowndrives.h"
-const char * ataprint_cpp_cvsid = "$Id: ataprint.cpp 3065 2010-02-10 22:16:50Z chrfranke $"
+const char * ataprint_cpp_cvsid = "$Id: ataprint.cpp 3081 2010-04-03 19:39:11Z chrfranke $"
ATAPRINT_H_CVSID;
// for passing global control variables
}
// Print SMART Extended Self-test Log (GP Log 0x07)
-static void PrintSmartExtSelfTestLog(const ata_smart_extselftestlog * log,
+static bool PrintSmartExtSelfTestLog(const ata_smart_extselftestlog * log,
unsigned nsectors, unsigned max_entries)
{
pout("SMART Extended Self-test Log Version: %u (%u sectors)\n",
if (!log->log_desc_index){
pout("No self-tests have been logged. [To run self-tests, use: smartctl -t]\n\n");
- return;
+ return true;
}
// Check index
unsigned logidx = log->log_desc_index;
if (logidx > nentries) {
pout("Invalid Self-test Log index = 0x%04x (reserved = 0x%02x)\n", logidx, log->reserved1);
- return;
+ return false;
}
// Index base is not clearly specified by ATA8-ACS (T13/1699-D Revision 6a),
false /*!print_error_only*/, print_header);
}
pout("\n");
+ return true;
}
static void ataPrintSelectiveSelfTestLog(const ata_selective_self_test_log * log, const ata_smart_values * sv)
EXIT(returnvalue|FAILCMD);
}
-// Initialize to zero just in case some SMART routines don't work
-static ata_identify_device drive;
-static ata_smart_values smartval;
-static ata_smart_thresholds_pvt smartthres;
-static ata_smart_errorlog smarterror;
-static ata_smart_selftestlog smartselftest;
-
int ataPrintMain (ata_device * device, const ata_print_options & options)
{
- int timewait,code;
- int returnval=0, retid=0, supported=0, needupdate=0;
- const char * powername = 0; char powerchg = 0;
+ int returnval = 0;
// If requested, check power mode first
+ const char * powername = 0;
+ bool powerchg = false;
if (options.powermode) {
unsigned char powerlimit = 0xff;
int powermode = ataCheckPowerMode(device);
}
}
+ // SMART values needed ?
+ bool need_smart_val = (
+ options.smart_check_status
+ || options.smart_general_values
+ || options.smart_vendor_attrib
+ || options.smart_error_log
+ || options.smart_selftest_log
+ || options.smart_selective_selftest_log
+ || options.smart_ext_error_log
+ || options.smart_ext_selftest_log
+ || options.smart_auto_offl_enable
+ || options.smart_auto_offl_disable
+ || options.smart_selftest_type != -1
+ );
+
+ // SMART must be enabled ?
+ bool need_smart_enabled = (
+ need_smart_val
+ || options.smart_auto_save_enable
+ || options.smart_auto_save_disable
+ );
+
+ // SMART feature set needed ?
+ bool need_smart_support = (
+ need_smart_enabled
+ || options.smart_enable
+ || options.smart_disable
+ );
+
// Start by getting Drive ID information. We need this, to know if SMART is supported.
- if ((retid=ataReadHDIdentity(device,&drive))<0){
+ ata_identify_device drive; memset(&drive, 0, sizeof(drive));
+ int retid = ataReadHDIdentity(device,&drive);
+ if (retid < 0) {
pout("Smartctl: Device Read Identity Failed (not an ATA/ATAPI device)\n\n");
failuretest(MANDATORY_CMD, returnval|=FAILID);
}
known = PrintDriveInfo(&drive, options.fix_swapped_id);
}
- // Was this a packet device?
- if (retid>0){
- pout("SMART support is: Unavailable - Packet Interface Devices [this device: %s] don't support ATA SMART\n", packetdevicetype(retid-1));
- failuretest(MANDATORY_CMD, returnval|=FAILSMART);
- }
-
- // if drive does not supports SMART it's time to exit
- supported=ataSmartSupport(&drive);
- if (supported != 1){
- if (supported==0) {
- pout("SMART support is: Unavailable - device lacks SMART capability.\n");
- failuretest(MANDATORY_CMD, returnval|=FAILSMART);
- pout(" Checking to be sure by trying SMART ENABLE command.\n");
- }
- else {
- pout("SMART support is: Ambiguous - ATA IDENTIFY DEVICE words 82-83 don't show if SMART supported.\n");
- if (!known) failuretest(MANDATORY_CMD, returnval|=FAILSMART);
- pout(" Checking for SMART support by trying SMART ENABLE command.\n");
- }
+ // Check and print SMART support and state
+ int smart_supported = -1, smart_enabled = -1;
+ if (need_smart_support || options.drive_info) {
- if (ataEnableSmart(device)){
- pout(" SMART ENABLE failed - this establishes that this device lacks SMART functionality.\n");
- failuretest(MANDATORY_CMD, returnval|=FAILSMART);
- supported=0;
- }
- else {
- pout(" SMART ENABLE appeared to work! Continuing.\n");
- supported=1;
- }
- if (!options.drive_info)
- pout("\n");
- }
-
- // Now print remaining drive info: is SMART enabled?
- if (options.drive_info) {
- int ison=ataIsSmartEnabled(&drive),isenabled=ison;
-
- if (ison==-1) {
- pout("SMART support is: Ambiguous - ATA IDENTIFY DEVICE words 85-87 don't show if SMART is enabled.\n");
- failuretest(MANDATORY_CMD, returnval|=FAILSMART);
- // check SMART support by trying a command
- pout(" Checking to be sure by trying SMART RETURN STATUS command.\n");
- isenabled=ataDoesSmartWork(device);
+ // Packet device ?
+ if (retid > 0) {
+ pout("SMART support is: Unavailable - Packet Interface Devices [this device: %s] don't support ATA SMART\n",
+ packetdevicetype(retid-1));
}
else {
- pout("SMART support is: Available - device has SMART capability.\n");
- if (device->ata_identify_is_cached()) {
- pout(" %sabled status cached by OS, trying SMART RETURN STATUS cmd.\n",
- (isenabled?"En":"Dis"));
- isenabled=ataDoesSmartWork(device);
+ // Disk device: SMART supported and enabled ?
+ smart_supported = ataSmartSupport(&drive);
+ smart_enabled = ataIsSmartEnabled(&drive);
+
+ if (smart_supported < 0)
+ pout("SMART support is: Ambiguous - ATA IDENTIFY DEVICE words 82-83 don't show if SMART supported.\n");
+ if (smart_supported && smart_enabled < 0) {
+ pout("SMART support is: Ambiguous - ATA IDENTIFY DEVICE words 85-87 don't show if SMART is enabled.\n");
+ if (need_smart_support) {
+ failuretest(MANDATORY_CMD, returnval|=FAILSMART);
+ // check SMART support by trying a command
+ pout(" Checking to be sure by trying SMART RETURN STATUS command.\n");
+ if (ataDoesSmartWork(device))
+ smart_supported = smart_enabled = 1;
+ }
+ }
+ else if (smart_supported < 0 && (smart_enabled > 0 || known))
+ // Assume supported if enabled or in drive database
+ smart_supported = 1;
+
+ if (smart_supported < 0)
+ pout("SMART support is: Unknown - Try option -s with argument 'on' to enable it.");
+ else if (!smart_supported)
+ pout("SMART support is: Unavailable - device lacks SMART capability.\n");
+ else {
+ if (options.drive_info)
+ pout("SMART support is: Available - device has SMART capability.\n");
+ if (smart_enabled >= 0) {
+ if (device->ata_identify_is_cached()) {
+ if (options.drive_info)
+ pout(" %sabled status cached by OS, trying SMART RETURN STATUS cmd.\n",
+ (smart_enabled?"En":"Dis"));
+ smart_enabled = ataDoesSmartWork(device);
+ }
+ if (options.drive_info)
+ pout("SMART support is: %s\n",
+ (smart_enabled ? "Enabled" : "Disabled"));
+ }
}
}
+ }
- if (isenabled)
- pout("SMART support is: Enabled\n");
- else {
- if (ison==-1)
- pout("SMART support is: Unavailable\n");
- else
- pout("SMART support is: Disabled\n");
- }
+ // Print remaining drive info
+ if (options.drive_info) {
// Print the (now possibly changed) power mode if available
if (powername)
pout("Power mode %s %s\n", (powerchg?"was:":"is: "), powername);
pout("\n");
}
-
+
+ // Exit if SMART is not supported but must be available to proceed
+ if (smart_supported <= 0 && need_smart_support)
+ failuretest(MANDATORY_CMD, returnval|=FAILSMART);
+
// START OF THE ENABLE/DISABLE SECTION OF THE CODE
if ( options.smart_disable || options.smart_enable
|| options.smart_auto_save_disable || options.smart_auto_save_enable
pout("Smartctl: SMART Enable Failed.\n\n");
failuretest(MANDATORY_CMD, returnval|=FAILSMART);
}
- else
+ else {
pout("SMART Enabled.\n");
+ smart_enabled = 1;
+ }
}
-
- // From here on, every command requires that SMART be enabled...
- if (!ataDoesSmartWork(device)) {
- pout("SMART Disabled. Use option -s with argument 'on' to enable it.\n");
- return returnval;
- }
-
+
// Turn off SMART on device
if (options.smart_disable) {
if (ataDisableSmart(device)) {
pout( "Smartctl: SMART Disable Failed.\n\n");
failuretest(MANDATORY_CMD,returnval|=FAILSMART);
}
+ }
+
+ // Exit if SMART is disabled but must be enabled to proceed
+ if (options.smart_disable || (smart_enabled <= 0 && need_smart_enabled)) {
pout("SMART Disabled. Use option -s with argument 'on' to enable it.\n");
- return returnval;
+ return returnval;
}
-
- // Let's ALWAYS issue this command to get the SMART status
- code=ataSmartStatus2(device);
- if (code==-1)
- failuretest(MANDATORY_CMD, returnval|=FAILSMART);
// Enable/Disable Auto-save attributes
if (options.smart_auto_save_enable) {
else
pout("SMART Attribute Autosave Disabled.\n");
}
-
- // for everything else read values and thresholds are needed
- if (ataReadSmartValues(device, &smartval)){
- pout("Smartctl: SMART Read Values failed.\n\n");
- failuretest(OPTIONAL_CMD, returnval|=FAILSMART);
- }
- if (ataReadSmartThresholds(device, &smartthres)){
- pout("Smartctl: SMART Read Thresholds failed.\n\n");
- failuretest(OPTIONAL_CMD, returnval|=FAILSMART);
+
+ // Read SMART values and thresholds if necessary
+ ata_smart_values smartval; memset(&smartval, 0, sizeof(smartval));
+ ata_smart_thresholds_pvt smartthres; memset(&smartthres, 0, sizeof(smartthres));
+ bool smart_val_ok = false, smart_thres_ok = false;
+
+ if (need_smart_val) {
+ if (ataReadSmartValues(device, &smartval)) {
+ pout("Smartctl: SMART Read Values failed.\n\n");
+ failuretest(OPTIONAL_CMD, returnval|=FAILSMART);
+ }
+ else {
+ smart_val_ok = true;
+
+ if (options.smart_check_status || options.smart_vendor_attrib) {
+ if (ataReadSmartThresholds(device, &smartthres)){
+ pout("Smartctl: SMART Read Thresholds failed.\n\n");
+ failuretest(OPTIONAL_CMD, returnval|=FAILSMART);
+ }
+ else
+ smart_thres_ok = true;
+ }
+ }
}
// Enable/Disable Off-line testing
+ bool needupdate = false;
if (options.smart_auto_offl_enable) {
if (!isSupportAutomaticTimer(&smartval)){
pout("Warning: device does not support SMART Automatic Timers.\n\n");
failuretest(OPTIONAL_CMD, returnval|=FAILSMART);
}
- needupdate=1;
+ needupdate = smart_val_ok;
if (ataEnableAutoOffline(device)){
pout( "Smartctl: SMART Enable Automatic Offline Failed.\n\n");
failuretest(OPTIONAL_CMD, returnval|=FAILSMART);
pout("Warning: device does not support SMART Automatic Timers.\n\n");
failuretest(OPTIONAL_CMD, returnval|=FAILSMART);
}
- needupdate=1;
+ needupdate = smart_val_ok;
if (ataDisableAutoOffline(device)){
pout("Smartctl: SMART Disable Automatic Offline Failed.\n\n");
failuretest(OPTIONAL_CMD, returnval|=FAILSMART);
if (needupdate && ataReadSmartValues(device, &smartval)){
pout("Smartctl: SMART Read Values failed.\n\n");
failuretest(OPTIONAL_CMD, returnval|=FAILSMART);
+ smart_val_ok = false;
}
// all this for a newline!
|| options.sct_temp_sts || options.sct_temp_hist )
pout("=== START OF READ SMART DATA SECTION ===\n");
- // Check SMART status (use previously returned value)
+ // Check SMART status
if (options.smart_check_status) {
- switch (code) {
+
+ switch (ataSmartStatus2(device)) {
case 0:
// The case where the disk health is OK
pout("SMART overall-health self-assessment test result: PASSED\n");
- if (find_failed_attr(&smartval, &smartthres, options.attribute_defs, 0)){
+ if (smart_thres_ok && find_failed_attr(&smartval, &smartthres, attribute_defs, 0)) {
if (options.smart_vendor_attrib)
pout("See vendor-specific Attribute list for marginal Attributes.\n\n");
else {
pout("SMART overall-health self-assessment test result: FAILED!\n"
"Drive failure expected in less than 24 hours. SAVE ALL DATA.\n");
PRINT_OFF(con);
- if (find_failed_attr(&smartval, &smartthres, options.attribute_defs, 1)){
+ if (smart_thres_ok && find_failed_attr(&smartval, &smartthres, attribute_defs, 1)) {
returnval|=FAILATTR;
if (options.smart_vendor_attrib)
pout("See vendor-specific Attribute list for failed Attributes.\n\n");
case -1:
default:
- // The case where something went wrong with HDIO_DRIVE_TASK ioctl()
- if (find_failed_attr(&smartval, &smartthres, options.attribute_defs, 1)){
+ // Something went wrong with the SMART STATUS command.
+ // The ATA SMART RETURN STATUS command provides the result in the ATA output
+ // registers. Buggy ATA/SATA drivers and SAT Layers often do not properly
+ // return the registers values.
+ failuretest(OPTIONAL_CMD, returnval|=FAILSMART);
+ if (!(smart_val_ok && smart_thres_ok)) {
+ PRINT_ON(con);
+ pout("SMART overall-health self-assessment test result: UNKNOWN!\n"
+ "SMART Status, Attributes and Thresholds cannot be read.\n\n");
+ }
+ else if (find_failed_attr(&smartval, &smartthres, attribute_defs, 1)) {
PRINT_ON(con);
pout("SMART overall-health self-assessment test result: FAILED!\n"
"Drive failure expected in less than 24 hours. SAVE ALL DATA.\n");
}
else {
pout("SMART overall-health self-assessment test result: PASSED\n");
- if (find_failed_attr(&smartval, &smartthres, options.attribute_defs, 0)){
+ pout("Warning: This result is based on an Attribute check.\n");
+ if (find_failed_attr(&smartval, &smartthres, attribute_defs, 0)) {
if (options.smart_vendor_attrib)
pout("See vendor-specific Attribute list for marginal Attributes.\n\n");
else {
} // end of checking SMART Status
// Print general SMART values
- if (options.smart_general_values)
+ if (smart_val_ok && options.smart_general_values)
PrintGeneralSmartValues(&smartval, &drive, fix_firmwarebug);
// Print vendor-specific attributes
- if (options.smart_vendor_attrib) {
+ if (smart_val_ok && options.smart_vendor_attrib) {
PRINT_ON(con);
PrintSmartAttribWithThres(&smartval, &smartthres, attribute_defs,
(con->printing_switchable ? 2 : 0));
pout("Warning: device does not support Error Logging\n");
failuretest(OPTIONAL_CMD, returnval|=FAILSMART);
}
+ ata_smart_errorlog smarterror; memset(&smarterror, 0, sizeof(smarterror));
if (ataReadErrorLog(device, &smarterror, fix_firmwarebug)){
pout("Smartctl: SMART Error Log Read Failed\n");
failuretest(OPTIONAL_CMD, returnval|=FAILSMART);
if (!ataReadExtSelfTestLog(device, log_07, nsectors))
failuretest(OPTIONAL_CMD, returnval|=FAILSMART);
else {
- PrintSmartExtSelfTestLog(log_07, nsectors, options.smart_ext_selftest_log);
- ok = true;
+ if (!PrintSmartExtSelfTestLog(log_07, nsectors, options.smart_ext_selftest_log))
+ returnval |= FAILLOG;
+ else
+ ok = true;
}
}
pout("Warning: device does not support Self Test Logging\n");
failuretest(OPTIONAL_CMD, returnval|=FAILSMART);
}
+ ata_smart_selftestlog smartselftest; memset(&smartselftest, 0, sizeof(smartselftest));
if(ataReadSelfTestLog(device, &smartselftest, fix_firmwarebug)){
pout("Smartctl: SMART Self Test Log Read Failed\n");
failuretest(OPTIONAL_CMD, returnval|=FAILSMART);
}
// START OF THE TESTING SECTION OF THE CODE. IF NO TESTING, RETURN
- if (options.smart_selftest_type == -1)
+ if (!smart_val_ok || options.smart_selftest_type == -1)
return returnval;
pout("=== START OF OFFLINE IMMEDIATE AND SELF-TEST SECTION ===\n");
}
// Now say how long the test will take to complete
- if ((timewait = TestTime(&smartval, options.smart_selftest_type))) {
+ int timewait = TestTime(&smartval, options.smart_selftest_type);
+ if (timewait) {
time_t t=time(NULL);
if (options.smart_selftest_type == OFFLINE_FULL_SCAN) {
t+=timewait;
# define _HAVE_CCISS
# endif
# include <asm/byteorder.h>
-# define be32toh __be32_to_cpu
+# ifndef be32toh
+# define be32toh __be32_to_cpu
+# endif
#elif defined(__FreeBSD__) && defined(HAVE_DEV_CISS_CISSIO_H)
# include <sys/endian.h>
# include <dev/ciss/cissio.h>
#
-# $Id: configure.in 3074 2010-03-05 23:00:30Z chrfranke $
+# $Id: configure.in 3116 2010-06-03 11:03:29Z chrfranke $
#
dnl Process this file with autoconf to produce a configure script.
AC_PREREQ(2.50)
AC_CONFIG_SRCDIR(smartctl.cpp)
smartmontools_configure_date=`date -u +'%Y-%m-%d %T %Z'`
-smartmontools_cvs_tag=`echo '$Id: configure.in 3074 2010-03-05 23:00:30Z chrfranke $'`
+smartmontools_cvs_tag=`echo '$Id: configure.in 3116 2010-06-03 11:03:29Z chrfranke $'`
smartmontools_release_date=2009-12-09
smartmontools_release_time="21:00:32 UTC"
CPPFLAGS="$CPPFLAGS -mno-cygwin"
LDFLAGS="$LDFLAGS -mno-cygwin"
fi
- CPPFLAGS="$CPPFLAGS -idirafter ${srcdir}/posix -idirafter ${srcdir}/os_win32"
+ CPPFLAGS="$CPPFLAGS -I$srcdir/os_win32"
;;
*-*-freebsd*)
CPPFLAGS="$CPPFLAGS -I/usr/src/sys"
AC_CHECK_TYPES([int64_t, uint64_t])
dnl Checks for library functions.
-AC_CHECK_FUNCS([getopt_long], , [
- AC_MSG_NOTICE([smartmontools does no longer support platforms without getopt_long().])
- AC_MSG_NOTICE([Please inform ${PACKAGE_BUGREPORT},])
- AC_MSG_NOTICE([including details about your build environment.])
- AC_MSG_ERROR([function getopt_long() not found])
-])
+AC_CHECK_FUNCS([getopt_long], [need_getopt_long=no], [need_getopt_long=yes])
+AM_CONDITIONAL(NEED_GETOPT_LONG, [test "$need_getopt_long" = "yes"])
+AC_CHECK_FUNCS([regcomp], [need_regex=no], [need_regex=yes])
+AM_CONDITIONAL(NEED_REGEX, [test "$need_regex" = "yes"])
AC_CHECK_FUNCS([getdomainname])
AC_CHECK_FUNCS([gethostname])
AC_CHECK_FUNCS([strtoull])
AC_CHECK_FUNCS([uname])
+# Check byte ordering (defines WORDS_BIGENDIAN)
+AC_C_BIGENDIAN
+
# Check whether snprintf appends null char and returns expected length on overflow
AH_TEMPLATE(HAVE_WORKING_SNPRINTF, [Define to 1 if the `snprintf' function is sane])
AC_MSG_CHECKING([for working snprintf])
AC_OUTPUT(Makefile examplescripts/Makefile)
AC_PROG_MAKE_SET
-# Print note that the docdir default value has changed
-# TODO: Remove this after next release
-if test "$docdir_is_default" = "yes"; then
- old_def_docdir='${prefix}/share/doc/${PACKAGE}-${VERSION}'
- old_def_docdir_eval="`eval eval eval echo $old_def_docdir`"
- docdir_eval="`eval eval eval echo $docdir`"
- AC_MSG_NOTICE([********** PLEASE NOTE **********])
- AC_MSG_NOTICE(['docdir' default has changed])
- AC_MSG_NOTICE([from: $old_def_docdir_eval])
- AC_MSG_NOTICE([to: $docdir_eval])
- AC_MSG_NOTICE([*********************************])
-fi
+echo "-----------------------------------------------------------------------------" >&AS_MESSAGE_FD
+echo "${PACKAGE}-${VERSION} configuration:" >&AS_MESSAGE_FD
+echo "host operating system: $host" >&AS_MESSAGE_FD
+echo "C++ compiler: $CXX" >&AS_MESSAGE_FD
+echo "preprocessor flags: $CPPFLAGS" >&AS_MESSAGE_FD
+echo "C++ compiler flags: $CXXFLAGS" >&AS_MESSAGE_FD
+echo "linker flags: $LDFLAGS" >&AS_MESSAGE_FD
+
+case "$host_os" in
+ mingw*)
+ if test -n "$drivedbdir"; then
+ echo "drive database file: EXEDIR/drivedb.h" >&AS_MESSAGE_FD
+ else
+ echo "drive database file: [[disabled]]" >&AS_MESSAGE_FD
+ fi
+ if test -n "$savestates"; then
+ echo "smartd save files: `eval eval eval echo $savestates`MODEL-SERIAL.TYPE.state" >&AS_MESSAGE_FD
+ fi
+ if test -n "$attributelog"; then
+ echo "smartd attribute logs: `eval eval eval echo $attributelog`MODEL-SERIAL.TYPE.csv" >&AS_MESSAGE_FD
+ fi
+ ;;
+
+ *)
+ echo "binary install path: `eval eval eval echo $sbindir`" >&AS_MESSAGE_FD
+ echo "man page install path: `eval eval eval echo $mandir`" >&AS_MESSAGE_FD
+ echo "doc file install path: `eval eval eval echo $docdir`" >&AS_MESSAGE_FD
+ if test "$docdir_is_default" = "yes"; then
+ echo "(NOTE: old default was: `eval eval eval echo ${prefix}/share/doc/${PACKAGE}-${VERSION}`)" >&AS_MESSAGE_FD
+ fi
+ if test -n "$drivedbdir"; then
+ echo "drive database file: `eval eval eval echo $drivedbdir`/drivedb.h" >&AS_MESSAGE_FD
+ echo "database update script: `eval eval eval echo $sbindir`/update-smart-drivedb" >&AS_MESSAGE_FD
+ else
+ echo "drive database file: [[disabled]]" >&AS_MESSAGE_FD
+ echo "database update script: [[disabled]]" >&AS_MESSAGE_FD
+ fi
+ echo "local drive database: `eval eval eval echo $sysconfdir`/smart_drivedb.h" >&AS_MESSAGE_FD
+ echo "smartd config file: `eval eval eval echo $sysconfdir`/smartd.conf${smartd_suffix}" >&AS_MESSAGE_FD
+ echo "smartd initd script: `eval eval eval echo $initddir`/smartd${smartd_suffix}" >&AS_MESSAGE_FD
+ if test -n "$savestates"; then
+ echo "smartd save files: `eval eval eval echo $savestates`MODEL-SERIAL.TYPE.state" >&AS_MESSAGE_FD
+ else
+ echo "smartd save files: [[disabled]]" >&AS_MESSAGE_FD
+ fi
+ if test -n "$attributelog"; then
+ echo "smartd attribute logs: `eval eval eval echo $attributelog`MODEL-SERIAL.TYPE.csv" >&AS_MESSAGE_FD
+ else
+ echo "smartd attribute logs: [[disabled]]" >&AS_MESSAGE_FD
+ fi
+ echo "libcap-ng support: $use_libcap_ng" >&AS_MESSAGE_FD
+ case "$host_os" in
+ linux*) echo "SELinux support: ${with_selinux-no}" >&AS_MESSAGE_FD ;;
+ esac
+ ;;
+esac
+echo "-----------------------------------------------------------------------------" >&AS_MESSAGE_FD
*
* Home page of code is: http://smartmontools.sourceforge.net
*
- * Copyright (C) 2008-9 Christian Franke <smartmontools-support@lists.sourceforge.net>
+ * Copyright (C) 2008-10 Christian Franke <smartmontools-support@lists.sourceforge.net>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
#include "dev_interface.h"
#include "dev_ata_cmd_set.h"
-const char * dev_legacy_cpp_cvsid = "$Id: dev_legacy.cpp 2973 2009-10-26 22:38:19Z chrfranke $"
+const char * dev_legacy_cpp_cvsid = "$Id: dev_legacy.cpp 3098 2010-04-30 17:35:35Z chrfranke $"
DEV_INTERFACE_H_CVSID;
extern smartmonctrl * con; // con->reportscsiioctl
static void free_devnames(char * * devnames, int numdevs)
{
- static const char version[] = "$Id: dev_legacy.cpp 2973 2009-10-26 22:38:19Z chrfranke $";
+ static const char version[] = "$Id: dev_legacy.cpp 3098 2010-04-30 17:35:35Z chrfranke $";
for (int i = 0; i < numdevs; i++)
FreeNonZero(devnames[i], -1,__LINE__, version);
FreeNonZero(devnames, (sizeof (char*) * numdevs),__LINE__, version);
set_err(EINVAL, "Option -d cciss,N requires N to be a non-negative integer");
return 0;
}
- if (!(0 <= disknum && disknum <= 15)) {
- set_err(EINVAL, "Option -d cciss,N (N=%d) must have 0 <= N <= 15", disknum);
+ if (!(0 <= disknum && disknum <= 127)) {
+ set_err(EINVAL, "Option -d cciss,N (N=%d) must have 0 <= N <= 127", disknum);
return 0;
}
return new legacy_cciss_device(this, name, disknum);
* modelfamily Informal string about the model family/series of a
* device. Set to "" if no info (apart from device id)
* known. The entry is ignored if this string starts with
- * a dollar sign.
+ * a dollar sign. Must not start with "USB:", see below.
* modelregexp POSIX extended regular expression to match the model of
* a device. This should never be "".
* firmwareregexp POSIX extended regular expression to match a devices's
* The table will be searched from the start to end or until the first match,
* so the order in the table is important for distinct entries that could match
* the same drive.
+ *
+ *
+ * Format for USB ID entries:
+ *
+ * modelfamily String with format "USB: DEVICE; BRIDGE" where
+ * DEVICE is the name of the device and BRIDGE is
+ * the name of the USB bridge. Both may be empty
+ * if no info known.
+ * modelregexp POSIX extended regular expression to match the USB
+ * vendor:product ID in hex notation ("0x1234:0xabcd").
+ * This should never be "".
+ * firmwareregexp POSIX extended regular expression to match the USB
+ * bcdDevice info. Only compared during search if other
+ * entries with same USB vendor:product ID exist.
+ * warningmsg Not used yet.
+ * presets String with one device type ('-d') option.
+ *
*/
/*
const drive_settings builtin_knowndrives[] = {
*/
- { "$Id: drivedb.h 3071 2010-03-04 21:17:09Z manfred99 $",
+ { "$Id: drivedb.h 3124 2010-07-12 19:21:00Z chrfranke $",
"-", "-",
"This is a dummy entry to hold the SVN-Id of drivedb.h",
""
{ "SuperTalent UltraDrive GX SSD",
"STT_FT[MD](28|32|56|64)GX25H",
"", "",
+ " -v 1,raw64"
" -v 9,raw64"
" -v 12,raw64"
" -v 184,raw64,Initial_Bad_Block_Count"
" -v 207,raw64,Max_Erase_Count"
" -v 208,raw64,Average_Erase_Count"
" -v 209,raw64,Remaining_Lifetime_Perc"
+ " -v 210,raw64"
+ " -v 211,raw64"
+ " -v 212,raw64"
+ " -v 213,raw64"
},
{ "Patriot Torqx SSD",
"Patriot[ -]Torqx.*",
"", "",
+ " -v 1,raw64"
" -v 9,raw64"
" -v 12,raw64"
" -v 184,raw64,Initial_Bad_Block_Count"
" -v 207,raw64,Max_Erase_Count"
" -v 208,raw64,Average_Erase_Count"
" -v 209,raw64,Remaining_Lifetime_Perc"
+ " -v 210,raw64"
+ " -v 211,raw64"
+ " -v 212,raw64"
+ " -v 213,raw64"
},
{ "OCZ Vertex SSD",
"OCZ[ -]VERTEX.*",
"", "",
+ " -v 1,raw64"
" -v 9,raw64"
" -v 12,raw64"
" -v 184,raw64,Initial_Bad_Block_Count"
" -v 207,raw64,Max_Erase_Count"
" -v 208,raw64,Average_Erase_Count"
" -v 209,raw64,Remaining_Lifetime_Perc"
+ " -v 210,raw64"
+ " -v 211,raw64"
+ " -v 212,raw64"
+ " -v 213,raw64"
},
{ "OCZ Agility SSD",
- "OCZ[ -]AGILITY",
+ "OCZ[ -]AGILITY.*",
"", "",
+ " -v 1,raw64"
" -v 9,raw64"
" -v 12,raw64"
" -v 184,raw64,Initial_Bad_Block_Count"
" -v 207,raw64,Max_Erase_Count"
" -v 208,raw64,Average_Erase_Count"
" -v 209,raw64,Remaining_Lifetime_Perc"
+ " -v 210,raw64"
+ " -v 211,raw64"
+ " -v 212,raw64"
+ " -v 213,raw64"
},
{ "Crucial M225 SSD",
"CRUCIAL_CT(64|128|256)M225",
" -v 209,raw64,Remaining_Lifetime_Perc"
" -v 210,raw64"
" -v 211,raw64"
+ " -v 212,raw64"
+ " -v 213,raw64"
},
{ "Intel X25-E SSD",
"SSDSA2SH(032|064)G1.* INTEL", // G1 = first generation
"-v 225,raw48,Host_Writes_Count"
},
{ "Intel X25-M SSD",
- "INTEL SSDSA2M(080|160)G2.*", // G2 = second generation
+ "INTEL SSDSA2MH(080|160)G1.*", // G1 = first generation, 50nm
"", "",
"-v 225,raw48,Host_Writes_Count"
},
- { "Transcend Solid-State Drive",
+ { "Intel X25-M SSD",
+ "INTEL SSDSA2M(080|160)G2.*", // G2 = second generation, 34nm
+ "", "",
+ "-v 225,raw48,Host_Writes_Count"
+ },
+ { "Transcend IDE Solid State Drive",
"TS(8|16|32|64|128)GSSD25-(M|S)",
"", "", ""
},
- { "Transcend Solid-State Drive V series",
+ { "Transcend SATA Solid State Drive",
"TS(8|16|32|64|128|192)GSSD25S-(M|S)",
+ "", "",
+ "-v 229,hex64,Halt_System_ID "
+ "-v 232,hex64,Firmware_Version_information "
+ "-v 233,hex64,ECC_Fail_Record "
+ "-v 234,raw24/raw24,Erase_Count_Avg/Max "
+ "-v 235,raw24/raw24,Block_Count_Good/System"
+ },
+ { "Transcend Ultra Series Solid State Drive (SATA II)",
+ "TS(60|120)GSSD25D-M",
"", "", ""
},
{ "Marvell SSD SD88SA024BA0 (SUN branded)",
"http://www.ibm.com/pc/support/site.wss/MIGR-42215.html",
""
},
- { "", // ExcelStor J240, J340, J360, J680, and J880
- "ExcelStor Technology J(24|34|36|68|88)0",
+ { "", // ExcelStor J240, J340, J360, J680, J880 and J8160
+ "ExcelStor Technology J(24|34|36|68|88|816)0",
"", "", ""
},
{ "", // Fujitsu M1623TAU
"SAMSUNG HM((061|080)G|(121|160)H|250J)I",
"", "", ""
},
+ { "SAMSUNG SpinPoint M series", // tested with MP0402H/UC100-11
+ "SAMSUNG MP0(302|402|603|804)H",
+ "",
+ "",
+ "-v 9,halfminutes"
+ },
/*
// TODO: Make the entries below more specific.
// These entries produce misleading results, because newer
"(Hitachi )?(HTS4212(60|80|10|12)H9AT00|HTS421260G9AT00)",
"", "", ""
},
- { "Hitachi Travelstar 5K80 family",
+ { "Hitachi Travelstar 5K80",
"(Hitachi )?HTS5480[8642]0M9AT00",
"", "", ""
},
- { "Hitachi Travelstar 5K100 series",
+ { "Hitachi Travelstar 5K100",
"(Hitachi )?HTS5410[1864]0G9(AT|SA)00",
"", "", ""
},
- { "Hitachi Travelstar E5K100 series",
+ { "Hitachi Travelstar E5K100",
"(Hitachi )?HTE541040G9(AT|SA)00",
"", "", ""
},
"(Hitachi )?HTS5412(60|80|10|12)H9(AT|SA)00",
"", "", ""
},
- { "Hitachi Travelstar 5K160 series",
+ { "Hitachi Travelstar 5K160",
"(Hitachi |HITACHI )?HTS5416([468]0|1[26])J9(AT|SA)00",
"", "", ""
},
- { "Hitachi Travelstar E5K160 series",
+ { "Hitachi Travelstar E5K160",
"(Hitachi )?HTE5416(12|16|60|80)J9(AT|SA)00",
"", "", ""
},
- { "Hitachi Travelstar 5K250 series",
+ { "Hitachi Travelstar 5K250",
"(Hitachi |HITACHI )?HTS5425(80|12|16|20|25)K9(A3|SA)00",
"", "", ""
},
- { "Hitachi Travelstar 5K320 series",
+ { "Hitachi Travelstar 5K320",
"(Hitachi |HITACHI )?HT(S|E)5432(80|12|16|25|32)L9(A3(00)?|SA01)",
"", "", ""
},
+ { "Hitachi Travelstar 5K500.B",
+ "(Hitachi )?HT[ES]5450(12|16|25|32|40|50)B9A30[01]",
+ "", "", ""
+ },
{ "Hitachi Travelstar 7K60",
"(Hitachi )?HTS726060M9AT00",
"", "", ""
"(IBM-)?IC35L(030|060|090|120|180)AVV207-[01]",
"", "", ""
},
- { "Hitachi Deskstar 7K80 series",
+ { "Hitachi Deskstar 7K80",
"(Hitachi )?HDS7280([48]0PLAT20|(40)?PLA320|80PLA380).*",
"", "", ""
},
"(Hitachi )?HDS7216(80|16)PLA[3T]80.*",
"", "", ""
},
- { "Hitachi Deskstar 7K250 series",
+ { "Hitachi Deskstar 7K250",
"(Hitachi )?HDS7225((40|80|12|16)VLAT20|(12|16|25)VLAT80|(80|12|16|25)VLSA80)",
"", "", ""
},
"HITACHI HDS7225SBSUN250G.*",
"", "", ""
},
- { "Hitachi Deskstar T7K250 series",
+ { "Hitachi Deskstar T7K250",
"(Hitachi )?HDT7225((25|20|16)DLA(T80|380))",
"", "", ""
},
- { "Hitachi Deskstar 7K400 series",
+ { "Hitachi Deskstar 7K400",
"(Hitachi )?HDS724040KL(AT|SA)80",
"", "", ""
},
- { "Hitachi Deskstar 7K500 series",
+ { "Hitachi Deskstar 7K500",
"(Hitachi )?HDS725050KLA(360|T80)",
"", "", ""
},
- { "Hitachi Deskstar P7K500 series",
+ { "Hitachi Deskstar P7K500",
"(Hitachi )?HDP7250(16|25|32|40|50)GLA(36|38|T8)0",
"", "", ""
},
"(Hitachi )?HDT7210((16|25)SLA380|(32|50|64|75|10)SLA360)",
"", "", ""
},
+ { "Hitachi Deskstar 7K1000.C",
+ "(Hitachi )?HDS7210((16|25)CLA382|(32|50)CLA362|(64|75|10)CLA332)",
+ "", "", ""
+ },
{ "Hitachi Deskstar 7K2000",
"Hitachi HDS722020ALA330",
"", "", ""
"(Hitachi )?HUA7210(50|75|10)KLA330",
"", "", ""
},
+ { "Hitachi Ultrastar A7K2000",
+ "(Hitachi )?HUA7220((50|10)C|20A)LA33[01]",
+ "", "", ""
+ },
{ "Toshiba 2.5\" HDD series (10-20 GB)",
"TOSHIBA MK(101[67]GAP|15[67]GAP|20(1[678]GAP|(18|23)GAS))",
"", "", ""
"WDC WD((50|64|75)00AA(C|V)S|(50|64|75)00AADS|10EA(C|V)S|(10|15|20)EADS)-.*",
"", "", ""
},
+ { "Western Digital Caviar Green (Adv. Format) family",
+ "WDC WD((64|80)00A|(10|15|20)E)ARS-.*",
+ "", "", ""
+ },
{ "Western Digital Caviar Black family",
"WDC WD((500|640|750)1AA|1001FA)LS-.*",
"", "", ""
"", "", ""
},
{ "Western Digital AV-GP family",
- "WDC WD((16|25|32|50|64|75)00AVVS|(50|75)00AVCS|10EVVS|(10|20)EVCS|WD(10|15|20)EVDS)-.*",
+ "WDC WD((16|25|32|50|64|75)00AVVS|(50|75)00AVCS|10EVVS|(10|20)EVCS|(10|15|20)EVDS)-.*",
"", "", ""
},
{ "Western Digital Raptor family",
- "WDC WD((360|740|800)GD|(360|740|1500)ADF[DS])-.*",
+ "WDC WD((360|740|800)GD|(360|740|800|1500)ADF[DS])-.*",
"", "", ""
},
{ "Western Digital Raptor X",
"", "", ""
},
{ "Western Digital My Passport Essential SE hard drive (USB interface)",
- "WDC WD7500KMVV-.*",
+ "WDC WD(7500K|10T)MVV-.*",
"", "", ""
},
{ "Western Digital My Passport hard drive (USB interface)",
"QUANTUM FIREBALLP KA(9|10).1",
"", "", ""
},
+
+ ////////////////////////////////////////////////////
+ // USB ID entries
+ ////////////////////////////////////////////////////
+
+ // ALi
+ { "USB: ; ALi M5621", // USB->PATA
+ "0x0402:0x5621",
+ "",
+ "",
+ "" // unsupported
+ },
+ // Cypress
+ { "USB: ; Cypress CY7C68300A (AT2)",
+ "0x04b4:0x6830",
+ "0x0001",
+ "",
+ "" // unsupported
+ },
+ { "USB: ; Cypress CY7C68300B/C (AT2LP)",
+ "0x04b4:0x6830",
+ "0x0240",
+ "",
+ "-d usbcypress"
+ },
+ // Myson Century
+ { "USB: ; Myson Century CS8818",
+ "0x04cf:0x8818",
+ "0xb007",
+ "",
+ "" // unsupported
+ },
+ // Samsung
+ { "USB: Samsung Story Station; ",
+ "0x04e8:0x5f06",
+ "",
+ "",
+ "-d sat"
+ },
+ // Sunplus
+ { "USB: ; SunPlus SPDIF215",
+ "0x04fc:0x0c15",
+ "0xf615",
+ "",
+ "-d usbsunplus"
+ },
+ { "USB: ; SunPlus SPDIF225", // USB+SATA->SATA
+ "0x04fc:0x0c25",
+ "0x0103",
+ "",
+ "-d usbsunplus"
+ },
+ // Iomega
+ { "USB: Iomega LPHD080-0; ",
+ "0x059b:0x0272",
+ "",
+ "",
+ "-d usbcypress"
+ },
+ { "USB: Iomega MDHD500-U; ",
+ "0x059b:0x0275",
+ "0x0001",
+ "",
+ "" // unsupported
+ },
+ { "USB: Iomega LDHD-UP; Sunplus",
+ "0x059b:0x0370",
+ "",
+ "",
+ "-d usbsunplus"
+ },
+ // LaCie
+ { "USB: LaCie hard disk (FA Porsche design);",
+ "0x059f:0x0651",
+ "",
+ "",
+ "" // unsupported
+ },
+ { "USB: LaCie hard disk; JMicron",
+ "0x059f:0x0951",
+ "",
+ "",
+ "-d usbjmicron"
+ },
+ { "USB: LaCie hard disk (Neil Poulton design);",
+ "0x059f:0x1018",
+ "",
+ "",
+ "-d sat"
+ },
+ { "USB: LaCie Desktop Hard Drive; JMicron",
+ "0x059f:0x1019",
+ "",
+ "",
+ "-d usbjmicron"
+ },
+ { "USB: LaCie Rugged Hard Drive; JMicron",
+ "0x059f:0x101d",
+ "0x0001",
+ "",
+ "-d usbjmicron,x"
+ },
+ // In-System Design
+ { "USB: ; In-System/Cypress ISD-300A1",
+ "0x05ab:0x0060",
+ "0x1101",
+ "",
+ "-d usbcypress"
+ },
+ // Genesys Logic
+ { "USB: ; Genesys Logic GL881E",
+ "0x05e3:0x0702",
+ "",
+ "",
+ "" // unsupported
+ },
+ { "USB: ; Genesys Logic", // TODO: requires '-T permissive'
+ "0x05e3:0x0718",
+ "0x0041",
+ "",
+ "-d sat"
+ },
+ // Prolific
+ { "USB: ; Prolific PL2507", // USB->PATA
+ "0x067b:0x2507",
+ "",
+ "",
+ "" // unsupported
+ },
+ { "USB: ; Prolific PL3507", // USB+IEE1394->PATA
+ "0x067b:0x3507",
+ "0x0001",
+ "",
+ "" // unsupported
+ },
+ // Freecom
+ { "USB: Freecom Hard Drive XS; Sunplus",
+ "0x07ab:0xfc8e",
+ "0x010f",
+ "",
+ "-d usbsunplus"
+ },
+ // Toshiba
+ { "USB: Toshiba PX1270E-1G16; Sunplus",
+ "0x0930:0x0b03",
+ "",
+ "",
+ "-d usbsunplus"
+ },
+ { "USB: Toshiba PX1396E-3T01; Sunplus", // similar to Dura Micro 501
+ "0x0930:0x0b09",
+ "",
+ "",
+ "-d usbsunplus"
+ },
+ // Seagate
+ { "USB: Seagate FreeAgent Go; ",
+ "0x0bc2:0x2(000|100|101)",
+ "",
+ "",
+ "-d sat"
+ },
+ { "USB: Seagate FreeAgent Go FW; ",
+ "0x0bc2:0x2200",
+ "",
+ "",
+ "-d sat"
+ },
+ { "USB: Seagate Expansion Portable; ",
+ "0x0bc2:0x2300",
+ "",
+ "",
+ "-d sat"
+ },
+ { "USB: Seagate FreeAgent Desktop; ",
+ "0x0bc2:0x3000",
+ "",
+ "",
+ "-d sat"
+ },
+ { "USB: Seagate FreeAgent Desk; ",
+ "0x0bc2:0x3001",
+ "",
+ "",
+ "-d sat"
+ },
+ // Dura Micro
+ { "USB: Dura Micro 509; Sunplus",
+ "0x0c0b:0xb159",
+ "0x0103",
+ "",
+ "-d usbsunplus"
+ },
+ // Maxtor
+ { "USB: Maxtor OneTouch; ",
+ "0x0d49:0x7300",
+ "0x0121",
+ "",
+ "-d sat"
+ },
+ { "USB: Maxtor OneTouch 4; ",
+ "0x0d49:0x7310",
+ "0x0125",
+ "",
+ "-d sat"
+ },
+ { "USB: Maxtor OneTouch 4 Mini; ",
+ "0x0d49:0x7350",
+ "0x0125",
+ "",
+ "-d sat"
+ },
+ { "USB: Maxtor Basics Desktop; ",
+ "0x0d49:0x7410",
+ "0x0122",
+ "",
+ "-d sat"
+ },
+ { "USB: Maxtor Basics Portable; ",
+ "0x0d49:0x7450",
+ "0x0122",
+ "",
+ "-d sat"
+ },
+ // Western Digital
+ { "USB: WD My Passport (IDE); Cypress",
+ "0x1058:0x0701",
+ "0x0240",
+ "",
+ "-d usbcypress"
+ },
+ { "USB: WD My Passport Portable; ",
+ "0x1058:0x0702",
+ "0x0102",
+ "",
+ "-d sat"
+ },
+ { "USB: WD My Passport Essential; ",
+ "0x1058:0x0704",
+ "0x0175",
+ "",
+ "-d sat"
+ },
+ { "USB: WD My Passport Elite; ",
+ "0x1058:0x0705",
+ "0x0175",
+ "",
+ "-d sat"
+ },
+ { "USB: WD My Passport 070A; ",
+ "0x1058:0x070a",
+ "0x1028",
+ "",
+ "-d sat"
+ },
+ { "USB: WD My Book ES; ",
+ "0x1058:0x0906",
+ "0x0012",
+ "",
+ "-d sat"
+ },
+ { "USB: WD Elements Desktop; ",
+ "0x1058:0x1001",
+ "0x0104",
+ "",
+ "-d sat"
+ },
+ { "USB: WD Elements Desktop WDE1UBK...; ",
+ "0x1058:0x1003",
+ "0x0175",
+ "",
+ "-d sat"
+ },
+ { "USB: WD Elements; ",
+ "0x1058:0x1010",
+ "0x0105",
+ "",
+ "-d sat"
+ },
+ { "USB: WD Elements Desktop; ", // 2TB
+ "0x1058:0x1021",
+ "0x2002",
+ "",
+ "-d sat"
+ },
+ { "USB: WD My Book Essential; ",
+ "0x1058:0x1100",
+ "0x0165",
+ "",
+ "-d sat"
+ },
+ { "USB: WD My Book; ",
+ "0x1058:0x1102",
+ "0x1028",
+ "",
+ "-d sat"
+ },
+ { "USB: WD My Book Essential; ",
+ "0x1058:0x1110",
+ "0x1030",
+ "",
+ "-d sat"
+ },
+ // A-DATA
+ { "USB: A-DATA SH93; Cypress",
+ "0x125f:0xa93a",
+ "0x0150",
+ "",
+ "-d usbcypress"
+ },
+ // Initio
+ { "USB: ; Initio 316000",
+ "0x13fd:0x0540",
+ "",
+ "",
+ "" // unsupported
+ },
+ { "USB: ; Initio", // USB->SATA
+ "0x13fd:0x1240",
+ "0x0104",
+ "",
+ "-d sat"
+ },
+ { "USB: ; Initio", // USB+SATA->SATA
+ "0x13fd:0x1340",
+ "0x0208",
+ "",
+ "-d sat"
+ },
+ // JMicron
+ { "USB: ; JMicron JM20329", // USB->SATA
+ "0x152d:0x2329",
+ "0x0100",
+ "",
+ "-d usbjmicron"
+ },
+ { "USB: ; JMicron JM20336", // USB+SATA->SATA, USB->2xSATA
+ "0x152d:0x2336",
+ "0x0100",
+ "",
+ "-d usbjmicron,x"
+ },
+ { "USB: ; JMicron JM20337/8", // USB->SATA+PATA, USB+SATA->PATA
+ "0x152d:0x2338",
+ "0x0100",
+ "",
+ "-d usbjmicron"
+ },
+ { "USB: ; JMicron JM20339", // USB->SATA
+ "0x152d:0x2339",
+ "0x0100",
+ "",
+ "-d usbjmicron,x"
+ },
+ { "USB: ; JMicron", // USB->SATA
+ "0x152d:0x2352",
+ "0x0100",
+ "",
+ "-d usbjmicron,x"
+ },
+ // Verbatim
+ { "USB: Verbatim FW/USB160; Oxford OXUF934SSA-LQAG", // USB+IEE1394->SATA
+ "0x18a5:0x0215",
+ "0x0001",
+ "",
+ "-d sat"
+ },
+ { "USB: Verbatim External Hard Drive 47519; Sunplus", // USB->SATA
+ "0x18a5:0x0216",
+ "",
+ "",
+ "-d usbsunplus"
+ },
+ // SunplusIT
+ { "USB: ; SunplusIT",
+ "0x1bcf:0x0c31",
+ "",
+ "",
+ "-d usbsunplus"
+ },
+ // Hitachi/SimpleTech
+ { "USB: Hitachi/SimpleTech; JMicron", // 1TB
+ "0x4971:0xce17",
+ "",
+ "",
+ "-d usbjmicron,x"
+ },
+ // OnSpec
+ { "USB: ; OnSpec", // USB->PATA
+ "0x55aa:0x2b00",
+ "0x0100",
+ "",
+ "" // unsupported
+ },
/*
}; // builtin_knowndrives[]
*/
--- /dev/null
+/* Getopt for GNU.
+ NOTE: getopt is now part of the C library, so if you don't know what
+ "Keep this file name-space clean" means, talk to drepper@gnu.org
+ before changing it!
+ Copyright (C) 1987,88,89,90,91,92,93,94,95,96,98,99,2000,2001,2002
+ Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, write to the Free
+ Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ 02111-1307 USA. */
+\f
+/* This tells Alpha OSF/1 not to define a getopt prototype in <stdio.h>.
+ Ditto for AIX 3.2 and <stdlib.h>. */
+#ifndef _NO_PROTO
+# define _NO_PROTO
+#endif
+
+#ifdef HAVE_CONFIG_H
+# include <config.h>
+#endif
+
+#if !defined __STDC__ || !__STDC__
+/* This is a separate conditional since some stdc systems
+ reject `defined (const)'. */
+# ifndef const
+# define const
+# endif
+#endif
+
+#include <stdio.h>
+
+/* Comment out all this code if we are using the GNU C Library, and are not
+ actually compiling the library itself. This code is part of the GNU C
+ Library, but also included in many other GNU distributions. Compiling
+ and linking in this code is a waste when using the GNU C library
+ (especially if it is a shared library). Rather than having every GNU
+ program understand `configure --with-gnu-libc' and omit the object files,
+ it is simpler to just do this in the source for each such file. */
+
+#define GETOPT_INTERFACE_VERSION 2
+#if !defined _LIBC && defined __GLIBC__ && __GLIBC__ >= 2
+# include <gnu-versions.h>
+# if _GNU_GETOPT_INTERFACE_VERSION == GETOPT_INTERFACE_VERSION
+# define ELIDE_CODE
+# endif
+#endif
+
+#ifndef ELIDE_CODE
+
+
+/* This needs to come after some library #include
+ to get __GNU_LIBRARY__ defined. */
+#ifdef __GNU_LIBRARY__
+/* Don't include stdlib.h for non-GNU C libraries because some of them
+ contain conflicting prototypes for getopt. */
+# include <stdlib.h>
+# include <unistd.h>
+#endif /* GNU C library. */
+
+#ifdef VMS
+# include <unixlib.h>
+# if HAVE_STRING_H - 0
+# include <string.h>
+# endif
+#endif
+
+#ifndef _
+/* This is for other GNU distributions with internationalized messages. */
+# if (HAVE_LIBINTL_H && ENABLE_NLS) || defined _LIBC
+# include <libintl.h>
+# ifndef _
+# define _(msgid) gettext (msgid)
+# endif
+# else
+# define _(msgid) (msgid)
+# endif
+# if defined _LIBC && defined USE_IN_LIBIO
+# include <wchar.h>
+# endif
+#endif
+
+#ifndef attribute_hidden
+# define attribute_hidden
+#endif
+
+/* This version of `getopt' appears to the caller like standard Unix `getopt'
+ but it behaves differently for the user, since it allows the user
+ to intersperse the options with the other arguments.
+
+ As `getopt' works, it permutes the elements of ARGV so that,
+ when it is done, all the options precede everything else. Thus
+ all application programs are extended to handle flexible argument order.
+
+ Setting the environment variable POSIXLY_CORRECT disables permutation.
+ Then the behavior is completely standard.
+
+ GNU application programs can use a third alternative mode in which
+ they can distinguish the relative order of options and other arguments. */
+
+#include "getopt.h"
+
+/* For communication from `getopt' to the caller.
+ When `getopt' finds an option that takes an argument,
+ the argument value is returned here.
+ Also, when `ordering' is RETURN_IN_ORDER,
+ each non-option ARGV-element is returned here. */
+
+char *optarg;
+
+/* Index in ARGV of the next element to be scanned.
+ This is used for communication to and from the caller
+ and for communication between successive calls to `getopt'.
+
+ On entry to `getopt', zero means this is the first call; initialize.
+
+ When `getopt' returns -1, this is the index of the first of the
+ non-option elements that the caller should itself scan.
+
+ Otherwise, `optind' communicates from one call to the next
+ how much of ARGV has been scanned so far. */
+
+/* 1003.2 says this must be 1 before any call. */
+int optind = 1;
+
+/* Formerly, initialization of getopt depended on optind==0, which
+ causes problems with re-calling getopt as programs generally don't
+ know that. */
+
+int __getopt_initialized attribute_hidden;
+
+/* The next char to be scanned in the option-element
+ in which the last option character we returned was found.
+ This allows us to pick up the scan where we left off.
+
+ If this is zero, or a null string, it means resume the scan
+ by advancing to the next ARGV-element. */
+
+static char *nextchar;
+
+/* Callers store zero here to inhibit the error message
+ for unrecognized options. */
+
+int opterr = 1;
+
+/* Set to an option character which was unrecognized.
+ This must be initialized on some systems to avoid linking in the
+ system's own getopt implementation. */
+
+int optopt = '?';
+
+/* Describe how to deal with options that follow non-option ARGV-elements.
+
+ If the caller did not specify anything,
+ the default is REQUIRE_ORDER if the environment variable
+ POSIXLY_CORRECT is defined, PERMUTE otherwise.
+
+ REQUIRE_ORDER means don't recognize them as options;
+ stop option processing when the first non-option is seen.
+ This is what Unix does.
+ This mode of operation is selected by either setting the environment
+ variable POSIXLY_CORRECT, or using `+' as the first character
+ of the list of option characters.
+
+ PERMUTE is the default. We permute the contents of ARGV as we scan,
+ so that eventually all the non-options are at the end. This allows options
+ to be given in any order, even with programs that were not written to
+ expect this.
+
+ RETURN_IN_ORDER is an option available to programs that were written
+ to expect options and other ARGV-elements in any order and that care about
+ the ordering of the two. We describe each non-option ARGV-element
+ as if it were the argument of an option with character code 1.
+ Using `-' as the first character of the list of option characters
+ selects this mode of operation.
+
+ The special argument `--' forces an end of option-scanning regardless
+ of the value of `ordering'. In the case of RETURN_IN_ORDER, only
+ `--' can cause `getopt' to return -1 with `optind' != ARGC. */
+
+static enum
+{
+ REQUIRE_ORDER, PERMUTE, RETURN_IN_ORDER
+} ordering;
+
+/* Value of POSIXLY_CORRECT environment variable. */
+static char *posixly_correct;
+\f
+#ifdef __GNU_LIBRARY__
+/* We want to avoid inclusion of string.h with non-GNU libraries
+ because there are many ways it can cause trouble.
+ On some systems, it contains special magic macros that don't work
+ in GCC. */
+# include <string.h>
+# define my_index strchr
+#else
+
+# if HAVE_STRING_H
+# include <string.h>
+# else
+# include <strings.h>
+# endif
+
+/* Avoid depending on library functions or files
+ whose names are inconsistent. */
+
+#ifndef getenv
+extern char *getenv ();
+#endif
+
+static char *
+my_index (str, chr)
+ const char *str;
+ int chr;
+{
+ while (*str)
+ {
+ if (*str == chr)
+ return (char *) str;
+ str++;
+ }
+ return 0;
+}
+
+/* If using GCC, we can safely declare strlen this way.
+ If not using GCC, it is ok not to declare it. */
+#ifdef __GNUC__
+/* Note that Motorola Delta 68k R3V7 comes with GCC but not stddef.h.
+ That was relevant to code that was here before. */
+# if (!defined __STDC__ || !__STDC__) && !defined strlen
+/* gcc with -traditional declares the built-in strlen to return int,
+ and has done so at least since version 2.4.5. -- rms. */
+extern int strlen (const char *);
+# endif /* not __STDC__ */
+#endif /* __GNUC__ */
+
+#endif /* not __GNU_LIBRARY__ */
+\f
+/* Handle permutation of arguments. */
+
+/* Describe the part of ARGV that contains non-options that have
+ been skipped. `first_nonopt' is the index in ARGV of the first of them;
+ `last_nonopt' is the index after the last of them. */
+
+static int first_nonopt;
+static int last_nonopt;
+
+#ifdef _LIBC
+/* Stored original parameters.
+ XXX This is no good solution. We should rather copy the args so
+ that we can compare them later. But we must not use malloc(3). */
+extern int __libc_argc;
+extern char **__libc_argv;
+
+/* Bash 2.0 gives us an environment variable containing flags
+ indicating ARGV elements that should not be considered arguments. */
+
+# ifdef USE_NONOPTION_FLAGS
+/* Defined in getopt_init.c */
+extern char *__getopt_nonoption_flags;
+
+static int nonoption_flags_max_len;
+static int nonoption_flags_len;
+# endif
+
+# ifdef USE_NONOPTION_FLAGS
+# define SWAP_FLAGS(ch1, ch2) \
+ if (nonoption_flags_len > 0) \
+ { \
+ char __tmp = __getopt_nonoption_flags[ch1]; \
+ __getopt_nonoption_flags[ch1] = __getopt_nonoption_flags[ch2]; \
+ __getopt_nonoption_flags[ch2] = __tmp; \
+ }
+# else
+# define SWAP_FLAGS(ch1, ch2)
+# endif
+#else /* !_LIBC */
+# define SWAP_FLAGS(ch1, ch2)
+#endif /* _LIBC */
+
+/* Exchange two adjacent subsequences of ARGV.
+ One subsequence is elements [first_nonopt,last_nonopt)
+ which contains all the non-options that have been skipped so far.
+ The other is elements [last_nonopt,optind), which contains all
+ the options processed since those non-options were skipped.
+
+ `first_nonopt' and `last_nonopt' are relocated so that they describe
+ the new indices of the non-options in ARGV after they are moved. */
+
+#if defined __STDC__ && __STDC__
+static void exchange (char **);
+#endif
+
+static void
+exchange (argv)
+ char **argv;
+{
+ int bottom = first_nonopt;
+ int middle = last_nonopt;
+ int top = optind;
+ char *tem;
+
+ /* Exchange the shorter segment with the far end of the longer segment.
+ That puts the shorter segment into the right place.
+ It leaves the longer segment in the right place overall,
+ but it consists of two parts that need to be swapped next. */
+
+#if defined _LIBC && defined USE_NONOPTION_FLAGS
+ /* First make sure the handling of the `__getopt_nonoption_flags'
+ string can work normally. Our top argument must be in the range
+ of the string. */
+ if (nonoption_flags_len > 0 && top >= nonoption_flags_max_len)
+ {
+ /* We must extend the array. The user plays games with us and
+ presents new arguments. */
+ char *new_str = malloc (top + 1);
+ if (new_str == NULL)
+ nonoption_flags_len = nonoption_flags_max_len = 0;
+ else
+ {
+ memset (__mempcpy (new_str, __getopt_nonoption_flags,
+ nonoption_flags_max_len),
+ '\0', top + 1 - nonoption_flags_max_len);
+ nonoption_flags_max_len = top + 1;
+ __getopt_nonoption_flags = new_str;
+ }
+ }
+#endif
+
+ while (top > middle && middle > bottom)
+ {
+ if (top - middle > middle - bottom)
+ {
+ /* Bottom segment is the short one. */
+ int len = middle - bottom;
+ register int i;
+
+ /* Swap it with the top part of the top segment. */
+ for (i = 0; i < len; i++)
+ {
+ tem = argv[bottom + i];
+ argv[bottom + i] = argv[top - (middle - bottom) + i];
+ argv[top - (middle - bottom) + i] = tem;
+ SWAP_FLAGS (bottom + i, top - (middle - bottom) + i);
+ }
+ /* Exclude the moved bottom segment from further swapping. */
+ top -= len;
+ }
+ else
+ {
+ /* Top segment is the short one. */
+ int len = top - middle;
+ register int i;
+
+ /* Swap it with the bottom part of the bottom segment. */
+ for (i = 0; i < len; i++)
+ {
+ tem = argv[bottom + i];
+ argv[bottom + i] = argv[middle + i];
+ argv[middle + i] = tem;
+ SWAP_FLAGS (bottom + i, middle + i);
+ }
+ /* Exclude the moved top segment from further swapping. */
+ bottom += len;
+ }
+ }
+
+ /* Update records for the slots the non-options now occupy. */
+
+ first_nonopt += (optind - last_nonopt);
+ last_nonopt = optind;
+}
+
+/* Initialize the internal data when the first call is made. */
+
+#if defined __STDC__ && __STDC__
+static const char *_getopt_initialize (int, char *const *, const char *);
+#endif
+static const char *
+_getopt_initialize (argc, argv, optstring)
+ int argc;
+ char *const *argv;
+ const char *optstring;
+{
+ /* Start processing options with ARGV-element 1 (since ARGV-element 0
+ is the program name); the sequence of previously skipped
+ non-option ARGV-elements is empty. */
+
+ first_nonopt = last_nonopt = optind;
+
+ nextchar = NULL;
+
+ posixly_correct = getenv ("POSIXLY_CORRECT");
+
+ /* Determine how to handle the ordering of options and nonoptions. */
+
+ if (optstring[0] == '-')
+ {
+ ordering = RETURN_IN_ORDER;
+ ++optstring;
+ }
+ else if (optstring[0] == '+')
+ {
+ ordering = REQUIRE_ORDER;
+ ++optstring;
+ }
+ else if (posixly_correct != NULL)
+ ordering = REQUIRE_ORDER;
+ else
+ ordering = PERMUTE;
+
+#if defined _LIBC && defined USE_NONOPTION_FLAGS
+ if (posixly_correct == NULL
+ && argc == __libc_argc && argv == __libc_argv)
+ {
+ if (nonoption_flags_max_len == 0)
+ {
+ if (__getopt_nonoption_flags == NULL
+ || __getopt_nonoption_flags[0] == '\0')
+ nonoption_flags_max_len = -1;
+ else
+ {
+ const char *orig_str = __getopt_nonoption_flags;
+ int len = nonoption_flags_max_len = strlen (orig_str);
+ if (nonoption_flags_max_len < argc)
+ nonoption_flags_max_len = argc;
+ __getopt_nonoption_flags =
+ (char *) malloc (nonoption_flags_max_len);
+ if (__getopt_nonoption_flags == NULL)
+ nonoption_flags_max_len = -1;
+ else
+ memset (__mempcpy (__getopt_nonoption_flags, orig_str, len),
+ '\0', nonoption_flags_max_len - len);
+ }
+ }
+ nonoption_flags_len = nonoption_flags_max_len;
+ }
+ else
+ nonoption_flags_len = 0;
+#endif
+
+ return optstring;
+}
+\f
+/* Scan elements of ARGV (whose length is ARGC) for option characters
+ given in OPTSTRING.
+
+ If an element of ARGV starts with '-', and is not exactly "-" or "--",
+ then it is an option element. The characters of this element
+ (aside from the initial '-') are option characters. If `getopt'
+ is called repeatedly, it returns successively each of the option characters
+ from each of the option elements.
+
+ If `getopt' finds another option character, it returns that character,
+ updating `optind' and `nextchar' so that the next call to `getopt' can
+ resume the scan with the following option character or ARGV-element.
+
+ If there are no more option characters, `getopt' returns -1.
+ Then `optind' is the index in ARGV of the first ARGV-element
+ that is not an option. (The ARGV-elements have been permuted
+ so that those that are not options now come last.)
+
+ OPTSTRING is a string containing the legitimate option characters.
+ If an option character is seen that is not listed in OPTSTRING,
+ return '?' after printing an error message. If you set `opterr' to
+ zero, the error message is suppressed but we still return '?'.
+
+ If a char in OPTSTRING is followed by a colon, that means it wants an arg,
+ so the following text in the same ARGV-element, or the text of the following
+ ARGV-element, is returned in `optarg'. Two colons mean an option that
+ wants an optional arg; if there is text in the current ARGV-element,
+ it is returned in `optarg', otherwise `optarg' is set to zero.
+
+ If OPTSTRING starts with `-' or `+', it requests different methods of
+ handling the non-option ARGV-elements.
+ See the comments about RETURN_IN_ORDER and REQUIRE_ORDER, above.
+
+ Long-named options begin with `--' instead of `-'.
+ Their names may be abbreviated as long as the abbreviation is unique
+ or is an exact match for some defined option. If they have an
+ argument, it follows the option name in the same ARGV-element, separated
+ from the option name by a `=', or else the in next ARGV-element.
+ When `getopt' finds a long-named option, it returns 0 if that option's
+ `flag' field is nonzero, the value of the option's `val' field
+ if the `flag' field is zero.
+
+ The elements of ARGV aren't really const, because we permute them.
+ But we pretend they're const in the prototype to be compatible
+ with other systems.
+
+ LONGOPTS is a vector of `struct option' terminated by an
+ element containing a name which is zero.
+
+ LONGIND returns the index in LONGOPT of the long-named option found.
+ It is only valid when a long-named option has been found by the most
+ recent call.
+
+ If LONG_ONLY is nonzero, '-' as well as '--' can introduce
+ long-named options. */
+
+int
+_getopt_internal (argc, argv, optstring, longopts, longind, long_only)
+ int argc;
+ char *const *argv;
+ const char *optstring;
+ const struct option *longopts;
+ int *longind;
+ int long_only;
+{
+ int print_errors = opterr;
+ if (optstring[0] == ':')
+ print_errors = 0;
+
+ if (argc < 1)
+ return -1;
+
+ optarg = NULL;
+
+ if (optind == 0 || !__getopt_initialized)
+ {
+ if (optind == 0)
+ optind = 1; /* Don't scan ARGV[0], the program name. */
+ optstring = _getopt_initialize (argc, argv, optstring);
+ __getopt_initialized = 1;
+ }
+
+ /* Test whether ARGV[optind] points to a non-option argument.
+ Either it does not have option syntax, or there is an environment flag
+ from the shell indicating it is not an option. The later information
+ is only used when the used in the GNU libc. */
+#if defined _LIBC && defined USE_NONOPTION_FLAGS
+# define NONOPTION_P (argv[optind][0] != '-' || argv[optind][1] == '\0' \
+ || (optind < nonoption_flags_len \
+ && __getopt_nonoption_flags[optind] == '1'))
+#else
+# define NONOPTION_P (argv[optind][0] != '-' || argv[optind][1] == '\0')
+#endif
+
+ if (nextchar == NULL || *nextchar == '\0')
+ {
+ /* Advance to the next ARGV-element. */
+
+ /* Give FIRST_NONOPT & LAST_NONOPT rational values if OPTIND has been
+ moved back by the user (who may also have changed the arguments). */
+ if (last_nonopt > optind)
+ last_nonopt = optind;
+ if (first_nonopt > optind)
+ first_nonopt = optind;
+
+ if (ordering == PERMUTE)
+ {
+ /* If we have just processed some options following some non-options,
+ exchange them so that the options come first. */
+
+ if (first_nonopt != last_nonopt && last_nonopt != optind)
+ exchange ((char **) argv);
+ else if (last_nonopt != optind)
+ first_nonopt = optind;
+
+ /* Skip any additional non-options
+ and extend the range of non-options previously skipped. */
+
+ while (optind < argc && NONOPTION_P)
+ optind++;
+ last_nonopt = optind;
+ }
+
+ /* The special ARGV-element `--' means premature end of options.
+ Skip it like a null option,
+ then exchange with previous non-options as if it were an option,
+ then skip everything else like a non-option. */
+
+ if (optind != argc && !strcmp (argv[optind], "--"))
+ {
+ optind++;
+
+ if (first_nonopt != last_nonopt && last_nonopt != optind)
+ exchange ((char **) argv);
+ else if (first_nonopt == last_nonopt)
+ first_nonopt = optind;
+ last_nonopt = argc;
+
+ optind = argc;
+ }
+
+ /* If we have done all the ARGV-elements, stop the scan
+ and back over any non-options that we skipped and permuted. */
+
+ if (optind == argc)
+ {
+ /* Set the next-arg-index to point at the non-options
+ that we previously skipped, so the caller will digest them. */
+ if (first_nonopt != last_nonopt)
+ optind = first_nonopt;
+ return -1;
+ }
+
+ /* If we have come to a non-option and did not permute it,
+ either stop the scan or describe it to the caller and pass it by. */
+
+ if (NONOPTION_P)
+ {
+ if (ordering == REQUIRE_ORDER)
+ return -1;
+ optarg = argv[optind++];
+ return 1;
+ }
+
+ /* We have found another option-ARGV-element.
+ Skip the initial punctuation. */
+
+ nextchar = (argv[optind] + 1
+ + (longopts != NULL && argv[optind][1] == '-'));
+ }
+
+ /* Decode the current option-ARGV-element. */
+
+ /* Check whether the ARGV-element is a long option.
+
+ If long_only and the ARGV-element has the form "-f", where f is
+ a valid short option, don't consider it an abbreviated form of
+ a long option that starts with f. Otherwise there would be no
+ way to give the -f short option.
+
+ On the other hand, if there's a long option "fubar" and
+ the ARGV-element is "-fu", do consider that an abbreviation of
+ the long option, just like "--fu", and not "-f" with arg "u".
+
+ This distinction seems to be the most useful approach. */
+
+ if (longopts != NULL
+ && (argv[optind][1] == '-'
+ || (long_only && (argv[optind][2] || !my_index (optstring, argv[optind][1])))))
+ {
+ char *nameend;
+ const struct option *p;
+ const struct option *pfound = NULL;
+ int exact = 0;
+ int ambig = 0;
+ int indfound = -1;
+ int option_index;
+
+ for (nameend = nextchar; *nameend && *nameend != '='; nameend++)
+ /* Do nothing. */ ;
+
+ /* Test all long options for either exact match
+ or abbreviated matches. */
+ for (p = longopts, option_index = 0; p->name; p++, option_index++)
+ if (!strncmp (p->name, nextchar, nameend - nextchar))
+ {
+ if ((unsigned int) (nameend - nextchar)
+ == (unsigned int) strlen (p->name))
+ {
+ /* Exact match found. */
+ pfound = p;
+ indfound = option_index;
+ exact = 1;
+ break;
+ }
+ else if (pfound == NULL)
+ {
+ /* First nonexact match found. */
+ pfound = p;
+ indfound = option_index;
+ }
+ else if (long_only
+ || pfound->has_arg != p->has_arg
+ || pfound->flag != p->flag
+ || pfound->val != p->val)
+ /* Second or later nonexact match found. */
+ ambig = 1;
+ }
+
+ if (ambig && !exact)
+ {
+ if (print_errors)
+ {
+#if defined _LIBC && defined USE_IN_LIBIO
+ char *buf;
+
+ if (__asprintf (&buf, _("%s: option `%s' is ambiguous\n"),
+ argv[0], argv[optind]) >= 0)
+ {
+
+ if (_IO_fwide (stderr, 0) > 0)
+ __fwprintf (stderr, L"%s", buf);
+ else
+ fputs (buf, stderr);
+
+ free (buf);
+ }
+#else
+ fprintf (stderr, _("%s: option `%s' is ambiguous\n"),
+ argv[0], argv[optind]);
+#endif
+ }
+ nextchar += strlen (nextchar);
+ optind++;
+ optopt = 0;
+ return '?';
+ }
+
+ if (pfound != NULL)
+ {
+ option_index = indfound;
+ optind++;
+ if (*nameend)
+ {
+ /* Don't test has_arg with >, because some C compilers don't
+ allow it to be used on enums. */
+ if (pfound->has_arg)
+ optarg = nameend + 1;
+ else
+ {
+ if (print_errors)
+ {
+#if defined _LIBC && defined USE_IN_LIBIO
+ char *buf;
+ int n;
+#endif
+
+ if (argv[optind - 1][1] == '-')
+ {
+ /* --option */
+#if defined _LIBC && defined USE_IN_LIBIO
+ n = __asprintf (&buf, _("\
+%s: option `--%s' doesn't allow an argument\n"),
+ argv[0], pfound->name);
+#else
+ fprintf (stderr, _("\
+%s: option `--%s' doesn't allow an argument\n"),
+ argv[0], pfound->name);
+#endif
+ }
+ else
+ {
+ /* +option or -option */
+#if defined _LIBC && defined USE_IN_LIBIO
+ n = __asprintf (&buf, _("\
+%s: option `%c%s' doesn't allow an argument\n"),
+ argv[0], argv[optind - 1][0],
+ pfound->name);
+#else
+ fprintf (stderr, _("\
+%s: option `%c%s' doesn't allow an argument\n"),
+ argv[0], argv[optind - 1][0], pfound->name);
+#endif
+ }
+
+#if defined _LIBC && defined USE_IN_LIBIO
+ if (n >= 0)
+ {
+ if (_IO_fwide (stderr, 0) > 0)
+ __fwprintf (stderr, L"%s", buf);
+ else
+ fputs (buf, stderr);
+
+ free (buf);
+ }
+#endif
+ }
+
+ nextchar += strlen (nextchar);
+
+ optopt = pfound->val;
+ return '?';
+ }
+ }
+ else if (pfound->has_arg == 1)
+ {
+ if (optind < argc)
+ optarg = argv[optind++];
+ else
+ {
+ if (print_errors)
+ {
+#if defined _LIBC && defined USE_IN_LIBIO
+ char *buf;
+
+ if (__asprintf (&buf, _("\
+%s: option `%s' requires an argument\n"),
+ argv[0], argv[optind - 1]) >= 0)
+ {
+ if (_IO_fwide (stderr, 0) > 0)
+ __fwprintf (stderr, L"%s", buf);
+ else
+ fputs (buf, stderr);
+
+ free (buf);
+ }
+#else
+ fprintf (stderr,
+ _("%s: option `%s' requires an argument\n"),
+ argv[0], argv[optind - 1]);
+#endif
+ }
+ nextchar += strlen (nextchar);
+ optopt = pfound->val;
+ return optstring[0] == ':' ? ':' : '?';
+ }
+ }
+ nextchar += strlen (nextchar);
+ if (longind != NULL)
+ *longind = option_index;
+ if (pfound->flag)
+ {
+ *(pfound->flag) = pfound->val;
+ return 0;
+ }
+ return pfound->val;
+ }
+
+ /* Can't find it as a long option. If this is not getopt_long_only,
+ or the option starts with '--' or is not a valid short
+ option, then it's an error.
+ Otherwise interpret it as a short option. */
+ if (!long_only || argv[optind][1] == '-'
+ || my_index (optstring, *nextchar) == NULL)
+ {
+ if (print_errors)
+ {
+#if defined _LIBC && defined USE_IN_LIBIO
+ char *buf;
+ int n;
+#endif
+
+ if (argv[optind][1] == '-')
+ {
+ /* --option */
+#if defined _LIBC && defined USE_IN_LIBIO
+ n = __asprintf (&buf, _("%s: unrecognized option `--%s'\n"),
+ argv[0], nextchar);
+#else
+ fprintf (stderr, _("%s: unrecognized option `--%s'\n"),
+ argv[0], nextchar);
+#endif
+ }
+ else
+ {
+ /* +option or -option */
+#if defined _LIBC && defined USE_IN_LIBIO
+ n = __asprintf (&buf, _("%s: unrecognized option `%c%s'\n"),
+ argv[0], argv[optind][0], nextchar);
+#else
+ fprintf (stderr, _("%s: unrecognized option `%c%s'\n"),
+ argv[0], argv[optind][0], nextchar);
+#endif
+ }
+
+#if defined _LIBC && defined USE_IN_LIBIO
+ if (n >= 0)
+ {
+ if (_IO_fwide (stderr, 0) > 0)
+ __fwprintf (stderr, L"%s", buf);
+ else
+ fputs (buf, stderr);
+
+ free (buf);
+ }
+#endif
+ }
+ nextchar = (char *) "";
+ optind++;
+ optopt = 0;
+ return '?';
+ }
+ }
+
+ /* Look at and handle the next short option-character. */
+
+ {
+ char c = *nextchar++;
+ char *temp = my_index (optstring, c);
+
+ /* Increment `optind' when we start to process its last character. */
+ if (*nextchar == '\0')
+ ++optind;
+
+ if (temp == NULL || c == ':')
+ {
+ if (print_errors)
+ {
+#if defined _LIBC && defined USE_IN_LIBIO
+ char *buf;
+ int n;
+#endif
+
+ if (posixly_correct)
+ {
+ /* 1003.2 specifies the format of this message. */
+#if defined _LIBC && defined USE_IN_LIBIO
+ n = __asprintf (&buf, _("%s: illegal option -- %c\n"),
+ argv[0], c);
+#else
+ fprintf (stderr, _("%s: illegal option -- %c\n"), argv[0], c);
+#endif
+ }
+ else
+ {
+#if defined _LIBC && defined USE_IN_LIBIO
+ n = __asprintf (&buf, _("%s: invalid option -- %c\n"),
+ argv[0], c);
+#else
+ fprintf (stderr, _("%s: invalid option -- %c\n"), argv[0], c);
+#endif
+ }
+
+#if defined _LIBC && defined USE_IN_LIBIO
+ if (n >= 0)
+ {
+ if (_IO_fwide (stderr, 0) > 0)
+ __fwprintf (stderr, L"%s", buf);
+ else
+ fputs (buf, stderr);
+
+ free (buf);
+ }
+#endif
+ }
+ optopt = c;
+ return '?';
+ }
+ /* Convenience. Treat POSIX -W foo same as long option --foo */
+ if (temp[0] == 'W' && temp[1] == ';')
+ {
+ char *nameend;
+ const struct option *p;
+ const struct option *pfound = NULL;
+ int exact = 0;
+ int ambig = 0;
+ int indfound = 0;
+ int option_index;
+
+ /* This is an option that requires an argument. */
+ if (*nextchar != '\0')
+ {
+ optarg = nextchar;
+ /* If we end this ARGV-element by taking the rest as an arg,
+ we must advance to the next element now. */
+ optind++;
+ }
+ else if (optind == argc)
+ {
+ if (print_errors)
+ {
+ /* 1003.2 specifies the format of this message. */
+#if defined _LIBC && defined USE_IN_LIBIO
+ char *buf;
+
+ if (__asprintf (&buf,
+ _("%s: option requires an argument -- %c\n"),
+ argv[0], c) >= 0)
+ {
+ if (_IO_fwide (stderr, 0) > 0)
+ __fwprintf (stderr, L"%s", buf);
+ else
+ fputs (buf, stderr);
+
+ free (buf);
+ }
+#else
+ fprintf (stderr, _("%s: option requires an argument -- %c\n"),
+ argv[0], c);
+#endif
+ }
+ optopt = c;
+ if (optstring[0] == ':')
+ c = ':';
+ else
+ c = '?';
+ return c;
+ }
+ else
+ /* We already incremented `optind' once;
+ increment it again when taking next ARGV-elt as argument. */
+ optarg = argv[optind++];
+
+ /* optarg is now the argument, see if it's in the
+ table of longopts. */
+
+ for (nextchar = nameend = optarg; *nameend && *nameend != '='; nameend++)
+ /* Do nothing. */ ;
+
+ /* Test all long options for either exact match
+ or abbreviated matches. */
+ for (p = longopts, option_index = 0; p->name; p++, option_index++)
+ if (!strncmp (p->name, nextchar, nameend - nextchar))
+ {
+ if ((unsigned int) (nameend - nextchar) == strlen (p->name))
+ {
+ /* Exact match found. */
+ pfound = p;
+ indfound = option_index;
+ exact = 1;
+ break;
+ }
+ else if (pfound == NULL)
+ {
+ /* First nonexact match found. */
+ pfound = p;
+ indfound = option_index;
+ }
+ else
+ /* Second or later nonexact match found. */
+ ambig = 1;
+ }
+ if (ambig && !exact)
+ {
+ if (print_errors)
+ {
+#if defined _LIBC && defined USE_IN_LIBIO
+ char *buf;
+
+ if (__asprintf (&buf, _("%s: option `-W %s' is ambiguous\n"),
+ argv[0], argv[optind]) >= 0)
+ {
+ if (_IO_fwide (stderr, 0) > 0)
+ __fwprintf (stderr, L"%s", buf);
+ else
+ fputs (buf, stderr);
+
+ free (buf);
+ }
+#else
+ fprintf (stderr, _("%s: option `-W %s' is ambiguous\n"),
+ argv[0], argv[optind]);
+#endif
+ }
+ nextchar += strlen (nextchar);
+ optind++;
+ return '?';
+ }
+ if (pfound != NULL)
+ {
+ option_index = indfound;
+ if (*nameend)
+ {
+ /* Don't test has_arg with >, because some C compilers don't
+ allow it to be used on enums. */
+ if (pfound->has_arg)
+ optarg = nameend + 1;
+ else
+ {
+ if (print_errors)
+ {
+#if defined _LIBC && defined USE_IN_LIBIO
+ char *buf;
+
+ if (__asprintf (&buf, _("\
+%s: option `-W %s' doesn't allow an argument\n"),
+ argv[0], pfound->name) >= 0)
+ {
+ if (_IO_fwide (stderr, 0) > 0)
+ __fwprintf (stderr, L"%s", buf);
+ else
+ fputs (buf, stderr);
+
+ free (buf);
+ }
+#else
+ fprintf (stderr, _("\
+%s: option `-W %s' doesn't allow an argument\n"),
+ argv[0], pfound->name);
+#endif
+ }
+
+ nextchar += strlen (nextchar);
+ return '?';
+ }
+ }
+ else if (pfound->has_arg == 1)
+ {
+ if (optind < argc)
+ optarg = argv[optind++];
+ else
+ {
+ if (print_errors)
+ {
+#if defined _LIBC && defined USE_IN_LIBIO
+ char *buf;
+
+ if (__asprintf (&buf, _("\
+%s: option `%s' requires an argument\n"),
+ argv[0], argv[optind - 1]) >= 0)
+ {
+ if (_IO_fwide (stderr, 0) > 0)
+ __fwprintf (stderr, L"%s", buf);
+ else
+ fputs (buf, stderr);
+
+ free (buf);
+ }
+#else
+ fprintf (stderr,
+ _("%s: option `%s' requires an argument\n"),
+ argv[0], argv[optind - 1]);
+#endif
+ }
+ nextchar += strlen (nextchar);
+ return optstring[0] == ':' ? ':' : '?';
+ }
+ }
+ nextchar += strlen (nextchar);
+ if (longind != NULL)
+ *longind = option_index;
+ if (pfound->flag)
+ {
+ *(pfound->flag) = pfound->val;
+ return 0;
+ }
+ return pfound->val;
+ }
+ nextchar = NULL;
+ return 'W'; /* Let the application handle it. */
+ }
+ if (temp[1] == ':')
+ {
+ if (temp[2] == ':')
+ {
+ /* This is an option that accepts an argument optionally. */
+ if (*nextchar != '\0')
+ {
+ optarg = nextchar;
+ optind++;
+ }
+ else
+ optarg = NULL;
+ nextchar = NULL;
+ }
+ else
+ {
+ /* This is an option that requires an argument. */
+ if (*nextchar != '\0')
+ {
+ optarg = nextchar;
+ /* If we end this ARGV-element by taking the rest as an arg,
+ we must advance to the next element now. */
+ optind++;
+ }
+ else if (optind == argc)
+ {
+ if (print_errors)
+ {
+ /* 1003.2 specifies the format of this message. */
+#if defined _LIBC && defined USE_IN_LIBIO
+ char *buf;
+
+ if (__asprintf (&buf, _("\
+%s: option requires an argument -- %c\n"),
+ argv[0], c) >= 0)
+ {
+ if (_IO_fwide (stderr, 0) > 0)
+ __fwprintf (stderr, L"%s", buf);
+ else
+ fputs (buf, stderr);
+
+ free (buf);
+ }
+#else
+ fprintf (stderr,
+ _("%s: option requires an argument -- %c\n"),
+ argv[0], c);
+#endif
+ }
+ optopt = c;
+ if (optstring[0] == ':')
+ c = ':';
+ else
+ c = '?';
+ }
+ else
+ /* We already incremented `optind' once;
+ increment it again when taking next ARGV-elt as argument. */
+ optarg = argv[optind++];
+ nextchar = NULL;
+ }
+ }
+ return c;
+ }
+}
+
+int
+getopt (argc, argv, optstring)
+ int argc;
+ char *const *argv;
+ const char *optstring;
+{
+ return _getopt_internal (argc, argv, optstring,
+ (const struct option *) 0,
+ (int *) 0,
+ 0);
+}
+
+#endif /* Not ELIDE_CODE. */
+\f
+#ifdef TEST
+
+/* Compile with -DTEST to make an executable for use in testing
+ the above definition of `getopt'. */
+
+int
+main (argc, argv)
+ int argc;
+ char **argv;
+{
+ int c;
+ int digit_optind = 0;
+
+ while (1)
+ {
+ int this_option_optind = optind ? optind : 1;
+
+ c = getopt (argc, argv, "abc:d:0123456789");
+ if (c == -1)
+ break;
+
+ switch (c)
+ {
+ case '0':
+ case '1':
+ case '2':
+ case '3':
+ case '4':
+ case '5':
+ case '6':
+ case '7':
+ case '8':
+ case '9':
+ if (digit_optind != 0 && digit_optind != this_option_optind)
+ printf ("digits occur in two different argv-elements.\n");
+ digit_optind = this_option_optind;
+ printf ("option %c\n", c);
+ break;
+
+ case 'a':
+ printf ("option a\n");
+ break;
+
+ case 'b':
+ printf ("option b\n");
+ break;
+
+ case 'c':
+ printf ("option c with value `%s'\n", optarg);
+ break;
+
+ case '?':
+ break;
+
+ default:
+ printf ("?? getopt returned character code 0%o ??\n", c);
+ }
+ }
+
+ if (optind < argc)
+ {
+ printf ("non-option ARGV-elements: ");
+ while (optind < argc)
+ printf ("%s ", argv[optind++]);
+ printf ("\n");
+ }
+
+ exit (0);
+}
+
+#endif /* TEST */
--- /dev/null
+/* Declarations for getopt.
+ Copyright (C) 1989-1994, 1996-1999, 2001 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, write to the Free
+ Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ 02111-1307 USA. */
+
+#ifndef _GETOPT_H
+
+#ifndef __need_getopt
+# define _GETOPT_H 1
+#endif
+
+/* If __GNU_LIBRARY__ is not already defined, either we are being used
+ standalone, or this is the first header included in the source file.
+ If we are being used with glibc, we need to include <features.h>, but
+ that does not exist if we are standalone. So: if __GNU_LIBRARY__ is
+ not defined, include <ctype.h>, which will pull in <features.h> for us
+ if it's from glibc. (Why ctype.h? It's guaranteed to exist and it
+ doesn't flood the namespace with stuff the way some other headers do.) */
+#if !defined __GNU_LIBRARY__
+# include <ctype.h>
+#endif
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* For communication from `getopt' to the caller.
+ When `getopt' finds an option that takes an argument,
+ the argument value is returned here.
+ Also, when `ordering' is RETURN_IN_ORDER,
+ each non-option ARGV-element is returned here. */
+
+extern char *optarg;
+
+/* Index in ARGV of the next element to be scanned.
+ This is used for communication to and from the caller
+ and for communication between successive calls to `getopt'.
+
+ On entry to `getopt', zero means this is the first call; initialize.
+
+ When `getopt' returns -1, this is the index of the first of the
+ non-option elements that the caller should itself scan.
+
+ Otherwise, `optind' communicates from one call to the next
+ how much of ARGV has been scanned so far. */
+
+extern int optind;
+
+/* Callers store zero here to inhibit the error message `getopt' prints
+ for unrecognized options. */
+
+extern int opterr;
+
+/* Set to an option character which was unrecognized. */
+
+extern int optopt;
+
+#ifndef __need_getopt
+/* Describe the long-named options requested by the application.
+ The LONG_OPTIONS argument to getopt_long or getopt_long_only is a vector
+ of `struct option' terminated by an element containing a name which is
+ zero.
+
+ The field `has_arg' is:
+ no_argument (or 0) if the option does not take an argument,
+ required_argument (or 1) if the option requires an argument,
+ optional_argument (or 2) if the option takes an optional argument.
+
+ If the field `flag' is not NULL, it points to a variable that is set
+ to the value given in the field `val' when the option is found, but
+ left unchanged if the option is not found.
+
+ To have a long-named option do something other than set an `int' to
+ a compiled-in constant, such as set a value from `optarg', set the
+ option's `flag' field to zero and its `val' field to a nonzero
+ value (the equivalent single-letter option character, if there is
+ one). For long options that have a zero `flag' field, `getopt'
+ returns the contents of the `val' field. */
+
+struct option
+{
+# if (defined __STDC__ && __STDC__) || defined __cplusplus
+ const char *name;
+# else
+ char *name;
+# endif
+ /* has_arg can't be an enum because some compilers complain about
+ type mismatches in all the code that assumes it is an int. */
+ int has_arg;
+ int *flag;
+ int val;
+};
+
+/* Names for the values of the `has_arg' field of `struct option'. */
+
+# define no_argument 0
+# define required_argument 1
+# define optional_argument 2
+#endif /* need getopt */
+
+
+/* Get definitions and prototypes for functions to process the
+ arguments in ARGV (ARGC of them, minus the program name) for
+ options given in OPTS.
+
+ Return the option character from OPTS just read. Return -1 when
+ there are no more options. For unrecognized options, or options
+ missing arguments, `optopt' is set to the option letter, and '?' is
+ returned.
+
+ The OPTS string is a list of characters which are recognized option
+ letters, optionally followed by colons, specifying that that letter
+ takes an argument, to be placed in `optarg'.
+
+ If a letter in OPTS is followed by two colons, its argument is
+ optional. This behavior is specific to the GNU `getopt'.
+
+ The argument `--' causes premature termination of argument
+ scanning, explicitly telling `getopt' that there are no more
+ options.
+
+ If OPTS begins with `--', then non-option arguments are treated as
+ arguments to the option '\0'. This behavior is specific to the GNU
+ `getopt'. */
+
+#if (defined __STDC__ && __STDC__) || defined __cplusplus
+# ifdef __GNU_LIBRARY__
+/* Many other libraries have conflicting prototypes for getopt, with
+ differences in the consts, in stdlib.h. To avoid compilation
+ errors, only prototype getopt for the GNU C library. */
+extern int getopt (int ___argc, char *const *___argv, const char *__shortopts);
+# else /* not __GNU_LIBRARY__ */
+extern int getopt ();
+# endif /* __GNU_LIBRARY__ */
+
+# ifndef __need_getopt
+extern int getopt_long (int ___argc, char *const *___argv,
+ const char *__shortopts,
+ const struct option *__longopts, int *__longind);
+extern int getopt_long_only (int ___argc, char *const *___argv,
+ const char *__shortopts,
+ const struct option *__longopts, int *__longind);
+
+/* Internal only. Users should not call this directly. */
+extern int _getopt_internal (int ___argc, char *const *___argv,
+ const char *__shortopts,
+ const struct option *__longopts, int *__longind,
+ int __long_only);
+# endif
+#else /* not __STDC__ */
+extern int getopt ();
+# ifndef __need_getopt
+extern int getopt_long ();
+extern int getopt_long_only ();
+
+extern int _getopt_internal ();
+# endif
+#endif /* __STDC__ */
+
+#ifdef __cplusplus
+}
+#endif
+
+/* Make sure we later can get all the definitions and declarations. */
+#undef __need_getopt
+
+#endif /* getopt.h */
--- /dev/null
+/* getopt_long and getopt_long_only entry points for GNU getopt.
+ Copyright (C) 1987,88,89,90,91,92,93,94,96,97,98
+ Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, write to the Free
+ Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ 02111-1307 USA. */
+\f
+#ifdef HAVE_CONFIG_H
+#include <config.h>
+#endif
+
+#ifdef _LIBC
+# include <getopt.h>
+#else
+# include "getopt.h"
+#endif
+
+#if !defined __STDC__ || !__STDC__
+/* This is a separate conditional since some stdc systems
+ reject `defined (const)'. */
+#ifndef const
+#define const
+#endif
+#endif
+
+#include <stdio.h>
+
+/* Comment out all this code if we are using the GNU C Library, and are not
+ actually compiling the library itself. This code is part of the GNU C
+ Library, but also included in many other GNU distributions. Compiling
+ and linking in this code is a waste when using the GNU C library
+ (especially if it is a shared library). Rather than having every GNU
+ program understand `configure --with-gnu-libc' and omit the object files,
+ it is simpler to just do this in the source for each such file. */
+
+#define GETOPT_INTERFACE_VERSION 2
+#if !defined _LIBC && defined __GLIBC__ && __GLIBC__ >= 2
+#include <gnu-versions.h>
+#if _GNU_GETOPT_INTERFACE_VERSION == GETOPT_INTERFACE_VERSION
+#define ELIDE_CODE
+#endif
+#endif
+
+#ifndef ELIDE_CODE
+
+
+/* This needs to come after some library #include
+ to get __GNU_LIBRARY__ defined. */
+#ifdef __GNU_LIBRARY__
+#include <stdlib.h>
+#endif
+
+#ifndef NULL
+#define NULL 0
+#endif
+
+int
+getopt_long (argc, argv, options, long_options, opt_index)
+ int argc;
+ char *const *argv;
+ const char *options;
+ const struct option *long_options;
+ int *opt_index;
+{
+ return _getopt_internal (argc, argv, options, long_options, opt_index, 0);
+}
+
+/* Like getopt_long, but '-' as well as '--' can indicate a long option.
+ If an option that starts with '-' (not '--') doesn't match a long option,
+ but does match a short option, it is parsed as a short option
+ instead. */
+
+int
+getopt_long_only (argc, argv, options, long_options, opt_index)
+ int argc;
+ char *const *argv;
+ const char *options;
+ const struct option *long_options;
+ int *opt_index;
+{
+ return _getopt_internal (argc, argv, options, long_options, opt_index, 1);
+}
+
+# ifdef _LIBC
+libc_hidden_def (getopt_long)
+libc_hidden_def (getopt_long_only)
+# endif
+
+#endif /* Not ELIDE_CODE. */
+\f
+#ifdef TEST
+
+#include <stdio.h>
+
+int
+main (argc, argv)
+ int argc;
+ char **argv;
+{
+ int c;
+ int digit_optind = 0;
+
+ while (1)
+ {
+ int this_option_optind = optind ? optind : 1;
+ int option_index = 0;
+ static struct option long_options[] =
+ {
+ {"add", 1, 0, 0},
+ {"append", 0, 0, 0},
+ {"delete", 1, 0, 0},
+ {"verbose", 0, 0, 0},
+ {"create", 0, 0, 0},
+ {"file", 1, 0, 0},
+ {0, 0, 0, 0}
+ };
+
+ c = getopt_long (argc, argv, "abc:d:0123456789",
+ long_options, &option_index);
+ if (c == -1)
+ break;
+
+ switch (c)
+ {
+ case 0:
+ printf ("option %s", long_options[option_index].name);
+ if (optarg)
+ printf (" with arg %s", optarg);
+ printf ("\n");
+ break;
+
+ case '0':
+ case '1':
+ case '2':
+ case '3':
+ case '4':
+ case '5':
+ case '6':
+ case '7':
+ case '8':
+ case '9':
+ if (digit_optind != 0 && digit_optind != this_option_optind)
+ printf ("digits occur in two different argv-elements.\n");
+ digit_optind = this_option_optind;
+ printf ("option %c\n", c);
+ break;
+
+ case 'a':
+ printf ("option a\n");
+ break;
+
+ case 'b':
+ printf ("option b\n");
+ break;
+
+ case 'c':
+ printf ("option c with value `%s'\n", optarg);
+ break;
+
+ case 'd':
+ printf ("option d with value `%s'\n", optarg);
+ break;
+
+ case '?':
+ break;
+
+ default:
+ printf ("?? getopt returned character code 0%o ??\n", c);
+ }
+ }
+
+ if (optind < argc)
+ {
+ printf ("non-option ARGV-elements: ");
+ while (optind < argc)
+ printf ("%s ", argv[optind++]);
+ printf ("\n");
+ }
+
+ exit (0);
+}
+
+#endif /* TEST */
* Home page of code is: http://smartmontools.sourceforge.net
* Address of support mailing list: smartmontools-support@lists.sourceforge.net
*
- * Copyright (C) 2003-9 Philip Williams, Bruce Allen
- * Copyright (C) 2008-9 Christian Franke <smartmontools-support@lists.sourceforge.net>
+ * Copyright (C) 2003-10 Philip Williams, Bruce Allen
+ * Copyright (C) 2008-10 Christian Franke <smartmontools-support@lists.sourceforge.net>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
#include <stdexcept>
-const char * knowndrives_cpp_cvsid = "$Id: knowndrives.cpp 3004 2009-12-19 19:39:12Z chrfranke $"
+const char * knowndrives_cpp_cvsid = "$Id: knowndrives.cpp 3093 2010-04-30 09:57:36Z chrfranke $"
KNOWNDRIVES_H_CVSID;
#define MODEL_STRING_LENGTH 40
static drive_database knowndrives;
+// Return true if modelfamily string describes entry for USB ID
+static bool is_usb_modelfamily(const char * modelfamily)
+{
+ return !strncmp(modelfamily, "USB:", 4);
+}
+
+// Return true if entry for USB ID
+static inline bool is_usb_entry(const drive_settings * dbentry)
+{
+ return is_usb_modelfamily(dbentry->modelfamily);
+}
+
// Compile regular expression, print message on failure.
static bool compile(regular_expression & regex, const char *pattern)
{
firmware = "";
for (unsigned i = 0; i < knowndrives.size(); i++) {
+ // Skip USB entries
+ if (is_usb_entry(&knowndrives[i]))
+ continue;
+
// Check whether model matches the regular expression in knowndrives[i].
if (!match(knowndrives[i].modelregexp, model))
continue;
return 0;
}
-// Parse '-v' and '-F' options in preset string, return false on error.
-static bool parse_presets(const char * presets, ata_vendor_attr_defs & defs,
- unsigned char & fix_firmwarebug)
+
+// Parse drive or USB options in preset string, return false on error.
+static bool parse_db_presets(const char * presets, ata_vendor_attr_defs * defs,
+ unsigned char * fix_firmwarebug, std::string * type)
{
for (int i = 0; ; ) {
i += strspn(presets+i, " \t");
char opt, arg[40+1+13]; int len = -1;
if (!(sscanf(presets+i, "-%c %40[^ ]%n", &opt, arg, &len) >= 2 && len > 0))
return false;
- if (opt == 'v') {
+ if (opt == 'v' && defs) {
// Parse "-v N,format[,name]"
- if (!parse_attribute_def(arg, defs, PRIOR_DATABASE))
+ if (!parse_attribute_def(arg, *defs, PRIOR_DATABASE))
return false;
}
- else if (opt == 'F') {
+ else if (opt == 'F' && fix_firmwarebug) {
unsigned char fix;
if (!strcmp(arg, "samsung"))
fix = FIX_SAMSUNG;
else
return false;
// Set only if not set by user
- if (fix_firmwarebug == FIX_NOTSPECIFIED)
- fix_firmwarebug = fix;
+ if (*fix_firmwarebug == FIX_NOTSPECIFIED)
+ *fix_firmwarebug = fix;
+ }
+ else if (opt == 'd' && type) {
+ // TODO: Check valid types
+ *type = arg;
}
else
return false;
return true;
}
+// Parse '-v' and '-F' options in preset string, return false on error.
+static inline bool parse_presets(const char * presets,
+ ata_vendor_attr_defs & defs,
+ unsigned char & fix_firmwarebug)
+{
+ return parse_db_presets(presets, &defs, &fix_firmwarebug, 0);
+}
+
+// Parse '-d' option in preset string, return false on error.
+static inline bool parse_usb_type(const char * presets, std::string & type)
+{
+ return parse_db_presets(presets, 0, 0, &type);
+}
+
+// Parse "USB: [DEVICE] ; [BRIDGE]" string
+static void parse_usb_names(const char * names, usb_dev_info & info)
+{
+ int n1 = -1, n2 = -1, n3 = -1;
+ sscanf(names, "USB: %n%*[^;]%n; %n", &n1, &n2, &n3);
+ if (0 < n1 && n1 < n2)
+ info.usb_device.assign(names+n1, n2-n1);
+ else
+ sscanf(names, "USB: ; %n", &n3);
+ if (0 < n3)
+ info.usb_bridge = names+n3;
+}
+
+// Search drivedb for USB device with vendor:product ID.
+int lookup_usb_device(int vendor_id, int product_id, int bcd_device,
+ usb_dev_info & info, usb_dev_info & info2)
+{
+ // Format strings to match
+ char usb_id_str[16], bcd_dev_str[16];
+ snprintf(usb_id_str, sizeof(usb_id_str), "0x%04x:0x%04x", vendor_id, product_id);
+ if (bcd_device >= 0)
+ snprintf(bcd_dev_str, sizeof(bcd_dev_str), "0x%04x", bcd_device);
+ else
+ bcd_dev_str[0] = 0;
+
+ int found = 0;
+ bool bcd_match = false;
+ for (unsigned i = 0; i < knowndrives.size(); i++) {
+ const drive_settings & dbentry = knowndrives[i];
+
+ // Skip drive entries
+ if (!is_usb_entry(&dbentry))
+ continue;
+
+ // Check whether USB vendor:product ID matches
+ if (!match(dbentry.modelregexp, usb_id_str))
+ continue;
+
+ // Parse '-d type'
+ usb_dev_info d;
+ if (!parse_usb_type(dbentry.presets, d.usb_type))
+ return 0; // Syntax error
+ parse_usb_names(dbentry.modelfamily, d);
+
+ // If two entries with same vendor:product ID have different
+ // types, use bcd_device (if provided by OS) to select entry.
+ bool bm = ( *bcd_dev_str && *dbentry.firmwareregexp
+ && match(dbentry.firmwareregexp, bcd_dev_str));
+
+ if (found == 0 || bm > bcd_match) {
+ info = d; found = 1;
+ bcd_match = bm;
+ }
+ else if (info.usb_type != d.usb_type && bm == bcd_match) {
+ // two different entries found
+ info2 = d; found = 2;
+ break;
+ }
+ }
+
+ return found;
+}
+
// Shows one entry of knowndrives[], returns #errors.
static int showonepreset(const drive_settings * dbentry)
{
"this error to smartmontools developers at " PACKAGE_BUGREPORT ".\n");
return 1;
}
-
+
+ bool usb = is_usb_entry(dbentry);
+
// print and check model and firmware regular expressions
int errcnt = 0;
regular_expression regex;
- pout("%-*s %s\n", TABLEPRINTWIDTH, "MODEL REGEXP:", dbentry->modelregexp);
+ pout("%-*s %s\n", TABLEPRINTWIDTH, (!usb ? "MODEL REGEXP:" : "USB Vendor:Product:"),
+ dbentry->modelregexp);
if (!compile(regex, dbentry->modelregexp))
errcnt++;
- pout("%-*s %s\n", TABLEPRINTWIDTH, "FIRMWARE REGEXP:", *dbentry->firmwareregexp ?
- dbentry->firmwareregexp : ".*"); // preserve old output (TODO: Change)
+ pout("%-*s %s\n", TABLEPRINTWIDTH, (!usb ? "FIRMWARE REGEXP:" : "USB bcdDevice:"),
+ *dbentry->firmwareregexp ? dbentry->firmwareregexp : ".*"); // preserve old output (TODO: Change)
if (*dbentry->firmwareregexp && !compile(regex, dbentry->firmwareregexp))
errcnt++;
- pout("%-*s %s\n", TABLEPRINTWIDTH, "MODEL FAMILY:", dbentry->modelfamily);
-
- // if there are any presets, then show them
- unsigned char fix_firmwarebug = 0;
- bool first_preset = true;
- if (*dbentry->presets) {
- ata_vendor_attr_defs defs;
- if (!parse_presets(dbentry->presets, defs, fix_firmwarebug)) {
- pout("Syntax error in preset option string \"%s\"\n", dbentry->presets);
- errcnt++;
+ if (!usb) {
+ pout("%-*s %s\n", TABLEPRINTWIDTH, "MODEL FAMILY:", dbentry->modelfamily);
+
+ // if there are any presets, then show them
+ unsigned char fix_firmwarebug = 0;
+ bool first_preset = true;
+ if (*dbentry->presets) {
+ ata_vendor_attr_defs defs;
+ if (!parse_presets(dbentry->presets, defs, fix_firmwarebug)) {
+ pout("Syntax error in preset option string \"%s\"\n", dbentry->presets);
+ errcnt++;
+ }
+ for (int i = 0; i < MAX_ATTRIBUTE_NUM; i++) {
+ if (defs[i].priority != PRIOR_DEFAULT) {
+ // Use leading zeros instead of spaces so that everything lines up.
+ pout("%-*s %03d %s\n", TABLEPRINTWIDTH, first_preset ? "ATTRIBUTE OPTIONS:" : "",
+ i, ata_get_smart_attr_name(i, defs).c_str());
+ first_preset = false;
+ }
+ }
}
- for (int i = 0; i < MAX_ATTRIBUTE_NUM; i++) {
- if (defs[i].priority != PRIOR_DEFAULT) {
- // Use leading zeros instead of spaces so that everything lines up.
- pout("%-*s %03d %s\n", TABLEPRINTWIDTH, first_preset ? "ATTRIBUTE OPTIONS:" : "",
- i, ata_get_smart_attr_name(i, defs).c_str());
- first_preset = false;
+ if (first_preset)
+ pout("%-*s %s\n", TABLEPRINTWIDTH, "ATTRIBUTE OPTIONS:", "None preset; no -v options are required.");
+
+ // describe firmwarefix
+ if (fix_firmwarebug) {
+ const char * fixdesc;
+ switch (fix_firmwarebug) {
+ case FIX_SAMSUNG:
+ fixdesc = "Fixes byte order in some SMART data (same as -F samsung)";
+ break;
+ case FIX_SAMSUNG2:
+ fixdesc = "Fixes byte order in some SMART data (same as -F samsung2)";
+ break;
+ case FIX_SAMSUNG3:
+ fixdesc = "Fixes completed self-test reported as in progress (same as -F samsung3)";
+ break;
+ default:
+ fixdesc = "UNKNOWN"; errcnt++;
+ break;
}
+ pout("%-*s %s\n", TABLEPRINTWIDTH, "OTHER PRESETS:", fixdesc);
}
}
- if (first_preset)
- pout("%-*s %s\n", TABLEPRINTWIDTH, "ATTRIBUTE OPTIONS:", "None preset; no -v options are required.");
-
- // describe firmwarefix
- if (fix_firmwarebug) {
- const char * fixdesc;
- switch (fix_firmwarebug) {
- case FIX_SAMSUNG:
- fixdesc = "Fixes byte order in some SMART data (same as -F samsung)";
- break;
- case FIX_SAMSUNG2:
- fixdesc = "Fixes byte order in some SMART data (same as -F samsung2)";
- break;
- case FIX_SAMSUNG3:
- fixdesc = "Fixes completed self-test reported as in progress (same as -F samsung3)";
- break;
- default:
- fixdesc = "UNKNOWN"; errcnt++;
- break;
+ else {
+ // Print USB info
+ usb_dev_info info; parse_usb_names(dbentry->modelfamily, info);
+ pout("%-*s %s\n", TABLEPRINTWIDTH, "USB Device:",
+ (!info.usb_device.empty() ? info.usb_device.c_str() : "[unknown]"));
+ pout("%-*s %s\n", TABLEPRINTWIDTH, "USB Bridge:",
+ (!info.usb_bridge.empty() ? info.usb_bridge.c_str() : "[unknown]"));
+
+ if (*dbentry->presets && !parse_usb_type(dbentry->presets, info.usb_type)) {
+ pout("Syntax error in USB type string \"%s\"\n", dbentry->presets);
+ errcnt++;
}
- pout("%-*s %s\n", TABLEPRINTWIDTH, "OTHER PRESETS:", fixdesc);
+ pout("%-*s %s\n", TABLEPRINTWIDTH, "USB Type",
+ (!info.usb_type.empty() ? info.usb_type.c_str() : "[unsupported]"));
}
// Print any special warnings
break;
case 4:
if (!token.value.empty()) {
- ata_vendor_attr_defs defs; unsigned char fix = 0;
- if (!parse_presets(token.value.c_str(), defs, fix)) {
- pout("%s(%d): Syntax error in preset option string\n", path, token.line);
- ok = false;
+ if (!is_usb_modelfamily(values[0].c_str())) {
+ ata_vendor_attr_defs defs; unsigned char fix = 0;
+ if (!parse_presets(token.value.c_str(), defs, fix)) {
+ pout("%s(%d): Syntax error in preset option string\n", path, token.line);
+ ok = false;
+ }
+ }
+ else {
+ std::string type;
+ if (!parse_usb_type(token.value.c_str(), type)) {
+ pout("%s(%d): Syntax error in USB type string\n", path, token.line);
+ ok = false;
+ }
}
}
break;
return parse_drive_database(parse_ptr(f), knowndrives, path);
}
-// Read drive databases from standard places.
-bool read_default_drive_databases()
+// Get path for additional database file
+const char * get_drivedb_path_add()
{
#ifndef _WIN32
- // Read file for local additions: /{,usr/local/}etc/smart_drivedb.h
- static const char db1[] = SMARTMONTOOLS_SYSCONFDIR"/smart_drivedb.h";
+ return SMARTMONTOOLS_SYSCONFDIR"/smart_drivedb.h";
+#else
+ static std::string path = get_exe_dir() + "/drivedb-add.h";
+ return path.c_str();
+#endif
+}
+
+#ifdef SMARTMONTOOLS_DRIVEDBDIR
+
+// Get path for default database file
+const char * get_drivedb_path_default()
+{
+#ifndef _WIN32
+ return SMARTMONTOOLS_DRIVEDBDIR"/drivedb.h";
#else
- static const char db1[] = "./smart_drivedb.h";
+ static std::string path = get_exe_dir() + "/drivedb.h";
+ return path.c_str();
+#endif
+}
+
#endif
+
+// Read drive databases from standard places.
+bool read_default_drive_databases()
+{
+ // Read file for local additions: /{,usr/local/}etc/smart_drivedb.h
+ const char * db1 = get_drivedb_path_add();
if (!access(db1, 0)) {
if (!read_drive_database(db1))
return false;
}
#ifdef SMARTMONTOOLS_DRIVEDBDIR
- // Read file from package: // /usr/{,local/}share/smartmontools/drivedb.h
- static const char db2[] = SMARTMONTOOLS_DRIVEDBDIR"/drivedb.h";
+ // Read file from package: /usr/{,local/}share/smartmontools/drivedb.h
+ const char * db2 = get_drivedb_path_default();
if (!access(db2, 0)) {
if (!read_drive_database(db2))
return false;
* Home page of code is: http://smartmontools.sourceforge.net
* Address of support mailing list: smartmontools-support@lists.sourceforge.net
*
- * Copyright (C) 2003-9 Philip Williams, Bruce Allen
- * Copyright (C) 2008-9 Christian Franke <smartmontools-support@lists.sourceforge.net>
+ * Copyright (C) 2003-10 Philip Williams, Bruce Allen
+ * Copyright (C) 2008-10 Christian Franke <smartmontools-support@lists.sourceforge.net>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
#ifndef KNOWNDRIVES_H_
#define KNOWNDRIVES_H_
-#define KNOWNDRIVES_H_CVSID "$Id: knowndrives.h 2998 2009-12-11 22:51:04Z chrfranke $\n"
+#define KNOWNDRIVES_H_CVSID "$Id: knowndrives.h 3093 2010-04-30 09:57:36Z chrfranke $\n"
// Structure to store drive database entries, see drivedb.h for a description.
struct drive_settings {
// string.
const drive_settings * lookup_drive(const char * model, const char * firmware);
+// info returned by lookup_usb_device()
+struct usb_dev_info
+{
+ std::string usb_device; // Device name, empty if unknown
+ std::string usb_bridge; // USB bridge name, empty if unknown
+ std::string usb_type; // Type string ('-d' option).
+};
+
+// Search drivedb for USB device with vendor:product ID.
+int lookup_usb_device(int vendor_id, int product_id, int bcd_device,
+ usb_dev_info & info, usb_dev_info & info2);
+
// Shows the presets (if any) that are available for the given drive.
void show_presets(const ata_identify_device * drive, bool fix_swapped_id);
bool apply_presets(const ata_identify_device * drive, ata_vendor_attr_defs & defs,
unsigned char & fix_firmwarebug, bool fix_swapped_id);
+// Get path for additional database file
+const char * get_drivedb_path_add();
+
+#ifdef SMARTMONTOOLS_DRIVEDBDIR
+// Get path for default database file
+const char * get_drivedb_path_default();
+#endif
+
// Read drive database from file.
bool read_drive_database(const char * path);
uint8_t status;
} __attribute__((packed)) megacmd_t;
-typedef struct {
+typedef union {
uint8_t *pointer;
-#if BITS_PER_LONG == 32
- uint8_t pad[4];
-#endif
+ uint8_t pad[8];
} ptr_t;
+// The above definition assumes sizeof(void*) <= 8.
+// This assumption also exists in the linux megaraid device driver.
+// So define a macro to check expected size of ptr_t at compile time using
+// a dummy typedef. On size mismatch, compiler reports a negative array
+// size. If you see an error message of this form, it means that
+// you have an unexpected pointer size on your platform and can not
+// use megaraid support in smartmontools.
+typedef char assert_sizeof_ptr_t[sizeof(ptr_t) == 8 ? 1 : -1];
+
struct uioctl_t
{
uint32_t inlen;
#define PATHINQ_SETTINGS_SIZE 128
#endif
-static __unused const char *filenameandversion="$Id: os_freebsd.cpp 3066 2010-02-15 23:10:49Z samm2 $";
+static __unused const char *filenameandversion="$Id: os_freebsd.cpp 3098 2010-04-30 17:35:35Z chrfranke $";
-const char *os_XXXX_c_cvsid="$Id: os_freebsd.cpp 3066 2010-02-15 23:10:49Z samm2 $" \
+const char *os_XXXX_c_cvsid="$Id: os_freebsd.cpp 3098 2010-04-30 17:35:35Z chrfranke $" \
ATACMDS_H_CVSID CCISS_H_CVSID CONFIG_H_CVSID INT64_H_CVSID OS_FREEBSD_H_CVSID SCSICMDS_H_CVSID UTILITY_H_CVSID;
extern smartmonctrl * con;
// global variable holding byte count of allocated memory
long long bytes;
-const char * dev_freebsd_cpp_cvsid = "$Id: os_freebsd.cpp 3066 2010-02-15 23:10:49Z samm2 $"
+const char * dev_freebsd_cpp_cvsid = "$Id: os_freebsd.cpp 3098 2010-04-30 17:35:35Z chrfranke $"
DEV_INTERFACE_H_CVSID;
extern smartmonctrl * con; // con->reportscsiioctl
set_err(EINVAL, "Option -d cciss,N requires N to be a non-negative integer");
return 0;
}
- if (!(0 <= disknum && disknum <= 15)) {
- set_err(EINVAL, "Option -d cciss,N (N=%d) must have 0 <= N <= 15", disknum);
+ if (!(0 <= disknum && disknum <= 127)) {
+ set_err(EINVAL, "Option -d cciss,N (N=%d) must have 0 <= N <= 127", disknum);
return 0;
}
return new freebsd_cciss_device(this, name, disknum);
* Home page of code is: http://smartmontools.sourceforge.net
*
* Copyright (C) 2003-10 Bruce Allen <smartmontools-support@lists.sourceforge.net>
- * Copyright (C) 2003-10 Doug Gilbert <dougg@torque.net>
+ * Copyright (C) 2003-10 Doug Gilbert <dgilbert@interlog.com>
* Copyright (C) 2008 Hank Wu <hank@areca.com.tw>
* Copyright (C) 2008 Oliver Bock <brevilo@users.sourceforge.net>
* Copyright (C) 2008-10 Christian Franke <smartmontools-support@lists.sourceforge.net>
#define ARGUSED(x) ((void)(x))
-const char *os_XXXX_c_cvsid="$Id: os_linux.cpp 3076 2010-03-12 22:23:08Z chrfranke $" \
+const char *os_XXXX_c_cvsid="$Id: os_linux.cpp 3098 2010-04-30 17:35:35Z chrfranke $" \
ATACMDS_H_CVSID CONFIG_H_CVSID INT64_H_CVSID OS_LINUX_H_CVSID SCSICMDS_H_CVSID UTILITY_H_CVSID;
/* for passing global control variables */
set_err(EINVAL, "Option -d cciss,N requires N to be a non-negative integer");
return 0;
}
- if (!(0 <= disknum && disknum <= 15)) {
- set_err(EINVAL, "Option -d cciss,N (N=%d) must have 0 <= N <= 15", disknum);
+ if (!(0 <= disknum && disknum <= 127)) {
+ set_err(EINVAL, "Option -d cciss,N (N=%d) must have 0 <= N <= 127", disknum);
return 0;
}
return new linux_cciss_device(this, name, disknum);
// should have one *_H_CVSID macro appearing below for each file
// appearing with #include "*.h" above. Please list these (below) in
// alphabetic/dictionary order.
-const char *os_XXXX_c_cvsid="$Id: os_qnxnto.cpp,v 1.3 2008/06/12 21:46:31 ballen4705 Exp $" \
+const char *os_XXXX_c_cvsid="$Id: os_qnxnto.cpp 3110 2010-05-24 20:38:38Z chrfranke $" \
ATACMDS_H_CVSID CONFIG_H_CVSID INT64_H_CVSID OS_QNXNTO_H_CVSID SCSICMDS_H_CVSID UTILITY_H_CVSID;
cpt.cam_timeout=cpt.cam_timeout?cpt.cam_timeout:CAM_TIME_DEFAULT;
if(cpt.cam_sense_len)
{
- SETIOV(&iov[1],cpt.cam_sense_ptr,cpt.cam_sense_len);
+ SETIOV(&iov[1],(void *)cpt.cam_sense_ptr,cpt.cam_sense_len);
cpt.cam_sense_ptr=sizeof(cpt);
icnt++;
}
#define SELECT_WIN_32_64(x32, x64) (x64)
#endif
-const char * os_win32_cpp_cvsid = "$Id: os_win32.cpp 3062 2010-02-09 21:02:27Z chrfranke $";
+const char * os_win32_cpp_cvsid = "$Id: os_win32.cpp 3118 2010-06-08 17:30:46Z chrfranke $";
// Disable Win9x/ME specific code if no longer supported by compiler.
#ifdef _WIN64
" Revision: \"%s\"\n"
" Removable: %s\n"
" BusType: 0x%02x\n",
- (data->desc.VendorIdOffset ? data->raw+data->desc.VendorIdOffset : ""),
- (data->desc.ProductIdOffset ? data->raw+data->desc.ProductIdOffset : ""),
- (data->desc.ProductRevisionOffset ? data->raw+data->desc.ProductRevisionOffset : ""),
+ (data->desc.VendorIdOffset ? data->raw+data->desc.VendorIdOffset : "(null)"),
+ (data->desc.ProductIdOffset ? data->raw+data->desc.ProductIdOffset : "(null)"),
+ (data->desc.ProductRevisionOffset ? data->raw+data->desc.ProductRevisionOffset : "(null)"),
(data->desc.RemovableMedia? "Yes":"No"), data->desc.BusType
);
}
return -1;
memset(id, 0, sizeof(*id));
- if (data.desc.ProductIdOffset)
- copy_swapped(id->model, data.raw+data.desc.ProductIdOffset, sizeof(id->model));
+
+ // Some drivers split ATA model string into VendorId and ProductId,
+ // others return it as ProductId only.
+ char model[sizeof(id->model) + 1] = "";
+
+ unsigned i = 0;
+ if (data.desc.VendorIdOffset) {
+ for ( ;i < sizeof(model)-1 && data.raw[data.desc.VendorIdOffset+i]; i++)
+ model[i] = data.raw[data.desc.VendorIdOffset+i];
+ }
+
+ if (data.desc.ProductIdOffset) {
+ while (i > 1 && model[i-2] == ' ') // Keep last blank from VendorId
+ i--;
+ for (unsigned j = 0; i < sizeof(model)-1 && data.raw[data.desc.ProductIdOffset+j]; i++, j++)
+ model[i] = data.raw[data.desc.ProductIdOffset+j];
+ }
+
+ while (i > 0 && model[i-1] == ' ')
+ i--;
+ model[i] = 0;
+ copy_swapped(id->model, model, sizeof(id->model));
+
if (data.desc.ProductRevisionOffset)
copy_swapped(id->fw_rev, data.raw+data.desc.ProductRevisionOffset, sizeof(id->fw_rev));
+
id->command_set_1 = 0x0001; id->command_set_2 = 0x4000; // SMART supported, words 82,83 valid
id->cfs_enable_1 = 0x0001; id->csf_default = 0x4000; // SMART enabled, words 85,87 valid
return 0;
}
}
+
+#ifndef __CYGWIN__
+
+// Get exe directory
+// (prototype in utiliy.h)
+std::string get_exe_dir()
+{
+ char path[MAX_PATH];
+ // Get path of this exe
+ if (!GetModuleFileNameA(GetModuleHandleA(0), path, sizeof(path)))
+ throw std::runtime_error("GetModuleFileName() failed");
+ // Replace backslash by slash
+ int sl = -1;
+ for (int i = 0; path[i]; i++)
+ if (path[i] == '\\') {
+ path[i] = '/'; sl = i;
+ }
+ // Remove filename
+ if (sl >= 0)
+ path[sl] = 0;
+ return path;
+}
+
+#endif
<Tool
Name="VCCLCompilerTool"
Optimization="0"
- AdditionalIncludeDirectories=".,..\posix"
+ AdditionalIncludeDirectories=".,..\getopt,..\regex"
PreprocessorDefinitions="_DEBUG;HAVE_CONFIG_H;_ERRCODE_DEFINED;errno_t=int;_USE_32BIT_TIME_T;_CRT_NONSTDC_NO_DEPRECATE;_CRT_SECURE_NO_DEPRECATE"
MinimalRebuild="true"
BasicRuntimeChecks="3"
/>
<Tool
Name="VCCLCompilerTool"
- AdditionalIncludeDirectories=".,..\posix"
+ AdditionalIncludeDirectories=".,..\getopt,..\regex"
PreprocessorDefinitions="NDEBUG;HAVE_CONFIG_H;_ERRCODE_DEFINED;errno_t=int;_USE_32BIT_TIME_T;_CRT_NONSTDC_NO_DEPRECATE;_CRT_SECURE_NO_DEPRECATE"
RuntimeLibrary="2"
UsePrecompiledHeader="0"
</File>
</Filter>
<Filter
- Name="posix"
+ Name="regex"
>
<File
- RelativePath="..\posix\getopt.c"
- >
- </File>
- <File
- RelativePath="..\posix\getopt.h"
- >
- </File>
- <File
- RelativePath="..\posix\getopt1.c"
- >
- </File>
- <File
- RelativePath="..\posix\regcomp.c"
+ RelativePath="..\regex\regcomp.c"
>
<FileConfiguration
Name="Debug|Win32"
</FileConfiguration>
</File>
<File
- RelativePath="..\posix\regex.c"
+ RelativePath="..\regex\regex.c"
>
</File>
<File
- RelativePath="..\posix\regex.h"
+ RelativePath="..\regex\regex.h"
>
</File>
<File
- RelativePath="..\posix\regex_internal.c"
+ RelativePath="..\regex\regex_internal.c"
>
<FileConfiguration
Name="Debug|Win32"
</FileConfiguration>
</File>
<File
- RelativePath="..\posix\regex_internal.h"
+ RelativePath="..\regex\regex_internal.h"
>
</File>
<File
- RelativePath="..\posix\regexec.c"
+ RelativePath="..\regex\regexec.c"
>
<FileConfiguration
Name="Debug|Win32"
</FileConfiguration>
</File>
</Filter>
+ <Filter
+ Name="getopt"
+ >
+ <File
+ RelativePath="..\getopt\getopt.c"
+ >
+ </File>
+ <File
+ RelativePath="..\getopt\getopt.h"
+ >
+ </File>
+ <File
+ RelativePath="..\getopt\getopt1.c"
+ >
+ </File>
+ </Filter>
<File
RelativePath="..\atacmdnames.cpp"
>
<Tool
Name="VCCLCompilerTool"
Optimization="0"
- AdditionalIncludeDirectories=".,..\posix"
+ AdditionalIncludeDirectories=".,..\getopt,..\regex"
PreprocessorDefinitions="_DEBUG;HAVE_CONFIG_H;_ERRCODE_DEFINED;errno_t=int;_USE_32BIT_TIME_T;_CRT_NONSTDC_NO_DEPRECATE;_CRT_SECURE_NO_DEPRECATE"
MinimalRebuild="true"
BasicRuntimeChecks="3"
/>
<Tool
Name="VCCLCompilerTool"
- AdditionalIncludeDirectories=".,..\posix"
+ AdditionalIncludeDirectories=".,..\getopt,..\regex"
PreprocessorDefinitions="NDEBUG;HAVE_CONFIG_H;_ERRCODE_DEFINED;errno_t=int;_USE_32BIT_TIME_T;_CRT_NONSTDC_NO_DEPRECATE;_CRT_SECURE_NO_DEPRECATE"
RuntimeLibrary="2"
UsePrecompiledHeader="0"
</File>
</Filter>
<Filter
- Name="posix"
+ Name="regex"
>
<File
- RelativePath="..\posix\getopt.c"
- >
- </File>
- <File
- RelativePath="..\posix\getopt.h"
- >
- </File>
- <File
- RelativePath="..\posix\getopt1.c"
- >
- </File>
- <File
- RelativePath="..\posix\regcomp.c"
+ RelativePath="..\regex\regcomp.c"
>
<FileConfiguration
Name="Debug|Win32"
</FileConfiguration>
</File>
<File
- RelativePath="..\posix\regex.c"
+ RelativePath="..\regex\regex.c"
>
</File>
<File
- RelativePath="..\posix\regex.h"
+ RelativePath="..\regex\regex.h"
>
</File>
<File
- RelativePath="..\posix\regex_internal.c"
+ RelativePath="..\regex\regex_internal.c"
>
<FileConfiguration
Name="Debug|Win32"
</FileConfiguration>
</File>
<File
- RelativePath="..\posix\regex_internal.h"
+ RelativePath="..\regex\regex_internal.h"
>
</File>
<File
- RelativePath="..\posix\regexec.c"
+ RelativePath="..\regex\regexec.c"
>
<FileConfiguration
Name="Debug|Win32"
</FileConfiguration>
</File>
</Filter>
+ <Filter
+ Name="getopt"
+ >
+ <File
+ RelativePath="..\getopt\getopt.c"
+ >
+ </File>
+ <File
+ RelativePath="..\getopt\getopt.h"
+ >
+ </File>
+ <File
+ RelativePath="..\getopt\getopt1.c"
+ >
+ </File>
+ </Filter>
<File
RelativePath="..\atacmdnames.cpp"
>
+++ /dev/null
-/* Getopt for GNU.
- NOTE: getopt is now part of the C library, so if you don't know what
- "Keep this file name-space clean" means, talk to drepper@gnu.org
- before changing it!
- Copyright (C) 1987,88,89,90,91,92,93,94,95,96,98,99,2000,2001,2002
- Free Software Foundation, Inc.
- This file is part of the GNU C Library.
-
- The GNU C Library is free software; you can redistribute it and/or
- modify it under the terms of the GNU Lesser General Public
- License as published by the Free Software Foundation; either
- version 2.1 of the License, or (at your option) any later version.
-
- The GNU C Library is distributed in the hope that it will be useful,
- but WITHOUT ANY WARRANTY; without even the implied warranty of
- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- Lesser General Public License for more details.
-
- You should have received a copy of the GNU Lesser General Public
- License along with the GNU C Library; if not, write to the Free
- Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
- 02111-1307 USA. */
-\f
-/* This tells Alpha OSF/1 not to define a getopt prototype in <stdio.h>.
- Ditto for AIX 3.2 and <stdlib.h>. */
-#ifndef _NO_PROTO
-# define _NO_PROTO
-#endif
-
-#ifdef HAVE_CONFIG_H
-# include <config.h>
-#endif
-
-#if !defined __STDC__ || !__STDC__
-/* This is a separate conditional since some stdc systems
- reject `defined (const)'. */
-# ifndef const
-# define const
-# endif
-#endif
-
-#include <stdio.h>
-
-/* Comment out all this code if we are using the GNU C Library, and are not
- actually compiling the library itself. This code is part of the GNU C
- Library, but also included in many other GNU distributions. Compiling
- and linking in this code is a waste when using the GNU C library
- (especially if it is a shared library). Rather than having every GNU
- program understand `configure --with-gnu-libc' and omit the object files,
- it is simpler to just do this in the source for each such file. */
-
-#define GETOPT_INTERFACE_VERSION 2
-#if !defined _LIBC && defined __GLIBC__ && __GLIBC__ >= 2
-# include <gnu-versions.h>
-# if _GNU_GETOPT_INTERFACE_VERSION == GETOPT_INTERFACE_VERSION
-# define ELIDE_CODE
-# endif
-#endif
-
-#ifndef ELIDE_CODE
-
-
-/* This needs to come after some library #include
- to get __GNU_LIBRARY__ defined. */
-#ifdef __GNU_LIBRARY__
-/* Don't include stdlib.h for non-GNU C libraries because some of them
- contain conflicting prototypes for getopt. */
-# include <stdlib.h>
-# include <unistd.h>
-#endif /* GNU C library. */
-
-#ifdef VMS
-# include <unixlib.h>
-# if HAVE_STRING_H - 0
-# include <string.h>
-# endif
-#endif
-
-#ifndef _
-/* This is for other GNU distributions with internationalized messages. */
-# if (HAVE_LIBINTL_H && ENABLE_NLS) || defined _LIBC
-# include <libintl.h>
-# ifndef _
-# define _(msgid) gettext (msgid)
-# endif
-# else
-# define _(msgid) (msgid)
-# endif
-# if defined _LIBC && defined USE_IN_LIBIO
-# include <wchar.h>
-# endif
-#endif
-
-#ifndef attribute_hidden
-# define attribute_hidden
-#endif
-
-/* This version of `getopt' appears to the caller like standard Unix `getopt'
- but it behaves differently for the user, since it allows the user
- to intersperse the options with the other arguments.
-
- As `getopt' works, it permutes the elements of ARGV so that,
- when it is done, all the options precede everything else. Thus
- all application programs are extended to handle flexible argument order.
-
- Setting the environment variable POSIXLY_CORRECT disables permutation.
- Then the behavior is completely standard.
-
- GNU application programs can use a third alternative mode in which
- they can distinguish the relative order of options and other arguments. */
-
-#include "getopt.h"
-
-/* For communication from `getopt' to the caller.
- When `getopt' finds an option that takes an argument,
- the argument value is returned here.
- Also, when `ordering' is RETURN_IN_ORDER,
- each non-option ARGV-element is returned here. */
-
-char *optarg;
-
-/* Index in ARGV of the next element to be scanned.
- This is used for communication to and from the caller
- and for communication between successive calls to `getopt'.
-
- On entry to `getopt', zero means this is the first call; initialize.
-
- When `getopt' returns -1, this is the index of the first of the
- non-option elements that the caller should itself scan.
-
- Otherwise, `optind' communicates from one call to the next
- how much of ARGV has been scanned so far. */
-
-/* 1003.2 says this must be 1 before any call. */
-int optind = 1;
-
-/* Formerly, initialization of getopt depended on optind==0, which
- causes problems with re-calling getopt as programs generally don't
- know that. */
-
-int __getopt_initialized attribute_hidden;
-
-/* The next char to be scanned in the option-element
- in which the last option character we returned was found.
- This allows us to pick up the scan where we left off.
-
- If this is zero, or a null string, it means resume the scan
- by advancing to the next ARGV-element. */
-
-static char *nextchar;
-
-/* Callers store zero here to inhibit the error message
- for unrecognized options. */
-
-int opterr = 1;
-
-/* Set to an option character which was unrecognized.
- This must be initialized on some systems to avoid linking in the
- system's own getopt implementation. */
-
-int optopt = '?';
-
-/* Describe how to deal with options that follow non-option ARGV-elements.
-
- If the caller did not specify anything,
- the default is REQUIRE_ORDER if the environment variable
- POSIXLY_CORRECT is defined, PERMUTE otherwise.
-
- REQUIRE_ORDER means don't recognize them as options;
- stop option processing when the first non-option is seen.
- This is what Unix does.
- This mode of operation is selected by either setting the environment
- variable POSIXLY_CORRECT, or using `+' as the first character
- of the list of option characters.
-
- PERMUTE is the default. We permute the contents of ARGV as we scan,
- so that eventually all the non-options are at the end. This allows options
- to be given in any order, even with programs that were not written to
- expect this.
-
- RETURN_IN_ORDER is an option available to programs that were written
- to expect options and other ARGV-elements in any order and that care about
- the ordering of the two. We describe each non-option ARGV-element
- as if it were the argument of an option with character code 1.
- Using `-' as the first character of the list of option characters
- selects this mode of operation.
-
- The special argument `--' forces an end of option-scanning regardless
- of the value of `ordering'. In the case of RETURN_IN_ORDER, only
- `--' can cause `getopt' to return -1 with `optind' != ARGC. */
-
-static enum
-{
- REQUIRE_ORDER, PERMUTE, RETURN_IN_ORDER
-} ordering;
-
-/* Value of POSIXLY_CORRECT environment variable. */
-static char *posixly_correct;
-\f
-#ifdef __GNU_LIBRARY__
-/* We want to avoid inclusion of string.h with non-GNU libraries
- because there are many ways it can cause trouble.
- On some systems, it contains special magic macros that don't work
- in GCC. */
-# include <string.h>
-# define my_index strchr
-#else
-
-# if HAVE_STRING_H
-# include <string.h>
-# else
-# include <strings.h>
-# endif
-
-/* Avoid depending on library functions or files
- whose names are inconsistent. */
-
-#ifndef getenv
-extern char *getenv ();
-#endif
-
-static char *
-my_index (str, chr)
- const char *str;
- int chr;
-{
- while (*str)
- {
- if (*str == chr)
- return (char *) str;
- str++;
- }
- return 0;
-}
-
-/* If using GCC, we can safely declare strlen this way.
- If not using GCC, it is ok not to declare it. */
-#ifdef __GNUC__
-/* Note that Motorola Delta 68k R3V7 comes with GCC but not stddef.h.
- That was relevant to code that was here before. */
-# if (!defined __STDC__ || !__STDC__) && !defined strlen
-/* gcc with -traditional declares the built-in strlen to return int,
- and has done so at least since version 2.4.5. -- rms. */
-extern int strlen (const char *);
-# endif /* not __STDC__ */
-#endif /* __GNUC__ */
-
-#endif /* not __GNU_LIBRARY__ */
-\f
-/* Handle permutation of arguments. */
-
-/* Describe the part of ARGV that contains non-options that have
- been skipped. `first_nonopt' is the index in ARGV of the first of them;
- `last_nonopt' is the index after the last of them. */
-
-static int first_nonopt;
-static int last_nonopt;
-
-#ifdef _LIBC
-/* Stored original parameters.
- XXX This is no good solution. We should rather copy the args so
- that we can compare them later. But we must not use malloc(3). */
-extern int __libc_argc;
-extern char **__libc_argv;
-
-/* Bash 2.0 gives us an environment variable containing flags
- indicating ARGV elements that should not be considered arguments. */
-
-# ifdef USE_NONOPTION_FLAGS
-/* Defined in getopt_init.c */
-extern char *__getopt_nonoption_flags;
-
-static int nonoption_flags_max_len;
-static int nonoption_flags_len;
-# endif
-
-# ifdef USE_NONOPTION_FLAGS
-# define SWAP_FLAGS(ch1, ch2) \
- if (nonoption_flags_len > 0) \
- { \
- char __tmp = __getopt_nonoption_flags[ch1]; \
- __getopt_nonoption_flags[ch1] = __getopt_nonoption_flags[ch2]; \
- __getopt_nonoption_flags[ch2] = __tmp; \
- }
-# else
-# define SWAP_FLAGS(ch1, ch2)
-# endif
-#else /* !_LIBC */
-# define SWAP_FLAGS(ch1, ch2)
-#endif /* _LIBC */
-
-/* Exchange two adjacent subsequences of ARGV.
- One subsequence is elements [first_nonopt,last_nonopt)
- which contains all the non-options that have been skipped so far.
- The other is elements [last_nonopt,optind), which contains all
- the options processed since those non-options were skipped.
-
- `first_nonopt' and `last_nonopt' are relocated so that they describe
- the new indices of the non-options in ARGV after they are moved. */
-
-#if defined __STDC__ && __STDC__
-static void exchange (char **);
-#endif
-
-static void
-exchange (argv)
- char **argv;
-{
- int bottom = first_nonopt;
- int middle = last_nonopt;
- int top = optind;
- char *tem;
-
- /* Exchange the shorter segment with the far end of the longer segment.
- That puts the shorter segment into the right place.
- It leaves the longer segment in the right place overall,
- but it consists of two parts that need to be swapped next. */
-
-#if defined _LIBC && defined USE_NONOPTION_FLAGS
- /* First make sure the handling of the `__getopt_nonoption_flags'
- string can work normally. Our top argument must be in the range
- of the string. */
- if (nonoption_flags_len > 0 && top >= nonoption_flags_max_len)
- {
- /* We must extend the array. The user plays games with us and
- presents new arguments. */
- char *new_str = malloc (top + 1);
- if (new_str == NULL)
- nonoption_flags_len = nonoption_flags_max_len = 0;
- else
- {
- memset (__mempcpy (new_str, __getopt_nonoption_flags,
- nonoption_flags_max_len),
- '\0', top + 1 - nonoption_flags_max_len);
- nonoption_flags_max_len = top + 1;
- __getopt_nonoption_flags = new_str;
- }
- }
-#endif
-
- while (top > middle && middle > bottom)
- {
- if (top - middle > middle - bottom)
- {
- /* Bottom segment is the short one. */
- int len = middle - bottom;
- register int i;
-
- /* Swap it with the top part of the top segment. */
- for (i = 0; i < len; i++)
- {
- tem = argv[bottom + i];
- argv[bottom + i] = argv[top - (middle - bottom) + i];
- argv[top - (middle - bottom) + i] = tem;
- SWAP_FLAGS (bottom + i, top - (middle - bottom) + i);
- }
- /* Exclude the moved bottom segment from further swapping. */
- top -= len;
- }
- else
- {
- /* Top segment is the short one. */
- int len = top - middle;
- register int i;
-
- /* Swap it with the bottom part of the bottom segment. */
- for (i = 0; i < len; i++)
- {
- tem = argv[bottom + i];
- argv[bottom + i] = argv[middle + i];
- argv[middle + i] = tem;
- SWAP_FLAGS (bottom + i, middle + i);
- }
- /* Exclude the moved top segment from further swapping. */
- bottom += len;
- }
- }
-
- /* Update records for the slots the non-options now occupy. */
-
- first_nonopt += (optind - last_nonopt);
- last_nonopt = optind;
-}
-
-/* Initialize the internal data when the first call is made. */
-
-#if defined __STDC__ && __STDC__
-static const char *_getopt_initialize (int, char *const *, const char *);
-#endif
-static const char *
-_getopt_initialize (argc, argv, optstring)
- int argc;
- char *const *argv;
- const char *optstring;
-{
- /* Start processing options with ARGV-element 1 (since ARGV-element 0
- is the program name); the sequence of previously skipped
- non-option ARGV-elements is empty. */
-
- first_nonopt = last_nonopt = optind;
-
- nextchar = NULL;
-
- posixly_correct = getenv ("POSIXLY_CORRECT");
-
- /* Determine how to handle the ordering of options and nonoptions. */
-
- if (optstring[0] == '-')
- {
- ordering = RETURN_IN_ORDER;
- ++optstring;
- }
- else if (optstring[0] == '+')
- {
- ordering = REQUIRE_ORDER;
- ++optstring;
- }
- else if (posixly_correct != NULL)
- ordering = REQUIRE_ORDER;
- else
- ordering = PERMUTE;
-
-#if defined _LIBC && defined USE_NONOPTION_FLAGS
- if (posixly_correct == NULL
- && argc == __libc_argc && argv == __libc_argv)
- {
- if (nonoption_flags_max_len == 0)
- {
- if (__getopt_nonoption_flags == NULL
- || __getopt_nonoption_flags[0] == '\0')
- nonoption_flags_max_len = -1;
- else
- {
- const char *orig_str = __getopt_nonoption_flags;
- int len = nonoption_flags_max_len = strlen (orig_str);
- if (nonoption_flags_max_len < argc)
- nonoption_flags_max_len = argc;
- __getopt_nonoption_flags =
- (char *) malloc (nonoption_flags_max_len);
- if (__getopt_nonoption_flags == NULL)
- nonoption_flags_max_len = -1;
- else
- memset (__mempcpy (__getopt_nonoption_flags, orig_str, len),
- '\0', nonoption_flags_max_len - len);
- }
- }
- nonoption_flags_len = nonoption_flags_max_len;
- }
- else
- nonoption_flags_len = 0;
-#endif
-
- return optstring;
-}
-\f
-/* Scan elements of ARGV (whose length is ARGC) for option characters
- given in OPTSTRING.
-
- If an element of ARGV starts with '-', and is not exactly "-" or "--",
- then it is an option element. The characters of this element
- (aside from the initial '-') are option characters. If `getopt'
- is called repeatedly, it returns successively each of the option characters
- from each of the option elements.
-
- If `getopt' finds another option character, it returns that character,
- updating `optind' and `nextchar' so that the next call to `getopt' can
- resume the scan with the following option character or ARGV-element.
-
- If there are no more option characters, `getopt' returns -1.
- Then `optind' is the index in ARGV of the first ARGV-element
- that is not an option. (The ARGV-elements have been permuted
- so that those that are not options now come last.)
-
- OPTSTRING is a string containing the legitimate option characters.
- If an option character is seen that is not listed in OPTSTRING,
- return '?' after printing an error message. If you set `opterr' to
- zero, the error message is suppressed but we still return '?'.
-
- If a char in OPTSTRING is followed by a colon, that means it wants an arg,
- so the following text in the same ARGV-element, or the text of the following
- ARGV-element, is returned in `optarg'. Two colons mean an option that
- wants an optional arg; if there is text in the current ARGV-element,
- it is returned in `optarg', otherwise `optarg' is set to zero.
-
- If OPTSTRING starts with `-' or `+', it requests different methods of
- handling the non-option ARGV-elements.
- See the comments about RETURN_IN_ORDER and REQUIRE_ORDER, above.
-
- Long-named options begin with `--' instead of `-'.
- Their names may be abbreviated as long as the abbreviation is unique
- or is an exact match for some defined option. If they have an
- argument, it follows the option name in the same ARGV-element, separated
- from the option name by a `=', or else the in next ARGV-element.
- When `getopt' finds a long-named option, it returns 0 if that option's
- `flag' field is nonzero, the value of the option's `val' field
- if the `flag' field is zero.
-
- The elements of ARGV aren't really const, because we permute them.
- But we pretend they're const in the prototype to be compatible
- with other systems.
-
- LONGOPTS is a vector of `struct option' terminated by an
- element containing a name which is zero.
-
- LONGIND returns the index in LONGOPT of the long-named option found.
- It is only valid when a long-named option has been found by the most
- recent call.
-
- If LONG_ONLY is nonzero, '-' as well as '--' can introduce
- long-named options. */
-
-int
-_getopt_internal (argc, argv, optstring, longopts, longind, long_only)
- int argc;
- char *const *argv;
- const char *optstring;
- const struct option *longopts;
- int *longind;
- int long_only;
-{
- int print_errors = opterr;
- if (optstring[0] == ':')
- print_errors = 0;
-
- if (argc < 1)
- return -1;
-
- optarg = NULL;
-
- if (optind == 0 || !__getopt_initialized)
- {
- if (optind == 0)
- optind = 1; /* Don't scan ARGV[0], the program name. */
- optstring = _getopt_initialize (argc, argv, optstring);
- __getopt_initialized = 1;
- }
-
- /* Test whether ARGV[optind] points to a non-option argument.
- Either it does not have option syntax, or there is an environment flag
- from the shell indicating it is not an option. The later information
- is only used when the used in the GNU libc. */
-#if defined _LIBC && defined USE_NONOPTION_FLAGS
-# define NONOPTION_P (argv[optind][0] != '-' || argv[optind][1] == '\0' \
- || (optind < nonoption_flags_len \
- && __getopt_nonoption_flags[optind] == '1'))
-#else
-# define NONOPTION_P (argv[optind][0] != '-' || argv[optind][1] == '\0')
-#endif
-
- if (nextchar == NULL || *nextchar == '\0')
- {
- /* Advance to the next ARGV-element. */
-
- /* Give FIRST_NONOPT & LAST_NONOPT rational values if OPTIND has been
- moved back by the user (who may also have changed the arguments). */
- if (last_nonopt > optind)
- last_nonopt = optind;
- if (first_nonopt > optind)
- first_nonopt = optind;
-
- if (ordering == PERMUTE)
- {
- /* If we have just processed some options following some non-options,
- exchange them so that the options come first. */
-
- if (first_nonopt != last_nonopt && last_nonopt != optind)
- exchange ((char **) argv);
- else if (last_nonopt != optind)
- first_nonopt = optind;
-
- /* Skip any additional non-options
- and extend the range of non-options previously skipped. */
-
- while (optind < argc && NONOPTION_P)
- optind++;
- last_nonopt = optind;
- }
-
- /* The special ARGV-element `--' means premature end of options.
- Skip it like a null option,
- then exchange with previous non-options as if it were an option,
- then skip everything else like a non-option. */
-
- if (optind != argc && !strcmp (argv[optind], "--"))
- {
- optind++;
-
- if (first_nonopt != last_nonopt && last_nonopt != optind)
- exchange ((char **) argv);
- else if (first_nonopt == last_nonopt)
- first_nonopt = optind;
- last_nonopt = argc;
-
- optind = argc;
- }
-
- /* If we have done all the ARGV-elements, stop the scan
- and back over any non-options that we skipped and permuted. */
-
- if (optind == argc)
- {
- /* Set the next-arg-index to point at the non-options
- that we previously skipped, so the caller will digest them. */
- if (first_nonopt != last_nonopt)
- optind = first_nonopt;
- return -1;
- }
-
- /* If we have come to a non-option and did not permute it,
- either stop the scan or describe it to the caller and pass it by. */
-
- if (NONOPTION_P)
- {
- if (ordering == REQUIRE_ORDER)
- return -1;
- optarg = argv[optind++];
- return 1;
- }
-
- /* We have found another option-ARGV-element.
- Skip the initial punctuation. */
-
- nextchar = (argv[optind] + 1
- + (longopts != NULL && argv[optind][1] == '-'));
- }
-
- /* Decode the current option-ARGV-element. */
-
- /* Check whether the ARGV-element is a long option.
-
- If long_only and the ARGV-element has the form "-f", where f is
- a valid short option, don't consider it an abbreviated form of
- a long option that starts with f. Otherwise there would be no
- way to give the -f short option.
-
- On the other hand, if there's a long option "fubar" and
- the ARGV-element is "-fu", do consider that an abbreviation of
- the long option, just like "--fu", and not "-f" with arg "u".
-
- This distinction seems to be the most useful approach. */
-
- if (longopts != NULL
- && (argv[optind][1] == '-'
- || (long_only && (argv[optind][2] || !my_index (optstring, argv[optind][1])))))
- {
- char *nameend;
- const struct option *p;
- const struct option *pfound = NULL;
- int exact = 0;
- int ambig = 0;
- int indfound = -1;
- int option_index;
-
- for (nameend = nextchar; *nameend && *nameend != '='; nameend++)
- /* Do nothing. */ ;
-
- /* Test all long options for either exact match
- or abbreviated matches. */
- for (p = longopts, option_index = 0; p->name; p++, option_index++)
- if (!strncmp (p->name, nextchar, nameend - nextchar))
- {
- if ((unsigned int) (nameend - nextchar)
- == (unsigned int) strlen (p->name))
- {
- /* Exact match found. */
- pfound = p;
- indfound = option_index;
- exact = 1;
- break;
- }
- else if (pfound == NULL)
- {
- /* First nonexact match found. */
- pfound = p;
- indfound = option_index;
- }
- else if (long_only
- || pfound->has_arg != p->has_arg
- || pfound->flag != p->flag
- || pfound->val != p->val)
- /* Second or later nonexact match found. */
- ambig = 1;
- }
-
- if (ambig && !exact)
- {
- if (print_errors)
- {
-#if defined _LIBC && defined USE_IN_LIBIO
- char *buf;
-
- if (__asprintf (&buf, _("%s: option `%s' is ambiguous\n"),
- argv[0], argv[optind]) >= 0)
- {
-
- if (_IO_fwide (stderr, 0) > 0)
- __fwprintf (stderr, L"%s", buf);
- else
- fputs (buf, stderr);
-
- free (buf);
- }
-#else
- fprintf (stderr, _("%s: option `%s' is ambiguous\n"),
- argv[0], argv[optind]);
-#endif
- }
- nextchar += strlen (nextchar);
- optind++;
- optopt = 0;
- return '?';
- }
-
- if (pfound != NULL)
- {
- option_index = indfound;
- optind++;
- if (*nameend)
- {
- /* Don't test has_arg with >, because some C compilers don't
- allow it to be used on enums. */
- if (pfound->has_arg)
- optarg = nameend + 1;
- else
- {
- if (print_errors)
- {
-#if defined _LIBC && defined USE_IN_LIBIO
- char *buf;
- int n;
-#endif
-
- if (argv[optind - 1][1] == '-')
- {
- /* --option */
-#if defined _LIBC && defined USE_IN_LIBIO
- n = __asprintf (&buf, _("\
-%s: option `--%s' doesn't allow an argument\n"),
- argv[0], pfound->name);
-#else
- fprintf (stderr, _("\
-%s: option `--%s' doesn't allow an argument\n"),
- argv[0], pfound->name);
-#endif
- }
- else
- {
- /* +option or -option */
-#if defined _LIBC && defined USE_IN_LIBIO
- n = __asprintf (&buf, _("\
-%s: option `%c%s' doesn't allow an argument\n"),
- argv[0], argv[optind - 1][0],
- pfound->name);
-#else
- fprintf (stderr, _("\
-%s: option `%c%s' doesn't allow an argument\n"),
- argv[0], argv[optind - 1][0], pfound->name);
-#endif
- }
-
-#if defined _LIBC && defined USE_IN_LIBIO
- if (n >= 0)
- {
- if (_IO_fwide (stderr, 0) > 0)
- __fwprintf (stderr, L"%s", buf);
- else
- fputs (buf, stderr);
-
- free (buf);
- }
-#endif
- }
-
- nextchar += strlen (nextchar);
-
- optopt = pfound->val;
- return '?';
- }
- }
- else if (pfound->has_arg == 1)
- {
- if (optind < argc)
- optarg = argv[optind++];
- else
- {
- if (print_errors)
- {
-#if defined _LIBC && defined USE_IN_LIBIO
- char *buf;
-
- if (__asprintf (&buf, _("\
-%s: option `%s' requires an argument\n"),
- argv[0], argv[optind - 1]) >= 0)
- {
- if (_IO_fwide (stderr, 0) > 0)
- __fwprintf (stderr, L"%s", buf);
- else
- fputs (buf, stderr);
-
- free (buf);
- }
-#else
- fprintf (stderr,
- _("%s: option `%s' requires an argument\n"),
- argv[0], argv[optind - 1]);
-#endif
- }
- nextchar += strlen (nextchar);
- optopt = pfound->val;
- return optstring[0] == ':' ? ':' : '?';
- }
- }
- nextchar += strlen (nextchar);
- if (longind != NULL)
- *longind = option_index;
- if (pfound->flag)
- {
- *(pfound->flag) = pfound->val;
- return 0;
- }
- return pfound->val;
- }
-
- /* Can't find it as a long option. If this is not getopt_long_only,
- or the option starts with '--' or is not a valid short
- option, then it's an error.
- Otherwise interpret it as a short option. */
- if (!long_only || argv[optind][1] == '-'
- || my_index (optstring, *nextchar) == NULL)
- {
- if (print_errors)
- {
-#if defined _LIBC && defined USE_IN_LIBIO
- char *buf;
- int n;
-#endif
-
- if (argv[optind][1] == '-')
- {
- /* --option */
-#if defined _LIBC && defined USE_IN_LIBIO
- n = __asprintf (&buf, _("%s: unrecognized option `--%s'\n"),
- argv[0], nextchar);
-#else
- fprintf (stderr, _("%s: unrecognized option `--%s'\n"),
- argv[0], nextchar);
-#endif
- }
- else
- {
- /* +option or -option */
-#if defined _LIBC && defined USE_IN_LIBIO
- n = __asprintf (&buf, _("%s: unrecognized option `%c%s'\n"),
- argv[0], argv[optind][0], nextchar);
-#else
- fprintf (stderr, _("%s: unrecognized option `%c%s'\n"),
- argv[0], argv[optind][0], nextchar);
-#endif
- }
-
-#if defined _LIBC && defined USE_IN_LIBIO
- if (n >= 0)
- {
- if (_IO_fwide (stderr, 0) > 0)
- __fwprintf (stderr, L"%s", buf);
- else
- fputs (buf, stderr);
-
- free (buf);
- }
-#endif
- }
- nextchar = (char *) "";
- optind++;
- optopt = 0;
- return '?';
- }
- }
-
- /* Look at and handle the next short option-character. */
-
- {
- char c = *nextchar++;
- char *temp = my_index (optstring, c);
-
- /* Increment `optind' when we start to process its last character. */
- if (*nextchar == '\0')
- ++optind;
-
- if (temp == NULL || c == ':')
- {
- if (print_errors)
- {
-#if defined _LIBC && defined USE_IN_LIBIO
- char *buf;
- int n;
-#endif
-
- if (posixly_correct)
- {
- /* 1003.2 specifies the format of this message. */
-#if defined _LIBC && defined USE_IN_LIBIO
- n = __asprintf (&buf, _("%s: illegal option -- %c\n"),
- argv[0], c);
-#else
- fprintf (stderr, _("%s: illegal option -- %c\n"), argv[0], c);
-#endif
- }
- else
- {
-#if defined _LIBC && defined USE_IN_LIBIO
- n = __asprintf (&buf, _("%s: invalid option -- %c\n"),
- argv[0], c);
-#else
- fprintf (stderr, _("%s: invalid option -- %c\n"), argv[0], c);
-#endif
- }
-
-#if defined _LIBC && defined USE_IN_LIBIO
- if (n >= 0)
- {
- if (_IO_fwide (stderr, 0) > 0)
- __fwprintf (stderr, L"%s", buf);
- else
- fputs (buf, stderr);
-
- free (buf);
- }
-#endif
- }
- optopt = c;
- return '?';
- }
- /* Convenience. Treat POSIX -W foo same as long option --foo */
- if (temp[0] == 'W' && temp[1] == ';')
- {
- char *nameend;
- const struct option *p;
- const struct option *pfound = NULL;
- int exact = 0;
- int ambig = 0;
- int indfound = 0;
- int option_index;
-
- /* This is an option that requires an argument. */
- if (*nextchar != '\0')
- {
- optarg = nextchar;
- /* If we end this ARGV-element by taking the rest as an arg,
- we must advance to the next element now. */
- optind++;
- }
- else if (optind == argc)
- {
- if (print_errors)
- {
- /* 1003.2 specifies the format of this message. */
-#if defined _LIBC && defined USE_IN_LIBIO
- char *buf;
-
- if (__asprintf (&buf,
- _("%s: option requires an argument -- %c\n"),
- argv[0], c) >= 0)
- {
- if (_IO_fwide (stderr, 0) > 0)
- __fwprintf (stderr, L"%s", buf);
- else
- fputs (buf, stderr);
-
- free (buf);
- }
-#else
- fprintf (stderr, _("%s: option requires an argument -- %c\n"),
- argv[0], c);
-#endif
- }
- optopt = c;
- if (optstring[0] == ':')
- c = ':';
- else
- c = '?';
- return c;
- }
- else
- /* We already incremented `optind' once;
- increment it again when taking next ARGV-elt as argument. */
- optarg = argv[optind++];
-
- /* optarg is now the argument, see if it's in the
- table of longopts. */
-
- for (nextchar = nameend = optarg; *nameend && *nameend != '='; nameend++)
- /* Do nothing. */ ;
-
- /* Test all long options for either exact match
- or abbreviated matches. */
- for (p = longopts, option_index = 0; p->name; p++, option_index++)
- if (!strncmp (p->name, nextchar, nameend - nextchar))
- {
- if ((unsigned int) (nameend - nextchar) == strlen (p->name))
- {
- /* Exact match found. */
- pfound = p;
- indfound = option_index;
- exact = 1;
- break;
- }
- else if (pfound == NULL)
- {
- /* First nonexact match found. */
- pfound = p;
- indfound = option_index;
- }
- else
- /* Second or later nonexact match found. */
- ambig = 1;
- }
- if (ambig && !exact)
- {
- if (print_errors)
- {
-#if defined _LIBC && defined USE_IN_LIBIO
- char *buf;
-
- if (__asprintf (&buf, _("%s: option `-W %s' is ambiguous\n"),
- argv[0], argv[optind]) >= 0)
- {
- if (_IO_fwide (stderr, 0) > 0)
- __fwprintf (stderr, L"%s", buf);
- else
- fputs (buf, stderr);
-
- free (buf);
- }
-#else
- fprintf (stderr, _("%s: option `-W %s' is ambiguous\n"),
- argv[0], argv[optind]);
-#endif
- }
- nextchar += strlen (nextchar);
- optind++;
- return '?';
- }
- if (pfound != NULL)
- {
- option_index = indfound;
- if (*nameend)
- {
- /* Don't test has_arg with >, because some C compilers don't
- allow it to be used on enums. */
- if (pfound->has_arg)
- optarg = nameend + 1;
- else
- {
- if (print_errors)
- {
-#if defined _LIBC && defined USE_IN_LIBIO
- char *buf;
-
- if (__asprintf (&buf, _("\
-%s: option `-W %s' doesn't allow an argument\n"),
- argv[0], pfound->name) >= 0)
- {
- if (_IO_fwide (stderr, 0) > 0)
- __fwprintf (stderr, L"%s", buf);
- else
- fputs (buf, stderr);
-
- free (buf);
- }
-#else
- fprintf (stderr, _("\
-%s: option `-W %s' doesn't allow an argument\n"),
- argv[0], pfound->name);
-#endif
- }
-
- nextchar += strlen (nextchar);
- return '?';
- }
- }
- else if (pfound->has_arg == 1)
- {
- if (optind < argc)
- optarg = argv[optind++];
- else
- {
- if (print_errors)
- {
-#if defined _LIBC && defined USE_IN_LIBIO
- char *buf;
-
- if (__asprintf (&buf, _("\
-%s: option `%s' requires an argument\n"),
- argv[0], argv[optind - 1]) >= 0)
- {
- if (_IO_fwide (stderr, 0) > 0)
- __fwprintf (stderr, L"%s", buf);
- else
- fputs (buf, stderr);
-
- free (buf);
- }
-#else
- fprintf (stderr,
- _("%s: option `%s' requires an argument\n"),
- argv[0], argv[optind - 1]);
-#endif
- }
- nextchar += strlen (nextchar);
- return optstring[0] == ':' ? ':' : '?';
- }
- }
- nextchar += strlen (nextchar);
- if (longind != NULL)
- *longind = option_index;
- if (pfound->flag)
- {
- *(pfound->flag) = pfound->val;
- return 0;
- }
- return pfound->val;
- }
- nextchar = NULL;
- return 'W'; /* Let the application handle it. */
- }
- if (temp[1] == ':')
- {
- if (temp[2] == ':')
- {
- /* This is an option that accepts an argument optionally. */
- if (*nextchar != '\0')
- {
- optarg = nextchar;
- optind++;
- }
- else
- optarg = NULL;
- nextchar = NULL;
- }
- else
- {
- /* This is an option that requires an argument. */
- if (*nextchar != '\0')
- {
- optarg = nextchar;
- /* If we end this ARGV-element by taking the rest as an arg,
- we must advance to the next element now. */
- optind++;
- }
- else if (optind == argc)
- {
- if (print_errors)
- {
- /* 1003.2 specifies the format of this message. */
-#if defined _LIBC && defined USE_IN_LIBIO
- char *buf;
-
- if (__asprintf (&buf, _("\
-%s: option requires an argument -- %c\n"),
- argv[0], c) >= 0)
- {
- if (_IO_fwide (stderr, 0) > 0)
- __fwprintf (stderr, L"%s", buf);
- else
- fputs (buf, stderr);
-
- free (buf);
- }
-#else
- fprintf (stderr,
- _("%s: option requires an argument -- %c\n"),
- argv[0], c);
-#endif
- }
- optopt = c;
- if (optstring[0] == ':')
- c = ':';
- else
- c = '?';
- }
- else
- /* We already incremented `optind' once;
- increment it again when taking next ARGV-elt as argument. */
- optarg = argv[optind++];
- nextchar = NULL;
- }
- }
- return c;
- }
-}
-
-int
-getopt (argc, argv, optstring)
- int argc;
- char *const *argv;
- const char *optstring;
-{
- return _getopt_internal (argc, argv, optstring,
- (const struct option *) 0,
- (int *) 0,
- 0);
-}
-
-#endif /* Not ELIDE_CODE. */
-\f
-#ifdef TEST
-
-/* Compile with -DTEST to make an executable for use in testing
- the above definition of `getopt'. */
-
-int
-main (argc, argv)
- int argc;
- char **argv;
-{
- int c;
- int digit_optind = 0;
-
- while (1)
- {
- int this_option_optind = optind ? optind : 1;
-
- c = getopt (argc, argv, "abc:d:0123456789");
- if (c == -1)
- break;
-
- switch (c)
- {
- case '0':
- case '1':
- case '2':
- case '3':
- case '4':
- case '5':
- case '6':
- case '7':
- case '8':
- case '9':
- if (digit_optind != 0 && digit_optind != this_option_optind)
- printf ("digits occur in two different argv-elements.\n");
- digit_optind = this_option_optind;
- printf ("option %c\n", c);
- break;
-
- case 'a':
- printf ("option a\n");
- break;
-
- case 'b':
- printf ("option b\n");
- break;
-
- case 'c':
- printf ("option c with value `%s'\n", optarg);
- break;
-
- case '?':
- break;
-
- default:
- printf ("?? getopt returned character code 0%o ??\n", c);
- }
- }
-
- if (optind < argc)
- {
- printf ("non-option ARGV-elements: ");
- while (optind < argc)
- printf ("%s ", argv[optind++]);
- printf ("\n");
- }
-
- exit (0);
-}
-
-#endif /* TEST */
+++ /dev/null
-/* Declarations for getopt.
- Copyright (C) 1989-1994, 1996-1999, 2001 Free Software Foundation, Inc.
- This file is part of the GNU C Library.
-
- The GNU C Library is free software; you can redistribute it and/or
- modify it under the terms of the GNU Lesser General Public
- License as published by the Free Software Foundation; either
- version 2.1 of the License, or (at your option) any later version.
-
- The GNU C Library is distributed in the hope that it will be useful,
- but WITHOUT ANY WARRANTY; without even the implied warranty of
- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- Lesser General Public License for more details.
-
- You should have received a copy of the GNU Lesser General Public
- License along with the GNU C Library; if not, write to the Free
- Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
- 02111-1307 USA. */
-
-#ifndef _GETOPT_H
-
-#ifndef __need_getopt
-# define _GETOPT_H 1
-#endif
-
-/* If __GNU_LIBRARY__ is not already defined, either we are being used
- standalone, or this is the first header included in the source file.
- If we are being used with glibc, we need to include <features.h>, but
- that does not exist if we are standalone. So: if __GNU_LIBRARY__ is
- not defined, include <ctype.h>, which will pull in <features.h> for us
- if it's from glibc. (Why ctype.h? It's guaranteed to exist and it
- doesn't flood the namespace with stuff the way some other headers do.) */
-#if !defined __GNU_LIBRARY__
-# include <ctype.h>
-#endif
-
-#ifdef __cplusplus
-extern "C" {
-#endif
-
-/* For communication from `getopt' to the caller.
- When `getopt' finds an option that takes an argument,
- the argument value is returned here.
- Also, when `ordering' is RETURN_IN_ORDER,
- each non-option ARGV-element is returned here. */
-
-extern char *optarg;
-
-/* Index in ARGV of the next element to be scanned.
- This is used for communication to and from the caller
- and for communication between successive calls to `getopt'.
-
- On entry to `getopt', zero means this is the first call; initialize.
-
- When `getopt' returns -1, this is the index of the first of the
- non-option elements that the caller should itself scan.
-
- Otherwise, `optind' communicates from one call to the next
- how much of ARGV has been scanned so far. */
-
-extern int optind;
-
-/* Callers store zero here to inhibit the error message `getopt' prints
- for unrecognized options. */
-
-extern int opterr;
-
-/* Set to an option character which was unrecognized. */
-
-extern int optopt;
-
-#ifndef __need_getopt
-/* Describe the long-named options requested by the application.
- The LONG_OPTIONS argument to getopt_long or getopt_long_only is a vector
- of `struct option' terminated by an element containing a name which is
- zero.
-
- The field `has_arg' is:
- no_argument (or 0) if the option does not take an argument,
- required_argument (or 1) if the option requires an argument,
- optional_argument (or 2) if the option takes an optional argument.
-
- If the field `flag' is not NULL, it points to a variable that is set
- to the value given in the field `val' when the option is found, but
- left unchanged if the option is not found.
-
- To have a long-named option do something other than set an `int' to
- a compiled-in constant, such as set a value from `optarg', set the
- option's `flag' field to zero and its `val' field to a nonzero
- value (the equivalent single-letter option character, if there is
- one). For long options that have a zero `flag' field, `getopt'
- returns the contents of the `val' field. */
-
-struct option
-{
-# if (defined __STDC__ && __STDC__) || defined __cplusplus
- const char *name;
-# else
- char *name;
-# endif
- /* has_arg can't be an enum because some compilers complain about
- type mismatches in all the code that assumes it is an int. */
- int has_arg;
- int *flag;
- int val;
-};
-
-/* Names for the values of the `has_arg' field of `struct option'. */
-
-# define no_argument 0
-# define required_argument 1
-# define optional_argument 2
-#endif /* need getopt */
-
-
-/* Get definitions and prototypes for functions to process the
- arguments in ARGV (ARGC of them, minus the program name) for
- options given in OPTS.
-
- Return the option character from OPTS just read. Return -1 when
- there are no more options. For unrecognized options, or options
- missing arguments, `optopt' is set to the option letter, and '?' is
- returned.
-
- The OPTS string is a list of characters which are recognized option
- letters, optionally followed by colons, specifying that that letter
- takes an argument, to be placed in `optarg'.
-
- If a letter in OPTS is followed by two colons, its argument is
- optional. This behavior is specific to the GNU `getopt'.
-
- The argument `--' causes premature termination of argument
- scanning, explicitly telling `getopt' that there are no more
- options.
-
- If OPTS begins with `--', then non-option arguments are treated as
- arguments to the option '\0'. This behavior is specific to the GNU
- `getopt'. */
-
-#if (defined __STDC__ && __STDC__) || defined __cplusplus
-# ifdef __GNU_LIBRARY__
-/* Many other libraries have conflicting prototypes for getopt, with
- differences in the consts, in stdlib.h. To avoid compilation
- errors, only prototype getopt for the GNU C library. */
-extern int getopt (int ___argc, char *const *___argv, const char *__shortopts);
-# else /* not __GNU_LIBRARY__ */
-extern int getopt ();
-# endif /* __GNU_LIBRARY__ */
-
-# ifndef __need_getopt
-extern int getopt_long (int ___argc, char *const *___argv,
- const char *__shortopts,
- const struct option *__longopts, int *__longind);
-extern int getopt_long_only (int ___argc, char *const *___argv,
- const char *__shortopts,
- const struct option *__longopts, int *__longind);
-
-/* Internal only. Users should not call this directly. */
-extern int _getopt_internal (int ___argc, char *const *___argv,
- const char *__shortopts,
- const struct option *__longopts, int *__longind,
- int __long_only);
-# endif
-#else /* not __STDC__ */
-extern int getopt ();
-# ifndef __need_getopt
-extern int getopt_long ();
-extern int getopt_long_only ();
-
-extern int _getopt_internal ();
-# endif
-#endif /* __STDC__ */
-
-#ifdef __cplusplus
-}
-#endif
-
-/* Make sure we later can get all the definitions and declarations. */
-#undef __need_getopt
-
-#endif /* getopt.h */
+++ /dev/null
-/* getopt_long and getopt_long_only entry points for GNU getopt.
- Copyright (C) 1987,88,89,90,91,92,93,94,96,97,98
- Free Software Foundation, Inc.
- This file is part of the GNU C Library.
-
- The GNU C Library is free software; you can redistribute it and/or
- modify it under the terms of the GNU Lesser General Public
- License as published by the Free Software Foundation; either
- version 2.1 of the License, or (at your option) any later version.
-
- The GNU C Library is distributed in the hope that it will be useful,
- but WITHOUT ANY WARRANTY; without even the implied warranty of
- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- Lesser General Public License for more details.
-
- You should have received a copy of the GNU Lesser General Public
- License along with the GNU C Library; if not, write to the Free
- Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
- 02111-1307 USA. */
-\f
-#ifdef HAVE_CONFIG_H
-#include <config.h>
-#endif
-
-#ifdef _LIBC
-# include <getopt.h>
-#else
-# include "getopt.h"
-#endif
-
-#if !defined __STDC__ || !__STDC__
-/* This is a separate conditional since some stdc systems
- reject `defined (const)'. */
-#ifndef const
-#define const
-#endif
-#endif
-
-#include <stdio.h>
-
-/* Comment out all this code if we are using the GNU C Library, and are not
- actually compiling the library itself. This code is part of the GNU C
- Library, but also included in many other GNU distributions. Compiling
- and linking in this code is a waste when using the GNU C library
- (especially if it is a shared library). Rather than having every GNU
- program understand `configure --with-gnu-libc' and omit the object files,
- it is simpler to just do this in the source for each such file. */
-
-#define GETOPT_INTERFACE_VERSION 2
-#if !defined _LIBC && defined __GLIBC__ && __GLIBC__ >= 2
-#include <gnu-versions.h>
-#if _GNU_GETOPT_INTERFACE_VERSION == GETOPT_INTERFACE_VERSION
-#define ELIDE_CODE
-#endif
-#endif
-
-#ifndef ELIDE_CODE
-
-
-/* This needs to come after some library #include
- to get __GNU_LIBRARY__ defined. */
-#ifdef __GNU_LIBRARY__
-#include <stdlib.h>
-#endif
-
-#ifndef NULL
-#define NULL 0
-#endif
-
-int
-getopt_long (argc, argv, options, long_options, opt_index)
- int argc;
- char *const *argv;
- const char *options;
- const struct option *long_options;
- int *opt_index;
-{
- return _getopt_internal (argc, argv, options, long_options, opt_index, 0);
-}
-
-/* Like getopt_long, but '-' as well as '--' can indicate a long option.
- If an option that starts with '-' (not '--') doesn't match a long option,
- but does match a short option, it is parsed as a short option
- instead. */
-
-int
-getopt_long_only (argc, argv, options, long_options, opt_index)
- int argc;
- char *const *argv;
- const char *options;
- const struct option *long_options;
- int *opt_index;
-{
- return _getopt_internal (argc, argv, options, long_options, opt_index, 1);
-}
-
-# ifdef _LIBC
-libc_hidden_def (getopt_long)
-libc_hidden_def (getopt_long_only)
-# endif
-
-#endif /* Not ELIDE_CODE. */
-\f
-#ifdef TEST
-
-#include <stdio.h>
-
-int
-main (argc, argv)
- int argc;
- char **argv;
-{
- int c;
- int digit_optind = 0;
-
- while (1)
- {
- int this_option_optind = optind ? optind : 1;
- int option_index = 0;
- static struct option long_options[] =
- {
- {"add", 1, 0, 0},
- {"append", 0, 0, 0},
- {"delete", 1, 0, 0},
- {"verbose", 0, 0, 0},
- {"create", 0, 0, 0},
- {"file", 1, 0, 0},
- {0, 0, 0, 0}
- };
-
- c = getopt_long (argc, argv, "abc:d:0123456789",
- long_options, &option_index);
- if (c == -1)
- break;
-
- switch (c)
- {
- case 0:
- printf ("option %s", long_options[option_index].name);
- if (optarg)
- printf (" with arg %s", optarg);
- printf ("\n");
- break;
-
- case '0':
- case '1':
- case '2':
- case '3':
- case '4':
- case '5':
- case '6':
- case '7':
- case '8':
- case '9':
- if (digit_optind != 0 && digit_optind != this_option_optind)
- printf ("digits occur in two different argv-elements.\n");
- digit_optind = this_option_optind;
- printf ("option %c\n", c);
- break;
-
- case 'a':
- printf ("option a\n");
- break;
-
- case 'b':
- printf ("option b\n");
- break;
-
- case 'c':
- printf ("option c with value `%s'\n", optarg);
- break;
-
- case 'd':
- printf ("option d with value `%s'\n", optarg);
- break;
-
- case '?':
- break;
-
- default:
- printf ("?? getopt returned character code 0%o ??\n", c);
- }
- }
-
- if (optind < argc)
- {
- printf ("non-option ARGV-elements: ");
- while (optind < argc)
- printf ("%s ", argv[optind++]);
- printf ("\n");
- }
-
- exit (0);
-}
-
-#endif /* TEST */
+++ /dev/null
-/* Extended regular expression matching and search library.
- Copyright (C) 2002, 2003 Free Software Foundation, Inc.
- This file is part of the GNU C Library.
- Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
-
- The GNU C Library is free software; you can redistribute it and/or
- modify it under the terms of the GNU Lesser General Public
- License as published by the Free Software Foundation; either
- version 2.1 of the License, or (at your option) any later version.
-
- The GNU C Library is distributed in the hope that it will be useful,
- but WITHOUT ANY WARRANTY; without even the implied warranty of
- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- Lesser General Public License for more details.
-
- You should have received a copy of the GNU Lesser General Public
- License along with the GNU C Library; if not, write to the Free
- Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
- 02111-1307 USA. */
-
-static reg_errcode_t re_compile_internal (regex_t *preg, const char * pattern,
- int length, reg_syntax_t syntax);
-static void re_compile_fastmap_iter (regex_t *bufp,
- const re_dfastate_t *init_state,
- char *fastmap);
-static reg_errcode_t init_dfa (re_dfa_t *dfa, int pat_len);
-static reg_errcode_t init_word_char (re_dfa_t *dfa);
-#ifdef RE_ENABLE_I18N
-static void free_charset (re_charset_t *cset);
-#endif /* RE_ENABLE_I18N */
-static void free_workarea_compile (regex_t *preg);
-static reg_errcode_t create_initial_state (re_dfa_t *dfa);
-static reg_errcode_t analyze (re_dfa_t *dfa);
-static reg_errcode_t analyze_tree (re_dfa_t *dfa, bin_tree_t *node);
-static void calc_first (re_dfa_t *dfa, bin_tree_t *node);
-static void calc_next (re_dfa_t *dfa, bin_tree_t *node);
-static void calc_epsdest (re_dfa_t *dfa, bin_tree_t *node);
-static reg_errcode_t duplicate_node_closure (re_dfa_t *dfa, int top_org_node,
- int top_clone_node, int root_node,
- unsigned int constraint);
-static reg_errcode_t duplicate_node (int *new_idx, re_dfa_t *dfa, int org_idx,
- unsigned int constraint);
-static int search_duplicated_node (re_dfa_t *dfa, int org_node,
- unsigned int constraint);
-static reg_errcode_t calc_eclosure (re_dfa_t *dfa);
-static reg_errcode_t calc_eclosure_iter (re_node_set *new_set, re_dfa_t *dfa,
- int node, int root);
-static void calc_inveclosure (re_dfa_t *dfa);
-static int fetch_number (re_string_t *input, re_token_t *token,
- reg_syntax_t syntax);
-static re_token_t fetch_token (re_string_t *input, reg_syntax_t syntax);
-static int peek_token (re_token_t *token, re_string_t *input,
- reg_syntax_t syntax);
-static int peek_token_bracket (re_token_t *token, re_string_t *input,
- reg_syntax_t syntax);
-static bin_tree_t *parse (re_string_t *regexp, regex_t *preg,
- reg_syntax_t syntax, reg_errcode_t *err);
-static bin_tree_t *parse_reg_exp (re_string_t *regexp, regex_t *preg,
- re_token_t *token, reg_syntax_t syntax,
- int nest, reg_errcode_t *err);
-static bin_tree_t *parse_branch (re_string_t *regexp, regex_t *preg,
- re_token_t *token, reg_syntax_t syntax,
- int nest, reg_errcode_t *err);
-static bin_tree_t *parse_expression (re_string_t *regexp, regex_t *preg,
- re_token_t *token, reg_syntax_t syntax,
- int nest, reg_errcode_t *err);
-static bin_tree_t *parse_sub_exp (re_string_t *regexp, regex_t *preg,
- re_token_t *token, reg_syntax_t syntax,
- int nest, reg_errcode_t *err);
-static bin_tree_t *parse_dup_op (bin_tree_t *dup_elem, re_string_t *regexp,
- re_dfa_t *dfa, re_token_t *token,
- reg_syntax_t syntax, reg_errcode_t *err);
-static bin_tree_t *parse_bracket_exp (re_string_t *regexp, re_dfa_t *dfa,
- re_token_t *token, reg_syntax_t syntax,
- reg_errcode_t *err);
-static reg_errcode_t parse_bracket_element (bracket_elem_t *elem,
- re_string_t *regexp,
- re_token_t *token, int token_len,
- re_dfa_t *dfa,
- reg_syntax_t syntax);
-static reg_errcode_t parse_bracket_symbol (bracket_elem_t *elem,
- re_string_t *regexp,
- re_token_t *token);
-#ifndef _LIBC
-# ifdef RE_ENABLE_I18N
-static reg_errcode_t build_range_exp (re_bitset_ptr_t sbcset,
- re_charset_t *mbcset, int *range_alloc,
- bracket_elem_t *start_elem,
- bracket_elem_t *end_elem);
-static reg_errcode_t build_collating_symbol (re_bitset_ptr_t sbcset,
- re_charset_t *mbcset,
- int *coll_sym_alloc,
- const unsigned char *name);
-# else /* not RE_ENABLE_I18N */
-static reg_errcode_t build_range_exp (re_bitset_ptr_t sbcset,
- bracket_elem_t *start_elem,
- bracket_elem_t *end_elem);
-static reg_errcode_t build_collating_symbol (re_bitset_ptr_t sbcset,
- const unsigned char *name);
-# endif /* not RE_ENABLE_I18N */
-#endif /* not _LIBC */
-#ifdef RE_ENABLE_I18N
-static reg_errcode_t build_equiv_class (re_bitset_ptr_t sbcset,
- re_charset_t *mbcset,
- int *equiv_class_alloc,
- const unsigned char *name);
-static reg_errcode_t build_charclass (re_bitset_ptr_t sbcset,
- re_charset_t *mbcset,
- int *char_class_alloc,
- const unsigned char *class_name,
- reg_syntax_t syntax);
-#else /* not RE_ENABLE_I18N */
-static reg_errcode_t build_equiv_class (re_bitset_ptr_t sbcset,
- const unsigned char *name);
-static reg_errcode_t build_charclass (re_bitset_ptr_t sbcset,
- const unsigned char *class_name,
- reg_syntax_t syntax);
-#endif /* not RE_ENABLE_I18N */
-static bin_tree_t *build_word_op (re_dfa_t *dfa, int not, reg_errcode_t *err);
-static void free_bin_tree (bin_tree_t *tree);
-static bin_tree_t *create_tree (bin_tree_t *left, bin_tree_t *right,
- re_token_type_t type, int index);
-static bin_tree_t *duplicate_tree (const bin_tree_t *src, re_dfa_t *dfa);
-\f
-/* This table gives an error message for each of the error codes listed
- in regex.h. Obviously the order here has to be same as there.
- POSIX doesn't require that we do anything for REG_NOERROR,
- but why not be nice? */
-
-const char __re_error_msgid[] attribute_hidden =
- {
-#define REG_NOERROR_IDX 0
- gettext_noop ("Success") /* REG_NOERROR */
- "\0"
-#define REG_NOMATCH_IDX (REG_NOERROR_IDX + sizeof "Success")
- gettext_noop ("No match") /* REG_NOMATCH */
- "\0"
-#define REG_BADPAT_IDX (REG_NOMATCH_IDX + sizeof "No match")
- gettext_noop ("Invalid regular expression") /* REG_BADPAT */
- "\0"
-#define REG_ECOLLATE_IDX (REG_BADPAT_IDX + sizeof "Invalid regular expression")
- gettext_noop ("Invalid collation character") /* REG_ECOLLATE */
- "\0"
-#define REG_ECTYPE_IDX (REG_ECOLLATE_IDX + sizeof "Invalid collation character")
- gettext_noop ("Invalid character class name") /* REG_ECTYPE */
- "\0"
-#define REG_EESCAPE_IDX (REG_ECTYPE_IDX + sizeof "Invalid character class name")
- gettext_noop ("Trailing backslash") /* REG_EESCAPE */
- "\0"
-#define REG_ESUBREG_IDX (REG_EESCAPE_IDX + sizeof "Trailing backslash")
- gettext_noop ("Invalid back reference") /* REG_ESUBREG */
- "\0"
-#define REG_EBRACK_IDX (REG_ESUBREG_IDX + sizeof "Invalid back reference")
- gettext_noop ("Unmatched [ or [^") /* REG_EBRACK */
- "\0"
-#define REG_EPAREN_IDX (REG_EBRACK_IDX + sizeof "Unmatched [ or [^")
- gettext_noop ("Unmatched ( or \\(") /* REG_EPAREN */
- "\0"
-#define REG_EBRACE_IDX (REG_EPAREN_IDX + sizeof "Unmatched ( or \\(")
- gettext_noop ("Unmatched \\{") /* REG_EBRACE */
- "\0"
-#define REG_BADBR_IDX (REG_EBRACE_IDX + sizeof "Unmatched \\{")
- gettext_noop ("Invalid content of \\{\\}") /* REG_BADBR */
- "\0"
-#define REG_ERANGE_IDX (REG_BADBR_IDX + sizeof "Invalid content of \\{\\}")
- gettext_noop ("Invalid range end") /* REG_ERANGE */
- "\0"
-#define REG_ESPACE_IDX (REG_ERANGE_IDX + sizeof "Invalid range end")
- gettext_noop ("Memory exhausted") /* REG_ESPACE */
- "\0"
-#define REG_BADRPT_IDX (REG_ESPACE_IDX + sizeof "Memory exhausted")
- gettext_noop ("Invalid preceding regular expression") /* REG_BADRPT */
- "\0"
-#define REG_EEND_IDX (REG_BADRPT_IDX + sizeof "Invalid preceding regular expression")
- gettext_noop ("Premature end of regular expression") /* REG_EEND */
- "\0"
-#define REG_ESIZE_IDX (REG_EEND_IDX + sizeof "Premature end of regular expression")
- gettext_noop ("Regular expression too big") /* REG_ESIZE */
- "\0"
-#define REG_ERPAREN_IDX (REG_ESIZE_IDX + sizeof "Regular expression too big")
- gettext_noop ("Unmatched ) or \\)") /* REG_ERPAREN */
- };
-
-const size_t __re_error_msgid_idx[] attribute_hidden =
- {
- REG_NOERROR_IDX,
- REG_NOMATCH_IDX,
- REG_BADPAT_IDX,
- REG_ECOLLATE_IDX,
- REG_ECTYPE_IDX,
- REG_EESCAPE_IDX,
- REG_ESUBREG_IDX,
- REG_EBRACK_IDX,
- REG_EPAREN_IDX,
- REG_EBRACE_IDX,
- REG_BADBR_IDX,
- REG_ERANGE_IDX,
- REG_ESPACE_IDX,
- REG_BADRPT_IDX,
- REG_EEND_IDX,
- REG_ESIZE_IDX,
- REG_ERPAREN_IDX
- };
-\f
-/* Entry points for GNU code. */
-
-/* re_compile_pattern is the GNU regular expression compiler: it
- compiles PATTERN (of length LENGTH) and puts the result in BUFP.
- Returns 0 if the pattern was valid, otherwise an error string.
-
- Assumes the `allocated' (and perhaps `buffer') and `translate' fields
- are set in BUFP on entry. */
-
-const char *
-re_compile_pattern (pattern, length, bufp)
- const char *pattern;
- size_t length;
- struct re_pattern_buffer *bufp;
-{
- reg_errcode_t ret;
-
- /* And GNU code determines whether or not to get register information
- by passing null for the REGS argument to re_match, etc., not by
- setting no_sub. */
- bufp->no_sub = 0;
-
- /* Match anchors at newline. */
- bufp->newline_anchor = 1;
-
- ret = re_compile_internal (bufp, pattern, length, re_syntax_options);
-
- if (!ret)
- return NULL;
- return gettext (__re_error_msgid + __re_error_msgid_idx[(int) ret]);
-}
-#ifdef _LIBC
-weak_alias (__re_compile_pattern, re_compile_pattern)
-#endif
-
-/* Set by `re_set_syntax' to the current regexp syntax to recognize. Can
- also be assigned to arbitrarily: each pattern buffer stores its own
- syntax, so it can be changed between regex compilations. */
-/* This has no initializer because initialized variables in Emacs
- become read-only after dumping. */
-reg_syntax_t re_syntax_options;
-
-
-/* Specify the precise syntax of regexps for compilation. This provides
- for compatibility for various utilities which historically have
- different, incompatible syntaxes.
-
- The argument SYNTAX is a bit mask comprised of the various bits
- defined in regex.h. We return the old syntax. */
-
-reg_syntax_t
-re_set_syntax (syntax)
- reg_syntax_t syntax;
-{
- reg_syntax_t ret = re_syntax_options;
-
- re_syntax_options = syntax;
- return ret;
-}
-#ifdef _LIBC
-weak_alias (__re_set_syntax, re_set_syntax)
-#endif
-
-int
-re_compile_fastmap (bufp)
- struct re_pattern_buffer *bufp;
-{
- re_dfa_t *dfa = (re_dfa_t *) bufp->buffer;
- char *fastmap = bufp->fastmap;
-
- memset (fastmap, '\0', sizeof (char) * SBC_MAX);
- re_compile_fastmap_iter (bufp, dfa->init_state, fastmap);
- if (dfa->init_state != dfa->init_state_word)
- re_compile_fastmap_iter (bufp, dfa->init_state_word, fastmap);
- if (dfa->init_state != dfa->init_state_nl)
- re_compile_fastmap_iter (bufp, dfa->init_state_nl, fastmap);
- if (dfa->init_state != dfa->init_state_begbuf)
- re_compile_fastmap_iter (bufp, dfa->init_state_begbuf, fastmap);
- bufp->fastmap_accurate = 1;
- return 0;
-}
-#ifdef _LIBC
-weak_alias (__re_compile_fastmap, re_compile_fastmap)
-#endif
-
-static inline void
-re_set_fastmap (char *fastmap, int icase, int ch)
-{
- fastmap[ch] = 1;
- if (icase)
- fastmap[tolower (ch)] = 1;
-}
-
-/* Helper function for re_compile_fastmap.
- Compile fastmap for the initial_state INIT_STATE. */
-
-static void
-re_compile_fastmap_iter (bufp, init_state, fastmap)
- regex_t *bufp;
- const re_dfastate_t *init_state;
- char *fastmap;
-{
- re_dfa_t *dfa = (re_dfa_t *) bufp->buffer;
- int node_cnt;
- int icase = (MB_CUR_MAX == 1 && (bufp->syntax & RE_ICASE));
- for (node_cnt = 0; node_cnt < init_state->nodes.nelem; ++node_cnt)
- {
- int node = init_state->nodes.elems[node_cnt];
- re_token_type_t type = dfa->nodes[node].type;
-
- if (type == CHARACTER)
- re_set_fastmap (fastmap, icase, dfa->nodes[node].opr.c);
- else if (type == SIMPLE_BRACKET)
- {
- int i, j, ch;
- for (i = 0, ch = 0; i < BITSET_UINTS; ++i)
- for (j = 0; j < UINT_BITS; ++j, ++ch)
- if (dfa->nodes[node].opr.sbcset[i] & (1 << j))
- re_set_fastmap (fastmap, icase, ch);
- }
-#ifdef RE_ENABLE_I18N
- else if (type == COMPLEX_BRACKET)
- {
- int i;
- re_charset_t *cset = dfa->nodes[node].opr.mbcset;
- if (cset->non_match || cset->ncoll_syms || cset->nequiv_classes
- || cset->nranges || cset->nchar_classes)
- {
-# ifdef _LIBC
- if (_NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES) != 0)
- {
- /* In this case we want to catch the bytes which are
- the first byte of any collation elements.
- e.g. In da_DK, we want to catch 'a' since "aa"
- is a valid collation element, and don't catch
- 'b' since 'b' is the only collation element
- which starts from 'b'. */
- int j, ch;
- const int32_t *table = (const int32_t *)
- _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB);
- for (i = 0, ch = 0; i < BITSET_UINTS; ++i)
- for (j = 0; j < UINT_BITS; ++j, ++ch)
- if (table[ch] < 0)
- re_set_fastmap (fastmap, icase, ch);
- }
-# else
- if (MB_CUR_MAX > 1)
- for (i = 0; i < SBC_MAX; ++i)
- if (__btowc (i) == WEOF)
- re_set_fastmap (fastmap, icase, i);
-# endif /* not _LIBC */
- }
- for (i = 0; i < cset->nmbchars; ++i)
- {
- char buf[256];
- mbstate_t state;
- memset (&state, '\0', sizeof (state));
- __wcrtomb (buf, cset->mbchars[i], &state);
- re_set_fastmap (fastmap, icase, *(unsigned char *) buf);
- }
- }
-#endif /* RE_ENABLE_I18N */
- else if (type == END_OF_RE || type == OP_PERIOD)
- {
- memset (fastmap, '\1', sizeof (char) * SBC_MAX);
- if (type == END_OF_RE)
- bufp->can_be_null = 1;
- return;
- }
- }
-}
-\f
-/* Entry point for POSIX code. */
-/* regcomp takes a regular expression as a string and compiles it.
-
- PREG is a regex_t *. We do not expect any fields to be initialized,
- since POSIX says we shouldn't. Thus, we set
-
- `buffer' to the compiled pattern;
- `used' to the length of the compiled pattern;
- `syntax' to RE_SYNTAX_POSIX_EXTENDED if the
- REG_EXTENDED bit in CFLAGS is set; otherwise, to
- RE_SYNTAX_POSIX_BASIC;
- `newline_anchor' to REG_NEWLINE being set in CFLAGS;
- `fastmap' to an allocated space for the fastmap;
- `fastmap_accurate' to zero;
- `re_nsub' to the number of subexpressions in PATTERN.
-
- PATTERN is the address of the pattern string.
-
- CFLAGS is a series of bits which affect compilation.
-
- If REG_EXTENDED is set, we use POSIX extended syntax; otherwise, we
- use POSIX basic syntax.
-
- If REG_NEWLINE is set, then . and [^...] don't match newline.
- Also, regexec will try a match beginning after every newline.
-
- If REG_ICASE is set, then we considers upper- and lowercase
- versions of letters to be equivalent when matching.
-
- If REG_NOSUB is set, then when PREG is passed to regexec, that
- routine will report only success or failure, and nothing about the
- registers.
-
- It returns 0 if it succeeds, nonzero if it doesn't. (See regex.h for
- the return codes and their meanings.) */
-
-int
-regcomp (preg, pattern, cflags)
- regex_t *__restrict preg;
- const char *__restrict pattern;
- int cflags;
-{
- reg_errcode_t ret;
- reg_syntax_t syntax = ((cflags & REG_EXTENDED) ? RE_SYNTAX_POSIX_EXTENDED
- : RE_SYNTAX_POSIX_BASIC);
-
- preg->buffer = NULL;
- preg->allocated = 0;
- preg->used = 0;
-
- /* Try to allocate space for the fastmap. */
- preg->fastmap = re_malloc (char, SBC_MAX);
- if (BE (preg->fastmap == NULL, 0))
- return REG_ESPACE;
-
- syntax |= (cflags & REG_ICASE) ? RE_ICASE : 0;
-
- /* If REG_NEWLINE is set, newlines are treated differently. */
- if (cflags & REG_NEWLINE)
- { /* REG_NEWLINE implies neither . nor [^...] match newline. */
- syntax &= ~RE_DOT_NEWLINE;
- syntax |= RE_HAT_LISTS_NOT_NEWLINE;
- /* It also changes the matching behavior. */
- preg->newline_anchor = 1;
- }
- else
- preg->newline_anchor = 0;
- preg->no_sub = !!(cflags & REG_NOSUB);
- preg->translate = NULL;
-
- ret = re_compile_internal (preg, pattern, strlen (pattern), syntax);
-
- /* POSIX doesn't distinguish between an unmatched open-group and an
- unmatched close-group: both are REG_EPAREN. */
- if (ret == REG_ERPAREN)
- ret = REG_EPAREN;
-
- /* We have already checked preg->fastmap != NULL. */
- if (BE (ret == REG_NOERROR, 1))
- /* Compute the fastmap now, since regexec cannot modify the pattern
- buffer. This function nevers fails in this implementation. */
- (void) re_compile_fastmap (preg);
- else
- {
- /* Some error occurred while compiling the expression. */
- re_free (preg->fastmap);
- preg->fastmap = NULL;
- }
-
- return (int) ret;
-}
-#ifdef _LIBC
-weak_alias (__regcomp, regcomp)
-#endif
-
-/* Returns a message corresponding to an error code, ERRCODE, returned
- from either regcomp or regexec. We don't use PREG here. */
-
-size_t
-regerror (errcode, preg, errbuf, errbuf_size)
- int errcode;
- const regex_t *preg;
- char *errbuf;
- size_t errbuf_size;
-{
- const char *msg;
- size_t msg_size;
-
- if (BE (errcode < 0
- || errcode >= (int) (sizeof (__re_error_msgid_idx)
- / sizeof (__re_error_msgid_idx[0])), 0))
- /* Only error codes returned by the rest of the code should be passed
- to this routine. If we are given anything else, or if other regex
- code generates an invalid error code, then the program has a bug.
- Dump core so we can fix it. */
- abort ();
-
- msg = gettext (__re_error_msgid + __re_error_msgid_idx[errcode]);
-
- msg_size = strlen (msg) + 1; /* Includes the null. */
-
- if (BE (errbuf_size != 0, 1))
- {
- if (BE (msg_size > errbuf_size, 0))
- {
-#if defined HAVE_MEMPCPY || defined _LIBC
- *((char *) __mempcpy (errbuf, msg, errbuf_size - 1)) = '\0';
-#else
- memcpy (errbuf, msg, errbuf_size - 1);
- errbuf[errbuf_size - 1] = 0;
-#endif
- }
- else
- memcpy (errbuf, msg, msg_size);
- }
-
- return msg_size;
-}
-#ifdef _LIBC
-weak_alias (__regerror, regerror)
-#endif
-
-
-static void
-free_dfa_content (re_dfa_t *dfa)
-{
- int i, j;
-
- re_free (dfa->subexps);
-
- for (i = 0; i < dfa->nodes_len; ++i)
- {
- re_token_t *node = dfa->nodes + i;
-#ifdef RE_ENABLE_I18N
- if (node->type == COMPLEX_BRACKET && node->duplicated == 0)
- free_charset (node->opr.mbcset);
- else
-#endif /* RE_ENABLE_I18N */
- if (node->type == SIMPLE_BRACKET && node->duplicated == 0)
- re_free (node->opr.sbcset);
- }
- re_free (dfa->nexts);
- for (i = 0; i < dfa->nodes_len; ++i)
- {
- if (dfa->eclosures != NULL)
- re_node_set_free (dfa->eclosures + i);
- if (dfa->inveclosures != NULL)
- re_node_set_free (dfa->inveclosures + i);
- if (dfa->edests != NULL)
- re_node_set_free (dfa->edests + i);
- }
- re_free (dfa->edests);
- re_free (dfa->eclosures);
- re_free (dfa->inveclosures);
- re_free (dfa->nodes);
-
- for (i = 0; i <= dfa->state_hash_mask; ++i)
- {
- struct re_state_table_entry *entry = dfa->state_table + i;
- for (j = 0; j < entry->num; ++j)
- {
- re_dfastate_t *state = entry->array[j];
- free_state (state);
- }
- re_free (entry->array);
- }
- re_free (dfa->state_table);
-
- if (dfa->word_char != NULL)
- re_free (dfa->word_char);
-#ifdef DEBUG
- re_free (dfa->re_str);
-#endif
-
- re_free (dfa);
-}
-
-
-/* Free dynamically allocated space used by PREG. */
-
-void
-regfree (preg)
- regex_t *preg;
-{
- re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- if (BE (dfa != NULL, 1))
- free_dfa_content (dfa);
-
- re_free (preg->fastmap);
-}
-#ifdef _LIBC
-weak_alias (__regfree, regfree)
-#endif
-\f
-/* Entry points compatible with 4.2 BSD regex library. We don't define
- them unless specifically requested. */
-
-#if defined _REGEX_RE_COMP || defined _LIBC
-
-/* BSD has one and only one pattern buffer. */
-static struct re_pattern_buffer re_comp_buf;
-
-char *
-# ifdef _LIBC
-/* Make these definitions weak in libc, so POSIX programs can redefine
- these names if they don't use our functions, and still use
- regcomp/regexec above without link errors. */
-weak_function
-# endif
-re_comp (s)
- const char *s;
-{
- reg_errcode_t ret;
- char *fastmap;
-
- if (!s)
- {
- if (!re_comp_buf.buffer)
- return gettext ("No previous regular expression");
- return 0;
- }
-
- if (re_comp_buf.buffer)
- {
- fastmap = re_comp_buf.fastmap;
- re_comp_buf.fastmap = NULL;
- __regfree (&re_comp_buf);
- memset (&re_comp_buf, '\0', sizeof (re_comp_buf));
- re_comp_buf.fastmap = fastmap;
- }
-
- if (re_comp_buf.fastmap == NULL)
- {
- re_comp_buf.fastmap = (char *) malloc (SBC_MAX);
- if (re_comp_buf.fastmap == NULL)
- return (char *) gettext (__re_error_msgid
- + __re_error_msgid_idx[(int) REG_ESPACE]);
- }
-
- /* Since `re_exec' always passes NULL for the `regs' argument, we
- don't need to initialize the pattern buffer fields which affect it. */
-
- /* Match anchors at newlines. */
- re_comp_buf.newline_anchor = 1;
-
- ret = re_compile_internal (&re_comp_buf, s, strlen (s), re_syntax_options);
-
- if (!ret)
- return NULL;
-
- /* Yes, we're discarding `const' here if !HAVE_LIBINTL. */
- return (char *) gettext (__re_error_msgid + __re_error_msgid_idx[(int) ret]);
-}
-
-#ifdef _LIBC
-libc_freeres_fn (free_mem)
-{
- __regfree (&re_comp_buf);
-}
-#endif
-
-#endif /* _REGEX_RE_COMP */
-\f
-/* Internal entry point.
- Compile the regular expression PATTERN, whose length is LENGTH.
- SYNTAX indicate regular expression's syntax. */
-
-static reg_errcode_t
-re_compile_internal (preg, pattern, length, syntax)
- regex_t *preg;
- const char * pattern;
- int length;
- reg_syntax_t syntax;
-{
- reg_errcode_t err = REG_NOERROR;
- re_dfa_t *dfa;
- re_string_t regexp;
-
- /* Initialize the pattern buffer. */
- preg->fastmap_accurate = 0;
- preg->syntax = syntax;
- preg->not_bol = preg->not_eol = 0;
- preg->used = 0;
- preg->re_nsub = 0;
- preg->can_be_null = 0;
- preg->regs_allocated = REGS_UNALLOCATED;
-
- /* Initialize the dfa. */
- dfa = (re_dfa_t *) preg->buffer;
- if (preg->allocated < sizeof (re_dfa_t))
- {
- /* If zero allocated, but buffer is non-null, try to realloc
- enough space. This loses if buffer's address is bogus, but
- that is the user's responsibility. If ->buffer is NULL this
- is a simple allocation. */
- dfa = re_realloc (preg->buffer, re_dfa_t, 1);
- if (dfa == NULL)
- return REG_ESPACE;
- preg->allocated = sizeof (re_dfa_t);
- }
- preg->buffer = (unsigned char *) dfa;
- preg->used = sizeof (re_dfa_t);
-
- err = init_dfa (dfa, length);
- if (BE (err != REG_NOERROR, 0))
- {
- re_free (dfa);
- preg->buffer = NULL;
- preg->allocated = 0;
- return err;
- }
-#ifdef DEBUG
- dfa->re_str = re_malloc (char, length + 1);
- strncpy (dfa->re_str, pattern, length + 1);
-#endif
-
- err = re_string_construct (®exp, pattern, length, preg->translate,
- syntax & RE_ICASE);
- if (BE (err != REG_NOERROR, 0))
- {
- re_free (dfa);
- preg->buffer = NULL;
- preg->allocated = 0;
- return err;
- }
-
- /* Parse the regular expression, and build a structure tree. */
- preg->re_nsub = 0;
- dfa->str_tree = parse (®exp, preg, syntax, &err);
- if (BE (dfa->str_tree == NULL, 0))
- goto re_compile_internal_free_return;
-
- /* Analyze the tree and collect information which is necessary to
- create the dfa. */
- err = analyze (dfa);
- if (BE (err != REG_NOERROR, 0))
- goto re_compile_internal_free_return;
-
- /* Then create the initial state of the dfa. */
- err = create_initial_state (dfa);
-
- /* Release work areas. */
- free_workarea_compile (preg);
- re_string_destruct (®exp);
-
- if (BE (err != REG_NOERROR, 0))
- {
- re_compile_internal_free_return:
- free_dfa_content (dfa);
- preg->buffer = NULL;
- preg->allocated = 0;
- }
-
- return err;
-}
-
-/* Initialize DFA. We use the length of the regular expression PAT_LEN
- as the initial length of some arrays. */
-
-static reg_errcode_t
-init_dfa (dfa, pat_len)
- re_dfa_t *dfa;
- int pat_len;
-{
- int table_size;
-
- memset (dfa, '\0', sizeof (re_dfa_t));
-
- dfa->nodes_alloc = pat_len + 1;
- dfa->nodes = re_malloc (re_token_t, dfa->nodes_alloc);
-
- dfa->states_alloc = pat_len + 1;
-
- /* table_size = 2 ^ ceil(log pat_len) */
- for (table_size = 1; table_size > 0; table_size <<= 1)
- if (table_size > pat_len)
- break;
-
- dfa->state_table = calloc (sizeof (struct re_state_table_entry), table_size);
- dfa->state_hash_mask = table_size - 1;
-
- dfa->subexps_alloc = 1;
- dfa->subexps = re_malloc (re_subexp_t, dfa->subexps_alloc);
- dfa->word_char = NULL;
-
- if (BE (dfa->nodes == NULL || dfa->state_table == NULL
- || dfa->subexps == NULL, 0))
- {
- /* We don't bother to free anything which was allocated. Very
- soon the process will go down anyway. */
- dfa->subexps = NULL;
- dfa->state_table = NULL;
- dfa->nodes = NULL;
- return REG_ESPACE;
- }
- return REG_NOERROR;
-}
-
-/* Initialize WORD_CHAR table, which indicate which character is
- "word". In this case "word" means that it is the word construction
- character used by some operators like "\<", "\>", etc. */
-
-static reg_errcode_t
-init_word_char (dfa)
- re_dfa_t *dfa;
-{
- int i, j, ch;
- dfa->word_char = (re_bitset_ptr_t) calloc (sizeof (bitset), 1);
- if (BE (dfa->word_char == NULL, 0))
- return REG_ESPACE;
- for (i = 0, ch = 0; i < BITSET_UINTS; ++i)
- for (j = 0; j < UINT_BITS; ++j, ++ch)
- if (isalnum (ch) || ch == '_')
- dfa->word_char[i] |= 1 << j;
- return REG_NOERROR;
-}
-
-/* Free the work area which are only used while compiling. */
-
-static void
-free_workarea_compile (preg)
- regex_t *preg;
-{
- re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- free_bin_tree (dfa->str_tree);
- dfa->str_tree = NULL;
- re_free (dfa->org_indices);
- dfa->org_indices = NULL;
-}
-
-/* Create initial states for all contexts. */
-
-static reg_errcode_t
-create_initial_state (dfa)
- re_dfa_t *dfa;
-{
- int first, i;
- reg_errcode_t err;
- re_node_set init_nodes;
-
- /* Initial states have the epsilon closure of the node which is
- the first node of the regular expression. */
- first = dfa->str_tree->first;
- dfa->init_node = first;
- err = re_node_set_init_copy (&init_nodes, dfa->eclosures + first);
- if (BE (err != REG_NOERROR, 0))
- return err;
-
- /* The back-references which are in initial states can epsilon transit,
- since in this case all of the subexpressions can be null.
- Then we add epsilon closures of the nodes which are the next nodes of
- the back-references. */
- if (dfa->nbackref > 0)
- for (i = 0; i < init_nodes.nelem; ++i)
- {
- int node_idx = init_nodes.elems[i];
- re_token_type_t type = dfa->nodes[node_idx].type;
-
- int clexp_idx;
- if (type != OP_BACK_REF)
- continue;
- for (clexp_idx = 0; clexp_idx < init_nodes.nelem; ++clexp_idx)
- {
- re_token_t *clexp_node;
- clexp_node = dfa->nodes + init_nodes.elems[clexp_idx];
- if (clexp_node->type == OP_CLOSE_SUBEXP
- && clexp_node->opr.idx + 1 == dfa->nodes[node_idx].opr.idx)
- break;
- }
- if (clexp_idx == init_nodes.nelem)
- continue;
-
- if (type == OP_BACK_REF)
- {
- int dest_idx = dfa->edests[node_idx].elems[0];
- if (!re_node_set_contains (&init_nodes, dest_idx))
- {
- re_node_set_merge (&init_nodes, dfa->eclosures + dest_idx);
- i = 0;
- }
- }
- }
-
- /* It must be the first time to invoke acquire_state. */
- dfa->init_state = re_acquire_state_context (&err, dfa, &init_nodes, 0);
- /* We don't check ERR here, since the initial state must not be NULL. */
- if (BE (dfa->init_state == NULL, 0))
- return err;
- if (dfa->init_state->has_constraint)
- {
- dfa->init_state_word = re_acquire_state_context (&err, dfa, &init_nodes,
- CONTEXT_WORD);
- dfa->init_state_nl = re_acquire_state_context (&err, dfa, &init_nodes,
- CONTEXT_NEWLINE);
- dfa->init_state_begbuf = re_acquire_state_context (&err, dfa,
- &init_nodes,
- CONTEXT_NEWLINE
- | CONTEXT_BEGBUF);
- if (BE (dfa->init_state_word == NULL || dfa->init_state_nl == NULL
- || dfa->init_state_begbuf == NULL, 0))
- return err;
- }
- else
- dfa->init_state_word = dfa->init_state_nl
- = dfa->init_state_begbuf = dfa->init_state;
-
- re_node_set_free (&init_nodes);
- return REG_NOERROR;
-}
-\f
-/* Analyze the structure tree, and calculate "first", "next", "edest",
- "eclosure", and "inveclosure". */
-
-static reg_errcode_t
-analyze (dfa)
- re_dfa_t *dfa;
-{
- int i;
- reg_errcode_t ret;
-
- /* Allocate arrays. */
- dfa->nexts = re_malloc (int, dfa->nodes_alloc);
- dfa->org_indices = re_malloc (int, dfa->nodes_alloc);
- dfa->edests = re_malloc (re_node_set, dfa->nodes_alloc);
- dfa->eclosures = re_malloc (re_node_set, dfa->nodes_alloc);
- dfa->inveclosures = re_malloc (re_node_set, dfa->nodes_alloc);
- if (BE (dfa->nexts == NULL || dfa->org_indices == NULL || dfa->edests == NULL
- || dfa->eclosures == NULL || dfa->inveclosures == NULL, 0))
- return REG_ESPACE;
- /* Initialize them. */
- for (i = 0; i < dfa->nodes_len; ++i)
- {
- dfa->nexts[i] = -1;
- re_node_set_init_empty (dfa->edests + i);
- re_node_set_init_empty (dfa->eclosures + i);
- re_node_set_init_empty (dfa->inveclosures + i);
- }
-
- ret = analyze_tree (dfa, dfa->str_tree);
- if (BE (ret == REG_NOERROR, 1))
- {
- ret = calc_eclosure (dfa);
- if (ret == REG_NOERROR)
- calc_inveclosure (dfa);
- }
- return ret;
-}
-
-/* Helper functions for analyze.
- This function calculate "first", "next", and "edest" for the subtree
- whose root is NODE. */
-
-static reg_errcode_t
-analyze_tree (dfa, node)
- re_dfa_t *dfa;
- bin_tree_t *node;
-{
- reg_errcode_t ret;
- if (node->first == -1)
- calc_first (dfa, node);
- if (node->next == -1)
- calc_next (dfa, node);
- if (node->eclosure.nelem == 0)
- calc_epsdest (dfa, node);
- /* Calculate "first" etc. for the left child. */
- if (node->left != NULL)
- {
- ret = analyze_tree (dfa, node->left);
- if (BE (ret != REG_NOERROR, 0))
- return ret;
- }
- /* Calculate "first" etc. for the right child. */
- if (node->right != NULL)
- {
- ret = analyze_tree (dfa, node->right);
- if (BE (ret != REG_NOERROR, 0))
- return ret;
- }
- return REG_NOERROR;
-}
-
-/* Calculate "first" for the node NODE. */
-static void
-calc_first (dfa, node)
- re_dfa_t *dfa;
- bin_tree_t *node;
-{
- int idx, type;
- idx = node->node_idx;
- type = (node->type == 0) ? dfa->nodes[idx].type : node->type;
-
- switch (type)
- {
-#ifdef DEBUG
- case OP_OPEN_BRACKET:
- case OP_CLOSE_BRACKET:
- case OP_OPEN_DUP_NUM:
- case OP_CLOSE_DUP_NUM:
- case OP_NON_MATCH_LIST:
- case OP_OPEN_COLL_ELEM:
- case OP_CLOSE_COLL_ELEM:
- case OP_OPEN_EQUIV_CLASS:
- case OP_CLOSE_EQUIV_CLASS:
- case OP_OPEN_CHAR_CLASS:
- case OP_CLOSE_CHAR_CLASS:
- /* These must not be appeared here. */
- assert (0);
-#endif
- case END_OF_RE:
- case CHARACTER:
- case OP_PERIOD:
- case OP_DUP_ASTERISK:
- case OP_DUP_QUESTION:
-#ifdef RE_ENABLE_I18N
- case COMPLEX_BRACKET:
-#endif /* RE_ENABLE_I18N */
- case SIMPLE_BRACKET:
- case OP_BACK_REF:
- case ANCHOR:
- case OP_OPEN_SUBEXP:
- case OP_CLOSE_SUBEXP:
- node->first = idx;
- break;
- case OP_DUP_PLUS:
-#ifdef DEBUG
- assert (node->left != NULL);
-#endif
- if (node->left->first == -1)
- calc_first (dfa, node->left);
- node->first = node->left->first;
- break;
- case OP_ALT:
- node->first = idx;
- break;
- /* else fall through */
- default:
-#ifdef DEBUG
- assert (node->left != NULL);
-#endif
- if (node->left->first == -1)
- calc_first (dfa, node->left);
- node->first = node->left->first;
- break;
- }
-}
-
-/* Calculate "next" for the node NODE. */
-
-static void
-calc_next (dfa, node)
- re_dfa_t *dfa;
- bin_tree_t *node;
-{
- int idx, type;
- bin_tree_t *parent = node->parent;
- if (parent == NULL)
- {
- node->next = -1;
- idx = node->node_idx;
- if (node->type == 0)
- dfa->nexts[idx] = node->next;
- return;
- }
-
- idx = parent->node_idx;
- type = (parent->type == 0) ? dfa->nodes[idx].type : parent->type;
-
- switch (type)
- {
- case OP_DUP_ASTERISK:
- case OP_DUP_PLUS:
- node->next = idx;
- break;
- case CONCAT:
- if (parent->left == node)
- {
- if (parent->right->first == -1)
- calc_first (dfa, parent->right);
- node->next = parent->right->first;
- break;
- }
- /* else fall through */
- default:
- if (parent->next == -1)
- calc_next (dfa, parent);
- node->next = parent->next;
- break;
- }
- idx = node->node_idx;
- if (node->type == 0)
- dfa->nexts[idx] = node->next;
-}
-
-/* Calculate "edest" for the node NODE. */
-
-static void
-calc_epsdest (dfa, node)
- re_dfa_t *dfa;
- bin_tree_t *node;
-{
- int idx;
- idx = node->node_idx;
- if (node->type == 0)
- {
- if (dfa->nodes[idx].type == OP_DUP_ASTERISK
- || dfa->nodes[idx].type == OP_DUP_PLUS
- || dfa->nodes[idx].type == OP_DUP_QUESTION)
- {
- if (node->left->first == -1)
- calc_first (dfa, node->left);
- if (node->next == -1)
- calc_next (dfa, node);
- re_node_set_init_2 (dfa->edests + idx, node->left->first,
- node->next);
- }
- else if (dfa->nodes[idx].type == OP_ALT)
- {
- int left, right;
- if (node->left != NULL)
- {
- if (node->left->first == -1)
- calc_first (dfa, node->left);
- left = node->left->first;
- }
- else
- {
- if (node->next == -1)
- calc_next (dfa, node);
- left = node->next;
- }
- if (node->right != NULL)
- {
- if (node->right->first == -1)
- calc_first (dfa, node->right);
- right = node->right->first;
- }
- else
- {
- if (node->next == -1)
- calc_next (dfa, node);
- right = node->next;
- }
- re_node_set_init_2 (dfa->edests + idx, left, right);
- }
- else if (dfa->nodes[idx].type == ANCHOR
- || dfa->nodes[idx].type == OP_OPEN_SUBEXP
- || dfa->nodes[idx].type == OP_CLOSE_SUBEXP
- || dfa->nodes[idx].type == OP_BACK_REF)
- re_node_set_init_1 (dfa->edests + idx, node->next);
- }
-}
-
-/* Duplicate the epsilon closure of the node ROOT_NODE.
- Note that duplicated nodes have constraint INIT_CONSTRAINT in addition
- to their own constraint. */
-
-static reg_errcode_t
-duplicate_node_closure (dfa, top_org_node, top_clone_node, root_node,
- init_constraint)
- re_dfa_t *dfa;
- int top_org_node, top_clone_node, root_node;
- unsigned int init_constraint;
-{
- reg_errcode_t err;
- int org_node, clone_node, ret;
- unsigned int constraint = init_constraint;
- for (org_node = top_org_node, clone_node = top_clone_node;;)
- {
- int org_dest, clone_dest;
- if (dfa->nodes[org_node].type == OP_BACK_REF)
- {
- /* If the back reference epsilon-transit, its destination must
- also have the constraint. Then duplicate the epsilon closure
- of the destination of the back reference, and store it in
- edests of the back reference. */
- org_dest = dfa->nexts[org_node];
- re_node_set_empty (dfa->edests + clone_node);
- err = duplicate_node (&clone_dest, dfa, org_dest, constraint);
- if (BE (err != REG_NOERROR, 0))
- return err;
- dfa->nexts[clone_node] = dfa->nexts[org_node];
- ret = re_node_set_insert (dfa->edests + clone_node, clone_dest);
- if (BE (ret < 0, 0))
- return REG_ESPACE;
- }
- else if (dfa->edests[org_node].nelem == 0)
- {
- /* In case of the node can't epsilon-transit, don't duplicate the
- destination and store the original destination as the
- destination of the node. */
- dfa->nexts[clone_node] = dfa->nexts[org_node];
- break;
- }
- else if (dfa->edests[org_node].nelem == 1)
- {
- /* In case of the node can epsilon-transit, and it has only one
- destination. */
- org_dest = dfa->edests[org_node].elems[0];
- re_node_set_empty (dfa->edests + clone_node);
- if (dfa->nodes[org_node].type == ANCHOR)
- {
- /* In case of the node has another constraint, append it. */
- if (org_node == root_node && clone_node != org_node)
- {
- /* ...but if the node is root_node itself, it means the
- epsilon closure have a loop, then tie it to the
- destination of the root_node. */
- ret = re_node_set_insert (dfa->edests + clone_node,
- org_dest);
- if (BE (ret < 0, 0))
- return REG_ESPACE;
- break;
- }
- constraint |= dfa->nodes[org_node].opr.ctx_type;
- }
- err = duplicate_node (&clone_dest, dfa, org_dest, constraint);
- if (BE (err != REG_NOERROR, 0))
- return err;
- ret = re_node_set_insert (dfa->edests + clone_node, clone_dest);
- if (BE (ret < 0, 0))
- return REG_ESPACE;
- }
- else /* dfa->edests[org_node].nelem == 2 */
- {
- /* In case of the node can epsilon-transit, and it has two
- destinations. E.g. '|', '*', '+', '?'. */
- org_dest = dfa->edests[org_node].elems[0];
- re_node_set_empty (dfa->edests + clone_node);
- /* Search for a duplicated node which satisfies the constraint. */
- clone_dest = search_duplicated_node (dfa, org_dest, constraint);
- if (clone_dest == -1)
- {
- /* There are no such a duplicated node, create a new one. */
- err = duplicate_node (&clone_dest, dfa, org_dest, constraint);
- if (BE (err != REG_NOERROR, 0))
- return err;
- ret = re_node_set_insert (dfa->edests + clone_node, clone_dest);
- if (BE (ret < 0, 0))
- return REG_ESPACE;
- err = duplicate_node_closure (dfa, org_dest, clone_dest,
- root_node, constraint);
- if (BE (err != REG_NOERROR, 0))
- return err;
- }
- else
- {
- /* There are a duplicated node which satisfy the constraint,
- use it to avoid infinite loop. */
- ret = re_node_set_insert (dfa->edests + clone_node, clone_dest);
- if (BE (ret < 0, 0))
- return REG_ESPACE;
- }
-
- org_dest = dfa->edests[org_node].elems[1];
- err = duplicate_node (&clone_dest, dfa, org_dest, constraint);
- if (BE (err != REG_NOERROR, 0))
- return err;
- ret = re_node_set_insert (dfa->edests + clone_node, clone_dest);
- if (BE (ret < 0, 0))
- return REG_ESPACE;
- }
- org_node = org_dest;
- clone_node = clone_dest;
- }
- return REG_NOERROR;
-}
-
-/* Search for a node which is duplicated from the node ORG_NODE, and
- satisfies the constraint CONSTRAINT. */
-
-static int
-search_duplicated_node (dfa, org_node, constraint)
- re_dfa_t *dfa;
- int org_node;
- unsigned int constraint;
-{
- int idx;
- for (idx = dfa->nodes_len - 1; dfa->nodes[idx].duplicated && idx > 0; --idx)
- {
- if (org_node == dfa->org_indices[idx]
- && constraint == dfa->nodes[idx].constraint)
- return idx; /* Found. */
- }
- return -1; /* Not found. */
-}
-
-/* Duplicate the node whose index is ORG_IDX and set the constraint CONSTRAINT.
- The new index will be stored in NEW_IDX and return REG_NOERROR if succeeded,
- otherwise return the error code. */
-
-static reg_errcode_t
-duplicate_node (new_idx, dfa, org_idx, constraint)
- re_dfa_t *dfa;
- int *new_idx, org_idx;
- unsigned int constraint;
-{
- re_token_t dup;
- int dup_idx;
-
- dup = dfa->nodes[org_idx];
- dup_idx = re_dfa_add_node (dfa, dup, 1);
- if (BE (dup_idx == -1, 0))
- return REG_ESPACE;
- dfa->nodes[dup_idx].constraint = constraint;
- if (dfa->nodes[org_idx].type == ANCHOR)
- dfa->nodes[dup_idx].constraint |= dfa->nodes[org_idx].opr.ctx_type;
- dfa->nodes[dup_idx].duplicated = 1;
- re_node_set_init_empty (dfa->edests + dup_idx);
- re_node_set_init_empty (dfa->eclosures + dup_idx);
- re_node_set_init_empty (dfa->inveclosures + dup_idx);
-
- /* Store the index of the original node. */
- dfa->org_indices[dup_idx] = org_idx;
- *new_idx = dup_idx;
- return REG_NOERROR;
-}
-
-static void
-calc_inveclosure (dfa)
- re_dfa_t *dfa;
-{
- int src, idx, dest;
- for (src = 0; src < dfa->nodes_len; ++src)
- {
- for (idx = 0; idx < dfa->eclosures[src].nelem; ++idx)
- {
- dest = dfa->eclosures[src].elems[idx];
- re_node_set_insert (dfa->inveclosures + dest, src);
- }
- }
-}
-
-/* Calculate "eclosure" for all the node in DFA. */
-
-static reg_errcode_t
-calc_eclosure (dfa)
- re_dfa_t *dfa;
-{
- int node_idx, incomplete;
-#ifdef DEBUG
- assert (dfa->nodes_len > 0);
-#endif
- incomplete = 0;
- /* For each nodes, calculate epsilon closure. */
- for (node_idx = 0; ; ++node_idx)
- {
- reg_errcode_t err;
- re_node_set eclosure_elem;
- if (node_idx == dfa->nodes_len)
- {
- if (!incomplete)
- break;
- incomplete = 0;
- node_idx = 0;
- }
-
-#ifdef DEBUG
- assert (dfa->eclosures[node_idx].nelem != -1);
-#endif
- /* If we have already calculated, skip it. */
- if (dfa->eclosures[node_idx].nelem != 0)
- continue;
- /* Calculate epsilon closure of `node_idx'. */
- err = calc_eclosure_iter (&eclosure_elem, dfa, node_idx, 1);
- if (BE (err != REG_NOERROR, 0))
- return err;
-
- if (dfa->eclosures[node_idx].nelem == 0)
- {
- incomplete = 1;
- re_node_set_free (&eclosure_elem);
- }
- }
- return REG_NOERROR;
-}
-
-/* Calculate epsilon closure of NODE. */
-
-static reg_errcode_t
-calc_eclosure_iter (new_set, dfa, node, root)
- re_node_set *new_set;
- re_dfa_t *dfa;
- int node, root;
-{
- reg_errcode_t err;
- unsigned int constraint;
- int i, incomplete;
- re_node_set eclosure;
- incomplete = 0;
- err = re_node_set_alloc (&eclosure, dfa->edests[node].nelem + 1);
- if (BE (err != REG_NOERROR, 0))
- return err;
-
- /* This indicates that we are calculating this node now.
- We reference this value to avoid infinite loop. */
- dfa->eclosures[node].nelem = -1;
-
- constraint = ((dfa->nodes[node].type == ANCHOR)
- ? dfa->nodes[node].opr.ctx_type : 0);
- /* If the current node has constraints, duplicate all nodes.
- Since they must inherit the constraints. */
- if (constraint && !dfa->nodes[dfa->edests[node].elems[0]].duplicated)
- {
- int org_node, cur_node;
- org_node = cur_node = node;
- err = duplicate_node_closure (dfa, node, node, node, constraint);
- if (BE (err != REG_NOERROR, 0))
- return err;
- }
-
- /* Expand each epsilon destination nodes. */
- if (IS_EPSILON_NODE(dfa->nodes[node].type))
- for (i = 0; i < dfa->edests[node].nelem; ++i)
- {
- re_node_set eclosure_elem;
- int edest = dfa->edests[node].elems[i];
- /* If calculating the epsilon closure of `edest' is in progress,
- return intermediate result. */
- if (dfa->eclosures[edest].nelem == -1)
- {
- incomplete = 1;
- continue;
- }
- /* If we haven't calculated the epsilon closure of `edest' yet,
- calculate now. Otherwise use calculated epsilon closure. */
- if (dfa->eclosures[edest].nelem == 0)
- {
- err = calc_eclosure_iter (&eclosure_elem, dfa, edest, 0);
- if (BE (err != REG_NOERROR, 0))
- return err;
- }
- else
- eclosure_elem = dfa->eclosures[edest];
- /* Merge the epsilon closure of `edest'. */
- re_node_set_merge (&eclosure, &eclosure_elem);
- /* If the epsilon closure of `edest' is incomplete,
- the epsilon closure of this node is also incomplete. */
- if (dfa->eclosures[edest].nelem == 0)
- {
- incomplete = 1;
- re_node_set_free (&eclosure_elem);
- }
- }
-
- /* Epsilon closures include itself. */
- re_node_set_insert (&eclosure, node);
- if (incomplete && !root)
- dfa->eclosures[node].nelem = 0;
- else
- dfa->eclosures[node] = eclosure;
- *new_set = eclosure;
- return REG_NOERROR;
-}
-\f
-/* Functions for token which are used in the parser. */
-
-/* Fetch a token from INPUT.
- We must not use this function inside bracket expressions. */
-
-static re_token_t
-fetch_token (input, syntax)
- re_string_t *input;
- reg_syntax_t syntax;
-{
- re_token_t token;
- int consumed_byte;
- consumed_byte = peek_token (&token, input, syntax);
- re_string_skip_bytes (input, consumed_byte);
- return token;
-}
-
-/* Peek a token from INPUT, and return the length of the token.
- We must not use this function inside bracket expressions. */
-
-static int
-peek_token (token, input, syntax)
- re_token_t *token;
- re_string_t *input;
- reg_syntax_t syntax;
-{
- unsigned char c;
-
- if (re_string_eoi (input))
- {
- token->type = END_OF_RE;
- return 0;
- }
-
- c = re_string_peek_byte (input, 0);
- token->opr.c = c;
-
-#ifdef RE_ENABLE_I18N
- token->mb_partial = 0;
- if (MB_CUR_MAX > 1 &&
- !re_string_first_byte (input, re_string_cur_idx (input)))
- {
- token->type = CHARACTER;
- token->mb_partial = 1;
- return 1;
- }
-#endif
- if (c == '\\')
- {
- unsigned char c2;
- if (re_string_cur_idx (input) + 1 >= re_string_length (input))
- {
- token->type = BACK_SLASH;
- return 1;
- }
-
- c2 = re_string_peek_byte_case (input, 1);
- token->opr.c = c2;
- token->type = CHARACTER;
- switch (c2)
- {
- case '|':
- if (!(syntax & RE_LIMITED_OPS) && !(syntax & RE_NO_BK_VBAR))
- token->type = OP_ALT;
- break;
- case '1': case '2': case '3': case '4': case '5':
- case '6': case '7': case '8': case '9':
- if (!(syntax & RE_NO_BK_REFS))
- {
- token->type = OP_BACK_REF;
- token->opr.idx = c2 - '0';
- }
- break;
- case '<':
- if (!(syntax & RE_NO_GNU_OPS))
- {
- token->type = ANCHOR;
- token->opr.idx = WORD_FIRST;
- }
- break;
- case '>':
- if (!(syntax & RE_NO_GNU_OPS))
- {
- token->type = ANCHOR;
- token->opr.idx = WORD_LAST;
- }
- break;
- case 'b':
- if (!(syntax & RE_NO_GNU_OPS))
- {
- token->type = ANCHOR;
- token->opr.idx = WORD_DELIM;
- }
- break;
- case 'B':
- if (!(syntax & RE_NO_GNU_OPS))
- {
- token->type = ANCHOR;
- token->opr.idx = INSIDE_WORD;
- }
- break;
- case 'w':
- if (!(syntax & RE_NO_GNU_OPS))
- token->type = OP_WORD;
- break;
- case 'W':
- if (!(syntax & RE_NO_GNU_OPS))
- token->type = OP_NOTWORD;
- break;
- case '`':
- if (!(syntax & RE_NO_GNU_OPS))
- {
- token->type = ANCHOR;
- token->opr.idx = BUF_FIRST;
- }
- break;
- case '\'':
- if (!(syntax & RE_NO_GNU_OPS))
- {
- token->type = ANCHOR;
- token->opr.idx = BUF_LAST;
- }
- break;
- case '(':
- if (!(syntax & RE_NO_BK_PARENS))
- token->type = OP_OPEN_SUBEXP;
- break;
- case ')':
- if (!(syntax & RE_NO_BK_PARENS))
- token->type = OP_CLOSE_SUBEXP;
- break;
- case '+':
- if (!(syntax & RE_LIMITED_OPS) && (syntax & RE_BK_PLUS_QM))
- token->type = OP_DUP_PLUS;
- break;
- case '?':
- if (!(syntax & RE_LIMITED_OPS) && (syntax & RE_BK_PLUS_QM))
- token->type = OP_DUP_QUESTION;
- break;
- case '{':
- if ((syntax & RE_INTERVALS) && (!(syntax & RE_NO_BK_BRACES)))
- token->type = OP_OPEN_DUP_NUM;
- break;
- case '}':
- if ((syntax & RE_INTERVALS) && (!(syntax & RE_NO_BK_BRACES)))
- token->type = OP_CLOSE_DUP_NUM;
- break;
- default:
- break;
- }
- return 2;
- }
-
- token->type = CHARACTER;
- switch (c)
- {
- case '\n':
- if (syntax & RE_NEWLINE_ALT)
- token->type = OP_ALT;
- break;
- case '|':
- if (!(syntax & RE_LIMITED_OPS) && (syntax & RE_NO_BK_VBAR))
- token->type = OP_ALT;
- break;
- case '*':
- token->type = OP_DUP_ASTERISK;
- break;
- case '+':
- if (!(syntax & RE_LIMITED_OPS) && !(syntax & RE_BK_PLUS_QM))
- token->type = OP_DUP_PLUS;
- break;
- case '?':
- if (!(syntax & RE_LIMITED_OPS) && !(syntax & RE_BK_PLUS_QM))
- token->type = OP_DUP_QUESTION;
- break;
- case '{':
- if ((syntax & RE_INTERVALS) && (syntax & RE_NO_BK_BRACES))
- token->type = OP_OPEN_DUP_NUM;
- break;
- case '}':
- if ((syntax & RE_INTERVALS) && (syntax & RE_NO_BK_BRACES))
- token->type = OP_CLOSE_DUP_NUM;
- break;
- case '(':
- if (syntax & RE_NO_BK_PARENS)
- token->type = OP_OPEN_SUBEXP;
- break;
- case ')':
- if (syntax & RE_NO_BK_PARENS)
- token->type = OP_CLOSE_SUBEXP;
- break;
- case '[':
- token->type = OP_OPEN_BRACKET;
- break;
- case '.':
- token->type = OP_PERIOD;
- break;
- case '^':
- if (!(syntax & RE_CONTEXT_INDEP_ANCHORS) &&
- re_string_cur_idx (input) != 0)
- {
- char prev = re_string_peek_byte (input, -1);
- if (prev != '|' && prev != '(' &&
- (!(syntax & RE_NEWLINE_ALT) || prev != '\n'))
- break;
- }
- token->type = ANCHOR;
- token->opr.idx = LINE_FIRST;
- break;
- case '$':
- if (!(syntax & RE_CONTEXT_INDEP_ANCHORS) &&
- re_string_cur_idx (input) + 1 != re_string_length (input))
- {
- re_token_t next;
- re_string_skip_bytes (input, 1);
- peek_token (&next, input, syntax);
- re_string_skip_bytes (input, -1);
- if (next.type != OP_ALT && next.type != OP_CLOSE_SUBEXP)
- break;
- }
- token->type = ANCHOR;
- token->opr.idx = LINE_LAST;
- break;
- default:
- break;
- }
- return 1;
-}
-
-/* Peek a token from INPUT, and return the length of the token.
- We must not use this function out of bracket expressions. */
-
-static int
-peek_token_bracket (token, input, syntax)
- re_token_t *token;
- re_string_t *input;
- reg_syntax_t syntax;
-{
- unsigned char c;
- if (re_string_eoi (input))
- {
- token->type = END_OF_RE;
- return 0;
- }
- c = re_string_peek_byte (input, 0);
- token->opr.c = c;
-
-#ifdef RE_ENABLE_I18N
- if (MB_CUR_MAX > 1 &&
- !re_string_first_byte (input, re_string_cur_idx (input)))
- {
- token->type = CHARACTER;
- return 1;
- }
-#endif /* RE_ENABLE_I18N */
-
- if (c == '\\' && (syntax & RE_BACKSLASH_ESCAPE_IN_LISTS))
- {
- /* In this case, '\' escape a character. */
- unsigned char c2;
- re_string_skip_bytes (input, 1);
- c2 = re_string_peek_byte (input, 0);
- token->opr.c = c2;
- token->type = CHARACTER;
- return 1;
- }
- if (c == '[') /* '[' is a special char in a bracket exps. */
- {
- unsigned char c2;
- int token_len;
- c2 = re_string_peek_byte (input, 1);
- token->opr.c = c2;
- token_len = 2;
- switch (c2)
- {
- case '.':
- token->type = OP_OPEN_COLL_ELEM;
- break;
- case '=':
- token->type = OP_OPEN_EQUIV_CLASS;
- break;
- case ':':
- if (syntax & RE_CHAR_CLASSES)
- {
- token->type = OP_OPEN_CHAR_CLASS;
- break;
- }
- /* else fall through. */
- default:
- token->type = CHARACTER;
- token->opr.c = c;
- token_len = 1;
- break;
- }
- return token_len;
- }
- switch (c)
- {
- case '-':
- token->type = OP_CHARSET_RANGE;
- break;
- case ']':
- token->type = OP_CLOSE_BRACKET;
- break;
- case '^':
- token->type = OP_NON_MATCH_LIST;
- break;
- default:
- token->type = CHARACTER;
- }
- return 1;
-}
-\f
-/* Functions for parser. */
-
-/* Entry point of the parser.
- Parse the regular expression REGEXP and return the structure tree.
- If an error is occured, ERR is set by error code, and return NULL.
- This function build the following tree, from regular expression <reg_exp>:
- CAT
- / \
- / \
- <reg_exp> EOR
-
- CAT means concatenation.
- EOR means end of regular expression. */
-
-static bin_tree_t *
-parse (regexp, preg, syntax, err)
- re_string_t *regexp;
- regex_t *preg;
- reg_syntax_t syntax;
- reg_errcode_t *err;
-{
- re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- bin_tree_t *tree, *eor, *root;
- re_token_t current_token;
- int new_idx;
- current_token = fetch_token (regexp, syntax);
- tree = parse_reg_exp (regexp, preg, ¤t_token, syntax, 0, err);
- if (BE (*err != REG_NOERROR && tree == NULL, 0))
- return NULL;
- new_idx = re_dfa_add_node (dfa, current_token, 0);
- eor = create_tree (NULL, NULL, 0, new_idx);
- if (tree != NULL)
- root = create_tree (tree, eor, CONCAT, 0);
- else
- root = eor;
- if (BE (new_idx == -1 || eor == NULL || root == NULL, 0))
- {
- *err = REG_ESPACE;
- return NULL;
- }
- return root;
-}
-
-/* This function build the following tree, from regular expression
- <branch1>|<branch2>:
- ALT
- / \
- / \
- <branch1> <branch2>
-
- ALT means alternative, which represents the operator `|'. */
-
-static bin_tree_t *
-parse_reg_exp (regexp, preg, token, syntax, nest, err)
- re_string_t *regexp;
- regex_t *preg;
- re_token_t *token;
- reg_syntax_t syntax;
- int nest;
- reg_errcode_t *err;
-{
- re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- bin_tree_t *tree, *branch = NULL;
- int new_idx;
- tree = parse_branch (regexp, preg, token, syntax, nest, err);
- if (BE (*err != REG_NOERROR && tree == NULL, 0))
- return NULL;
-
- while (token->type == OP_ALT)
- {
- re_token_t alt_token = *token;
- new_idx = re_dfa_add_node (dfa, alt_token, 0);
- *token = fetch_token (regexp, syntax);
- if (token->type != OP_ALT && token->type != END_OF_RE
- && (nest == 0 || token->type != OP_CLOSE_SUBEXP))
- {
- branch = parse_branch (regexp, preg, token, syntax, nest, err);
- if (BE (*err != REG_NOERROR && branch == NULL, 0))
- {
- free_bin_tree (tree);
- return NULL;
- }
- }
- else
- branch = NULL;
- tree = create_tree (tree, branch, 0, new_idx);
- if (BE (new_idx == -1 || tree == NULL, 0))
- {
- *err = REG_ESPACE;
- return NULL;
- }
- dfa->has_plural_match = 1;
- }
- return tree;
-}
-
-/* This function build the following tree, from regular expression
- <exp1><exp2>:
- CAT
- / \
- / \
- <exp1> <exp2>
-
- CAT means concatenation. */
-
-static bin_tree_t *
-parse_branch (regexp, preg, token, syntax, nest, err)
- re_string_t *regexp;
- regex_t *preg;
- re_token_t *token;
- reg_syntax_t syntax;
- int nest;
- reg_errcode_t *err;
-{
- bin_tree_t *tree, *exp;
- tree = parse_expression (regexp, preg, token, syntax, nest, err);
- if (BE (*err != REG_NOERROR && tree == NULL, 0))
- return NULL;
-
- while (token->type != OP_ALT && token->type != END_OF_RE
- && (nest == 0 || token->type != OP_CLOSE_SUBEXP))
- {
- exp = parse_expression (regexp, preg, token, syntax, nest, err);
- if (BE (*err != REG_NOERROR && exp == NULL, 0))
- {
- free_bin_tree (tree);
- return NULL;
- }
- if (tree != NULL && exp != NULL)
- {
- tree = create_tree (tree, exp, CONCAT, 0);
- if (tree == NULL)
- {
- *err = REG_ESPACE;
- return NULL;
- }
- }
- else if (tree == NULL)
- tree = exp;
- /* Otherwise exp == NULL, we don't need to create new tree. */
- }
- return tree;
-}
-
-/* This function build the following tree, from regular expression a*:
- *
- |
- a
-*/
-
-static bin_tree_t *
-parse_expression (regexp, preg, token, syntax, nest, err)
- re_string_t *regexp;
- regex_t *preg;
- re_token_t *token;
- reg_syntax_t syntax;
- int nest;
- reg_errcode_t *err;
-{
- re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- bin_tree_t *tree;
- int new_idx;
- switch (token->type)
- {
- case CHARACTER:
- new_idx = re_dfa_add_node (dfa, *token, 0);
- tree = create_tree (NULL, NULL, 0, new_idx);
- if (BE (new_idx == -1 || tree == NULL, 0))
- {
- *err = REG_ESPACE;
- return NULL;
- }
-#ifdef RE_ENABLE_I18N
- if (MB_CUR_MAX > 1)
- {
- while (!re_string_eoi (regexp)
- && !re_string_first_byte (regexp, re_string_cur_idx (regexp)))
- {
- bin_tree_t *mbc_remain;
- *token = fetch_token (regexp, syntax);
- new_idx = re_dfa_add_node (dfa, *token, 0);
- mbc_remain = create_tree (NULL, NULL, 0, new_idx);
- tree = create_tree (tree, mbc_remain, CONCAT, 0);
- if (BE (new_idx == -1 || mbc_remain == NULL || tree == NULL, 0))
- {
- *err = REG_ESPACE;
- return NULL;
- }
- }
- }
-#endif
- break;
- case OP_OPEN_SUBEXP:
- tree = parse_sub_exp (regexp, preg, token, syntax, nest + 1, err);
- if (BE (*err != REG_NOERROR && tree == NULL, 0))
- return NULL;
- break;
- case OP_OPEN_BRACKET:
- tree = parse_bracket_exp (regexp, dfa, token, syntax, err);
- if (BE (*err != REG_NOERROR && tree == NULL, 0))
- return NULL;
- break;
- case OP_BACK_REF:
- if (BE (preg->re_nsub < token->opr.idx
- || dfa->subexps[token->opr.idx - 1].end == -1, 0))
- {
- *err = REG_ESUBREG;
- return NULL;
- }
- dfa->used_bkref_map |= 1 << (token->opr.idx - 1);
- new_idx = re_dfa_add_node (dfa, *token, 0);
- tree = create_tree (NULL, NULL, 0, new_idx);
- if (BE (new_idx == -1 || tree == NULL, 0))
- {
- *err = REG_ESPACE;
- return NULL;
- }
- ++dfa->nbackref;
- dfa->has_mb_node = 1;
- break;
- case OP_DUP_ASTERISK:
- case OP_DUP_PLUS:
- case OP_DUP_QUESTION:
- case OP_OPEN_DUP_NUM:
- if (syntax & RE_CONTEXT_INVALID_OPS)
- {
- *err = REG_BADRPT;
- return NULL;
- }
- else if (syntax & RE_CONTEXT_INDEP_OPS)
- {
- *token = fetch_token (regexp, syntax);
- return parse_expression (regexp, preg, token, syntax, nest, err);
- }
- /* else fall through */
- case OP_CLOSE_SUBEXP:
- if ((token->type == OP_CLOSE_SUBEXP) &&
- !(syntax & RE_UNMATCHED_RIGHT_PAREN_ORD))
- {
- *err = REG_ERPAREN;
- return NULL;
- }
- /* else fall through */
- case OP_CLOSE_DUP_NUM:
- /* We treat it as a normal character. */
-
- /* Then we can these characters as normal characters. */
- token->type = CHARACTER;
- new_idx = re_dfa_add_node (dfa, *token, 0);
- tree = create_tree (NULL, NULL, 0, new_idx);
- if (BE (new_idx == -1 || tree == NULL, 0))
- {
- *err = REG_ESPACE;
- return NULL;
- }
- break;
- case ANCHOR:
- if (dfa->word_char == NULL)
- {
- *err = init_word_char (dfa);
- if (BE (*err != REG_NOERROR, 0))
- return NULL;
- }
- if (token->opr.ctx_type == WORD_DELIM)
- {
- bin_tree_t *tree_first, *tree_last;
- int idx_first, idx_last;
- token->opr.ctx_type = WORD_FIRST;
- idx_first = re_dfa_add_node (dfa, *token, 0);
- tree_first = create_tree (NULL, NULL, 0, idx_first);
- token->opr.ctx_type = WORD_LAST;
- idx_last = re_dfa_add_node (dfa, *token, 0);
- tree_last = create_tree (NULL, NULL, 0, idx_last);
- token->type = OP_ALT;
- new_idx = re_dfa_add_node (dfa, *token, 0);
- tree = create_tree (tree_first, tree_last, 0, new_idx);
- if (BE (idx_first == -1 || idx_last == -1 || new_idx == -1
- || tree_first == NULL || tree_last == NULL
- || tree == NULL, 0))
- {
- *err = REG_ESPACE;
- return NULL;
- }
- }
- else
- {
- new_idx = re_dfa_add_node (dfa, *token, 0);
- tree = create_tree (NULL, NULL, 0, new_idx);
- if (BE (new_idx == -1 || tree == NULL, 0))
- {
- *err = REG_ESPACE;
- return NULL;
- }
- }
- /* We must return here, since ANCHORs can't be followed
- by repetition operators.
- eg. RE"^*" is invalid or "<ANCHOR(^)><CHAR(*)>",
- it must not be "<ANCHOR(^)><REPEAT(*)>". */
- *token = fetch_token (regexp, syntax);
- return tree;
- case OP_PERIOD:
- new_idx = re_dfa_add_node (dfa, *token, 0);
- tree = create_tree (NULL, NULL, 0, new_idx);
- if (BE (new_idx == -1 || tree == NULL, 0))
- {
- *err = REG_ESPACE;
- return NULL;
- }
- if (MB_CUR_MAX > 1)
- dfa->has_mb_node = 1;
- break;
- case OP_WORD:
- tree = build_word_op (dfa, 0, err);
- if (BE (*err != REG_NOERROR && tree == NULL, 0))
- return NULL;
- break;
- case OP_NOTWORD:
- tree = build_word_op (dfa, 1, err);
- if (BE (*err != REG_NOERROR && tree == NULL, 0))
- return NULL;
- break;
- case OP_ALT:
- case END_OF_RE:
- return NULL;
- case BACK_SLASH:
- *err = REG_EESCAPE;
- return NULL;
- default:
- /* Must not happen? */
-#ifdef DEBUG
- assert (0);
-#endif
- return NULL;
- }
- *token = fetch_token (regexp, syntax);
-
- while (token->type == OP_DUP_ASTERISK || token->type == OP_DUP_PLUS
- || token->type == OP_DUP_QUESTION || token->type == OP_OPEN_DUP_NUM)
- {
- tree = parse_dup_op (tree, regexp, dfa, token, syntax, err);
- if (BE (*err != REG_NOERROR && tree == NULL, 0))
- return NULL;
- dfa->has_plural_match = 1;
- }
-
- return tree;
-}
-
-/* This function build the following tree, from regular expression
- (<reg_exp>):
- SUBEXP
- |
- <reg_exp>
-*/
-
-static bin_tree_t *
-parse_sub_exp (regexp, preg, token, syntax, nest, err)
- re_string_t *regexp;
- regex_t *preg;
- re_token_t *token;
- reg_syntax_t syntax;
- int nest;
- reg_errcode_t *err;
-{
- re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- bin_tree_t *tree, *left_par, *right_par;
- size_t cur_nsub;
- int new_idx;
- cur_nsub = preg->re_nsub++;
- if (dfa->subexps_alloc < preg->re_nsub)
- {
- re_subexp_t *new_array;
- dfa->subexps_alloc *= 2;
- new_array = re_realloc (dfa->subexps, re_subexp_t, dfa->subexps_alloc);
- if (BE (new_array == NULL, 0))
- {
- dfa->subexps_alloc /= 2;
- *err = REG_ESPACE;
- return NULL;
- }
- dfa->subexps = new_array;
- }
- dfa->subexps[cur_nsub].start = dfa->nodes_len;
- dfa->subexps[cur_nsub].end = -1;
-
- new_idx = re_dfa_add_node (dfa, *token, 0);
- left_par = create_tree (NULL, NULL, 0, new_idx);
- if (BE (new_idx == -1 || left_par == NULL, 0))
- {
- *err = REG_ESPACE;
- return NULL;
- }
- dfa->nodes[new_idx].opr.idx = cur_nsub;
- *token = fetch_token (regexp, syntax);
-
- /* The subexpression may be a null string. */
- if (token->type == OP_CLOSE_SUBEXP)
- tree = NULL;
- else
- {
- tree = parse_reg_exp (regexp, preg, token, syntax, nest, err);
- if (BE (*err != REG_NOERROR && tree == NULL, 0))
- return NULL;
- }
- if (BE (token->type != OP_CLOSE_SUBEXP, 0))
- {
- free_bin_tree (tree);
- *err = REG_BADPAT;
- return NULL;
- }
- new_idx = re_dfa_add_node (dfa, *token, 0);
- dfa->subexps[cur_nsub].end = dfa->nodes_len;
- right_par = create_tree (NULL, NULL, 0, new_idx);
- tree = ((tree == NULL) ? right_par
- : create_tree (tree, right_par, CONCAT, 0));
- tree = create_tree (left_par, tree, CONCAT, 0);
- if (BE (new_idx == -1 || right_par == NULL || tree == NULL, 0))
- {
- *err = REG_ESPACE;
- return NULL;
- }
- dfa->nodes[new_idx].opr.idx = cur_nsub;
-
- return tree;
-}
-
-/* This function parse repetition operators like "*", "+", "{1,3}" etc. */
-
-static bin_tree_t *
-parse_dup_op (dup_elem, regexp, dfa, token, syntax, err)
- bin_tree_t *dup_elem;
- re_string_t *regexp;
- re_dfa_t *dfa;
- re_token_t *token;
- reg_syntax_t syntax;
- reg_errcode_t *err;
-{
- re_token_t dup_token;
- bin_tree_t *tree = dup_elem, *work_tree;
- int new_idx, start_idx = re_string_cur_idx (regexp);
- re_token_t start_token = *token;
- if (token->type == OP_OPEN_DUP_NUM)
- {
- int i;
- int end = 0;
- int start = fetch_number (regexp, token, syntax);
- bin_tree_t *elem;
- if (start == -1)
- {
- if (token->type == CHARACTER && token->opr.c == ',')
- start = 0; /* We treat "{,m}" as "{0,m}". */
- else
- {
- *err = REG_BADBR; /* <re>{} is invalid. */
- return NULL;
- }
- }
- if (BE (start != -2, 1))
- {
- /* We treat "{n}" as "{n,n}". */
- end = ((token->type == OP_CLOSE_DUP_NUM) ? start
- : ((token->type == CHARACTER && token->opr.c == ',')
- ? fetch_number (regexp, token, syntax) : -2));
- }
- if (BE (start == -2 || end == -2, 0))
- {
- /* Invalid sequence. */
- if (token->type == OP_CLOSE_DUP_NUM)
- goto parse_dup_op_invalid_interval;
- else
- goto parse_dup_op_ebrace;
- }
- if (BE (start == 0 && end == 0, 0))
- {
- /* We treat "<re>{0}" and "<re>{0,0}" as null string. */
- *token = fetch_token (regexp, syntax);
- free_bin_tree (dup_elem);
- return NULL;
- }
-
- /* Extract "<re>{n,m}" to "<re><re>...<re><re>{0,<m-n>}". */
- elem = tree;
- for (i = 0; i < start; ++i)
- if (i != 0)
- {
- work_tree = duplicate_tree (elem, dfa);
- tree = create_tree (tree, work_tree, CONCAT, 0);
- if (BE (work_tree == NULL || tree == NULL, 0))
- goto parse_dup_op_espace;
- }
-
- if (end == -1)
- {
- /* We treat "<re>{0,}" as "<re>*". */
- dup_token.type = OP_DUP_ASTERISK;
- if (start > 0)
- {
- elem = duplicate_tree (elem, dfa);
- new_idx = re_dfa_add_node (dfa, dup_token, 0);
- work_tree = create_tree (elem, NULL, 0, new_idx);
- tree = create_tree (tree, work_tree, CONCAT, 0);
- if (BE (elem == NULL || new_idx == -1 || work_tree == NULL
- || tree == NULL, 0))
- goto parse_dup_op_espace;
- }
- else
- {
- new_idx = re_dfa_add_node (dfa, dup_token, 0);
- tree = create_tree (elem, NULL, 0, new_idx);
- if (BE (new_idx == -1 || tree == NULL, 0))
- goto parse_dup_op_espace;
- }
- }
- else if (end - start > 0)
- {
- /* Then extract "<re>{0,m}" to "<re>?<re>?...<re>?". */
- dup_token.type = OP_DUP_QUESTION;
- if (start > 0)
- {
- elem = duplicate_tree (elem, dfa);
- new_idx = re_dfa_add_node (dfa, dup_token, 0);
- elem = create_tree (elem, NULL, 0, new_idx);
- tree = create_tree (tree, elem, CONCAT, 0);
- if (BE (elem == NULL || new_idx == -1 || tree == NULL, 0))
- goto parse_dup_op_espace;
- }
- else
- {
- new_idx = re_dfa_add_node (dfa, dup_token, 0);
- tree = elem = create_tree (elem, NULL, 0, new_idx);
- if (BE (new_idx == -1 || tree == NULL, 0))
- goto parse_dup_op_espace;
- }
- for (i = 1; i < end - start; ++i)
- {
- work_tree = duplicate_tree (elem, dfa);
- tree = create_tree (tree, work_tree, CONCAT, 0);
- if (BE (work_tree == NULL || tree == NULL, 0))
- {
- *err = REG_ESPACE;
- return NULL;
- }
- }
- }
- }
- else
- {
- new_idx = re_dfa_add_node (dfa, *token, 0);
- tree = create_tree (tree, NULL, 0, new_idx);
- if (BE (new_idx == -1 || tree == NULL, 0))
- {
- *err = REG_ESPACE;
- return NULL;
- }
- }
- *token = fetch_token (regexp, syntax);
- return tree;
-
- parse_dup_op_espace:
- free_bin_tree (tree);
- *err = REG_ESPACE;
- return NULL;
-
- parse_dup_op_ebrace:
- if (BE (!(syntax & RE_INVALID_INTERVAL_ORD), 0))
- {
- *err = REG_EBRACE;
- return NULL;
- }
- goto parse_dup_op_rollback;
- parse_dup_op_invalid_interval:
- if (BE (!(syntax & RE_INVALID_INTERVAL_ORD), 0))
- {
- *err = REG_BADBR;
- return NULL;
- }
- parse_dup_op_rollback:
- re_string_set_index (regexp, start_idx);
- *token = start_token;
- token->type = CHARACTER;
- return dup_elem;
-}
-
-/* Size of the names for collating symbol/equivalence_class/character_class.
- I'm not sure, but maybe enough. */
-#define BRACKET_NAME_BUF_SIZE 32
-
-#ifndef _LIBC
- /* Local function for parse_bracket_exp only used in case of NOT _LIBC.
- Build the range expression which starts from START_ELEM, and ends
- at END_ELEM. The result are written to MBCSET and SBCSET.
- RANGE_ALLOC is the allocated size of mbcset->range_starts, and
- mbcset->range_ends, is a pointer argument sinse we may
- update it. */
-
-static reg_errcode_t
-# ifdef RE_ENABLE_I18N
-build_range_exp (sbcset, mbcset, range_alloc, start_elem, end_elem)
- re_charset_t *mbcset;
- int *range_alloc;
-# else /* not RE_ENABLE_I18N */
-build_range_exp (sbcset, start_elem, end_elem)
-# endif /* not RE_ENABLE_I18N */
- re_bitset_ptr_t sbcset;
- bracket_elem_t *start_elem, *end_elem;
-{
- unsigned int start_ch, end_ch;
- /* Equivalence Classes and Character Classes can't be a range start/end. */
- if (BE (start_elem->type == EQUIV_CLASS || start_elem->type == CHAR_CLASS
- || end_elem->type == EQUIV_CLASS || end_elem->type == CHAR_CLASS,
- 0))
- return REG_ERANGE;
-
- /* We can handle no multi character collating elements without libc
- support. */
- if (BE ((start_elem->type == COLL_SYM
- && strlen ((char *) start_elem->opr.name) > 1)
- || (end_elem->type == COLL_SYM
- && strlen ((char *) end_elem->opr.name) > 1), 0))
- return REG_ECOLLATE;
-
-# ifdef RE_ENABLE_I18N
- {
- wchar_t wc, start_wc, end_wc;
- wchar_t cmp_buf[6] = {L'\0', L'\0', L'\0', L'\0', L'\0', L'\0'};
-
- start_ch = ((start_elem->type == SB_CHAR) ? start_elem->opr.ch
- : ((start_elem->type == COLL_SYM) ? start_elem->opr.name[0]
- : 0));
- end_ch = ((end_elem->type == SB_CHAR) ? end_elem->opr.ch
- : ((end_elem->type == COLL_SYM) ? end_elem->opr.name[0]
- : 0));
- start_wc = ((start_elem->type == SB_CHAR || start_elem->type == COLL_SYM)
- ? __btowc (start_ch) : start_elem->opr.wch);
- end_wc = ((end_elem->type == SB_CHAR || end_elem->type == COLL_SYM)
- ? __btowc (end_ch) : end_elem->opr.wch);
- cmp_buf[0] = start_wc;
- cmp_buf[4] = end_wc;
- if (wcscoll (cmp_buf, cmp_buf + 4) > 0)
- return REG_ERANGE;
-
- /* Check the space of the arrays. */
- if (*range_alloc == mbcset->nranges)
- {
- /* There are not enough space, need realloc. */
- wchar_t *new_array_start, *new_array_end;
- int new_nranges;
-
- /* +1 in case of mbcset->nranges is 0. */
- new_nranges = 2 * mbcset->nranges + 1;
- /* Use realloc since mbcset->range_starts and mbcset->range_ends
- are NULL if *range_alloc == 0. */
- new_array_start = re_realloc (mbcset->range_starts, wchar_t,
- new_nranges);
- new_array_end = re_realloc (mbcset->range_ends, wchar_t,
- new_nranges);
-
- if (BE (new_array_start == NULL || new_array_end == NULL, 0))
- return REG_ESPACE;
-
- mbcset->range_starts = new_array_start;
- mbcset->range_ends = new_array_end;
- *range_alloc = new_nranges;
- }
-
- mbcset->range_starts[mbcset->nranges] = start_wc;
- mbcset->range_ends[mbcset->nranges++] = end_wc;
-
- /* Build the table for single byte characters. */
- for (wc = 0; wc <= SBC_MAX; ++wc)
- {
- cmp_buf[2] = wc;
- if (wcscoll (cmp_buf, cmp_buf + 2) <= 0
- && wcscoll (cmp_buf + 2, cmp_buf + 4) <= 0)
- bitset_set (sbcset, wc);
- }
- }
-# else /* not RE_ENABLE_I18N */
- {
- unsigned int ch;
- start_ch = ((start_elem->type == SB_CHAR ) ? start_elem->opr.ch
- : ((start_elem->type == COLL_SYM) ? start_elem->opr.name[0]
- : 0));
- end_ch = ((end_elem->type == SB_CHAR ) ? end_elem->opr.ch
- : ((end_elem->type == COLL_SYM) ? end_elem->opr.name[0]
- : 0));
- if (start_ch > end_ch)
- return REG_ERANGE;
- /* Build the table for single byte characters. */
- for (ch = 0; ch <= SBC_MAX; ++ch)
- if (start_ch <= ch && ch <= end_ch)
- bitset_set (sbcset, ch);
- }
-# endif /* not RE_ENABLE_I18N */
- return REG_NOERROR;
-}
-#endif /* not _LIBC */
-
-#ifndef _LIBC
-/* Helper function for parse_bracket_exp only used in case of NOT _LIBC..
- Build the collating element which is represented by NAME.
- The result are written to MBCSET and SBCSET.
- COLL_SYM_ALLOC is the allocated size of mbcset->coll_sym, is a
- pointer argument since we may update it. */
-
-static reg_errcode_t
-# ifdef RE_ENABLE_I18N
-build_collating_symbol (sbcset, mbcset, coll_sym_alloc, name)
- re_charset_t *mbcset;
- int *coll_sym_alloc;
-# else /* not RE_ENABLE_I18N */
-build_collating_symbol (sbcset, name)
-# endif /* not RE_ENABLE_I18N */
- re_bitset_ptr_t sbcset;
- const unsigned char *name;
-{
- size_t name_len = strlen ((const char *) name);
- if (BE (name_len != 1, 0))
- return REG_ECOLLATE;
- else
- {
- bitset_set (sbcset, name[0]);
- return REG_NOERROR;
- }
-}
-#endif /* not _LIBC */
-
-/* This function parse bracket expression like "[abc]", "[a-c]",
- "[[.a-a.]]" etc. */
-
-static bin_tree_t *
-parse_bracket_exp (regexp, dfa, token, syntax, err)
- re_string_t *regexp;
- re_dfa_t *dfa;
- re_token_t *token;
- reg_syntax_t syntax;
- reg_errcode_t *err;
-{
-#ifdef _LIBC
- const unsigned char *collseqmb;
- const char *collseqwc;
- uint32_t nrules;
- int32_t table_size;
- const int32_t *symb_table;
- const unsigned char *extra;
-
- /* Local function for parse_bracket_exp used in _LIBC environement.
- Seek the collating symbol entry correspondings to NAME.
- Return the index of the symbol in the SYMB_TABLE. */
-
- static inline int32_t
- seek_collating_symbol_entry (name, name_len)
- const unsigned char *name;
- size_t name_len;
- {
- int32_t hash = elem_hash ((const char *) name, name_len);
- int32_t elem = hash % table_size;
- int32_t second = hash % (table_size - 2);
- while (symb_table[2 * elem] != 0)
- {
- /* First compare the hashing value. */
- if (symb_table[2 * elem] == hash
- /* Compare the length of the name. */
- && name_len == extra[symb_table[2 * elem + 1]]
- /* Compare the name. */
- && memcmp (name, &extra[symb_table[2 * elem + 1] + 1],
- name_len) == 0)
- {
- /* Yep, this is the entry. */
- break;
- }
-
- /* Next entry. */
- elem += second;
- }
- return elem;
- }
-
- /* Local function for parse_bracket_exp used in _LIBC environement.
- Look up the collation sequence value of BR_ELEM.
- Return the value if succeeded, UINT_MAX otherwise. */
-
- static inline unsigned int
- lookup_collation_sequence_value (br_elem)
- bracket_elem_t *br_elem;
- {
- if (br_elem->type == SB_CHAR)
- {
- /*
- if (MB_CUR_MAX == 1)
- */
- if (nrules == 0)
- return collseqmb[br_elem->opr.ch];
- else
- {
- wint_t wc = __btowc (br_elem->opr.ch);
- return collseq_table_lookup (collseqwc, wc);
- }
- }
- else if (br_elem->type == MB_CHAR)
- {
- return collseq_table_lookup (collseqwc, br_elem->opr.wch);
- }
- else if (br_elem->type == COLL_SYM)
- {
- size_t sym_name_len = strlen ((char *) br_elem->opr.name);
- if (nrules != 0)
- {
- int32_t elem, idx;
- elem = seek_collating_symbol_entry (br_elem->opr.name,
- sym_name_len);
- if (symb_table[2 * elem] != 0)
- {
- /* We found the entry. */
- idx = symb_table[2 * elem + 1];
- /* Skip the name of collating element name. */
- idx += 1 + extra[idx];
- /* Skip the byte sequence of the collating element. */
- idx += 1 + extra[idx];
- /* Adjust for the alignment. */
- idx = (idx + 3) & ~3;
- /* Skip the multibyte collation sequence value. */
- idx += sizeof (unsigned int);
- /* Skip the wide char sequence of the collating element. */
- idx += sizeof (unsigned int) *
- (1 + *(unsigned int *) (extra + idx));
- /* Return the collation sequence value. */
- return *(unsigned int *) (extra + idx);
- }
- else if (symb_table[2 * elem] == 0 && sym_name_len == 1)
- {
- /* No valid character. Match it as a single byte
- character. */
- return collseqmb[br_elem->opr.name[0]];
- }
- }
- else if (sym_name_len == 1)
- return collseqmb[br_elem->opr.name[0]];
- }
- return UINT_MAX;
- }
-
- /* Local function for parse_bracket_exp used in _LIBC environement.
- Build the range expression which starts from START_ELEM, and ends
- at END_ELEM. The result are written to MBCSET and SBCSET.
- RANGE_ALLOC is the allocated size of mbcset->range_starts, and
- mbcset->range_ends, is a pointer argument sinse we may
- update it. */
-
- static inline reg_errcode_t
-# ifdef RE_ENABLE_I18N
- build_range_exp (sbcset, mbcset, range_alloc, start_elem, end_elem)
- re_charset_t *mbcset;
- int *range_alloc;
-# else /* not RE_ENABLE_I18N */
- build_range_exp (sbcset, start_elem, end_elem)
-# endif /* not RE_ENABLE_I18N */
- re_bitset_ptr_t sbcset;
- bracket_elem_t *start_elem, *end_elem;
- {
- unsigned int ch;
- uint32_t start_collseq;
- uint32_t end_collseq;
-
-# ifdef RE_ENABLE_I18N
- /* Check the space of the arrays. */
- if (*range_alloc == mbcset->nranges)
- {
- /* There are not enough space, need realloc. */
- uint32_t *new_array_start;
- uint32_t *new_array_end;
- int new_nranges;
-
- /* +1 in case of mbcset->nranges is 0. */
- new_nranges = 2 * mbcset->nranges + 1;
- /* Use realloc since mbcset->range_starts and mbcset->range_ends
- are NULL if *range_alloc == 0. */
- new_array_start = re_realloc (mbcset->range_starts, uint32_t,
- new_nranges);
- new_array_end = re_realloc (mbcset->range_ends, uint32_t,
- new_nranges);
-
- if (BE (new_array_start == NULL || new_array_end == NULL, 0))
- return REG_ESPACE;
-
- mbcset->range_starts = new_array_start;
- mbcset->range_ends = new_array_end;
- *range_alloc = new_nranges;
- }
-# endif /* RE_ENABLE_I18N */
-
- /* Equivalence Classes and Character Classes can't be a range
- start/end. */
- if (BE (start_elem->type == EQUIV_CLASS || start_elem->type == CHAR_CLASS
- || end_elem->type == EQUIV_CLASS || end_elem->type == CHAR_CLASS,
- 0))
- return REG_ERANGE;
-
- start_collseq = lookup_collation_sequence_value (start_elem);
- end_collseq = lookup_collation_sequence_value (end_elem);
- /* Check start/end collation sequence values. */
- if (BE (start_collseq == UINT_MAX || end_collseq == UINT_MAX, 0))
- return REG_ECOLLATE;
- if (BE ((syntax & RE_NO_EMPTY_RANGES) && start_collseq > end_collseq, 0))
- return REG_ERANGE;
-
-# ifdef RE_ENABLE_I18N
- /* Got valid collation sequence values, add them as a new entry. */
- mbcset->range_starts[mbcset->nranges] = start_collseq;
- mbcset->range_ends[mbcset->nranges++] = end_collseq;
-# endif /* RE_ENABLE_I18N */
-
- /* Build the table for single byte characters. */
- for (ch = 0; ch <= SBC_MAX; ch++)
- {
- uint32_t ch_collseq;
- /*
- if (MB_CUR_MAX == 1)
- */
- if (nrules == 0)
- ch_collseq = collseqmb[ch];
- else
- ch_collseq = collseq_table_lookup (collseqwc, __btowc (ch));
- if (start_collseq <= ch_collseq && ch_collseq <= end_collseq)
- bitset_set (sbcset, ch);
- }
- return REG_NOERROR;
- }
-
- /* Local function for parse_bracket_exp used in _LIBC environement.
- Build the collating element which is represented by NAME.
- The result are written to MBCSET and SBCSET.
- COLL_SYM_ALLOC is the allocated size of mbcset->coll_sym, is a
- pointer argument sinse we may update it. */
-
- static inline reg_errcode_t
-# ifdef RE_ENABLE_I18N
- build_collating_symbol (sbcset, mbcset, coll_sym_alloc, name)
- re_charset_t *mbcset;
- int *coll_sym_alloc;
-# else /* not RE_ENABLE_I18N */
- build_collating_symbol (sbcset, name)
-# endif /* not RE_ENABLE_I18N */
- re_bitset_ptr_t sbcset;
- const unsigned char *name;
- {
- int32_t elem, idx;
- size_t name_len = strlen ((const char *) name);
- if (nrules != 0)
- {
- elem = seek_collating_symbol_entry (name, name_len);
- if (symb_table[2 * elem] != 0)
- {
- /* We found the entry. */
- idx = symb_table[2 * elem + 1];
- /* Skip the name of collating element name. */
- idx += 1 + extra[idx];
- }
- else if (symb_table[2 * elem] == 0 && name_len == 1)
- {
- /* No valid character, treat it as a normal
- character. */
- bitset_set (sbcset, name[0]);
- return REG_NOERROR;
- }
- else
- return REG_ECOLLATE;
-
-# ifdef RE_ENABLE_I18N
- /* Got valid collation sequence, add it as a new entry. */
- /* Check the space of the arrays. */
- if (*coll_sym_alloc == mbcset->ncoll_syms)
- {
- /* Not enough, realloc it. */
- /* +1 in case of mbcset->ncoll_syms is 0. */
- *coll_sym_alloc = 2 * mbcset->ncoll_syms + 1;
- /* Use realloc since mbcset->coll_syms is NULL
- if *alloc == 0. */
- mbcset->coll_syms = re_realloc (mbcset->coll_syms, int32_t,
- *coll_sym_alloc);
- if (BE (mbcset->coll_syms == NULL, 0))
- return REG_ESPACE;
- }
- mbcset->coll_syms[mbcset->ncoll_syms++] = idx;
-# endif /* RE_ENABLE_I18N */
- return REG_NOERROR;
- }
- else
- {
- if (BE (name_len != 1, 0))
- return REG_ECOLLATE;
- else
- {
- bitset_set (sbcset, name[0]);
- return REG_NOERROR;
- }
- }
- }
-#endif
-
- re_token_t br_token;
- re_bitset_ptr_t sbcset;
-#ifdef RE_ENABLE_I18N
- re_charset_t *mbcset;
- int coll_sym_alloc = 0, range_alloc = 0, mbchar_alloc = 0;
- int equiv_class_alloc = 0, char_class_alloc = 0;
-#else /* not RE_ENABLE_I18N */
- int non_match = 0;
-#endif /* not RE_ENABLE_I18N */
- bin_tree_t *work_tree;
- int token_len, new_idx;
-#ifdef _LIBC
- collseqmb = (const unsigned char *)
- _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQMB);
- nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES);
- if (nrules)
- {
- /*
- if (MB_CUR_MAX > 1)
- */
- collseqwc = _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQWC);
- table_size = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_SYMB_HASH_SIZEMB);
- symb_table = (const int32_t *) _NL_CURRENT (LC_COLLATE,
- _NL_COLLATE_SYMB_TABLEMB);
- extra = (const unsigned char *) _NL_CURRENT (LC_COLLATE,
- _NL_COLLATE_SYMB_EXTRAMB);
- }
-#endif
- sbcset = (re_bitset_ptr_t) calloc (sizeof (unsigned int), BITSET_UINTS);
-#ifdef RE_ENABLE_I18N
- mbcset = (re_charset_t *) calloc (sizeof (re_charset_t), 1);
-#endif /* RE_ENABLE_I18N */
-#ifdef RE_ENABLE_I18N
- if (BE (sbcset == NULL || mbcset == NULL, 0))
-#else
- if (BE (sbcset == NULL, 0))
-#endif /* RE_ENABLE_I18N */
- {
- *err = REG_ESPACE;
- return NULL;
- }
-
- token_len = peek_token_bracket (token, regexp, syntax);
- if (BE (token->type == END_OF_RE, 0))
- {
- *err = REG_BADPAT;
- goto parse_bracket_exp_free_return;
- }
- if (token->type == OP_NON_MATCH_LIST)
- {
-#ifdef RE_ENABLE_I18N
- int i;
- mbcset->non_match = 1;
-#else /* not RE_ENABLE_I18N */
- non_match = 1;
-#endif /* not RE_ENABLE_I18N */
- if (syntax & RE_HAT_LISTS_NOT_NEWLINE)
- bitset_set (sbcset, '\0');
- re_string_skip_bytes (regexp, token_len); /* Skip a token. */
- token_len = peek_token_bracket (token, regexp, syntax);
- if (BE (token->type == END_OF_RE, 0))
- {
- *err = REG_BADPAT;
- goto parse_bracket_exp_free_return;
- }
-#ifdef RE_ENABLE_I18N
- if (MB_CUR_MAX > 1)
- for (i = 0; i < SBC_MAX; ++i)
- if (__btowc (i) == WEOF)
- bitset_set (sbcset, i);
-#endif /* RE_ENABLE_I18N */
- }
-
- /* We treat the first ']' as a normal character. */
- if (token->type == OP_CLOSE_BRACKET)
- token->type = CHARACTER;
-
- while (1)
- {
- bracket_elem_t start_elem, end_elem;
- unsigned char start_name_buf[BRACKET_NAME_BUF_SIZE];
- unsigned char end_name_buf[BRACKET_NAME_BUF_SIZE];
- reg_errcode_t ret;
- int token_len2 = 0, is_range_exp = 0;
- re_token_t token2;
-
- start_elem.opr.name = start_name_buf;
- ret = parse_bracket_element (&start_elem, regexp, token, token_len, dfa,
- syntax);
- if (BE (ret != REG_NOERROR, 0))
- {
- *err = ret;
- goto parse_bracket_exp_free_return;
- }
-
- token_len = peek_token_bracket (token, regexp, syntax);
- if (BE (token->type == END_OF_RE, 0))
- {
- *err = REG_BADPAT;
- goto parse_bracket_exp_free_return;
- }
- if (token->type == OP_CHARSET_RANGE)
- {
- re_string_skip_bytes (regexp, token_len); /* Skip '-'. */
- token_len2 = peek_token_bracket (&token2, regexp, syntax);
- if (BE (token->type == END_OF_RE, 0))
- {
- *err = REG_BADPAT;
- goto parse_bracket_exp_free_return;
- }
- if (token2.type == OP_CLOSE_BRACKET)
- {
- /* We treat the last '-' as a normal character. */
- re_string_skip_bytes (regexp, -token_len);
- token->type = CHARACTER;
- }
- else
- is_range_exp = 1;
- }
-
- if (is_range_exp == 1)
- {
- end_elem.opr.name = end_name_buf;
- ret = parse_bracket_element (&end_elem, regexp, &token2, token_len2,
- dfa, syntax);
- if (BE (ret != REG_NOERROR, 0))
- {
- *err = ret;
- goto parse_bracket_exp_free_return;
- }
-
- token_len = peek_token_bracket (token, regexp, syntax);
- if (BE (token->type == END_OF_RE, 0))
- {
- *err = REG_BADPAT;
- goto parse_bracket_exp_free_return;
- }
- *err = build_range_exp (sbcset,
-#ifdef RE_ENABLE_I18N
- mbcset, &range_alloc,
-#endif /* RE_ENABLE_I18N */
- &start_elem, &end_elem);
- if (BE (*err != REG_NOERROR, 0))
- goto parse_bracket_exp_free_return;
- }
- else
- {
- switch (start_elem.type)
- {
- case SB_CHAR:
- bitset_set (sbcset, start_elem.opr.ch);
- break;
-#ifdef RE_ENABLE_I18N
- case MB_CHAR:
- /* Check whether the array has enough space. */
- if (mbchar_alloc == mbcset->nmbchars)
- {
- /* Not enough, realloc it. */
- /* +1 in case of mbcset->nmbchars is 0. */
- mbchar_alloc = 2 * mbcset->nmbchars + 1;
- /* Use realloc since array is NULL if *alloc == 0. */
- mbcset->mbchars = re_realloc (mbcset->mbchars, wchar_t,
- mbchar_alloc);
- if (BE (mbcset->mbchars == NULL, 0))
- goto parse_bracket_exp_espace;
- }
- mbcset->mbchars[mbcset->nmbchars++] = start_elem.opr.wch;
- break;
-#endif /* RE_ENABLE_I18N */
- case EQUIV_CLASS:
- *err = build_equiv_class (sbcset,
-#ifdef RE_ENABLE_I18N
- mbcset, &equiv_class_alloc,
-#endif /* RE_ENABLE_I18N */
- start_elem.opr.name);
- if (BE (*err != REG_NOERROR, 0))
- goto parse_bracket_exp_free_return;
- break;
- case COLL_SYM:
- *err = build_collating_symbol (sbcset,
-#ifdef RE_ENABLE_I18N
- mbcset, &coll_sym_alloc,
-#endif /* RE_ENABLE_I18N */
- start_elem.opr.name);
- if (BE (*err != REG_NOERROR, 0))
- goto parse_bracket_exp_free_return;
- break;
- case CHAR_CLASS:
- *err = build_charclass (sbcset,
-#ifdef RE_ENABLE_I18N
- mbcset, &char_class_alloc,
-#endif /* RE_ENABLE_I18N */
- start_elem.opr.name, syntax);
- if (BE (*err != REG_NOERROR, 0))
- goto parse_bracket_exp_free_return;
- break;
- default:
- assert (0);
- break;
- }
- }
- if (token->type == OP_CLOSE_BRACKET)
- break;
- }
-
- re_string_skip_bytes (regexp, token_len); /* Skip a token. */
-
- /* If it is non-matching list. */
-#ifdef RE_ENABLE_I18N
- if (mbcset->non_match)
-#else /* not RE_ENABLE_I18N */
- if (non_match)
-#endif /* not RE_ENABLE_I18N */
- bitset_not (sbcset);
-
- /* Build a tree for simple bracket. */
- br_token.type = SIMPLE_BRACKET;
- br_token.opr.sbcset = sbcset;
- new_idx = re_dfa_add_node (dfa, br_token, 0);
- work_tree = create_tree (NULL, NULL, 0, new_idx);
- if (BE (new_idx == -1 || work_tree == NULL, 0))
- goto parse_bracket_exp_espace;
-
-#ifdef RE_ENABLE_I18N
- if (mbcset->nmbchars || mbcset->ncoll_syms || mbcset->nequiv_classes
- || mbcset->nranges || (MB_CUR_MAX > 1 && (mbcset->nchar_classes
- || mbcset->non_match)))
- {
- re_token_t alt_token;
- bin_tree_t *mbc_tree;
- /* Build a tree for complex bracket. */
- br_token.type = COMPLEX_BRACKET;
- br_token.opr.mbcset = mbcset;
- dfa->has_mb_node = 1;
- new_idx = re_dfa_add_node (dfa, br_token, 0);
- mbc_tree = create_tree (NULL, NULL, 0, new_idx);
- if (BE (new_idx == -1 || mbc_tree == NULL, 0))
- goto parse_bracket_exp_espace;
- /* Then join them by ALT node. */
- dfa->has_plural_match = 1;
- alt_token.type = OP_ALT;
- new_idx = re_dfa_add_node (dfa, alt_token, 0);
- work_tree = create_tree (work_tree, mbc_tree, 0, new_idx);
- if (BE (new_idx != -1 && mbc_tree != NULL, 1))
- return work_tree;
- }
- else
- {
- free_charset (mbcset);
- return work_tree;
- }
-#else /* not RE_ENABLE_I18N */
- return work_tree;
-#endif /* not RE_ENABLE_I18N */
-
- parse_bracket_exp_espace:
- *err = REG_ESPACE;
- parse_bracket_exp_free_return:
- re_free (sbcset);
-#ifdef RE_ENABLE_I18N
- free_charset (mbcset);
-#endif /* RE_ENABLE_I18N */
- return NULL;
-}
-
-/* Parse an element in the bracket expression. */
-
-static reg_errcode_t
-parse_bracket_element (elem, regexp, token, token_len, dfa, syntax)
- bracket_elem_t *elem;
- re_string_t *regexp;
- re_token_t *token;
- int token_len;
- re_dfa_t *dfa;
- reg_syntax_t syntax;
-{
-#ifdef RE_ENABLE_I18N
- int cur_char_size;
- cur_char_size = re_string_char_size_at (regexp, re_string_cur_idx (regexp));
- if (cur_char_size > 1)
- {
- elem->type = MB_CHAR;
- elem->opr.wch = re_string_wchar_at (regexp, re_string_cur_idx (regexp));
- re_string_skip_bytes (regexp, cur_char_size);
- return REG_NOERROR;
- }
-#endif /* RE_ENABLE_I18N */
- re_string_skip_bytes (regexp, token_len); /* Skip a token. */
- if (token->type == OP_OPEN_COLL_ELEM || token->type == OP_OPEN_CHAR_CLASS
- || token->type == OP_OPEN_EQUIV_CLASS)
- return parse_bracket_symbol (elem, regexp, token);
- elem->type = SB_CHAR;
- elem->opr.ch = token->opr.c;
- return REG_NOERROR;
-}
-
-/* Parse a bracket symbol in the bracket expression. Bracket symbols are
- such as [:<character_class>:], [.<collating_element>.], and
- [=<equivalent_class>=]. */
-
-static reg_errcode_t
-parse_bracket_symbol (elem, regexp, token)
- bracket_elem_t *elem;
- re_string_t *regexp;
- re_token_t *token;
-{
- unsigned char ch, delim = token->opr.c;
- int i = 0;
- for (;; ++i)
- {
- if (re_string_eoi(regexp) || i >= BRACKET_NAME_BUF_SIZE)
- return REG_EBRACK;
- if (token->type == OP_OPEN_CHAR_CLASS)
- ch = re_string_fetch_byte_case (regexp);
- else
- ch = re_string_fetch_byte (regexp);
- if (ch == delim && re_string_peek_byte (regexp, 0) == ']')
- break;
- elem->opr.name[i] = ch;
- }
- re_string_skip_bytes (regexp, 1);
- elem->opr.name[i] = '\0';
- switch (token->type)
- {
- case OP_OPEN_COLL_ELEM:
- elem->type = COLL_SYM;
- break;
- case OP_OPEN_EQUIV_CLASS:
- elem->type = EQUIV_CLASS;
- break;
- case OP_OPEN_CHAR_CLASS:
- elem->type = CHAR_CLASS;
- break;
- default:
- break;
- }
- return REG_NOERROR;
-}
-
- /* Helper function for parse_bracket_exp.
- Build the equivalence class which is represented by NAME.
- The result are written to MBCSET and SBCSET.
- EQUIV_CLASS_ALLOC is the allocated size of mbcset->equiv_classes,
- is a pointer argument sinse we may update it. */
-
-static reg_errcode_t
-#ifdef RE_ENABLE_I18N
-build_equiv_class (sbcset, mbcset, equiv_class_alloc, name)
- re_charset_t *mbcset;
- int *equiv_class_alloc;
-#else /* not RE_ENABLE_I18N */
-build_equiv_class (sbcset, name)
-#endif /* not RE_ENABLE_I18N */
- re_bitset_ptr_t sbcset;
- const unsigned char *name;
-{
-#if defined _LIBC && defined RE_ENABLE_I18N
- uint32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES);
- if (nrules != 0)
- {
- const int32_t *table, *indirect;
- const unsigned char *weights, *extra, *cp;
- unsigned char char_buf[2];
- int32_t idx1, idx2;
- unsigned int ch;
- size_t len;
- /* This #include defines a local function! */
-# include <locale/weight.h>
- /* Calculate the index for equivalence class. */
- cp = name;
- table = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB);
- weights = (const unsigned char *) _NL_CURRENT (LC_COLLATE,
- _NL_COLLATE_WEIGHTMB);
- extra = (const unsigned char *) _NL_CURRENT (LC_COLLATE,
- _NL_COLLATE_EXTRAMB);
- indirect = (const int32_t *) _NL_CURRENT (LC_COLLATE,
- _NL_COLLATE_INDIRECTMB);
- idx1 = findidx (&cp);
- if (BE (idx1 == 0 || cp < name + strlen ((const char *) name), 0))
- /* This isn't a valid character. */
- return REG_ECOLLATE;
-
- /* Build single byte matcing table for this equivalence class. */
- char_buf[1] = (unsigned char) '\0';
- len = weights[idx1];
- for (ch = 0; ch < SBC_MAX; ++ch)
- {
- char_buf[0] = ch;
- cp = char_buf;
- idx2 = findidx (&cp);
-/*
- idx2 = table[ch];
-*/
- if (idx2 == 0)
- /* This isn't a valid character. */
- continue;
- if (len == weights[idx2])
- {
- int cnt = 0;
- while (cnt <= len &&
- weights[idx1 + 1 + cnt] == weights[idx2 + 1 + cnt])
- ++cnt;
-
- if (cnt > len)
- bitset_set (sbcset, ch);
- }
- }
- /* Check whether the array has enough space. */
- if (*equiv_class_alloc == mbcset->nequiv_classes)
- {
- /* Not enough, realloc it. */
- /* +1 in case of mbcset->nequiv_classes is 0. */
- *equiv_class_alloc = 2 * mbcset->nequiv_classes + 1;
- /* Use realloc since the array is NULL if *alloc == 0. */
- mbcset->equiv_classes = re_realloc (mbcset->equiv_classes, int32_t,
- *equiv_class_alloc);
- if (BE (mbcset->equiv_classes == NULL, 0))
- return REG_ESPACE;
- }
- mbcset->equiv_classes[mbcset->nequiv_classes++] = idx1;
- }
- else
-#endif /* _LIBC && RE_ENABLE_I18N */
- {
- if (BE (strlen ((const char *) name) != 1, 0))
- return REG_ECOLLATE;
- bitset_set (sbcset, *name);
- }
- return REG_NOERROR;
-}
-
- /* Helper function for parse_bracket_exp.
- Build the character class which is represented by NAME.
- The result are written to MBCSET and SBCSET.
- CHAR_CLASS_ALLOC is the allocated size of mbcset->char_classes,
- is a pointer argument sinse we may update it. */
-
-static reg_errcode_t
-#ifdef RE_ENABLE_I18N
-build_charclass (sbcset, mbcset, char_class_alloc, class_name, syntax)
- re_charset_t *mbcset;
- int *char_class_alloc;
-#else /* not RE_ENABLE_I18N */
-build_charclass (sbcset, class_name, syntax)
-#endif /* not RE_ENABLE_I18N */
- re_bitset_ptr_t sbcset;
- const unsigned char *class_name;
- reg_syntax_t syntax;
-{
- int i;
- const char *name = (const char *) class_name;
-
- /* In case of REG_ICASE "upper" and "lower" match the both of
- upper and lower cases. */
- if ((syntax & RE_ICASE)
- && (strcmp (name, "upper") == 0 || strcmp (name, "lower") == 0))
- name = "alpha";
-
-#ifdef RE_ENABLE_I18N
- /* Check the space of the arrays. */
- if (*char_class_alloc == mbcset->nchar_classes)
- {
- /* Not enough, realloc it. */
- /* +1 in case of mbcset->nchar_classes is 0. */
- *char_class_alloc = 2 * mbcset->nchar_classes + 1;
- /* Use realloc since array is NULL if *alloc == 0. */
- mbcset->char_classes = re_realloc (mbcset->char_classes, wctype_t,
- *char_class_alloc);
- if (BE (mbcset->char_classes == NULL, 0))
- return REG_ESPACE;
- }
- mbcset->char_classes[mbcset->nchar_classes++] = __wctype (name);
-#endif /* RE_ENABLE_I18N */
-
-#define BUILD_CHARCLASS_LOOP(ctype_func)\
- for (i = 0; i < SBC_MAX; ++i) \
- { \
- if (ctype_func (i)) \
- bitset_set (sbcset, i); \
- }
-
- if (strcmp (name, "alnum") == 0)
- BUILD_CHARCLASS_LOOP (isalnum)
- else if (strcmp (name, "cntrl") == 0)
- BUILD_CHARCLASS_LOOP (iscntrl)
- else if (strcmp (name, "lower") == 0)
- BUILD_CHARCLASS_LOOP (islower)
- else if (strcmp (name, "space") == 0)
- BUILD_CHARCLASS_LOOP (isspace)
- else if (strcmp (name, "alpha") == 0)
- BUILD_CHARCLASS_LOOP (isalpha)
- else if (strcmp (name, "digit") == 0)
- BUILD_CHARCLASS_LOOP (isdigit)
- else if (strcmp (name, "print") == 0)
- BUILD_CHARCLASS_LOOP (isprint)
- else if (strcmp (name, "upper") == 0)
- BUILD_CHARCLASS_LOOP (isupper)
- else if (strcmp (name, "blank") == 0)
- BUILD_CHARCLASS_LOOP (isblank)
- else if (strcmp (name, "graph") == 0)
- BUILD_CHARCLASS_LOOP (isgraph)
- else if (strcmp (name, "punct") == 0)
- BUILD_CHARCLASS_LOOP (ispunct)
- else if (strcmp (name, "xdigit") == 0)
- BUILD_CHARCLASS_LOOP (isxdigit)
- else
- return REG_ECTYPE;
-
- return REG_NOERROR;
-}
-
-static bin_tree_t *
-build_word_op (dfa, not, err)
- re_dfa_t *dfa;
- int not;
- reg_errcode_t *err;
-{
- re_bitset_ptr_t sbcset;
-#ifdef RE_ENABLE_I18N
- re_charset_t *mbcset;
- int alloc = 0;
-#else /* not RE_ENABLE_I18N */
- int non_match = 0;
-#endif /* not RE_ENABLE_I18N */
- reg_errcode_t ret;
- re_token_t br_token;
- bin_tree_t *tree;
- int new_idx;
-
- sbcset = (re_bitset_ptr_t) calloc (sizeof (unsigned int), BITSET_UINTS);
-#ifdef RE_ENABLE_I18N
- mbcset = (re_charset_t *) calloc (sizeof (re_charset_t), 1);
-#endif /* RE_ENABLE_I18N */
-
-#ifdef RE_ENABLE_I18N
- if (BE (sbcset == NULL || mbcset == NULL, 0))
-#else /* not RE_ENABLE_I18N */
- if (BE (sbcset == NULL, 0))
-#endif /* not RE_ENABLE_I18N */
- {
- *err = REG_ESPACE;
- return NULL;
- }
-
- if (not)
- {
-#ifdef RE_ENABLE_I18N
- int i;
- /*
- if (syntax & RE_HAT_LISTS_NOT_NEWLINE)
- bitset_set(cset->sbcset, '\0');
- */
- mbcset->non_match = 1;
- if (MB_CUR_MAX > 1)
- for (i = 0; i < SBC_MAX; ++i)
- if (__btowc (i) == WEOF)
- bitset_set (sbcset, i);
-#else /* not RE_ENABLE_I18N */
- non_match = 1;
-#endif /* not RE_ENABLE_I18N */
- }
-
- /* We don't care the syntax in this case. */
- ret = build_charclass (sbcset,
-#ifdef RE_ENABLE_I18N
- mbcset, &alloc,
-#endif /* RE_ENABLE_I18N */
- (const unsigned char *) "alpha", 0);
-
- if (BE (ret != REG_NOERROR, 0))
- {
- re_free (sbcset);
-#ifdef RE_ENABLE_I18N
- free_charset (mbcset);
-#endif /* RE_ENABLE_I18N */
- *err = ret;
- return NULL;
- }
- /* \w match '_' also. */
- bitset_set (sbcset, '_');
-
- /* If it is non-matching list. */
-#ifdef RE_ENABLE_I18N
- if (mbcset->non_match)
-#else /* not RE_ENABLE_I18N */
- if (non_match)
-#endif /* not RE_ENABLE_I18N */
- bitset_not (sbcset);
-
- /* Build a tree for simple bracket. */
- br_token.type = SIMPLE_BRACKET;
- br_token.opr.sbcset = sbcset;
- new_idx = re_dfa_add_node (dfa, br_token, 0);
- tree = create_tree (NULL, NULL, 0, new_idx);
- if (BE (new_idx == -1 || tree == NULL, 0))
- goto build_word_op_espace;
-
-#ifdef RE_ENABLE_I18N
- if (MB_CUR_MAX > 1)
- {
- re_token_t alt_token;
- bin_tree_t *mbc_tree;
- /* Build a tree for complex bracket. */
- br_token.type = COMPLEX_BRACKET;
- br_token.opr.mbcset = mbcset;
- dfa->has_mb_node = 1;
- new_idx = re_dfa_add_node (dfa, br_token, 0);
- mbc_tree = create_tree (NULL, NULL, 0, new_idx);
- if (BE (new_idx == -1 || mbc_tree == NULL, 0))
- goto build_word_op_espace;
- /* Then join them by ALT node. */
- alt_token.type = OP_ALT;
- new_idx = re_dfa_add_node (dfa, alt_token, 0);
- tree = create_tree (tree, mbc_tree, 0, new_idx);
- if (BE (new_idx != -1 && mbc_tree != NULL, 1))
- return tree;
- }
- else
- {
- free_charset (mbcset);
- return tree;
- }
-#else /* not RE_ENABLE_I18N */
- return tree;
-#endif /* not RE_ENABLE_I18N */
-
- build_word_op_espace:
- re_free (sbcset);
-#ifdef RE_ENABLE_I18N
- free_charset (mbcset);
-#endif /* RE_ENABLE_I18N */
- *err = REG_ESPACE;
- return NULL;
-}
-
-/* This is intended for the expressions like "a{1,3}".
- Fetch a number from `input', and return the number.
- Return -1, if the number field is empty like "{,1}".
- Return -2, If an error is occured. */
-
-static int
-fetch_number (input, token, syntax)
- re_string_t *input;
- re_token_t *token;
- reg_syntax_t syntax;
-{
- int num = -1;
- unsigned char c;
- while (1)
- {
- *token = fetch_token (input, syntax);
- c = token->opr.c;
- if (BE (token->type == END_OF_RE, 0))
- return -2;
- if (token->type == OP_CLOSE_DUP_NUM || c == ',')
- break;
- num = ((token->type != CHARACTER || c < '0' || '9' < c || num == -2)
- ? -2 : ((num == -1) ? c - '0' : num * 10 + c - '0'));
- num = (num > RE_DUP_MAX) ? -2 : num;
- }
- return num;
-}
-\f
-#ifdef RE_ENABLE_I18N
-static void
-free_charset (re_charset_t *cset)
-{
- re_free (cset->mbchars);
-# ifdef _LIBC
- re_free (cset->coll_syms);
- re_free (cset->equiv_classes);
- re_free (cset->range_starts);
- re_free (cset->range_ends);
-# endif
- re_free (cset->char_classes);
- re_free (cset);
-}
-#endif /* RE_ENABLE_I18N */
-\f
-/* Functions for binary tree operation. */
-
-/* Create a node of tree.
- Note: This function automatically free left and right if malloc fails. */
-
-static bin_tree_t *
-create_tree (left, right, type, index)
- bin_tree_t *left;
- bin_tree_t *right;
- re_token_type_t type;
- int index;
-{
- bin_tree_t *tree;
- tree = re_malloc (bin_tree_t, 1);
- if (BE (tree == NULL, 0))
- {
- free_bin_tree (left);
- free_bin_tree (right);
- return NULL;
- }
- tree->parent = NULL;
- tree->left = left;
- tree->right = right;
- tree->type = type;
- tree->node_idx = index;
- tree->first = -1;
- tree->next = -1;
- re_node_set_init_empty (&tree->eclosure);
-
- if (left != NULL)
- left->parent = tree;
- if (right != NULL)
- right->parent = tree;
- return tree;
-}
-
-/* Free the sub tree pointed by TREE. */
-
-static void
-free_bin_tree (tree)
- bin_tree_t *tree;
-{
- if (tree == NULL)
- return;
- /*re_node_set_free (&tree->eclosure);*/
- free_bin_tree (tree->left);
- free_bin_tree (tree->right);
- re_free (tree);
-}
-
-/* Duplicate the node SRC, and return new node. */
-
-static bin_tree_t *
-duplicate_tree (src, dfa)
- const bin_tree_t *src;
- re_dfa_t *dfa;
-{
- bin_tree_t *left = NULL, *right = NULL, *new_tree;
- int new_node_idx;
- /* Since node indies must be according to Post-order of the tree,
- we must duplicate the left at first. */
- if (src->left != NULL)
- {
- left = duplicate_tree (src->left, dfa);
- if (left == NULL)
- return NULL;
- }
-
- /* Secondaly, duplicate the right. */
- if (src->right != NULL)
- {
- right = duplicate_tree (src->right, dfa);
- if (right == NULL)
- {
- free_bin_tree (left);
- return NULL;
- }
- }
-
- /* At last, duplicate itself. */
- if (src->type == NON_TYPE)
- {
- new_node_idx = re_dfa_add_node (dfa, dfa->nodes[src->node_idx], 0);
- dfa->nodes[new_node_idx].duplicated = 1;
- if (BE (new_node_idx == -1, 0))
- {
- free_bin_tree (left);
- free_bin_tree (right);
- return NULL;
- }
- }
- else
- new_node_idx = src->type;
-
- new_tree = create_tree (left, right, src->type, new_node_idx);
- if (BE (new_tree == NULL, 0))
- {
- free_bin_tree (left);
- free_bin_tree (right);
- }
- return new_tree;
-}
+++ /dev/null
-/* Extended regular expression matching and search library.
- Copyright (C) 2002, 2003 Free Software Foundation, Inc.
- This file is part of the GNU C Library.
- Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
-
- The GNU C Library is free software; you can redistribute it and/or
- modify it under the terms of the GNU Lesser General Public
- License as published by the Free Software Foundation; either
- version 2.1 of the License, or (at your option) any later version.
-
- The GNU C Library is distributed in the hope that it will be useful,
- but WITHOUT ANY WARRANTY; without even the implied warranty of
- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- Lesser General Public License for more details.
-
- You should have received a copy of the GNU Lesser General Public
- License along with the GNU C Library; if not, write to the Free
- Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
- 02111-1307 USA. */
-
-#ifdef _LIBC
-/* We have to keep the namespace clean. */
-# define regfree(preg) __regfree (preg)
-# define regexec(pr, st, nm, pm, ef) __regexec (pr, st, nm, pm, ef)
-# define regcomp(preg, pattern, cflags) __regcomp (preg, pattern, cflags)
-# define regerror(errcode, preg, errbuf, errbuf_size) \
- __regerror(errcode, preg, errbuf, errbuf_size)
-# define re_set_registers(bu, re, nu, st, en) \
- __re_set_registers (bu, re, nu, st, en)
-# define re_match_2(bufp, string1, size1, string2, size2, pos, regs, stop) \
- __re_match_2 (bufp, string1, size1, string2, size2, pos, regs, stop)
-# define re_match(bufp, string, size, pos, regs) \
- __re_match (bufp, string, size, pos, regs)
-# define re_search(bufp, string, size, startpos, range, regs) \
- __re_search (bufp, string, size, startpos, range, regs)
-# define re_compile_pattern(pattern, length, bufp) \
- __re_compile_pattern (pattern, length, bufp)
-# define re_set_syntax(syntax) __re_set_syntax (syntax)
-# define re_search_2(bufp, st1, s1, st2, s2, startpos, range, regs, stop) \
- __re_search_2 (bufp, st1, s1, st2, s2, startpos, range, regs, stop)
-# define re_compile_fastmap(bufp) __re_compile_fastmap (bufp)
-#endif
-
-/* POSIX says that <sys/types.h> must be included (by the caller) before
- <regex.h>. */
-#include <sys/types.h>
-#include <regex.h>
-#include "regex_internal.h"
-
-#include "regex_internal.c"
-#include "regcomp.c"
-#include "regexec.c"
-
-/* Binary backward compatibility. */
-#if _LIBC
-# include <shlib-compat.h>
-# if SHLIB_COMPAT (libc, GLIBC_2_0, GLIBC_2_3)
-link_warning (re_max_failures, "the 're_max_failures' variable is obsolete and will go away.")
-int re_max_failures = 2000;
-# endif
-#endif
+++ /dev/null
-/* Definitions for data structures and routines for the regular
- expression library.
- Copyright (C) 1985,1989-93,1995-98,2000,2001,2002
- Free Software Foundation, Inc.
- This file is part of the GNU C Library.
-
- The GNU C Library is free software; you can redistribute it and/or
- modify it under the terms of the GNU Lesser General Public
- License as published by the Free Software Foundation; either
- version 2.1 of the License, or (at your option) any later version.
-
- The GNU C Library is distributed in the hope that it will be useful,
- but WITHOUT ANY WARRANTY; without even the implied warranty of
- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- Lesser General Public License for more details.
-
- You should have received a copy of the GNU Lesser General Public
- License along with the GNU C Library; if not, write to the Free
- Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
- 02111-1307 USA. */
-
-#ifndef _REGEX_H
-#define _REGEX_H 1
-
-/* Allow the use in C++ code. */
-#ifdef __cplusplus
-extern "C" {
-#endif
-
-/* POSIX says that <sys/types.h> must be included (by the caller) before
- <regex.h>. */
-
-#if !defined _POSIX_C_SOURCE && !defined _POSIX_SOURCE && (defined VMS || defined _MSC_VER)
-/* VMS doesn't have `size_t' in <sys/types.h>, even though POSIX says it
- should be there. Same for Microsoft Visual C++ 6.0 */
-# include <stddef.h>
-#endif
-
-/* The following two types have to be signed and unsigned integer type
- wide enough to hold a value of a pointer. For most ANSI compilers
- ptrdiff_t and size_t should be likely OK. Still size of these two
- types is 2 for Microsoft C. Ugh... */
-typedef long int s_reg_t;
-typedef unsigned long int active_reg_t;
-
-/* The following bits are used to determine the regexp syntax we
- recognize. The set/not-set meanings are chosen so that Emacs syntax
- remains the value 0. The bits are given in alphabetical order, and
- the definitions shifted by one from the previous bit; thus, when we
- add or remove a bit, only one other definition need change. */
-typedef unsigned long int reg_syntax_t;
-
-/* If this bit is not set, then \ inside a bracket expression is literal.
- If set, then such a \ quotes the following character. */
-#define RE_BACKSLASH_ESCAPE_IN_LISTS ((unsigned long int) 1)
-
-/* If this bit is not set, then + and ? are operators, and \+ and \? are
- literals.
- If set, then \+ and \? are operators and + and ? are literals. */
-#define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1)
-
-/* If this bit is set, then character classes are supported. They are:
- [:alpha:], [:upper:], [:lower:], [:digit:], [:alnum:], [:xdigit:],
- [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:].
- If not set, then character classes are not supported. */
-#define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1)
-
-/* If this bit is set, then ^ and $ are always anchors (outside bracket
- expressions, of course).
- If this bit is not set, then it depends:
- ^ is an anchor if it is at the beginning of a regular
- expression or after an open-group or an alternation operator;
- $ is an anchor if it is at the end of a regular expression, or
- before a close-group or an alternation operator.
-
- This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because
- POSIX draft 11.2 says that * etc. in leading positions is undefined.
- We already implemented a previous draft which made those constructs
- invalid, though, so we haven't changed the code back. */
-#define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1)
-
-/* If this bit is set, then special characters are always special
- regardless of where they are in the pattern.
- If this bit is not set, then special characters are special only in
- some contexts; otherwise they are ordinary. Specifically,
- * + ? and intervals are only special when not after the beginning,
- open-group, or alternation operator. */
-#define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1)
-
-/* If this bit is set, then *, +, ?, and { cannot be first in an re or
- immediately after an alternation or begin-group operator. */
-#define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1)
-
-/* If this bit is set, then . matches newline.
- If not set, then it doesn't. */
-#define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1)
-
-/* If this bit is set, then . doesn't match NUL.
- If not set, then it does. */
-#define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1)
-
-/* If this bit is set, nonmatching lists [^...] do not match newline.
- If not set, they do. */
-#define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1)
-
-/* If this bit is set, either \{...\} or {...} defines an
- interval, depending on RE_NO_BK_BRACES.
- If not set, \{, \}, {, and } are literals. */
-#define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1)
-
-/* If this bit is set, +, ? and | aren't recognized as operators.
- If not set, they are. */
-#define RE_LIMITED_OPS (RE_INTERVALS << 1)
-
-/* If this bit is set, newline is an alternation operator.
- If not set, newline is literal. */
-#define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1)
-
-/* If this bit is set, then `{...}' defines an interval, and \{ and \}
- are literals.
- If not set, then `\{...\}' defines an interval. */
-#define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1)
-
-/* If this bit is set, (...) defines a group, and \( and \) are literals.
- If not set, \(...\) defines a group, and ( and ) are literals. */
-#define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1)
-
-/* If this bit is set, then \<digit> matches <digit>.
- If not set, then \<digit> is a back-reference. */
-#define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1)
-
-/* If this bit is set, then | is an alternation operator, and \| is literal.
- If not set, then \| is an alternation operator, and | is literal. */
-#define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1)
-
-/* If this bit is set, then an ending range point collating higher
- than the starting range point, as in [z-a], is invalid.
- If not set, then when ending range point collates higher than the
- starting range point, the range is ignored. */
-#define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1)
-
-/* If this bit is set, then an unmatched ) is ordinary.
- If not set, then an unmatched ) is invalid. */
-#define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1)
-
-/* If this bit is set, succeed as soon as we match the whole pattern,
- without further backtracking. */
-#define RE_NO_POSIX_BACKTRACKING (RE_UNMATCHED_RIGHT_PAREN_ORD << 1)
-
-/* If this bit is set, do not process the GNU regex operators.
- If not set, then the GNU regex operators are recognized. */
-#define RE_NO_GNU_OPS (RE_NO_POSIX_BACKTRACKING << 1)
-
-/* If this bit is set, turn on internal regex debugging.
- If not set, and debugging was on, turn it off.
- This only works if regex.c is compiled -DDEBUG.
- We define this bit always, so that all that's needed to turn on
- debugging is to recompile regex.c; the calling code can always have
- this bit set, and it won't affect anything in the normal case. */
-#define RE_DEBUG (RE_NO_GNU_OPS << 1)
-
-/* If this bit is set, a syntactically invalid interval is treated as
- a string of ordinary characters. For example, the ERE 'a{1' is
- treated as 'a\{1'. */
-#define RE_INVALID_INTERVAL_ORD (RE_DEBUG << 1)
-
-/* If this bit is set, then ignore case when matching.
- If not set, then case is significant. */
-#define RE_ICASE (RE_INVALID_INTERVAL_ORD << 1)
-
-/* This global variable defines the particular regexp syntax to use (for
- some interfaces). When a regexp is compiled, the syntax used is
- stored in the pattern buffer, so changing this does not affect
- already-compiled regexps. */
-extern reg_syntax_t re_syntax_options;
-\f
-/* Define combinations of the above bits for the standard possibilities.
- (The [[[ comments delimit what gets put into the Texinfo file, so
- don't delete them!) */
-/* [[[begin syntaxes]]] */
-#define RE_SYNTAX_EMACS 0
-
-#define RE_SYNTAX_AWK \
- (RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL \
- | RE_NO_BK_PARENS | RE_NO_BK_REFS \
- | RE_NO_BK_VBAR | RE_NO_EMPTY_RANGES \
- | RE_DOT_NEWLINE | RE_CONTEXT_INDEP_ANCHORS \
- | RE_UNMATCHED_RIGHT_PAREN_ORD | RE_NO_GNU_OPS)
-
-#define RE_SYNTAX_GNU_AWK \
- ((RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DEBUG) \
- & ~(RE_DOT_NOT_NULL | RE_INTERVALS | RE_CONTEXT_INDEP_OPS \
- | RE_CONTEXT_INVALID_OPS ))
-
-#define RE_SYNTAX_POSIX_AWK \
- (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS \
- | RE_INTERVALS | RE_NO_GNU_OPS)
-
-#define RE_SYNTAX_GREP \
- (RE_BK_PLUS_QM | RE_CHAR_CLASSES \
- | RE_HAT_LISTS_NOT_NEWLINE | RE_INTERVALS \
- | RE_NEWLINE_ALT)
-
-#define RE_SYNTAX_EGREP \
- (RE_CHAR_CLASSES | RE_CONTEXT_INDEP_ANCHORS \
- | RE_CONTEXT_INDEP_OPS | RE_HAT_LISTS_NOT_NEWLINE \
- | RE_NEWLINE_ALT | RE_NO_BK_PARENS \
- | RE_NO_BK_VBAR)
-
-#define RE_SYNTAX_POSIX_EGREP \
- (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES \
- | RE_INVALID_INTERVAL_ORD)
-
-/* P1003.2/D11.2, section 4.20.7.1, lines 5078ff. */
-#define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC
-
-#define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC
-
-/* Syntax bits common to both basic and extended POSIX regex syntax. */
-#define _RE_SYNTAX_POSIX_COMMON \
- (RE_CHAR_CLASSES | RE_DOT_NEWLINE | RE_DOT_NOT_NULL \
- | RE_INTERVALS | RE_NO_EMPTY_RANGES)
-
-#define RE_SYNTAX_POSIX_BASIC \
- (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM)
-
-/* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes
- RE_LIMITED_OPS, i.e., \? \+ \| are not recognized. Actually, this
- isn't minimal, since other operators, such as \`, aren't disabled. */
-#define RE_SYNTAX_POSIX_MINIMAL_BASIC \
- (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS)
-
-#define RE_SYNTAX_POSIX_EXTENDED \
- (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \
- | RE_CONTEXT_INDEP_OPS | RE_NO_BK_BRACES \
- | RE_NO_BK_PARENS | RE_NO_BK_VBAR \
- | RE_CONTEXT_INVALID_OPS | RE_UNMATCHED_RIGHT_PAREN_ORD)
-
-/* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INDEP_OPS is
- removed and RE_NO_BK_REFS is added. */
-#define RE_SYNTAX_POSIX_MINIMAL_EXTENDED \
- (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \
- | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES \
- | RE_NO_BK_PARENS | RE_NO_BK_REFS \
- | RE_NO_BK_VBAR | RE_UNMATCHED_RIGHT_PAREN_ORD)
-/* [[[end syntaxes]]] */
-\f
-/* Maximum number of duplicates an interval can allow. Some systems
- (erroneously) define this in other header files, but we want our
- value, so remove any previous define. */
-#ifdef RE_DUP_MAX
-# undef RE_DUP_MAX
-#endif
-/* If sizeof(int) == 2, then ((1 << 15) - 1) overflows. */
-#define RE_DUP_MAX (0x7fff)
-
-
-/* POSIX `cflags' bits (i.e., information for `regcomp'). */
-
-/* If this bit is set, then use extended regular expression syntax.
- If not set, then use basic regular expression syntax. */
-#define REG_EXTENDED 1
-
-/* If this bit is set, then ignore case when matching.
- If not set, then case is significant. */
-#define REG_ICASE (REG_EXTENDED << 1)
-
-/* If this bit is set, then anchors do not match at newline
- characters in the string.
- If not set, then anchors do match at newlines. */
-#define REG_NEWLINE (REG_ICASE << 1)
-
-/* If this bit is set, then report only success or fail in regexec.
- If not set, then returns differ between not matching and errors. */
-#define REG_NOSUB (REG_NEWLINE << 1)
-
-
-/* POSIX `eflags' bits (i.e., information for regexec). */
-
-/* If this bit is set, then the beginning-of-line operator doesn't match
- the beginning of the string (presumably because it's not the
- beginning of a line).
- If not set, then the beginning-of-line operator does match the
- beginning of the string. */
-#define REG_NOTBOL 1
-
-/* Like REG_NOTBOL, except for the end-of-line. */
-#define REG_NOTEOL (1 << 1)
-
-
-/* If any error codes are removed, changed, or added, update the
- `re_error_msg' table in regex.c. */
-typedef enum
-{
-#ifdef _XOPEN_SOURCE
- REG_ENOSYS = -1, /* This will never happen for this implementation. */
-#endif
-
- REG_NOERROR = 0, /* Success. */
- REG_NOMATCH, /* Didn't find a match (for regexec). */
-
- /* POSIX regcomp return error codes. (In the order listed in the
- standard.) */
- REG_BADPAT, /* Invalid pattern. */
- REG_ECOLLATE, /* Not implemented. */
- REG_ECTYPE, /* Invalid character class name. */
- REG_EESCAPE, /* Trailing backslash. */
- REG_ESUBREG, /* Invalid back reference. */
- REG_EBRACK, /* Unmatched left bracket. */
- REG_EPAREN, /* Parenthesis imbalance. */
- REG_EBRACE, /* Unmatched \{. */
- REG_BADBR, /* Invalid contents of \{\}. */
- REG_ERANGE, /* Invalid range end. */
- REG_ESPACE, /* Ran out of memory. */
- REG_BADRPT, /* No preceding re for repetition op. */
-
- /* Error codes we've added. */
- REG_EEND, /* Premature end. */
- REG_ESIZE, /* Compiled pattern bigger than 2^16 bytes. */
- REG_ERPAREN /* Unmatched ) or \); not returned from regcomp. */
-} reg_errcode_t;
-\f
-/* This data structure represents a compiled pattern. Before calling
- the pattern compiler, the fields `buffer', `allocated', `fastmap',
- `translate', and `no_sub' can be set. After the pattern has been
- compiled, the `re_nsub' field is available. All other fields are
- private to the regex routines. */
-
-#ifndef RE_TRANSLATE_TYPE
-# define RE_TRANSLATE_TYPE char *
-#endif
-
-struct re_pattern_buffer
-{
-/* [[[begin pattern_buffer]]] */
- /* Space that holds the compiled pattern. It is declared as
- `unsigned char *' because its elements are
- sometimes used as array indexes. */
- unsigned char *buffer;
-
- /* Number of bytes to which `buffer' points. */
- unsigned long int allocated;
-
- /* Number of bytes actually used in `buffer'. */
- unsigned long int used;
-
- /* Syntax setting with which the pattern was compiled. */
- reg_syntax_t syntax;
-
- /* Pointer to a fastmap, if any, otherwise zero. re_search uses
- the fastmap, if there is one, to skip over impossible
- starting points for matches. */
- char *fastmap;
-
- /* Either a translate table to apply to all characters before
- comparing them, or zero for no translation. The translation
- is applied to a pattern when it is compiled and to a string
- when it is matched. */
- RE_TRANSLATE_TYPE translate;
-
- /* Number of subexpressions found by the compiler. */
- size_t re_nsub;
-
- /* Zero if this pattern cannot match the empty string, one else.
- Well, in truth it's used only in `re_search_2', to see
- whether or not we should use the fastmap, so we don't set
- this absolutely perfectly; see `re_compile_fastmap' (the
- `duplicate' case). */
- unsigned can_be_null : 1;
-
- /* If REGS_UNALLOCATED, allocate space in the `regs' structure
- for `max (RE_NREGS, re_nsub + 1)' groups.
- If REGS_REALLOCATE, reallocate space if necessary.
- If REGS_FIXED, use what's there. */
-#define REGS_UNALLOCATED 0
-#define REGS_REALLOCATE 1
-#define REGS_FIXED 2
- unsigned regs_allocated : 2;
-
- /* Set to zero when `regex_compile' compiles a pattern; set to one
- by `re_compile_fastmap' if it updates the fastmap. */
- unsigned fastmap_accurate : 1;
-
- /* If set, `re_match_2' does not return information about
- subexpressions. */
- unsigned no_sub : 1;
-
- /* If set, a beginning-of-line anchor doesn't match at the
- beginning of the string. */
- unsigned not_bol : 1;
-
- /* Similarly for an end-of-line anchor. */
- unsigned not_eol : 1;
-
- /* If true, an anchor at a newline matches. */
- unsigned newline_anchor : 1;
-
-/* [[[end pattern_buffer]]] */
-};
-
-typedef struct re_pattern_buffer regex_t;
-\f
-/* Type for byte offsets within the string. POSIX mandates this. */
-typedef int regoff_t;
-
-
-/* This is the structure we store register match data in. See
- regex.texinfo for a full description of what registers match. */
-struct re_registers
-{
- unsigned num_regs;
- regoff_t *start;
- regoff_t *end;
-};
-
-
-/* If `regs_allocated' is REGS_UNALLOCATED in the pattern buffer,
- `re_match_2' returns information about at least this many registers
- the first time a `regs' structure is passed. */
-#ifndef RE_NREGS
-# define RE_NREGS 30
-#endif
-
-
-/* POSIX specification for registers. Aside from the different names than
- `re_registers', POSIX uses an array of structures, instead of a
- structure of arrays. */
-typedef struct
-{
- regoff_t rm_so; /* Byte offset from string's start to substring's start. */
- regoff_t rm_eo; /* Byte offset from string's start to substring's end. */
-} regmatch_t;
-\f
-/* Declarations for routines. */
-
-/* To avoid duplicating every routine declaration -- once with a
- prototype (if we are ANSI), and once without (if we aren't) -- we
- use the following macro to declare argument types. This
- unfortunately clutters up the declarations a bit, but I think it's
- worth it. */
-
-#if defined(__STDC__) || defined(__cplusplus)
-
-# define _RE_ARGS(args) args
-
-#else /* not __STDC__ */
-
-# define _RE_ARGS(args) ()
-
-#endif /* not __STDC__ */
-
-/* Sets the current default syntax to SYNTAX, and return the old syntax.
- You can also simply assign to the `re_syntax_options' variable. */
-extern reg_syntax_t re_set_syntax _RE_ARGS ((reg_syntax_t syntax));
-
-/* Compile the regular expression PATTERN, with length LENGTH
- and syntax given by the global `re_syntax_options', into the buffer
- BUFFER. Return NULL if successful, and an error string if not. */
-extern const char *re_compile_pattern
- _RE_ARGS ((const char *pattern, size_t length,
- struct re_pattern_buffer *buffer));
-
-
-/* Compile a fastmap for the compiled pattern in BUFFER; used to
- accelerate searches. Return 0 if successful and -2 if was an
- internal error. */
-extern int re_compile_fastmap _RE_ARGS ((struct re_pattern_buffer *buffer));
-
-
-/* Search in the string STRING (with length LENGTH) for the pattern
- compiled into BUFFER. Start searching at position START, for RANGE
- characters. Return the starting position of the match, -1 for no
- match, or -2 for an internal error. Also return register
- information in REGS (if REGS and BUFFER->no_sub are nonzero). */
-extern int re_search
- _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string,
- int length, int start, int range, struct re_registers *regs));
-
-
-/* Like `re_search', but search in the concatenation of STRING1 and
- STRING2. Also, stop searching at index START + STOP. */
-extern int re_search_2
- _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1,
- int length1, const char *string2, int length2,
- int start, int range, struct re_registers *regs, int stop));
-
-
-/* Like `re_search', but return how many characters in STRING the regexp
- in BUFFER matched, starting at position START. */
-extern int re_match
- _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string,
- int length, int start, struct re_registers *regs));
-
-
-/* Relates to `re_match' as `re_search_2' relates to `re_search'. */
-extern int re_match_2
- _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1,
- int length1, const char *string2, int length2,
- int start, struct re_registers *regs, int stop));
-
-
-/* Set REGS to hold NUM_REGS registers, storing them in STARTS and
- ENDS. Subsequent matches using BUFFER and REGS will use this memory
- for recording register information. STARTS and ENDS must be
- allocated with malloc, and must each be at least `NUM_REGS * sizeof
- (regoff_t)' bytes long.
-
- If NUM_REGS == 0, then subsequent matches should allocate their own
- register data.
-
- Unless this function is called, the first search or match using
- PATTERN_BUFFER will allocate its own register data, without
- freeing the old data. */
-extern void re_set_registers
- _RE_ARGS ((struct re_pattern_buffer *buffer, struct re_registers *regs,
- unsigned num_regs, regoff_t *starts, regoff_t *ends));
-
-#if defined _REGEX_RE_COMP || defined _LIBC
-# ifndef _CRAY
-/* 4.2 bsd compatibility. */
-extern char *re_comp _RE_ARGS ((const char *));
-extern int re_exec _RE_ARGS ((const char *));
-# endif
-#endif
-
-/* GCC 2.95 and later have "__restrict"; C99 compilers have
- "restrict", and "configure" may have defined "restrict". */
-#ifndef __restrict
-# if ! (2 < __GNUC__ || (2 == __GNUC__ && 95 <= __GNUC_MINOR__))
-# if defined restrict || 199901L <= __STDC_VERSION__
-# define __restrict restrict
-# else
-# define __restrict
-# endif
-# endif
-#endif
-/* gcc 3.1 and up support the [restrict] syntax. */
-#ifndef __restrict_arr
-# if (__GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 1)) && !defined(__WIN32__)
-# define __restrict_arr __restrict
-# else
-# define __restrict_arr
-# endif
-#endif
-
-/* POSIX compatibility. */
-extern int regcomp _RE_ARGS ((regex_t *__restrict __preg,
- const char *__restrict __pattern,
- int __cflags));
-
-extern int regexec _RE_ARGS ((const regex_t *__restrict __preg,
- const char *__restrict __string, size_t __nmatch,
- regmatch_t __pmatch[__restrict_arr],
- int __eflags));
-
-extern size_t regerror _RE_ARGS ((int __errcode, const regex_t *__preg,
- char *__errbuf, size_t __errbuf_size));
-
-extern void regfree _RE_ARGS ((regex_t *__preg));
-
-
-#ifdef __cplusplus
-}
-#endif /* C++ */
-
-#endif /* regex.h */
-\f
-/*
-Local variables:
-make-backup-files: t
-version-control: t
-trim-versions-without-asking: nil
-End:
-*/
+++ /dev/null
-/* Extended regular expression matching and search library.
- Copyright (C) 2002, 2003 Free Software Foundation, Inc.
- This file is part of the GNU C Library.
- Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
-
- The GNU C Library is free software; you can redistribute it and/or
- modify it under the terms of the GNU Lesser General Public
- License as published by the Free Software Foundation; either
- version 2.1 of the License, or (at your option) any later version.
-
- The GNU C Library is distributed in the hope that it will be useful,
- but WITHOUT ANY WARRANTY; without even the implied warranty of
- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- Lesser General Public License for more details.
-
- You should have received a copy of the GNU Lesser General Public
- License along with the GNU C Library; if not, write to the Free
- Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
- 02111-1307 USA. */
-
-static void re_string_construct_common (const char *str, int len,
- re_string_t *pstr,
- RE_TRANSLATE_TYPE trans, int icase);
-#ifdef RE_ENABLE_I18N
-static int re_string_skip_chars (re_string_t *pstr, int new_raw_idx,
- wint_t *last_wc);
-#endif /* RE_ENABLE_I18N */
-static re_dfastate_t *create_newstate_common (re_dfa_t *dfa,
- const re_node_set *nodes,
- unsigned int hash);
-static reg_errcode_t register_state (re_dfa_t *dfa, re_dfastate_t *newstate,
- unsigned int hash);
-static re_dfastate_t *create_ci_newstate (re_dfa_t *dfa,
- const re_node_set *nodes,
- unsigned int hash);
-static re_dfastate_t *create_cd_newstate (re_dfa_t *dfa,
- const re_node_set *nodes,
- unsigned int context,
- unsigned int hash);
-static unsigned int inline calc_state_hash (const re_node_set *nodes,
- unsigned int context);
-\f
-/* Functions for string operation. */
-
-/* This function allocate the buffers. It is necessary to call
- re_string_reconstruct before using the object. */
-
-static reg_errcode_t
-re_string_allocate (pstr, str, len, init_len, trans, icase)
- re_string_t *pstr;
- const char *str;
- int len, init_len, icase;
- RE_TRANSLATE_TYPE trans;
-{
- reg_errcode_t ret;
- int init_buf_len = (len + 1 < init_len) ? len + 1: init_len;
- re_string_construct_common (str, len, pstr, trans, icase);
- pstr->stop = pstr->len;
-
- ret = re_string_realloc_buffers (pstr, init_buf_len);
- if (BE (ret != REG_NOERROR, 0))
- return ret;
-
- pstr->mbs_case = (MBS_CASE_ALLOCATED (pstr) ? pstr->mbs_case
- : (unsigned char *) str);
- pstr->mbs = MBS_ALLOCATED (pstr) ? pstr->mbs : pstr->mbs_case;
- pstr->valid_len = (MBS_CASE_ALLOCATED (pstr) || MBS_ALLOCATED (pstr)
- || MB_CUR_MAX > 1) ? pstr->valid_len : len;
- return REG_NOERROR;
-}
-
-/* This function allocate the buffers, and initialize them. */
-
-static reg_errcode_t
-re_string_construct (pstr, str, len, trans, icase)
- re_string_t *pstr;
- const char *str;
- int len, icase;
- RE_TRANSLATE_TYPE trans;
-{
- reg_errcode_t ret;
- re_string_construct_common (str, len, pstr, trans, icase);
- pstr->stop = pstr->len;
- /* Set 0 so that this function can initialize whole buffers. */
- pstr->valid_len = 0;
-
- if (len > 0)
- {
- ret = re_string_realloc_buffers (pstr, len + 1);
- if (BE (ret != REG_NOERROR, 0))
- return ret;
- }
- pstr->mbs_case = (MBS_CASE_ALLOCATED (pstr) ? pstr->mbs_case
- : (unsigned char *) str);
- pstr->mbs = MBS_ALLOCATED (pstr) ? pstr->mbs : pstr->mbs_case;
-
- if (icase)
- {
-#ifdef RE_ENABLE_I18N
- if (MB_CUR_MAX > 1)
- build_wcs_upper_buffer (pstr);
- else
-#endif /* RE_ENABLE_I18N */
- build_upper_buffer (pstr);
- }
- else
- {
-#ifdef RE_ENABLE_I18N
- if (MB_CUR_MAX > 1)
- build_wcs_buffer (pstr);
- else
-#endif /* RE_ENABLE_I18N */
- {
- if (trans != NULL)
- re_string_translate_buffer (pstr);
- else
- pstr->valid_len = len;
- }
- }
-
- /* Initialized whole buffers, then valid_len == bufs_len. */
- pstr->valid_len = pstr->bufs_len;
- return REG_NOERROR;
-}
-
-/* Helper functions for re_string_allocate, and re_string_construct. */
-
-static reg_errcode_t
-re_string_realloc_buffers (pstr, new_buf_len)
- re_string_t *pstr;
- int new_buf_len;
-{
-#ifdef RE_ENABLE_I18N
- if (MB_CUR_MAX > 1)
- {
- wint_t *new_array = re_realloc (pstr->wcs, wint_t, new_buf_len);
- if (BE (new_array == NULL, 0))
- return REG_ESPACE;
- pstr->wcs = new_array;
- }
-#endif /* RE_ENABLE_I18N */
- if (MBS_ALLOCATED (pstr))
- {
- unsigned char *new_array = re_realloc (pstr->mbs, unsigned char,
- new_buf_len);
- if (BE (new_array == NULL, 0))
- return REG_ESPACE;
- pstr->mbs = new_array;
- }
- if (MBS_CASE_ALLOCATED (pstr))
- {
- unsigned char *new_array = re_realloc (pstr->mbs_case, unsigned char,
- new_buf_len);
- if (BE (new_array == NULL, 0))
- return REG_ESPACE;
- pstr->mbs_case = new_array;
- if (!MBS_ALLOCATED (pstr))
- pstr->mbs = pstr->mbs_case;
- }
- pstr->bufs_len = new_buf_len;
- return REG_NOERROR;
-}
-
-
-static void
-re_string_construct_common (str, len, pstr, trans, icase)
- const char *str;
- int len;
- re_string_t *pstr;
- RE_TRANSLATE_TYPE trans;
- int icase;
-{
- memset (pstr, '\0', sizeof (re_string_t));
- pstr->raw_mbs = (const unsigned char *) str;
- pstr->len = len;
- pstr->trans = trans;
- pstr->icase = icase ? 1 : 0;
-}
-
-#ifdef RE_ENABLE_I18N
-
-/* Build wide character buffer PSTR->WCS.
- If the byte sequence of the string are:
- <mb1>(0), <mb1>(1), <mb2>(0), <mb2>(1), <sb3>
- Then wide character buffer will be:
- <wc1> , WEOF , <wc2> , WEOF , <wc3>
- We use WEOF for padding, they indicate that the position isn't
- a first byte of a multibyte character.
-
- Note that this function assumes PSTR->VALID_LEN elements are already
- built and starts from PSTR->VALID_LEN. */
-
-static void
-build_wcs_buffer (pstr)
- re_string_t *pstr;
-{
- mbstate_t prev_st;
- int byte_idx, end_idx, mbclen, remain_len;
- /* Build the buffers from pstr->valid_len to either pstr->len or
- pstr->bufs_len. */
- end_idx = (pstr->bufs_len > pstr->len)? pstr->len : pstr->bufs_len;
- for (byte_idx = pstr->valid_len; byte_idx < end_idx;)
- {
- wchar_t wc;
- remain_len = end_idx - byte_idx;
- prev_st = pstr->cur_state;
- mbclen = mbrtowc (&wc, ((const char *) pstr->raw_mbs + pstr->raw_mbs_idx
- + byte_idx), remain_len, &pstr->cur_state);
- if (BE (mbclen == (size_t) -2, 0))
- {
- /* The buffer doesn't have enough space, finish to build. */
- pstr->cur_state = prev_st;
- break;
- }
- else if (BE (mbclen == (size_t) -1 || mbclen == 0, 0))
- {
- /* We treat these cases as a singlebyte character. */
- mbclen = 1;
- wc = (wchar_t) pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx];
- pstr->cur_state = prev_st;
- }
-
- /* Apply the translateion if we need. */
- if (pstr->trans != NULL && mbclen == 1)
- {
- int ch = pstr->trans[pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx]];
- pstr->mbs_case[byte_idx] = ch;
- }
- /* Write wide character and padding. */
- pstr->wcs[byte_idx++] = wc;
- /* Write paddings. */
- for (remain_len = byte_idx + mbclen - 1; byte_idx < remain_len ;)
- pstr->wcs[byte_idx++] = WEOF;
- }
- pstr->valid_len = byte_idx;
-}
-
-/* Build wide character buffer PSTR->WCS like build_wcs_buffer,
- but for REG_ICASE. */
-
-static void
-build_wcs_upper_buffer (pstr)
- re_string_t *pstr;
-{
- mbstate_t prev_st;
- int byte_idx, end_idx, mbclen, remain_len;
- /* Build the buffers from pstr->valid_len to either pstr->len or
- pstr->bufs_len. */
- end_idx = (pstr->bufs_len > pstr->len)? pstr->len : pstr->bufs_len;
- for (byte_idx = pstr->valid_len; byte_idx < end_idx;)
- {
- wchar_t wc;
- remain_len = end_idx - byte_idx;
- prev_st = pstr->cur_state;
- mbclen = mbrtowc (&wc, ((const char *) pstr->raw_mbs + pstr->raw_mbs_idx
- + byte_idx), remain_len, &pstr->cur_state);
- if (BE (mbclen == (size_t) -2, 0))
- {
- /* The buffer doesn't have enough space, finish to build. */
- pstr->cur_state = prev_st;
- break;
- }
- else if (mbclen == 1 || mbclen == (size_t) -1 || mbclen == 0)
- {
- /* In case of a singlebyte character. */
- int ch = pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx];
- /* Apply the translateion if we need. */
- if (pstr->trans != NULL && mbclen == 1)
- {
- ch = pstr->trans[ch];
- pstr->mbs_case[byte_idx] = ch;
- }
- pstr->wcs[byte_idx] = iswlower (wc) ? toupper (wc) : wc;
- pstr->mbs[byte_idx++] = islower (ch) ? toupper (ch) : ch;
- if (BE (mbclen == (size_t) -1, 0))
- pstr->cur_state = prev_st;
- }
- else /* mbclen > 1 */
- {
- if (iswlower (wc))
- wcrtomb ((char *) pstr->mbs + byte_idx, towupper (wc), &prev_st);
- else
- memcpy (pstr->mbs + byte_idx,
- pstr->raw_mbs + pstr->raw_mbs_idx + byte_idx, mbclen);
- pstr->wcs[byte_idx++] = iswlower (wc) ? toupper (wc) : wc;
- /* Write paddings. */
- for (remain_len = byte_idx + mbclen - 1; byte_idx < remain_len ;)
- pstr->wcs[byte_idx++] = WEOF;
- }
- }
- pstr->valid_len = byte_idx;
-}
-
-/* Skip characters until the index becomes greater than NEW_RAW_IDX.
- Return the index. */
-
-static int
-re_string_skip_chars (pstr, new_raw_idx, last_wc)
- re_string_t *pstr;
- int new_raw_idx;
- wint_t *last_wc;
-{
- mbstate_t prev_st;
- int rawbuf_idx, mbclen;
- wchar_t wc = 0;
-
- /* Skip the characters which are not necessary to check. */
- for (rawbuf_idx = pstr->raw_mbs_idx + pstr->valid_len;
- rawbuf_idx < new_raw_idx;)
- {
- int remain_len;
- remain_len = pstr->len - rawbuf_idx;
- prev_st = pstr->cur_state;
- mbclen = mbrtowc (&wc, (const char *) pstr->raw_mbs + rawbuf_idx,
- remain_len, &pstr->cur_state);
- if (BE (mbclen == (size_t) -2 || mbclen == (size_t) -1 || mbclen == 0, 0))
- {
- /* We treat these cases as a singlebyte character. */
- mbclen = 1;
- pstr->cur_state = prev_st;
- }
- /* Then proceed the next character. */
- rawbuf_idx += mbclen;
- }
- *last_wc = (wint_t) wc;
- return rawbuf_idx;
-}
-#endif /* RE_ENABLE_I18N */
-
-/* Build the buffer PSTR->MBS, and apply the translation if we need.
- This function is used in case of REG_ICASE. */
-
-static void
-build_upper_buffer (pstr)
- re_string_t *pstr;
-{
- int char_idx, end_idx;
- end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len;
-
- for (char_idx = pstr->valid_len; char_idx < end_idx; ++char_idx)
- {
- int ch = pstr->raw_mbs[pstr->raw_mbs_idx + char_idx];
- if (pstr->trans != NULL)
- {
- ch = pstr->trans[ch];
- pstr->mbs_case[char_idx] = ch;
- }
- if (islower (ch))
- pstr->mbs[char_idx] = toupper (ch);
- else
- pstr->mbs[char_idx] = ch;
- }
- pstr->valid_len = char_idx;
-}
-
-/* Apply TRANS to the buffer in PSTR. */
-
-static void
-re_string_translate_buffer (pstr)
- re_string_t *pstr;
-{
- int buf_idx, end_idx;
- end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len;
-
- for (buf_idx = pstr->valid_len; buf_idx < end_idx; ++buf_idx)
- {
- int ch = pstr->raw_mbs[pstr->raw_mbs_idx + buf_idx];
- pstr->mbs_case[buf_idx] = pstr->trans[ch];
- }
-
- pstr->valid_len = buf_idx;
-}
-
-/* This function re-construct the buffers.
- Concretely, convert to wide character in case of MB_CUR_MAX > 1,
- convert to upper case in case of REG_ICASE, apply translation. */
-
-static reg_errcode_t
-re_string_reconstruct (pstr, idx, eflags, newline)
- re_string_t *pstr;
- int idx, eflags, newline;
-{
- int offset = idx - pstr->raw_mbs_idx;
- if (offset < 0)
- {
- /* Reset buffer. */
-#ifdef RE_ENABLE_I18N
- if (MB_CUR_MAX > 1)
- memset (&pstr->cur_state, '\0', sizeof (mbstate_t));
-#endif /* RE_ENABLE_I18N */
- pstr->len += pstr->raw_mbs_idx;
- pstr->stop += pstr->raw_mbs_idx;
- pstr->valid_len = pstr->raw_mbs_idx = 0;
- pstr->tip_context = ((eflags & REG_NOTBOL) ? CONTEXT_BEGBUF
- : CONTEXT_NEWLINE | CONTEXT_BEGBUF);
- if (!MBS_CASE_ALLOCATED (pstr))
- pstr->mbs_case = (unsigned char *) pstr->raw_mbs;
- if (!MBS_ALLOCATED (pstr) && !MBS_CASE_ALLOCATED (pstr))
- pstr->mbs = (unsigned char *) pstr->raw_mbs;
- offset = idx;
- }
-
- if (offset != 0)
- {
- /* Are the characters which are already checked remain? */
- if (offset < pstr->valid_len)
- {
- /* Yes, move them to the front of the buffer. */
- pstr->tip_context = re_string_context_at (pstr, offset - 1, eflags,
- newline);
-#ifdef RE_ENABLE_I18N
- if (MB_CUR_MAX > 1)
- memmove (pstr->wcs, pstr->wcs + offset,
- (pstr->valid_len - offset) * sizeof (wint_t));
-#endif /* RE_ENABLE_I18N */
- if (MBS_ALLOCATED (pstr))
- memmove (pstr->mbs, pstr->mbs + offset,
- pstr->valid_len - offset);
- if (MBS_CASE_ALLOCATED (pstr))
- memmove (pstr->mbs_case, pstr->mbs_case + offset,
- pstr->valid_len - offset);
- pstr->valid_len -= offset;
-#if DEBUG
- assert (pstr->valid_len > 0);
-#endif
- }
- else
- {
- /* No, skip all characters until IDX. */
- pstr->valid_len = 0;
-#ifdef RE_ENABLE_I18N
- if (MB_CUR_MAX > 1)
- {
- int wcs_idx;
- wint_t wc;
- pstr->valid_len = re_string_skip_chars (pstr, idx, &wc) - idx;
- for (wcs_idx = 0; wcs_idx < pstr->valid_len; ++wcs_idx)
- pstr->wcs[wcs_idx] = WEOF;
- if (pstr->trans && wc <= 0xff)
- wc = pstr->trans[wc];
- pstr->tip_context = (IS_WIDE_WORD_CHAR (wc) ? CONTEXT_WORD
- : ((newline && IS_WIDE_NEWLINE (wc))
- ? CONTEXT_NEWLINE : 0));
- }
- else
-#endif /* RE_ENABLE_I18N */
- {
- int c = pstr->raw_mbs[pstr->raw_mbs_idx + offset - 1];
- if (pstr->trans)
- c = pstr->trans[c];
- pstr->tip_context = (IS_WORD_CHAR (c) ? CONTEXT_WORD
- : ((newline && IS_NEWLINE (c))
- ? CONTEXT_NEWLINE : 0));
- }
- }
- if (!MBS_CASE_ALLOCATED (pstr))
- {
- pstr->mbs_case += offset;
- /* In case of !MBS_ALLOCATED && !MBS_CASE_ALLOCATED. */
- if (!MBS_ALLOCATED (pstr))
- pstr->mbs += offset;
- }
- }
- pstr->raw_mbs_idx = idx;
- pstr->len -= offset;
- pstr->stop -= offset;
-
- /* Then build the buffers. */
-#ifdef RE_ENABLE_I18N
- if (MB_CUR_MAX > 1)
- {
- if (pstr->icase)
- build_wcs_upper_buffer (pstr);
- else
- build_wcs_buffer (pstr);
- }
- else
-#endif /* RE_ENABLE_I18N */
- {
- if (pstr->icase)
- build_upper_buffer (pstr);
- else if (pstr->trans != NULL)
- re_string_translate_buffer (pstr);
- }
- pstr->cur_idx = 0;
-
- return REG_NOERROR;
-}
-
-static void
-re_string_destruct (pstr)
- re_string_t *pstr;
-{
-#ifdef RE_ENABLE_I18N
- re_free (pstr->wcs);
-#endif /* RE_ENABLE_I18N */
- if (MBS_ALLOCATED (pstr))
- re_free (pstr->mbs);
- if (MBS_CASE_ALLOCATED (pstr))
- re_free (pstr->mbs_case);
-}
-
-/* Return the context at IDX in INPUT. */
-
-static unsigned int
-re_string_context_at (input, idx, eflags, newline_anchor)
- const re_string_t *input;
- int idx, eflags, newline_anchor;
-{
- int c;
- if (idx < 0 || idx == input->len)
- {
- if (idx < 0)
- /* In this case, we use the value stored in input->tip_context,
- since we can't know the character in input->mbs[-1] here. */
- return input->tip_context;
- else /* (idx == input->len) */
- return ((eflags & REG_NOTEOL) ? CONTEXT_ENDBUF
- : CONTEXT_NEWLINE | CONTEXT_ENDBUF);
- }
-#ifdef RE_ENABLE_I18N
- if (MB_CUR_MAX > 1)
- {
- wint_t wc;
- int wc_idx = idx;
- while(input->wcs[wc_idx] == WEOF)
- {
-#ifdef DEBUG
- /* It must not happen. */
- assert (wc_idx >= 0);
-#endif
- --wc_idx;
- if (wc_idx < 0)
- return input->tip_context;
- }
- wc = input->wcs[wc_idx];
- if (IS_WIDE_WORD_CHAR (wc))
- return CONTEXT_WORD;
- return (newline_anchor && IS_WIDE_NEWLINE (wc)) ? CONTEXT_NEWLINE : 0;
- }
- else
-#endif
- {
- c = re_string_byte_at (input, idx);
- if (IS_WORD_CHAR (c))
- return CONTEXT_WORD;
- return (newline_anchor && IS_NEWLINE (c)) ? CONTEXT_NEWLINE : 0;
- }
-}
-\f
-/* Functions for set operation. */
-
-static reg_errcode_t
-re_node_set_alloc (set, size)
- re_node_set *set;
- int size;
-{
- set->alloc = size;
- set->nelem = 0;
- set->elems = re_malloc (int, size);
- if (BE (set->elems == NULL, 0))
- return REG_ESPACE;
- return REG_NOERROR;
-}
-
-static reg_errcode_t
-re_node_set_init_1 (set, elem)
- re_node_set *set;
- int elem;
-{
- set->alloc = 1;
- set->nelem = 1;
- set->elems = re_malloc (int, 1);
- if (BE (set->elems == NULL, 0))
- {
- set->alloc = set->nelem = 0;
- return REG_ESPACE;
- }
- set->elems[0] = elem;
- return REG_NOERROR;
-}
-
-static reg_errcode_t
-re_node_set_init_2 (set, elem1, elem2)
- re_node_set *set;
- int elem1, elem2;
-{
- set->alloc = 2;
- set->elems = re_malloc (int, 2);
- if (BE (set->elems == NULL, 0))
- return REG_ESPACE;
- if (elem1 == elem2)
- {
- set->nelem = 1;
- set->elems[0] = elem1;
- }
- else
- {
- set->nelem = 2;
- if (elem1 < elem2)
- {
- set->elems[0] = elem1;
- set->elems[1] = elem2;
- }
- else
- {
- set->elems[0] = elem2;
- set->elems[1] = elem1;
- }
- }
- return REG_NOERROR;
-}
-
-static reg_errcode_t
-re_node_set_init_copy (dest, src)
- re_node_set *dest;
- const re_node_set *src;
-{
- dest->nelem = src->nelem;
- if (src->nelem > 0)
- {
- dest->alloc = dest->nelem;
- dest->elems = re_malloc (int, dest->alloc);
- if (BE (dest->elems == NULL, 0))
- {
- dest->alloc = dest->nelem = 0;
- return REG_ESPACE;
- }
- memcpy (dest->elems, src->elems, src->nelem * sizeof (int));
- }
- else
- re_node_set_init_empty (dest);
- return REG_NOERROR;
-}
-
-/* Calculate the intersection of the sets SRC1 and SRC2. And merge it to
- DEST. Return value indicate the error code or REG_NOERROR if succeeded.
- Note: We assume dest->elems is NULL, when dest->alloc is 0. */
-
-static reg_errcode_t
-re_node_set_add_intersect (dest, src1, src2)
- re_node_set *dest;
- const re_node_set *src1, *src2;
-{
- int i1, i2, id;
- if (src1->nelem > 0 && src2->nelem > 0)
- {
- if (src1->nelem + src2->nelem + dest->nelem > dest->alloc)
- {
- dest->alloc = src1->nelem + src2->nelem + dest->nelem;
- dest->elems = re_realloc (dest->elems, int, dest->alloc);
- if (BE (dest->elems == NULL, 0))
- return REG_ESPACE;
- }
- }
- else
- return REG_NOERROR;
-
- for (i1 = i2 = id = 0 ; i1 < src1->nelem && i2 < src2->nelem ;)
- {
- if (src1->elems[i1] > src2->elems[i2])
- {
- ++i2;
- continue;
- }
- if (src1->elems[i1] == src2->elems[i2])
- {
- while (id < dest->nelem && dest->elems[id] < src2->elems[i2])
- ++id;
- if (id < dest->nelem && dest->elems[id] == src2->elems[i2])
- ++id;
- else
- {
- memmove (dest->elems + id + 1, dest->elems + id,
- sizeof (int) * (dest->nelem - id));
- dest->elems[id++] = src2->elems[i2++];
- ++dest->nelem;
- }
- }
- ++i1;
- }
- return REG_NOERROR;
-}
-
-/* Calculate the union set of the sets SRC1 and SRC2. And store it to
- DEST. Return value indicate the error code or REG_NOERROR if succeeded. */
-
-static reg_errcode_t
-re_node_set_init_union (dest, src1, src2)
- re_node_set *dest;
- const re_node_set *src1, *src2;
-{
- int i1, i2, id;
- if (src1 != NULL && src1->nelem > 0 && src2 != NULL && src2->nelem > 0)
- {
- dest->alloc = src1->nelem + src2->nelem;
- dest->elems = re_malloc (int, dest->alloc);
- if (BE (dest->elems == NULL, 0))
- return REG_ESPACE;
- }
- else
- {
- if (src1 != NULL && src1->nelem > 0)
- return re_node_set_init_copy (dest, src1);
- else if (src2 != NULL && src2->nelem > 0)
- return re_node_set_init_copy (dest, src2);
- else
- re_node_set_init_empty (dest);
- return REG_NOERROR;
- }
- for (i1 = i2 = id = 0 ; i1 < src1->nelem && i2 < src2->nelem ;)
- {
- if (src1->elems[i1] > src2->elems[i2])
- {
- dest->elems[id++] = src2->elems[i2++];
- continue;
- }
- if (src1->elems[i1] == src2->elems[i2])
- ++i2;
- dest->elems[id++] = src1->elems[i1++];
- }
- if (i1 < src1->nelem)
- {
- memcpy (dest->elems + id, src1->elems + i1,
- (src1->nelem - i1) * sizeof (int));
- id += src1->nelem - i1;
- }
- else if (i2 < src2->nelem)
- {
- memcpy (dest->elems + id, src2->elems + i2,
- (src2->nelem - i2) * sizeof (int));
- id += src2->nelem - i2;
- }
- dest->nelem = id;
- return REG_NOERROR;
-}
-
-/* Calculate the union set of the sets DEST and SRC. And store it to
- DEST. Return value indicate the error code or REG_NOERROR if succeeded. */
-
-static reg_errcode_t
-re_node_set_merge (dest, src)
- re_node_set *dest;
- const re_node_set *src;
-{
- int si, di;
- if (src == NULL || src->nelem == 0)
- return REG_NOERROR;
- if (dest->alloc < src->nelem + dest->nelem)
- {
- int *new_buffer;
- dest->alloc = 2 * (src->nelem + dest->alloc);
- new_buffer = re_realloc (dest->elems, int, dest->alloc);
- if (BE (new_buffer == NULL, 0))
- return REG_ESPACE;
- dest->elems = new_buffer;
- }
-
- for (si = 0, di = 0 ; si < src->nelem && di < dest->nelem ;)
- {
- int cp_from, ncp, mid, right, src_elem = src->elems[si];
- /* Binary search the spot we will add the new element. */
- right = dest->nelem;
- while (di < right)
- {
- mid = (di + right) / 2;
- if (dest->elems[mid] < src_elem)
- di = mid + 1;
- else
- right = mid;
- }
- if (di >= dest->nelem)
- break;
-
- if (dest->elems[di] == src_elem)
- {
- /* Skip since, DEST already has the element. */
- ++di;
- ++si;
- continue;
- }
-
- /* Skip the src elements which are less than dest->elems[di]. */
- cp_from = si;
- while (si < src->nelem && src->elems[si] < dest->elems[di])
- ++si;
- /* Copy these src elements. */
- ncp = si - cp_from;
- memmove (dest->elems + di + ncp, dest->elems + di,
- sizeof (int) * (dest->nelem - di));
- memcpy (dest->elems + di, src->elems + cp_from,
- sizeof (int) * ncp);
- /* Update counters. */
- di += ncp;
- dest->nelem += ncp;
- }
-
- /* Copy remaining src elements. */
- if (si < src->nelem)
- {
- memcpy (dest->elems + di, src->elems + si,
- sizeof (int) * (src->nelem - si));
- dest->nelem += src->nelem - si;
- }
- return REG_NOERROR;
-}
-
-/* Insert the new element ELEM to the re_node_set* SET.
- return 0 if SET already has ELEM,
- return -1 if an error is occured, return 1 otherwise. */
-
-static int
-re_node_set_insert (set, elem)
- re_node_set *set;
- int elem;
-{
- int idx, right, mid;
- /* In case of the set is empty. */
- if (set->elems == NULL || set->alloc == 0)
- {
- if (BE (re_node_set_init_1 (set, elem) == REG_NOERROR, 1))
- return 1;
- else
- return -1;
- }
-
- /* Binary search the spot we will add the new element. */
- idx = 0;
- right = set->nelem;
- while (idx < right)
- {
- mid = (idx + right) / 2;
- if (set->elems[mid] < elem)
- idx = mid + 1;
- else
- right = mid;
- }
-
- /* Realloc if we need. */
- if (set->alloc < set->nelem + 1)
- {
- int *new_array;
- set->alloc = set->alloc * 2;
- new_array = re_malloc (int, set->alloc);
- if (BE (new_array == NULL, 0))
- return -1;
- /* Copy the elements they are followed by the new element. */
- if (idx > 0)
- memcpy (new_array, set->elems, sizeof (int) * (idx));
- /* Copy the elements which follows the new element. */
- if (set->nelem - idx > 0)
- memcpy (new_array + idx + 1, set->elems + idx,
- sizeof (int) * (set->nelem - idx));
- re_free (set->elems);
- set->elems = new_array;
- }
- else
- {
- /* Move the elements which follows the new element. */
- if (set->nelem - idx > 0)
- memmove (set->elems + idx + 1, set->elems + idx,
- sizeof (int) * (set->nelem - idx));
- }
- /* Insert the new element. */
- set->elems[idx] = elem;
- ++set->nelem;
- return 1;
-}
-
-/* Compare two node sets SET1 and SET2.
- return 1 if SET1 and SET2 are equivalent, retrun 0 otherwise. */
-
-static int
-re_node_set_compare (set1, set2)
- const re_node_set *set1, *set2;
-{
- int i;
- if (set1 == NULL || set2 == NULL || set1->nelem != set2->nelem)
- return 0;
- for (i = 0 ; i < set1->nelem ; i++)
- if (set1->elems[i] != set2->elems[i])
- return 0;
- return 1;
-}
-
-/* Return (idx + 1) if SET contains the element ELEM, return 0 otherwise. */
-
-static int
-re_node_set_contains (set, elem)
- const re_node_set *set;
- int elem;
-{
- int idx, right, mid;
- if (set->nelem <= 0)
- return 0;
-
- /* Binary search the element. */
- idx = 0;
- right = set->nelem - 1;
- while (idx < right)
- {
- mid = (idx + right) / 2;
- if (set->elems[mid] < elem)
- idx = mid + 1;
- else
- right = mid;
- }
- return set->elems[idx] == elem ? idx + 1 : 0;
-}
-
-static void
-re_node_set_remove_at (set, idx)
- re_node_set *set;
- int idx;
-{
- if (idx < 0 || idx >= set->nelem)
- return;
- if (idx < set->nelem - 1)
- memmove (set->elems + idx, set->elems + idx + 1,
- sizeof (int) * (set->nelem - idx - 1));
- --set->nelem;
-}
-\f
-
-/* Add the token TOKEN to dfa->nodes, and return the index of the token.
- Or return -1, if an error will be occured. */
-
-static int
-re_dfa_add_node (dfa, token, mode)
- re_dfa_t *dfa;
- re_token_t token;
- int mode;
-{
- if (dfa->nodes_len >= dfa->nodes_alloc)
- {
- re_token_t *new_array;
- dfa->nodes_alloc *= 2;
- new_array = re_realloc (dfa->nodes, re_token_t, dfa->nodes_alloc);
- if (BE (new_array == NULL, 0))
- return -1;
- else
- dfa->nodes = new_array;
- if (mode)
- {
- int *new_nexts, *new_indices;
- re_node_set *new_edests, *new_eclosures, *new_inveclosures;
-
- new_nexts = re_realloc (dfa->nexts, int, dfa->nodes_alloc);
- new_indices = re_realloc (dfa->org_indices, int, dfa->nodes_alloc);
- new_edests = re_realloc (dfa->edests, re_node_set, dfa->nodes_alloc);
- new_eclosures = re_realloc (dfa->eclosures, re_node_set,
- dfa->nodes_alloc);
- new_inveclosures = re_realloc (dfa->inveclosures, re_node_set,
- dfa->nodes_alloc);
- if (BE (new_nexts == NULL || new_indices == NULL
- || new_edests == NULL || new_eclosures == NULL
- || new_inveclosures == NULL, 0))
- return -1;
- dfa->nexts = new_nexts;
- dfa->org_indices = new_indices;
- dfa->edests = new_edests;
- dfa->eclosures = new_eclosures;
- dfa->inveclosures = new_inveclosures;
- }
- }
- dfa->nodes[dfa->nodes_len] = token;
- dfa->nodes[dfa->nodes_len].duplicated = 0;
- dfa->nodes[dfa->nodes_len].constraint = 0;
- return dfa->nodes_len++;
-}
-
-static unsigned int inline
-calc_state_hash (nodes, context)
- const re_node_set *nodes;
- unsigned int context;
-{
- unsigned int hash = nodes->nelem + context;
- int i;
- for (i = 0 ; i < nodes->nelem ; i++)
- hash += nodes->elems[i];
- return hash;
-}
-
-/* Search for the state whose node_set is equivalent to NODES.
- Return the pointer to the state, if we found it in the DFA.
- Otherwise create the new one and return it. In case of an error
- return NULL and set the error code in ERR.
- Note: - We assume NULL as the invalid state, then it is possible that
- return value is NULL and ERR is REG_NOERROR.
- - We never return non-NULL value in case of any errors, it is for
- optimization. */
-
-static re_dfastate_t*
-re_acquire_state (err, dfa, nodes)
- reg_errcode_t *err;
- re_dfa_t *dfa;
- const re_node_set *nodes;
-{
- unsigned int hash;
- re_dfastate_t *new_state;
- struct re_state_table_entry *spot;
- int i;
- if (BE (nodes->nelem == 0, 0))
- {
- *err = REG_NOERROR;
- return NULL;
- }
- hash = calc_state_hash (nodes, 0);
- spot = dfa->state_table + (hash & dfa->state_hash_mask);
-
- for (i = 0 ; i < spot->num ; i++)
- {
- re_dfastate_t *state = spot->array[i];
- if (hash != state->hash)
- continue;
- if (re_node_set_compare (&state->nodes, nodes))
- return state;
- }
-
- /* There are no appropriate state in the dfa, create the new one. */
- new_state = create_ci_newstate (dfa, nodes, hash);
- if (BE (new_state != NULL, 1))
- return new_state;
- else
- {
- *err = REG_ESPACE;
- return NULL;
- }
-}
-
-/* Search for the state whose node_set is equivalent to NODES and
- whose context is equivalent to CONTEXT.
- Return the pointer to the state, if we found it in the DFA.
- Otherwise create the new one and return it. In case of an error
- return NULL and set the error code in ERR.
- Note: - We assume NULL as the invalid state, then it is possible that
- return value is NULL and ERR is REG_NOERROR.
- - We never return non-NULL value in case of any errors, it is for
- optimization. */
-
-static re_dfastate_t*
-re_acquire_state_context (err, dfa, nodes, context)
- reg_errcode_t *err;
- re_dfa_t *dfa;
- const re_node_set *nodes;
- unsigned int context;
-{
- unsigned int hash;
- re_dfastate_t *new_state;
- struct re_state_table_entry *spot;
- int i;
- if (nodes->nelem == 0)
- {
- *err = REG_NOERROR;
- return NULL;
- }
- hash = calc_state_hash (nodes, context);
- spot = dfa->state_table + (hash & dfa->state_hash_mask);
-
- for (i = 0 ; i < spot->num ; i++)
- {
- re_dfastate_t *state = spot->array[i];
- if (hash != state->hash)
- continue;
- if (re_node_set_compare (state->entrance_nodes, nodes)
- && state->context == context)
- return state;
- }
- /* There are no appropriate state in `dfa', create the new one. */
- new_state = create_cd_newstate (dfa, nodes, context, hash);
- if (BE (new_state != NULL, 1))
- return new_state;
- else
- {
- *err = REG_ESPACE;
- return NULL;
- }
-}
-
-/* Allocate memory for DFA state and initialize common properties.
- Return the new state if succeeded, otherwise return NULL. */
-
-static re_dfastate_t *
-create_newstate_common (dfa, nodes, hash)
- re_dfa_t *dfa;
- const re_node_set *nodes;
- unsigned int hash;
-{
- re_dfastate_t *newstate;
- reg_errcode_t err;
- newstate = (re_dfastate_t *) calloc (sizeof (re_dfastate_t), 1);
- if (BE (newstate == NULL, 0))
- return NULL;
- err = re_node_set_init_copy (&newstate->nodes, nodes);
- if (BE (err != REG_NOERROR, 0))
- {
- re_free (newstate);
- return NULL;
- }
- newstate->trtable = NULL;
- newstate->trtable_search = NULL;
- newstate->hash = hash;
- return newstate;
-}
-
-/* Store the new state NEWSTATE whose hash value is HASH in appropriate
- position. Return value indicate the error code if failed. */
-
-static reg_errcode_t
-register_state (dfa, newstate, hash)
- re_dfa_t *dfa;
- re_dfastate_t *newstate;
- unsigned int hash;
-{
- struct re_state_table_entry *spot;
- spot = dfa->state_table + (hash & dfa->state_hash_mask);
-
- if (spot->alloc <= spot->num)
- {
- re_dfastate_t **new_array;
- spot->alloc = 2 * spot->num + 2;
- new_array = re_realloc (spot->array, re_dfastate_t *, spot->alloc);
- if (BE (new_array == NULL, 0))
- return REG_ESPACE;
- spot->array = new_array;
- }
- spot->array[spot->num++] = newstate;
- return REG_NOERROR;
-}
-
-/* Create the new state which is independ of contexts.
- Return the new state if succeeded, otherwise return NULL. */
-
-static re_dfastate_t *
-create_ci_newstate (dfa, nodes, hash)
- re_dfa_t *dfa;
- const re_node_set *nodes;
- unsigned int hash;
-{
- int i;
- reg_errcode_t err;
- re_dfastate_t *newstate;
- newstate = create_newstate_common (dfa, nodes, hash);
- if (BE (newstate == NULL, 0))
- return NULL;
- newstate->entrance_nodes = &newstate->nodes;
-
- for (i = 0 ; i < nodes->nelem ; i++)
- {
- re_token_t *node = dfa->nodes + nodes->elems[i];
- re_token_type_t type = node->type;
- if (type == CHARACTER && !node->constraint)
- continue;
-
- /* If the state has the halt node, the state is a halt state. */
- else if (type == END_OF_RE)
- newstate->halt = 1;
-#ifdef RE_ENABLE_I18N
- else if (type == COMPLEX_BRACKET
- || (type == OP_PERIOD && MB_CUR_MAX > 1))
- newstate->accept_mb = 1;
-#endif /* RE_ENABLE_I18N */
- else if (type == OP_BACK_REF)
- newstate->has_backref = 1;
- else if (type == ANCHOR || node->constraint)
- newstate->has_constraint = 1;
- }
- err = register_state (dfa, newstate, hash);
- if (BE (err != REG_NOERROR, 0))
- {
- free_state (newstate);
- newstate = NULL;
- }
- return newstate;
-}
-
-/* Create the new state which is depend on the context CONTEXT.
- Return the new state if succeeded, otherwise return NULL. */
-
-static re_dfastate_t *
-create_cd_newstate (dfa, nodes, context, hash)
- re_dfa_t *dfa;
- const re_node_set *nodes;
- unsigned int context, hash;
-{
- int i, nctx_nodes = 0;
- reg_errcode_t err;
- re_dfastate_t *newstate;
-
- newstate = create_newstate_common (dfa, nodes, hash);
- if (BE (newstate == NULL, 0))
- return NULL;
- newstate->context = context;
- newstate->entrance_nodes = &newstate->nodes;
-
- for (i = 0 ; i < nodes->nelem ; i++)
- {
- unsigned int constraint = 0;
- re_token_t *node = dfa->nodes + nodes->elems[i];
- re_token_type_t type = node->type;
- if (node->constraint)
- constraint = node->constraint;
-
- if (type == CHARACTER && !constraint)
- continue;
- /* If the state has the halt node, the state is a halt state. */
- else if (type == END_OF_RE)
- newstate->halt = 1;
-#ifdef RE_ENABLE_I18N
- else if (type == COMPLEX_BRACKET
- || (type == OP_PERIOD && MB_CUR_MAX > 1))
- newstate->accept_mb = 1;
-#endif /* RE_ENABLE_I18N */
- else if (type == OP_BACK_REF)
- newstate->has_backref = 1;
- else if (type == ANCHOR)
- constraint = node->opr.ctx_type;
-
- if (constraint)
- {
- if (newstate->entrance_nodes == &newstate->nodes)
- {
- newstate->entrance_nodes = re_malloc (re_node_set, 1);
- if (BE (newstate->entrance_nodes == NULL, 0))
- {
- free_state (newstate);
- return NULL;
- }
- re_node_set_init_copy (newstate->entrance_nodes, nodes);
- nctx_nodes = 0;
- newstate->has_constraint = 1;
- }
-
- if (NOT_SATISFY_PREV_CONSTRAINT (constraint,context))
- {
- re_node_set_remove_at (&newstate->nodes, i - nctx_nodes);
- ++nctx_nodes;
- }
- }
- }
- err = register_state (dfa, newstate, hash);
- if (BE (err != REG_NOERROR, 0))
- {
- free_state (newstate);
- newstate = NULL;
- }
- return newstate;
-}
-
-static void
-free_state (state)
- re_dfastate_t *state;
-{
- if (state->entrance_nodes != &state->nodes)
- {
- re_node_set_free (state->entrance_nodes);
- re_free (state->entrance_nodes);
- }
- re_node_set_free (&state->nodes);
- re_free (state->trtable);
- re_free (state->trtable_search);
- re_free (state);
-}
+++ /dev/null
-/* Extended regular expression matching and search library.
- Copyright (C) 2002, 2003 Free Software Foundation, Inc.
- This file is part of the GNU C Library.
- Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
-
- The GNU C Library is free software; you can redistribute it and/or
- modify it under the terms of the GNU Lesser General Public
- License as published by the Free Software Foundation; either
- version 2.1 of the License, or (at your option) any later version.
-
- The GNU C Library is distributed in the hope that it will be useful,
- but WITHOUT ANY WARRANTY; without even the implied warranty of
- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- Lesser General Public License for more details.
-
- You should have received a copy of the GNU Lesser General Public
- License along with the GNU C Library; if not, write to the Free
- Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
- 02111-1307 USA. */
-
-#ifndef _REGEX_INTERNAL_H
-#define _REGEX_INTERNAL_H 1
-
-#ifdef HAVE_CONFIG_H
-#include "config.h"
-#endif
-
-#include <assert.h>
-#include <ctype.h>
-#include <limits.h>
-#include <stdio.h>
-#include <stdlib.h>
-#include <string.h>
-
-#if defined HAVE_LOCALE_H || defined _LIBC
-# include <locale.h>
-#endif
-#if defined HAVE_WCHAR_H || defined _LIBC
-# include <wchar.h>
-#endif /* HAVE_WCHAR_H || _LIBC */
-#if defined HAVE_WCTYPE_H || defined _LIBC
-# include <wctype.h>
-#endif /* HAVE_WCTYPE_H || _LIBC */
-
-/* In case that the system doesn't have isblank(). */
-#if !defined _LIBC && !defined HAVE_ISBLANK && !defined isblank
-# define isblank(ch) ((ch) == ' ' || (ch) == '\t')
-#endif
-
-#ifdef _LIBC
-# ifndef _RE_DEFINE_LOCALE_FUNCTIONS
-# define _RE_DEFINE_LOCALE_FUNCTIONS 1
-# include <locale/localeinfo.h>
-# include <locale/elem-hash.h>
-# include <locale/coll-lookup.h>
-# endif
-#endif
-
-/* This is for other GNU distributions with internationalized messages. */
-#if HAVE_LIBINTL_H || defined _LIBC
-# include <libintl.h>
-# ifdef _LIBC
-# undef gettext
-# define gettext(msgid) \
- INTUSE(__dcgettext) (INTUSE(_libc_intl_domainname), msgid, LC_MESSAGES)
-# endif
-#else
-# define gettext(msgid) (msgid)
-#endif
-
-#ifndef gettext_noop
-/* This define is so xgettext can find the internationalizable
- strings. */
-# define gettext_noop(String) String
-#endif
-
-#if (defined MB_CUR_MAX && HAVE_LOCALE_H && HAVE_WCTYPE_H && HAVE_WCHAR_H && HAVE_WCRTOMB && HAVE_MBRTOWC && HAVE_WCSCOLL) || _LIBC
-# define RE_ENABLE_I18N
-#endif
-
-#if __GNUC__ >= 3
-# define BE(expr, val) __builtin_expect (expr, val)
-#else
-# define BE(expr, val) (expr)
-# define inline
-#endif
-
-/* Number of bits in a byte. */
-#define BYTE_BITS 8
-/* Number of single byte character. */
-#define SBC_MAX 256
-
-#define COLL_ELEM_LEN_MAX 8
-
-/* The character which represents newline. */
-#define NEWLINE_CHAR '\n'
-#define WIDE_NEWLINE_CHAR L'\n'
-
-/* Rename to standard API for using out of glibc. */
-#ifndef _LIBC
-# define __wctype wctype
-# define __iswctype iswctype
-# define __btowc btowc
-# define __mempcpy mempcpy
-# define __wcrtomb wcrtomb
-# define attribute_hidden
-#endif /* not _LIBC */
-
-extern const char __re_error_msgid[] attribute_hidden;
-extern const size_t __re_error_msgid_idx[] attribute_hidden;
-
-/* Number of bits in an unsinged int. */
-#define UINT_BITS (sizeof (unsigned int) * BYTE_BITS)
-/* Number of unsigned int in an bit_set. */
-#define BITSET_UINTS ((SBC_MAX + UINT_BITS - 1) / UINT_BITS)
-typedef unsigned int bitset[BITSET_UINTS];
-typedef unsigned int *re_bitset_ptr_t;
-
-#define bitset_set(set,i) (set[i / UINT_BITS] |= 1 << i % UINT_BITS)
-#define bitset_clear(set,i) (set[i / UINT_BITS] &= ~(1 << i % UINT_BITS))
-#define bitset_contain(set,i) (set[i / UINT_BITS] & (1 << i % UINT_BITS))
-#define bitset_empty(set) memset (set, 0, sizeof (unsigned int) * BITSET_UINTS)
-#define bitset_set_all(set) \
- memset (set, 255, sizeof (unsigned int) * BITSET_UINTS)
-#define bitset_copy(dest,src) \
- memcpy (dest, src, sizeof (unsigned int) * BITSET_UINTS)
-static inline void bitset_not (bitset set);
-static inline void bitset_merge (bitset dest, const bitset src);
-static inline void bitset_not_merge (bitset dest, const bitset src);
-
-#define PREV_WORD_CONSTRAINT 0x0001
-#define PREV_NOTWORD_CONSTRAINT 0x0002
-#define NEXT_WORD_CONSTRAINT 0x0004
-#define NEXT_NOTWORD_CONSTRAINT 0x0008
-#define PREV_NEWLINE_CONSTRAINT 0x0010
-#define NEXT_NEWLINE_CONSTRAINT 0x0020
-#define PREV_BEGBUF_CONSTRAINT 0x0040
-#define NEXT_ENDBUF_CONSTRAINT 0x0080
-#define DUMMY_CONSTRAINT 0x0100
-
-typedef enum
-{
- INSIDE_WORD = PREV_WORD_CONSTRAINT | NEXT_WORD_CONSTRAINT,
- WORD_FIRST = PREV_NOTWORD_CONSTRAINT | NEXT_WORD_CONSTRAINT,
- WORD_LAST = PREV_WORD_CONSTRAINT | NEXT_NOTWORD_CONSTRAINT,
- LINE_FIRST = PREV_NEWLINE_CONSTRAINT,
- LINE_LAST = NEXT_NEWLINE_CONSTRAINT,
- BUF_FIRST = PREV_BEGBUF_CONSTRAINT,
- BUF_LAST = NEXT_ENDBUF_CONSTRAINT,
- WORD_DELIM = DUMMY_CONSTRAINT
-} re_context_type;
-
-typedef struct
-{
- int alloc;
- int nelem;
- int *elems;
-} re_node_set;
-
-typedef enum
-{
- NON_TYPE = 0,
-
- /* Token type, these are used only by token. */
- OP_OPEN_BRACKET,
- OP_CLOSE_BRACKET,
- OP_CHARSET_RANGE,
- OP_OPEN_DUP_NUM,
- OP_CLOSE_DUP_NUM,
- OP_NON_MATCH_LIST,
- OP_OPEN_COLL_ELEM,
- OP_CLOSE_COLL_ELEM,
- OP_OPEN_EQUIV_CLASS,
- OP_CLOSE_EQUIV_CLASS,
- OP_OPEN_CHAR_CLASS,
- OP_CLOSE_CHAR_CLASS,
- OP_WORD,
- OP_NOTWORD,
- BACK_SLASH,
-
- /* Tree type, these are used only by tree. */
- CONCAT,
- ALT,
- SUBEXP,
- SIMPLE_BRACKET,
-#ifdef RE_ENABLE_I18N
- COMPLEX_BRACKET,
-#endif /* RE_ENABLE_I18N */
-
- /* Node type, These are used by token, node, tree. */
- OP_OPEN_SUBEXP,
- OP_CLOSE_SUBEXP,
- OP_PERIOD,
- CHARACTER,
- END_OF_RE,
- OP_ALT,
- OP_DUP_ASTERISK,
- OP_DUP_PLUS,
- OP_DUP_QUESTION,
- OP_BACK_REF,
- ANCHOR,
-
- /* Dummy marker. */
- END_OF_RE_TOKEN_T
-} re_token_type_t;
-
-#ifdef RE_ENABLE_I18N
-typedef struct
-{
- /* Multibyte characters. */
- wchar_t *mbchars;
-
- /* Collating symbols. */
-# ifdef _LIBC
- int32_t *coll_syms;
-# endif
-
- /* Equivalence classes. */
-# ifdef _LIBC
- int32_t *equiv_classes;
-# endif
-
- /* Range expressions. */
-# ifdef _LIBC
- uint32_t *range_starts;
- uint32_t *range_ends;
-# else /* not _LIBC */
- wchar_t *range_starts;
- wchar_t *range_ends;
-# endif /* not _LIBC */
-
- /* Character classes. */
- wctype_t *char_classes;
-
- /* If this character set is the non-matching list. */
- unsigned int non_match : 1;
-
- /* # of multibyte characters. */
- int nmbchars;
-
- /* # of collating symbols. */
- int ncoll_syms;
-
- /* # of equivalence classes. */
- int nequiv_classes;
-
- /* # of range expressions. */
- int nranges;
-
- /* # of character classes. */
- int nchar_classes;
-} re_charset_t;
-#endif /* RE_ENABLE_I18N */
-
-typedef struct
-{
- union
- {
- unsigned char c; /* for CHARACTER */
- re_bitset_ptr_t sbcset; /* for SIMPLE_BRACKET */
-#ifdef RE_ENABLE_I18N
- re_charset_t *mbcset; /* for COMPLEX_BRACKET */
-#endif /* RE_ENABLE_I18N */
- int idx; /* for BACK_REF */
- re_context_type ctx_type; /* for ANCHOR */
- } opr;
-#if __GNUC__ >= 2
- re_token_type_t type : 8;
-#else
- re_token_type_t type;
-#endif
- unsigned int constraint : 10; /* context constraint */
- unsigned int duplicated : 1;
-#ifdef RE_ENABLE_I18N
- unsigned int mb_partial : 1;
-#endif
-} re_token_t;
-
-#define IS_EPSILON_NODE(type) \
- ((type) == OP_ALT || (type) == OP_DUP_ASTERISK || (type) == OP_DUP_PLUS \
- || (type) == OP_DUP_QUESTION || (type) == ANCHOR \
- || (type) == OP_OPEN_SUBEXP || (type) == OP_CLOSE_SUBEXP)
-
-#define ACCEPT_MB_NODE(type) \
- ((type) == COMPLEX_BRACKET || (type) == OP_PERIOD)
-
-struct re_string_t
-{
- /* Indicate the raw buffer which is the original string passed as an
- argument of regexec(), re_search(), etc.. */
- const unsigned char *raw_mbs;
- /* Store the multibyte string. In case of "case insensitive mode" like
- REG_ICASE, upper cases of the string are stored, otherwise MBS points
- the same address that RAW_MBS points. */
- unsigned char *mbs;
- /* Store the case sensitive multibyte string. In case of
- "case insensitive mode", the original string are stored,
- otherwise MBS_CASE points the same address that MBS points. */
- unsigned char *mbs_case;
-#ifdef RE_ENABLE_I18N
- /* Store the wide character string which is corresponding to MBS. */
- wint_t *wcs;
- mbstate_t cur_state;
-#endif
- /* Index in RAW_MBS. Each character mbs[i] corresponds to
- raw_mbs[raw_mbs_idx + i]. */
- int raw_mbs_idx;
- /* The length of the valid characters in the buffers. */
- int valid_len;
- /* The length of the buffers MBS, MBS_CASE, and WCS. */
- int bufs_len;
- /* The index in MBS, which is updated by re_string_fetch_byte. */
- int cur_idx;
- /* This is length_of_RAW_MBS - RAW_MBS_IDX. */
- int len;
- /* End of the buffer may be shorter than its length in the cases such
- as re_match_2, re_search_2. Then, we use STOP for end of the buffer
- instead of LEN. */
- int stop;
-
- /* The context of mbs[0]. We store the context independently, since
- the context of mbs[0] may be different from raw_mbs[0], which is
- the beginning of the input string. */
- unsigned int tip_context;
- /* The translation passed as a part of an argument of re_compile_pattern. */
- RE_TRANSLATE_TYPE trans;
- /* 1 if REG_ICASE. */
- unsigned int icase : 1;
-};
-typedef struct re_string_t re_string_t;
-/* In case of REG_ICASE, we allocate the buffer dynamically for mbs. */
-#define MBS_ALLOCATED(pstr) (pstr->icase)
-/* In case that we need translation, we allocate the buffer dynamically
- for mbs_case. Note that mbs == mbs_case if not REG_ICASE. */
-#define MBS_CASE_ALLOCATED(pstr) (pstr->trans != NULL)
-
-
-static reg_errcode_t re_string_allocate (re_string_t *pstr, const char *str,
- int len, int init_len,
- RE_TRANSLATE_TYPE trans, int icase);
-static reg_errcode_t re_string_construct (re_string_t *pstr, const char *str,
- int len, RE_TRANSLATE_TYPE trans,
- int icase);
-static reg_errcode_t re_string_reconstruct (re_string_t *pstr, int idx,
- int eflags, int newline);
-static reg_errcode_t re_string_realloc_buffers (re_string_t *pstr,
- int new_buf_len);
-#ifdef RE_ENABLE_I18N
-static void build_wcs_buffer (re_string_t *pstr);
-static void build_wcs_upper_buffer (re_string_t *pstr);
-#endif /* RE_ENABLE_I18N */
-static void build_upper_buffer (re_string_t *pstr);
-static void re_string_translate_buffer (re_string_t *pstr);
-static void re_string_destruct (re_string_t *pstr);
-#ifdef RE_ENABLE_I18N
-static int re_string_elem_size_at (const re_string_t *pstr, int idx);
-static inline int re_string_char_size_at (const re_string_t *pstr, int idx);
-static inline wint_t re_string_wchar_at (const re_string_t *pstr, int idx);
-#endif /* RE_ENABLE_I18N */
-static unsigned int re_string_context_at (const re_string_t *input, int idx,
- int eflags, int newline_anchor);
-#define re_string_peek_byte(pstr, offset) \
- ((pstr)->mbs[(pstr)->cur_idx + offset])
-#define re_string_peek_byte_case(pstr, offset) \
- ((pstr)->mbs_case[(pstr)->cur_idx + offset])
-#define re_string_fetch_byte(pstr) \
- ((pstr)->mbs[(pstr)->cur_idx++])
-#define re_string_fetch_byte_case(pstr) \
- ((pstr)->mbs_case[(pstr)->cur_idx++])
-#define re_string_first_byte(pstr, idx) \
- ((idx) == (pstr)->len || (pstr)->wcs[idx] != WEOF)
-#define re_string_is_single_byte_char(pstr, idx) \
- ((pstr)->wcs[idx] != WEOF && ((pstr)->len == (idx) \
- || (pstr)->wcs[(idx) + 1] != WEOF))
-#define re_string_eoi(pstr) ((pstr)->stop <= (pstr)->cur_idx)
-#define re_string_cur_idx(pstr) ((pstr)->cur_idx)
-#define re_string_get_buffer(pstr) ((pstr)->mbs)
-#define re_string_length(pstr) ((pstr)->len)
-#define re_string_byte_at(pstr,idx) ((pstr)->mbs[idx])
-#define re_string_skip_bytes(pstr,idx) ((pstr)->cur_idx += (idx))
-#define re_string_set_index(pstr,idx) ((pstr)->cur_idx = (idx))
-
-#define re_malloc(t,n) ((t *) malloc ((n) * sizeof (t)))
-#define re_realloc(p,t,n) ((t *) realloc (p, (n) * sizeof (t)))
-#define re_free(p) free (p)
-
-struct bin_tree_t
-{
- struct bin_tree_t *parent;
- struct bin_tree_t *left;
- struct bin_tree_t *right;
-
- /* `node_idx' is the index in dfa->nodes, if `type' == 0.
- Otherwise `type' indicate the type of this node. */
- re_token_type_t type;
- int node_idx;
-
- int first;
- int next;
- re_node_set eclosure;
-};
-typedef struct bin_tree_t bin_tree_t;
-
-
-#define CONTEXT_WORD 1
-#define CONTEXT_NEWLINE (CONTEXT_WORD << 1)
-#define CONTEXT_BEGBUF (CONTEXT_NEWLINE << 1)
-#define CONTEXT_ENDBUF (CONTEXT_BEGBUF << 1)
-
-#define IS_WORD_CONTEXT(c) ((c) & CONTEXT_WORD)
-#define IS_NEWLINE_CONTEXT(c) ((c) & CONTEXT_NEWLINE)
-#define IS_BEGBUF_CONTEXT(c) ((c) & CONTEXT_BEGBUF)
-#define IS_ENDBUF_CONTEXT(c) ((c) & CONTEXT_ENDBUF)
-#define IS_ORDINARY_CONTEXT(c) ((c) == 0)
-
-#define IS_WORD_CHAR(ch) (isalnum (ch) || (ch) == '_')
-#define IS_NEWLINE(ch) ((ch) == NEWLINE_CHAR)
-#define IS_WIDE_WORD_CHAR(ch) (iswalnum (ch) || (ch) == L'_')
-#define IS_WIDE_NEWLINE(ch) ((ch) == WIDE_NEWLINE_CHAR)
-
-#define NOT_SATISFY_PREV_CONSTRAINT(constraint,context) \
- ((((constraint) & PREV_WORD_CONSTRAINT) && !IS_WORD_CONTEXT (context)) \
- || ((constraint & PREV_NOTWORD_CONSTRAINT) && IS_WORD_CONTEXT (context)) \
- || ((constraint & PREV_NEWLINE_CONSTRAINT) && !IS_NEWLINE_CONTEXT (context))\
- || ((constraint & PREV_BEGBUF_CONSTRAINT) && !IS_BEGBUF_CONTEXT (context)))
-
-#define NOT_SATISFY_NEXT_CONSTRAINT(constraint,context) \
- ((((constraint) & NEXT_WORD_CONSTRAINT) && !IS_WORD_CONTEXT (context)) \
- || (((constraint) & NEXT_NOTWORD_CONSTRAINT) && IS_WORD_CONTEXT (context)) \
- || (((constraint) & NEXT_NEWLINE_CONSTRAINT) && !IS_NEWLINE_CONTEXT (context)) \
- || (((constraint) & NEXT_ENDBUF_CONSTRAINT) && !IS_ENDBUF_CONTEXT (context)))
-
-struct re_dfastate_t
-{
- unsigned int hash;
- re_node_set nodes;
- re_node_set *entrance_nodes;
- struct re_dfastate_t **trtable;
- struct re_dfastate_t **trtable_search;
- /* If this state is a special state.
- A state is a special state if the state is the halt state, or
- a anchor. */
- unsigned int context : 2;
- unsigned int halt : 1;
- /* If this state can accept `multi byte'.
- Note that we refer to multibyte characters, and multi character
- collating elements as `multi byte'. */
- unsigned int accept_mb : 1;
- /* If this state has backreference node(s). */
- unsigned int has_backref : 1;
- unsigned int has_constraint : 1;
-};
-typedef struct re_dfastate_t re_dfastate_t;
-
-typedef struct
-{
- /* start <= node < end */
- int start;
- int end;
-} re_subexp_t;
-
-struct re_state_table_entry
-{
- int num;
- int alloc;
- re_dfastate_t **array;
-};
-
-/* Array type used in re_sub_match_last_t and re_sub_match_top_t. */
-
-typedef struct
-{
- int next_idx;
- int alloc;
- re_dfastate_t **array;
-} state_array_t;
-
-/* Store information about the node NODE whose type is OP_CLOSE_SUBEXP. */
-
-typedef struct
-{
- int node;
- int str_idx; /* The position NODE match at. */
- state_array_t path;
-} re_sub_match_last_t;
-
-/* Store information about the node NODE whose type is OP_OPEN_SUBEXP.
- And information about the node, whose type is OP_CLOSE_SUBEXP,
- corresponding to NODE is stored in LASTS. */
-
-typedef struct
-{
- int str_idx;
- int node;
- int next_last_offset;
- state_array_t *path;
- int alasts; /* Allocation size of LASTS. */
- int nlasts; /* The number of LASTS. */
- re_sub_match_last_t **lasts;
-} re_sub_match_top_t;
-
-struct re_backref_cache_entry
-{
- int node;
- int str_idx;
- int subexp_from;
- int subexp_to;
- int flag;
-};
-
-typedef struct
-{
- /* EFLAGS of the argument of regexec. */
- int eflags;
- /* Where the matching ends. */
- int match_last;
- int last_node;
- /* The string object corresponding to the input string. */
- re_string_t *input;
- /* The state log used by the matcher. */
- re_dfastate_t **state_log;
- int state_log_top;
- /* Back reference cache. */
- int nbkref_ents;
- int abkref_ents;
- struct re_backref_cache_entry *bkref_ents;
- int max_mb_elem_len;
- int nsub_tops;
- int asub_tops;
- re_sub_match_top_t **sub_tops;
-} re_match_context_t;
-
-typedef struct
-{
- int cur_bkref;
- int cls_subexp_idx;
-
- re_dfastate_t **sifted_states;
- re_dfastate_t **limited_states;
-
- re_node_set limits;
-
- int last_node;
- int last_str_idx;
- int check_subexp;
-} re_sift_context_t;
-
-struct re_fail_stack_ent_t
-{
- int idx;
- int node;
- regmatch_t *regs;
- re_node_set eps_via_nodes;
-};
-
-struct re_fail_stack_t
-{
- int num;
- int alloc;
- struct re_fail_stack_ent_t *stack;
-};
-
-struct re_dfa_t
-{
- re_bitset_ptr_t word_char;
-
- /* number of subexpressions `re_nsub' is in regex_t. */
- int subexps_alloc;
- re_subexp_t *subexps;
-
- re_token_t *nodes;
- int nodes_alloc;
- int nodes_len;
- bin_tree_t *str_tree;
- int *nexts;
- int *org_indices;
- re_node_set *edests;
- re_node_set *eclosures;
- re_node_set *inveclosures;
- struct re_state_table_entry *state_table;
- unsigned int state_hash_mask;
- re_dfastate_t *init_state;
- re_dfastate_t *init_state_word;
- re_dfastate_t *init_state_nl;
- re_dfastate_t *init_state_begbuf;
- int states_alloc;
- int init_node;
- int nbackref; /* The number of backreference in this dfa. */
- /* Bitmap expressing which backreference is used. */
- unsigned int used_bkref_map;
-#ifdef DEBUG
- char* re_str;
-#endif
- unsigned int has_plural_match : 1;
- /* If this dfa has "multibyte node", which is a backreference or
- a node which can accept multibyte character or multi character
- collating element. */
- unsigned int has_mb_node : 1;
-};
-typedef struct re_dfa_t re_dfa_t;
-
-static reg_errcode_t re_node_set_alloc (re_node_set *set, int size);
-static reg_errcode_t re_node_set_init_1 (re_node_set *set, int elem);
-static reg_errcode_t re_node_set_init_2 (re_node_set *set, int elem1,
- int elem2);
-#define re_node_set_init_empty(set) memset (set, '\0', sizeof (re_node_set))
-static reg_errcode_t re_node_set_init_copy (re_node_set *dest,
- const re_node_set *src);
-static reg_errcode_t re_node_set_add_intersect (re_node_set *dest,
- const re_node_set *src1,
- const re_node_set *src2);
-static reg_errcode_t re_node_set_init_union (re_node_set *dest,
- const re_node_set *src1,
- const re_node_set *src2);
-static reg_errcode_t re_node_set_merge (re_node_set *dest,
- const re_node_set *src);
-static int re_node_set_insert (re_node_set *set, int elem);
-static int re_node_set_compare (const re_node_set *set1,
- const re_node_set *set2);
-static int re_node_set_contains (const re_node_set *set, int elem);
-static void re_node_set_remove_at (re_node_set *set, int idx);
-#define re_node_set_remove(set,id) \
- (re_node_set_remove_at (set, re_node_set_contains (set, id) - 1))
-#define re_node_set_empty(p) ((p)->nelem = 0)
-#define re_node_set_free(set) re_free ((set)->elems)
-static int re_dfa_add_node (re_dfa_t *dfa, re_token_t token, int mode);
-static re_dfastate_t *re_acquire_state (reg_errcode_t *err, re_dfa_t *dfa,
- const re_node_set *nodes);
-static re_dfastate_t *re_acquire_state_context (reg_errcode_t *err,
- re_dfa_t *dfa,
- const re_node_set *nodes,
- unsigned int context);
-static void free_state (re_dfastate_t *state);
-\f
-
-typedef enum
-{
- SB_CHAR,
- MB_CHAR,
- EQUIV_CLASS,
- COLL_SYM,
- CHAR_CLASS
-} bracket_elem_type;
-
-typedef struct
-{
- bracket_elem_type type;
- union
- {
- unsigned char ch;
- unsigned char *name;
- wchar_t wch;
- } opr;
-} bracket_elem_t;
-
-
-/* Inline functions for bitset operation. */
-static inline void
-bitset_not (set)
- bitset set;
-{
- int bitset_i;
- for (bitset_i = 0; bitset_i < BITSET_UINTS; ++bitset_i)
- set[bitset_i] = ~set[bitset_i];
-}
-
-static inline void
-bitset_merge (dest, src)
- bitset dest;
- const bitset src;
-{
- int bitset_i;
- for (bitset_i = 0; bitset_i < BITSET_UINTS; ++bitset_i)
- dest[bitset_i] |= src[bitset_i];
-}
-
-static inline void
-bitset_not_merge (dest, src)
- bitset dest;
- const bitset src;
-{
- int i;
- for (i = 0; i < BITSET_UINTS; ++i)
- dest[i] |= ~src[i];
-}
-
-#ifdef RE_ENABLE_I18N
-/* Inline functions for re_string. */
-static inline int
-re_string_char_size_at (pstr, idx)
- const re_string_t *pstr;
- int idx;
-{
- int byte_idx;
- if (MB_CUR_MAX == 1)
- return 1;
- for (byte_idx = 1; idx + byte_idx < pstr->len; ++byte_idx)
- if (pstr->wcs[idx + byte_idx] != WEOF)
- break;
- return byte_idx;
-}
-
-static inline wint_t
-re_string_wchar_at (pstr, idx)
- const re_string_t *pstr;
- int idx;
-{
- if (MB_CUR_MAX == 1)
- return (wint_t) pstr->mbs[idx];
- return (wint_t) pstr->wcs[idx];
-}
-
-static int
-re_string_elem_size_at (pstr, idx)
- const re_string_t *pstr;
- int idx;
-{
-#ifdef _LIBC
- const unsigned char *p, *extra;
- const int32_t *table, *indirect;
- int32_t tmp;
-# include <locale/weight.h>
- uint_fast32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES);
-
- if (nrules != 0)
- {
- table = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB);
- extra = (const unsigned char *)
- _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB);
- indirect = (const int32_t *) _NL_CURRENT (LC_COLLATE,
- _NL_COLLATE_INDIRECTMB);
- p = pstr->mbs + idx;
- tmp = findidx (&p);
- return p - pstr->mbs - idx;
- }
- else
-#endif /* _LIBC */
- return 1;
-}
-#endif /* RE_ENABLE_I18N */
-
-#endif /* _REGEX_INTERNAL_H */
+++ /dev/null
-/* Extended regular expression matching and search library.
- Copyright (C) 2002, 2003 Free Software Foundation, Inc.
- This file is part of the GNU C Library.
- Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
-
- The GNU C Library is free software; you can redistribute it and/or
- modify it under the terms of the GNU Lesser General Public
- License as published by the Free Software Foundation; either
- version 2.1 of the License, or (at your option) any later version.
-
- The GNU C Library is distributed in the hope that it will be useful,
- but WITHOUT ANY WARRANTY; without even the implied warranty of
- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- Lesser General Public License for more details.
-
- You should have received a copy of the GNU Lesser General Public
- License along with the GNU C Library; if not, write to the Free
- Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
- 02111-1307 USA. */
-
-static reg_errcode_t match_ctx_init (re_match_context_t *cache, int eflags,
- re_string_t *input, int n);
-static void match_ctx_clean (re_match_context_t *mctx);
-static void match_ctx_free (re_match_context_t *cache);
-static void match_ctx_free_subtops (re_match_context_t *mctx);
-static reg_errcode_t match_ctx_add_entry (re_match_context_t *cache, int node,
- int str_idx, int from, int to);
-static int search_cur_bkref_entry (re_match_context_t *mctx, int str_idx);
-static void match_ctx_clear_flag (re_match_context_t *mctx);
-static reg_errcode_t match_ctx_add_subtop (re_match_context_t *mctx, int node,
- int str_idx);
-static re_sub_match_last_t * match_ctx_add_sublast (re_sub_match_top_t *subtop,
- int node, int str_idx);
-static void sift_ctx_init (re_sift_context_t *sctx, re_dfastate_t **sifted_sts,
- re_dfastate_t **limited_sts, int last_node,
- int last_str_idx, int check_subexp);
-static reg_errcode_t re_search_internal (const regex_t *preg,
- const char *string, int length,
- int start, int range, int stop,
- size_t nmatch, regmatch_t pmatch[],
- int eflags);
-static int re_search_2_stub (struct re_pattern_buffer *bufp,
- const char *string1, int length1,
- const char *string2, int length2,
- int start, int range, struct re_registers *regs,
- int stop, int ret_len);
-static int re_search_stub (struct re_pattern_buffer *bufp,
- const char *string, int length, int start,
- int range, int stop, struct re_registers *regs,
- int ret_len);
-static unsigned re_copy_regs (struct re_registers *regs, regmatch_t *pmatch,
- int nregs, int regs_allocated);
-static inline re_dfastate_t *acquire_init_state_context (reg_errcode_t *err,
- const regex_t *preg,
- const re_match_context_t *mctx,
- int idx);
-static reg_errcode_t prune_impossible_nodes (const regex_t *preg,
- re_match_context_t *mctx);
-static int check_matching (const regex_t *preg, re_match_context_t *mctx,
- int fl_search, int fl_longest_match);
-static int check_halt_node_context (const re_dfa_t *dfa, int node,
- unsigned int context);
-static int check_halt_state_context (const regex_t *preg,
- const re_dfastate_t *state,
- const re_match_context_t *mctx, int idx);
-static void update_regs (re_dfa_t *dfa, regmatch_t *pmatch, int cur_node,
- int cur_idx, int nmatch);
-static int proceed_next_node (const regex_t *preg, int nregs, regmatch_t *regs,
- const re_match_context_t *mctx,
- int *pidx, int node, re_node_set *eps_via_nodes,
- struct re_fail_stack_t *fs);
-static reg_errcode_t push_fail_stack (struct re_fail_stack_t *fs,
- int str_idx, int *dests, int nregs,
- regmatch_t *regs,
- re_node_set *eps_via_nodes);
-static int pop_fail_stack (struct re_fail_stack_t *fs, int *pidx, int nregs,
- regmatch_t *regs, re_node_set *eps_via_nodes);
-static reg_errcode_t set_regs (const regex_t *preg,
- const re_match_context_t *mctx,
- size_t nmatch, regmatch_t *pmatch,
- int fl_backtrack);
-static reg_errcode_t free_fail_stack_return (struct re_fail_stack_t *fs);
-
-#ifdef RE_ENABLE_I18N
-static int sift_states_iter_mb (const regex_t *preg,
- const re_match_context_t *mctx,
- re_sift_context_t *sctx,
- int node_idx, int str_idx, int max_str_idx);
-#endif /* RE_ENABLE_I18N */
-static reg_errcode_t sift_states_backward (const regex_t *preg,
- re_match_context_t *mctx,
- re_sift_context_t *sctx);
-static reg_errcode_t update_cur_sifted_state (const regex_t *preg,
- re_match_context_t *mctx,
- re_sift_context_t *sctx,
- int str_idx,
- re_node_set *dest_nodes);
-static reg_errcode_t add_epsilon_src_nodes (re_dfa_t *dfa,
- re_node_set *dest_nodes,
- const re_node_set *candidates);
-static reg_errcode_t sub_epsilon_src_nodes (re_dfa_t *dfa, int node,
- re_node_set *dest_nodes,
- const re_node_set *and_nodes);
-static int check_dst_limits (re_dfa_t *dfa, re_node_set *limits,
- re_match_context_t *mctx, int dst_node,
- int dst_idx, int src_node, int src_idx);
-static int check_dst_limits_calc_pos (re_dfa_t *dfa, re_match_context_t *mctx,
- int limit, re_node_set *eclosures,
- int subexp_idx, int node, int str_idx);
-static reg_errcode_t check_subexp_limits (re_dfa_t *dfa,
- re_node_set *dest_nodes,
- const re_node_set *candidates,
- re_node_set *limits,
- struct re_backref_cache_entry *bkref_ents,
- int str_idx);
-static reg_errcode_t sift_states_bkref (const regex_t *preg,
- re_match_context_t *mctx,
- re_sift_context_t *sctx,
- int str_idx, re_node_set *dest_nodes);
-static reg_errcode_t clean_state_log_if_need (re_match_context_t *mctx,
- int next_state_log_idx);
-static reg_errcode_t merge_state_array (re_dfa_t *dfa, re_dfastate_t **dst,
- re_dfastate_t **src, int num);
-static re_dfastate_t *transit_state (reg_errcode_t *err, const regex_t *preg,
- re_match_context_t *mctx,
- re_dfastate_t *state, int fl_search);
-static reg_errcode_t check_subexp_matching_top (re_dfa_t *dfa,
- re_match_context_t *mctx,
- re_node_set *cur_nodes,
- int str_idx);
-static re_dfastate_t *transit_state_sb (reg_errcode_t *err, const regex_t *preg,
- re_dfastate_t *pstate,
- int fl_search,
- re_match_context_t *mctx);
-#ifdef RE_ENABLE_I18N
-static reg_errcode_t transit_state_mb (const regex_t *preg,
- re_dfastate_t *pstate,
- re_match_context_t *mctx);
-#endif /* RE_ENABLE_I18N */
-static reg_errcode_t transit_state_bkref (const regex_t *preg,
- re_node_set *nodes,
- re_match_context_t *mctx);
-static reg_errcode_t get_subexp (const regex_t *preg, re_match_context_t *mctx,
- int bkref_node, int bkref_str_idx);
-static reg_errcode_t get_subexp_sub (const regex_t *preg,
- re_match_context_t *mctx,
- re_sub_match_top_t *sub_top,
- re_sub_match_last_t *sub_last,
- int bkref_node, int bkref_str);
-static int find_subexp_node (re_dfa_t *dfa, re_node_set *nodes,
- int subexp_idx, int fl_open);
-static reg_errcode_t check_arrival (const regex_t *preg,
- re_match_context_t *mctx,
- state_array_t *path, int top_node,
- int top_str, int last_node, int last_str,
- int fl_open);
-static reg_errcode_t check_arrival_add_next_nodes (const regex_t *preg,
- re_dfa_t *dfa,
- re_match_context_t *mctx,
- int str_idx,
- re_node_set *cur_nodes,
- re_node_set *next_nodes);
-static reg_errcode_t check_arrival_expand_ecl (re_dfa_t *dfa,
- re_node_set *cur_nodes,
- int ex_subexp, int fl_open);
-static reg_errcode_t check_arrival_expand_ecl_sub (re_dfa_t *dfa,
- re_node_set *dst_nodes,
- int target, int ex_subexp,
- int fl_open);
-static reg_errcode_t expand_bkref_cache (const regex_t *preg,
- re_match_context_t *mctx,
- re_node_set *cur_nodes, int cur_str,
- int last_str, int subexp_num,
- int fl_open);
-static re_dfastate_t **build_trtable (const regex_t *dfa,
- const re_dfastate_t *state,
- int fl_search);
-#ifdef RE_ENABLE_I18N
-static int check_node_accept_bytes (const regex_t *preg, int node_idx,
- const re_string_t *input, int idx);
-# ifdef _LIBC
-static unsigned int find_collation_sequence_value (const unsigned char *mbs,
- size_t name_len);
-# endif /* _LIBC */
-#endif /* RE_ENABLE_I18N */
-static int group_nodes_into_DFAstates (const regex_t *dfa,
- const re_dfastate_t *state,
- re_node_set *states_node,
- bitset *states_ch);
-static int check_node_accept (const regex_t *preg, const re_token_t *node,
- const re_match_context_t *mctx, int idx);
-static reg_errcode_t extend_buffers (re_match_context_t *mctx);
-\f
-/* Entry point for POSIX code. */
-
-/* regexec searches for a given pattern, specified by PREG, in the
- string STRING.
-
- If NMATCH is zero or REG_NOSUB was set in the cflags argument to
- `regcomp', we ignore PMATCH. Otherwise, we assume PMATCH has at
- least NMATCH elements, and we set them to the offsets of the
- corresponding matched substrings.
-
- EFLAGS specifies `execution flags' which affect matching: if
- REG_NOTBOL is set, then ^ does not match at the beginning of the
- string; if REG_NOTEOL is set, then $ does not match at the end.
-
- We return 0 if we find a match and REG_NOMATCH if not. */
-
-int
-regexec (preg, string, nmatch, pmatch, eflags)
- const regex_t *__restrict preg;
- const char *__restrict string;
- size_t nmatch;
- regmatch_t pmatch[];
- int eflags;
-{
- reg_errcode_t err;
- int length = strlen (string);
- if (preg->no_sub)
- err = re_search_internal (preg, string, length, 0, length, length, 0,
- NULL, eflags);
- else
- err = re_search_internal (preg, string, length, 0, length, length, nmatch,
- pmatch, eflags);
- return err != REG_NOERROR;
-}
-#ifdef _LIBC
-weak_alias (__regexec, regexec)
-#endif
-
-/* Entry points for GNU code. */
-
-/* re_match, re_search, re_match_2, re_search_2
-
- The former two functions operate on STRING with length LENGTH,
- while the later two operate on concatenation of STRING1 and STRING2
- with lengths LENGTH1 and LENGTH2, respectively.
-
- re_match() matches the compiled pattern in BUFP against the string,
- starting at index START.
-
- re_search() first tries matching at index START, then it tries to match
- starting from index START + 1, and so on. The last start position tried
- is START + RANGE. (Thus RANGE = 0 forces re_search to operate the same
- way as re_match().)
-
- The parameter STOP of re_{match,search}_2 specifies that no match exceeding
- the first STOP characters of the concatenation of the strings should be
- concerned.
-
- If REGS is not NULL, and BUFP->no_sub is not set, the offsets of the match
- and all groups is stroed in REGS. (For the "_2" variants, the offsets are
- computed relative to the concatenation, not relative to the individual
- strings.)
-
- On success, re_match* functions return the length of the match, re_search*
- return the position of the start of the match. Return value -1 means no
- match was found and -2 indicates an internal error. */
-
-int
-re_match (bufp, string, length, start, regs)
- struct re_pattern_buffer *bufp;
- const char *string;
- int length, start;
- struct re_registers *regs;
-{
- return re_search_stub (bufp, string, length, start, 0, length, regs, 1);
-}
-#ifdef _LIBC
-weak_alias (__re_match, re_match)
-#endif
-
-int
-re_search (bufp, string, length, start, range, regs)
- struct re_pattern_buffer *bufp;
- const char *string;
- int length, start, range;
- struct re_registers *regs;
-{
- return re_search_stub (bufp, string, length, start, range, length, regs, 0);
-}
-#ifdef _LIBC
-weak_alias (__re_search, re_search)
-#endif
-
-int
-re_match_2 (bufp, string1, length1, string2, length2, start, regs, stop)
- struct re_pattern_buffer *bufp;
- const char *string1, *string2;
- int length1, length2, start, stop;
- struct re_registers *regs;
-{
- return re_search_2_stub (bufp, string1, length1, string2, length2,
- start, 0, regs, stop, 1);
-}
-#ifdef _LIBC
-weak_alias (__re_match_2, re_match_2)
-#endif
-
-int
-re_search_2 (bufp, string1, length1, string2, length2, start, range, regs, stop)
- struct re_pattern_buffer *bufp;
- const char *string1, *string2;
- int length1, length2, start, range, stop;
- struct re_registers *regs;
-{
- return re_search_2_stub (bufp, string1, length1, string2, length2,
- start, range, regs, stop, 0);
-}
-#ifdef _LIBC
-weak_alias (__re_search_2, re_search_2)
-#endif
-
-static int
-re_search_2_stub (bufp, string1, length1, string2, length2, start, range, regs,
- stop, ret_len)
- struct re_pattern_buffer *bufp;
- const char *string1, *string2;
- int length1, length2, start, range, stop, ret_len;
- struct re_registers *regs;
-{
- const char *str;
- int rval;
- int len = length1 + length2;
- int free_str = 0;
-
- if (BE (length1 < 0 || length2 < 0 || stop < 0, 0))
- return -2;
-
- /* Concatenate the strings. */
- if (length2 > 0)
- if (length1 > 0)
- {
- char *s = re_malloc (char, len);
-
- if (BE (s == NULL, 0))
- return -2;
- memcpy (s, string1, length1);
- memcpy (s + length1, string2, length2);
- str = s;
- free_str = 1;
- }
- else
- str = string2;
- else
- str = string1;
-
- rval = re_search_stub (bufp, str, len, start, range, stop, regs,
- ret_len);
- if (free_str)
- re_free ((char *) str);
- return rval;
-}
-
-/* The parameters have the same meaning as those of re_search.
- Additional parameters:
- If RET_LEN is nonzero the length of the match is returned (re_match style);
- otherwise the position of the match is returned. */
-
-static int
-re_search_stub (bufp, string, length, start, range, stop, regs, ret_len)
- struct re_pattern_buffer *bufp;
- const char *string;
- int length, start, range, stop, ret_len;
- struct re_registers *regs;
-{
- reg_errcode_t result;
- regmatch_t *pmatch;
- int nregs, rval;
- int eflags = 0;
-
- /* Check for out-of-range. */
- if (BE (start < 0 || start > length, 0))
- return -1;
- if (BE (start + range > length, 0))
- range = length - start;
- else if (BE (start + range < 0, 0))
- range = -start;
-
- eflags |= (bufp->not_bol) ? REG_NOTBOL : 0;
- eflags |= (bufp->not_eol) ? REG_NOTEOL : 0;
-
- /* Compile fastmap if we haven't yet. */
- if (range > 0 && bufp->fastmap != NULL && !bufp->fastmap_accurate)
- re_compile_fastmap (bufp);
-
- if (BE (bufp->no_sub, 0))
- regs = NULL;
-
- /* We need at least 1 register. */
- if (regs == NULL)
- nregs = 1;
- else if (BE (bufp->regs_allocated == REGS_FIXED &&
- regs->num_regs < bufp->re_nsub + 1, 0))
- {
- nregs = regs->num_regs;
- if (BE (nregs < 1, 0))
- {
- /* Nothing can be copied to regs. */
- regs = NULL;
- nregs = 1;
- }
- }
- else
- nregs = bufp->re_nsub + 1;
- pmatch = re_malloc (regmatch_t, nregs);
- if (BE (pmatch == NULL, 0))
- return -2;
-
- result = re_search_internal (bufp, string, length, start, range, stop,
- nregs, pmatch, eflags);
-
- rval = 0;
-
- /* I hope we needn't fill ther regs with -1's when no match was found. */
- if (result != REG_NOERROR)
- rval = -1;
- else if (regs != NULL)
- {
- /* If caller wants register contents data back, copy them. */
- bufp->regs_allocated = re_copy_regs (regs, pmatch, nregs,
- bufp->regs_allocated);
- if (BE (bufp->regs_allocated == REGS_UNALLOCATED, 0))
- rval = -2;
- }
-
- if (BE (rval == 0, 1))
- {
- if (ret_len)
- {
- assert (pmatch[0].rm_so == start);
- rval = pmatch[0].rm_eo - start;
- }
- else
- rval = pmatch[0].rm_so;
- }
- re_free (pmatch);
- return rval;
-}
-
-static unsigned
-re_copy_regs (regs, pmatch, nregs, regs_allocated)
- struct re_registers *regs;
- regmatch_t *pmatch;
- int nregs, regs_allocated;
-{
- int rval = REGS_REALLOCATE;
- int i;
- int need_regs = nregs + 1;
- /* We need one extra element beyond `num_regs' for the `-1' marker GNU code
- uses. */
-
- /* Have the register data arrays been allocated? */
- if (regs_allocated == REGS_UNALLOCATED)
- { /* No. So allocate them with malloc. */
- regs->start = re_malloc (regoff_t, need_regs);
- if (BE (regs->start == NULL, 0))
- return REGS_UNALLOCATED;
- regs->end = re_malloc (regoff_t, need_regs);
- if (BE (regs->end == NULL, 0))
- {
- re_free (regs->start);
- return REGS_UNALLOCATED;
- }
- regs->num_regs = need_regs;
- }
- else if (regs_allocated == REGS_REALLOCATE)
- { /* Yes. If we need more elements than were already
- allocated, reallocate them. If we need fewer, just
- leave it alone. */
- if (need_regs > regs->num_regs)
- {
- regs->start = re_realloc (regs->start, regoff_t, need_regs);
- if (BE (regs->start == NULL, 0))
- {
- if (regs->end != NULL)
- re_free (regs->end);
- return REGS_UNALLOCATED;
- }
- regs->end = re_realloc (regs->end, regoff_t, need_regs);
- if (BE (regs->end == NULL, 0))
- {
- re_free (regs->start);
- return REGS_UNALLOCATED;
- }
- regs->num_regs = need_regs;
- }
- }
- else
- {
- assert (regs_allocated == REGS_FIXED);
- /* This function may not be called with REGS_FIXED and nregs too big. */
- assert (regs->num_regs >= nregs);
- rval = REGS_FIXED;
- }
-
- /* Copy the regs. */
- for (i = 0; i < nregs; ++i)
- {
- regs->start[i] = pmatch[i].rm_so;
- regs->end[i] = pmatch[i].rm_eo;
- }
- for ( ; i < regs->num_regs; ++i)
- regs->start[i] = regs->end[i] = -1;
-
- return rval;
-}
-
-/* Set REGS to hold NUM_REGS registers, storing them in STARTS and
- ENDS. Subsequent matches using PATTERN_BUFFER and REGS will use
- this memory for recording register information. STARTS and ENDS
- must be allocated using the malloc library routine, and must each
- be at least NUM_REGS * sizeof (regoff_t) bytes long.
-
- If NUM_REGS == 0, then subsequent matches should allocate their own
- register data.
-
- Unless this function is called, the first search or match using
- PATTERN_BUFFER will allocate its own register data, without
- freeing the old data. */
-
-void
-re_set_registers (bufp, regs, num_regs, starts, ends)
- struct re_pattern_buffer *bufp;
- struct re_registers *regs;
- unsigned num_regs;
- regoff_t *starts, *ends;
-{
- if (num_regs)
- {
- bufp->regs_allocated = REGS_REALLOCATE;
- regs->num_regs = num_regs;
- regs->start = starts;
- regs->end = ends;
- }
- else
- {
- bufp->regs_allocated = REGS_UNALLOCATED;
- regs->num_regs = 0;
- regs->start = regs->end = (regoff_t *) 0;
- }
-}
-#ifdef _LIBC
-weak_alias (__re_set_registers, re_set_registers)
-#endif
-\f
-/* Entry points compatible with 4.2 BSD regex library. We don't define
- them unless specifically requested. */
-
-#if defined _REGEX_RE_COMP || defined _LIBC
-int
-# ifdef _LIBC
-weak_function
-# endif
-re_exec (s)
- const char *s;
-{
- return 0 == regexec (&re_comp_buf, s, 0, NULL, 0);
-}
-#endif /* _REGEX_RE_COMP */
-\f
-static re_node_set empty_set;
-
-/* Internal entry point. */
-
-/* Searches for a compiled pattern PREG in the string STRING, whose
- length is LENGTH. NMATCH, PMATCH, and EFLAGS have the same
- mingings with regexec. START, and RANGE have the same meanings
- with re_search.
- Return REG_NOERROR if we find a match, and REG_NOMATCH if not,
- otherwise return the error code.
- Note: We assume front end functions already check ranges.
- (START + RANGE >= 0 && START + RANGE <= LENGTH) */
-
-static reg_errcode_t
-re_search_internal (preg, string, length, start, range, stop, nmatch, pmatch,
- eflags)
- const regex_t *preg;
- const char *string;
- int length, start, range, stop, eflags;
- size_t nmatch;
- regmatch_t pmatch[];
-{
- reg_errcode_t err;
- re_dfa_t *dfa = (re_dfa_t *)preg->buffer;
- re_string_t input;
- int left_lim, right_lim, incr;
- int fl_longest_match, match_first, match_last = -1;
- int fast_translate, sb;
- re_match_context_t mctx;
- char *fastmap = ((preg->fastmap != NULL && preg->fastmap_accurate
- && range && !preg->can_be_null) ? preg->fastmap : NULL);
-
- /* Check if the DFA haven't been compiled. */
- if (BE (preg->used == 0 || dfa->init_state == NULL
- || dfa->init_state_word == NULL || dfa->init_state_nl == NULL
- || dfa->init_state_begbuf == NULL, 0))
- return REG_NOMATCH;
-
- re_node_set_init_empty (&empty_set);
- memset (&mctx, '\0', sizeof (re_match_context_t));
-
- /* We must check the longest matching, if nmatch > 0. */
- fl_longest_match = (nmatch != 0 || dfa->nbackref);
-
- err = re_string_allocate (&input, string, length, dfa->nodes_len + 1,
- preg->translate, preg->syntax & RE_ICASE);
- if (BE (err != REG_NOERROR, 0))
- goto free_return;
- input.stop = stop;
-
- err = match_ctx_init (&mctx, eflags, &input, dfa->nbackref * 2);
- if (BE (err != REG_NOERROR, 0))
- goto free_return;
-
- /* We will log all the DFA states through which the dfa pass,
- if nmatch > 1, or this dfa has "multibyte node", which is a
- back-reference or a node which can accept multibyte character or
- multi character collating element. */
- if (nmatch > 1 || dfa->has_mb_node)
- {
- mctx.state_log = re_malloc (re_dfastate_t *, dfa->nodes_len + 1);
- if (BE (mctx.state_log == NULL, 0))
- {
- err = REG_ESPACE;
- goto free_return;
- }
- }
- else
- mctx.state_log = NULL;
-
-#ifdef DEBUG
- /* We assume front-end functions already check them. */
- assert (start + range >= 0 && start + range <= length);
-#endif
-
- match_first = start;
- input.tip_context = ((eflags & REG_NOTBOL) ? CONTEXT_BEGBUF
- : CONTEXT_NEWLINE | CONTEXT_BEGBUF);
-
- /* Check incrementally whether of not the input string match. */
- incr = (range < 0) ? -1 : 1;
- left_lim = (range < 0) ? start + range : start;
- right_lim = (range < 0) ? start : start + range;
- sb = MB_CUR_MAX == 1;
- fast_translate = sb || !(preg->syntax & RE_ICASE || preg->translate);
-
- for (;;)
- {
- /* At first get the current byte from input string. */
- if (fastmap)
- {
- if (BE (fast_translate, 1))
- {
- unsigned RE_TRANSLATE_TYPE t
- = (unsigned RE_TRANSLATE_TYPE) preg->translate;
- if (BE (range >= 0, 1))
- {
- if (BE (t != NULL, 0))
- {
- while (BE (match_first < right_lim, 1)
- && !fastmap[t[(unsigned char) string[match_first]]])
- ++match_first;
- }
- else
- {
- while (BE (match_first < right_lim, 1)
- && !fastmap[(unsigned char) string[match_first]])
- ++match_first;
- }
- if (BE (match_first == right_lim, 0))
- {
- int ch = match_first >= length
- ? 0 : (unsigned char) string[match_first];
- if (!fastmap[t ? t[ch] : ch])
- break;
- }
- }
- else
- {
- while (match_first >= left_lim)
- {
- int ch = match_first >= length
- ? 0 : (unsigned char) string[match_first];
- if (fastmap[t ? t[ch] : ch])
- break;
- --match_first;
- }
- if (match_first < left_lim)
- break;
- }
- }
- else
- {
- int ch;
-
- do
- {
- /* In this case, we can't determine easily the current byte,
- since it might be a component byte of a multibyte
- character. Then we use the constructed buffer
- instead. */
- /* If MATCH_FIRST is out of the valid range, reconstruct the
- buffers. */
- if (input.raw_mbs_idx + input.valid_len <= match_first
- || match_first < input.raw_mbs_idx)
- {
- err = re_string_reconstruct (&input, match_first, eflags,
- preg->newline_anchor);
- if (BE (err != REG_NOERROR, 0))
- goto free_return;
- }
- /* If MATCH_FIRST is out of the buffer, leave it as '\0'.
- Note that MATCH_FIRST must not be smaller than 0. */
- ch = ((match_first >= length) ? 0
- : re_string_byte_at (&input,
- match_first - input.raw_mbs_idx));
- if (fastmap[ch])
- break;
- match_first += incr;
- }
- while (match_first >= left_lim && match_first <= right_lim);
- if (! fastmap[ch])
- break;
- }
- }
-
- /* Reconstruct the buffers so that the matcher can assume that
- the matching starts from the begining of the buffer. */
- err = re_string_reconstruct (&input, match_first, eflags,
- preg->newline_anchor);
- if (BE (err != REG_NOERROR, 0))
- goto free_return;
-#ifdef RE_ENABLE_I18N
- /* Eliminate it when it is a component of a multibyte character
- and isn't the head of a multibyte character. */
- if (sb || re_string_first_byte (&input, 0))
-#endif
- {
- /* It seems to be appropriate one, then use the matcher. */
- /* We assume that the matching starts from 0. */
- mctx.state_log_top = mctx.nbkref_ents = mctx.max_mb_elem_len = 0;
- match_last = check_matching (preg, &mctx, 0, fl_longest_match);
- if (match_last != -1)
- {
- if (BE (match_last == -2, 0))
- {
- err = REG_ESPACE;
- goto free_return;
- }
- else
- {
- mctx.match_last = match_last;
- if ((!preg->no_sub && nmatch > 1) || dfa->nbackref)
- {
- re_dfastate_t *pstate = mctx.state_log[match_last];
- mctx.last_node = check_halt_state_context (preg, pstate,
- &mctx, match_last);
- }
- if ((!preg->no_sub && nmatch > 1 && dfa->has_plural_match)
- || dfa->nbackref)
- {
- err = prune_impossible_nodes (preg, &mctx);
- if (err == REG_NOERROR)
- break;
- if (BE (err != REG_NOMATCH, 0))
- goto free_return;
- }
- else
- break; /* We found a matching. */
- }
- }
- match_ctx_clean (&mctx);
- }
- /* Update counter. */
- match_first += incr;
- if (match_first < left_lim || right_lim < match_first)
- break;
- }
-
- /* Set pmatch[] if we need. */
- if (match_last != -1 && nmatch > 0)
- {
- int reg_idx;
-
- /* Initialize registers. */
- for (reg_idx = 0; reg_idx < nmatch; ++reg_idx)
- pmatch[reg_idx].rm_so = pmatch[reg_idx].rm_eo = -1;
-
- /* Set the points where matching start/end. */
- pmatch[0].rm_so = 0;
- pmatch[0].rm_eo = mctx.match_last;
-
- if (!preg->no_sub && nmatch > 1)
- {
- err = set_regs (preg, &mctx, nmatch, pmatch,
- dfa->has_plural_match && dfa->nbackref > 0);
- if (BE (err != REG_NOERROR, 0))
- goto free_return;
- }
-
- /* At last, add the offset to the each registers, since we slided
- the buffers so that We can assume that the matching starts from 0. */
- for (reg_idx = 0; reg_idx < nmatch; ++reg_idx)
- if (pmatch[reg_idx].rm_so != -1)
- {
- pmatch[reg_idx].rm_so += match_first;
- pmatch[reg_idx].rm_eo += match_first;
- }
- }
- err = (match_last == -1) ? REG_NOMATCH : REG_NOERROR;
- free_return:
- re_free (mctx.state_log);
- if (dfa->nbackref)
- match_ctx_free (&mctx);
- re_string_destruct (&input);
- return err;
-}
-
-static reg_errcode_t
-prune_impossible_nodes (preg, mctx)
- const regex_t *preg;
- re_match_context_t *mctx;
-{
- int halt_node, match_last;
- reg_errcode_t ret;
- re_dfa_t *dfa = (re_dfa_t *)preg->buffer;
- re_dfastate_t **sifted_states;
- re_dfastate_t **lim_states = NULL;
- re_sift_context_t sctx;
-#ifdef DEBUG
- assert (mctx->state_log != NULL);
-#endif
- match_last = mctx->match_last;
- halt_node = mctx->last_node;
- sifted_states = re_malloc (re_dfastate_t *, match_last + 1);
- if (BE (sifted_states == NULL, 0))
- {
- ret = REG_ESPACE;
- goto free_return;
- }
- if (dfa->nbackref)
- {
- lim_states = re_malloc (re_dfastate_t *, match_last + 1);
- if (BE (lim_states == NULL, 0))
- {
- ret = REG_ESPACE;
- goto free_return;
- }
- while (1)
- {
- memset (lim_states, '\0',
- sizeof (re_dfastate_t *) * (match_last + 1));
- match_ctx_clear_flag (mctx);
- sift_ctx_init (&sctx, sifted_states, lim_states, halt_node,
- match_last, 0);
- ret = sift_states_backward (preg, mctx, &sctx);
- re_node_set_free (&sctx.limits);
- if (BE (ret != REG_NOERROR, 0))
- goto free_return;
- if (sifted_states[0] != NULL || lim_states[0] != NULL)
- break;
- do
- {
- --match_last;
- if (match_last < 0)
- {
- ret = REG_NOMATCH;
- goto free_return;
- }
- } while (!mctx->state_log[match_last]->halt);
- halt_node = check_halt_state_context (preg,
- mctx->state_log[match_last],
- mctx, match_last);
- }
- ret = merge_state_array (dfa, sifted_states, lim_states,
- match_last + 1);
- re_free (lim_states);
- lim_states = NULL;
- if (BE (ret != REG_NOERROR, 0))
- goto free_return;
- }
- else
- {
- sift_ctx_init (&sctx, sifted_states, lim_states, halt_node,
- match_last, 0);
- ret = sift_states_backward (preg, mctx, &sctx);
- re_node_set_free (&sctx.limits);
- if (BE (ret != REG_NOERROR, 0))
- goto free_return;
- }
- re_free (mctx->state_log);
- mctx->state_log = sifted_states;
- sifted_states = NULL;
- mctx->last_node = halt_node;
- mctx->match_last = match_last;
- ret = REG_NOERROR;
- free_return:
- re_free (sifted_states);
- re_free (lim_states);
- return ret;
-}
-
-/* Acquire an initial state and return it.
- We must select appropriate initial state depending on the context,
- since initial states may have constraints like "\<", "^", etc.. */
-
-static inline re_dfastate_t *
-acquire_init_state_context (err, preg, mctx, idx)
- reg_errcode_t *err;
- const regex_t *preg;
- const re_match_context_t *mctx;
- int idx;
-{
- re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
-
- *err = REG_NOERROR;
- if (dfa->init_state->has_constraint)
- {
- unsigned int context;
- context = re_string_context_at (mctx->input, idx - 1, mctx->eflags,
- preg->newline_anchor);
- if (IS_WORD_CONTEXT (context))
- return dfa->init_state_word;
- else if (IS_ORDINARY_CONTEXT (context))
- return dfa->init_state;
- else if (IS_BEGBUF_CONTEXT (context) && IS_NEWLINE_CONTEXT (context))
- return dfa->init_state_begbuf;
- else if (IS_NEWLINE_CONTEXT (context))
- return dfa->init_state_nl;
- else if (IS_BEGBUF_CONTEXT (context))
- {
- /* It is relatively rare case, then calculate on demand. */
- return re_acquire_state_context (err, dfa,
- dfa->init_state->entrance_nodes,
- context);
- }
- else
- /* Must not happen? */
- return dfa->init_state;
- }
- else
- return dfa->init_state;
-}
-
-/* Check whether the regular expression match input string INPUT or not,
- and return the index where the matching end, return -1 if not match,
- or return -2 in case of an error.
- FL_SEARCH means we must search where the matching starts,
- FL_LONGEST_MATCH means we want the POSIX longest matching.
- Note that the matcher assume that the maching starts from the current
- index of the buffer. */
-
-static int
-check_matching (preg, mctx, fl_search, fl_longest_match)
- const regex_t *preg;
- re_match_context_t *mctx;
- int fl_search, fl_longest_match;
-{
- re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- reg_errcode_t err;
- int match = 0;
- int match_last = -1;
- int cur_str_idx = re_string_cur_idx (mctx->input);
- re_dfastate_t *cur_state;
-
- cur_state = acquire_init_state_context (&err, preg, mctx, cur_str_idx);
- /* An initial state must not be NULL(invalid state). */
- if (BE (cur_state == NULL, 0))
- return -2;
- if (mctx->state_log != NULL)
- mctx->state_log[cur_str_idx] = cur_state;
-
- /* Check OP_OPEN_SUBEXP in the initial state in case that we use them
- later. E.g. Processing back references. */
- if (dfa->nbackref)
- {
- err = check_subexp_matching_top (dfa, mctx, &cur_state->nodes, 0);
- if (BE (err != REG_NOERROR, 0))
- return err;
- }
-
- if (cur_state->has_backref)
- {
- err = transit_state_bkref (preg, &cur_state->nodes, mctx);
- if (BE (err != REG_NOERROR, 0))
- return err;
- }
-
- /* If the RE accepts NULL string. */
- if (cur_state->halt)
- {
- if (!cur_state->has_constraint
- || check_halt_state_context (preg, cur_state, mctx, cur_str_idx))
- {
- if (!fl_longest_match)
- return cur_str_idx;
- else
- {
- match_last = cur_str_idx;
- match = 1;
- }
- }
- }
-
- while (!re_string_eoi (mctx->input))
- {
- cur_state = transit_state (&err, preg, mctx, cur_state,
- fl_search && !match);
- if (cur_state == NULL) /* Reached at the invalid state or an error. */
- {
- cur_str_idx = re_string_cur_idx (mctx->input);
- if (BE (err != REG_NOERROR, 0))
- return -2;
- if (fl_search && !match)
- {
- /* Restart from initial state, since we are searching
- the point from where matching start. */
-#ifdef RE_ENABLE_I18N
- if (MB_CUR_MAX == 1
- || re_string_first_byte (mctx->input, cur_str_idx))
-#endif /* RE_ENABLE_I18N */
- cur_state = acquire_init_state_context (&err, preg, mctx,
- cur_str_idx);
- if (BE (cur_state == NULL && err != REG_NOERROR, 0))
- return -2;
- if (mctx->state_log != NULL)
- mctx->state_log[cur_str_idx] = cur_state;
- }
- else if (!fl_longest_match && match)
- break;
- else /* (fl_longest_match && match) || (!fl_search && !match) */
- {
- if (mctx->state_log == NULL)
- break;
- else
- {
- int max = mctx->state_log_top;
- for (; cur_str_idx <= max; ++cur_str_idx)
- if (mctx->state_log[cur_str_idx] != NULL)
- break;
- if (cur_str_idx > max)
- break;
- }
- }
- }
-
- if (cur_state != NULL && cur_state->halt)
- {
- /* Reached at a halt state.
- Check the halt state can satisfy the current context. */
- if (!cur_state->has_constraint
- || check_halt_state_context (preg, cur_state, mctx,
- re_string_cur_idx (mctx->input)))
- {
- /* We found an appropriate halt state. */
- match_last = re_string_cur_idx (mctx->input);
- match = 1;
- if (!fl_longest_match)
- break;
- }
- }
- }
- return match_last;
-}
-
-/* Check NODE match the current context. */
-
-static int check_halt_node_context (dfa, node, context)
- const re_dfa_t *dfa;
- int node;
- unsigned int context;
-{
- re_token_type_t type = dfa->nodes[node].type;
- unsigned int constraint = dfa->nodes[node].constraint;
- if (type != END_OF_RE)
- return 0;
- if (!constraint)
- return 1;
- if (NOT_SATISFY_NEXT_CONSTRAINT (constraint, context))
- return 0;
- return 1;
-}
-
-/* Check the halt state STATE match the current context.
- Return 0 if not match, if the node, STATE has, is a halt node and
- match the context, return the node. */
-
-static int
-check_halt_state_context (preg, state, mctx, idx)
- const regex_t *preg;
- const re_dfastate_t *state;
- const re_match_context_t *mctx;
- int idx;
-{
- re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- int i;
- unsigned int context;
-#ifdef DEBUG
- assert (state->halt);
-#endif
- context = re_string_context_at (mctx->input, idx, mctx->eflags,
- preg->newline_anchor);
- for (i = 0; i < state->nodes.nelem; ++i)
- if (check_halt_node_context (dfa, state->nodes.elems[i], context))
- return state->nodes.elems[i];
- return 0;
-}
-
-/* Compute the next node to which "NFA" transit from NODE("NFA" is a NFA
- corresponding to the DFA).
- Return the destination node, and update EPS_VIA_NODES, return -1 in case
- of errors. */
-
-static int
-proceed_next_node (preg, nregs, regs, mctx, pidx, node, eps_via_nodes, fs)
- const regex_t *preg;
- regmatch_t *regs;
- const re_match_context_t *mctx;
- int nregs, *pidx, node;
- re_node_set *eps_via_nodes;
- struct re_fail_stack_t *fs;
-{
- re_dfa_t *dfa = (re_dfa_t *)preg->buffer;
- int i, err, dest_node;
- dest_node = -1;
- if (IS_EPSILON_NODE (dfa->nodes[node].type))
- {
- re_node_set *cur_nodes = &mctx->state_log[*pidx]->nodes;
- int ndest, dest_nodes[2];
- err = re_node_set_insert (eps_via_nodes, node);
- if (BE (err < 0, 0))
- return -1;
- /* Pick up valid destinations. */
- for (ndest = 0, i = 0; i < dfa->edests[node].nelem; ++i)
- {
- int candidate = dfa->edests[node].elems[i];
- if (!re_node_set_contains (cur_nodes, candidate))
- continue;
- dest_nodes[0] = (ndest == 0) ? candidate : dest_nodes[0];
- dest_nodes[1] = (ndest == 1) ? candidate : dest_nodes[1];
- ++ndest;
- }
- if (ndest <= 1)
- return ndest == 0 ? -1 : (ndest == 1 ? dest_nodes[0] : 0);
- /* In order to avoid infinite loop like "(a*)*". */
- if (re_node_set_contains (eps_via_nodes, dest_nodes[0]))
- return dest_nodes[1];
- if (fs != NULL)
- push_fail_stack (fs, *pidx, dest_nodes, nregs, regs, eps_via_nodes);
- return dest_nodes[0];
- }
- else
- {
- int naccepted = 0;
- re_token_type_t type = dfa->nodes[node].type;
-
-#ifdef RE_ENABLE_I18N
- if (ACCEPT_MB_NODE (type))
- naccepted = check_node_accept_bytes (preg, node, mctx->input, *pidx);
- else
-#endif /* RE_ENABLE_I18N */
- if (type == OP_BACK_REF)
- {
- int subexp_idx = dfa->nodes[node].opr.idx;
- naccepted = regs[subexp_idx].rm_eo - regs[subexp_idx].rm_so;
- if (fs != NULL)
- {
- if (regs[subexp_idx].rm_so == -1 || regs[subexp_idx].rm_eo == -1)
- return -1;
- else if (naccepted)
- {
- char *buf = (char *) re_string_get_buffer (mctx->input);
- if (memcmp (buf + regs[subexp_idx].rm_so, buf + *pidx,
- naccepted) != 0)
- return -1;
- }
- }
-
- if (naccepted == 0)
- {
- err = re_node_set_insert (eps_via_nodes, node);
- if (BE (err < 0, 0))
- return -2;
- dest_node = dfa->edests[node].elems[0];
- if (re_node_set_contains (&mctx->state_log[*pidx]->nodes,
- dest_node))
- return dest_node;
- }
- }
-
- if (naccepted != 0
- || check_node_accept (preg, dfa->nodes + node, mctx, *pidx))
- {
- dest_node = dfa->nexts[node];
- *pidx = (naccepted == 0) ? *pidx + 1 : *pidx + naccepted;
- if (fs && (*pidx > mctx->match_last || mctx->state_log[*pidx] == NULL
- || !re_node_set_contains (&mctx->state_log[*pidx]->nodes,
- dest_node)))
- return -1;
- re_node_set_empty (eps_via_nodes);
- return dest_node;
- }
- }
- return -1;
-}
-
-static reg_errcode_t
-push_fail_stack (fs, str_idx, dests, nregs, regs, eps_via_nodes)
- struct re_fail_stack_t *fs;
- int str_idx, *dests, nregs;
- regmatch_t *regs;
- re_node_set *eps_via_nodes;
-{
- reg_errcode_t err;
- int num = fs->num++;
- if (fs->num == fs->alloc)
- {
- struct re_fail_stack_ent_t *new_array;
- fs->alloc *= 2;
- new_array = realloc (fs->stack, (sizeof (struct re_fail_stack_ent_t)
- * fs->alloc));
- if (new_array == NULL)
- return REG_ESPACE;
- fs->stack = new_array;
- }
- fs->stack[num].idx = str_idx;
- fs->stack[num].node = dests[1];
- fs->stack[num].regs = re_malloc (regmatch_t, nregs);
- memcpy (fs->stack[num].regs, regs, sizeof (regmatch_t) * nregs);
- err = re_node_set_init_copy (&fs->stack[num].eps_via_nodes, eps_via_nodes);
- return err;
-}
-
-static int
-pop_fail_stack (fs, pidx, nregs, regs, eps_via_nodes)
- struct re_fail_stack_t *fs;
- int *pidx, nregs;
- regmatch_t *regs;
- re_node_set *eps_via_nodes;
-{
- int num = --fs->num;
- assert (num >= 0);
- *pidx = fs->stack[num].idx;
- memcpy (regs, fs->stack[num].regs, sizeof (regmatch_t) * nregs);
- re_node_set_free (eps_via_nodes);
- re_free (fs->stack[num].regs);
- *eps_via_nodes = fs->stack[num].eps_via_nodes;
- return fs->stack[num].node;
-}
-
-/* Set the positions where the subexpressions are starts/ends to registers
- PMATCH.
- Note: We assume that pmatch[0] is already set, and
- pmatch[i].rm_so == pmatch[i].rm_eo == -1 (i > 1). */
-
-static reg_errcode_t
-set_regs (preg, mctx, nmatch, pmatch, fl_backtrack)
- const regex_t *preg;
- const re_match_context_t *mctx;
- size_t nmatch;
- regmatch_t *pmatch;
- int fl_backtrack;
-{
- re_dfa_t *dfa = (re_dfa_t *)preg->buffer;
- int idx, cur_node, real_nmatch;
- re_node_set eps_via_nodes;
- struct re_fail_stack_t *fs;
- struct re_fail_stack_t fs_body = {0, 2, NULL};
-#ifdef DEBUG
- assert (nmatch > 1);
- assert (mctx->state_log != NULL);
-#endif
- if (fl_backtrack)
- {
- fs = &fs_body;
- fs->stack = re_malloc (struct re_fail_stack_ent_t, fs->alloc);
- }
- else
- fs = NULL;
- cur_node = dfa->init_node;
- real_nmatch = (nmatch <= preg->re_nsub) ? nmatch : preg->re_nsub + 1;
- re_node_set_init_empty (&eps_via_nodes);
- for (idx = pmatch[0].rm_so; idx <= pmatch[0].rm_eo ;)
- {
- update_regs (dfa, pmatch, cur_node, idx, real_nmatch);
- if (idx == pmatch[0].rm_eo && cur_node == mctx->last_node)
- {
- int reg_idx;
- if (fs)
- {
- for (reg_idx = 0; reg_idx < nmatch; ++reg_idx)
- if (pmatch[reg_idx].rm_so > -1 && pmatch[reg_idx].rm_eo == -1)
- break;
- if (reg_idx == nmatch)
- {
- re_node_set_free (&eps_via_nodes);
- return free_fail_stack_return (fs);
- }
- cur_node = pop_fail_stack (fs, &idx, nmatch, pmatch,
- &eps_via_nodes);
- }
- else
- {
- re_node_set_free (&eps_via_nodes);
- return REG_NOERROR;
- }
- }
-
- /* Proceed to next node. */
- cur_node = proceed_next_node (preg, nmatch, pmatch, mctx, &idx, cur_node,
- &eps_via_nodes, fs);
-
- if (BE (cur_node < 0, 0))
- {
- if (cur_node == -2)
- return REG_ESPACE;
- if (fs)
- cur_node = pop_fail_stack (fs, &idx, nmatch, pmatch,
- &eps_via_nodes);
- else
- {
- re_node_set_free (&eps_via_nodes);
- return REG_NOMATCH;
- }
- }
- }
- re_node_set_free (&eps_via_nodes);
- return free_fail_stack_return (fs);
-}
-
-static reg_errcode_t
-free_fail_stack_return (fs)
- struct re_fail_stack_t *fs;
-{
- if (fs)
- {
- int fs_idx;
- for (fs_idx = 0; fs_idx < fs->num; ++fs_idx)
- {
- re_node_set_free (&fs->stack[fs_idx].eps_via_nodes);
- re_free (fs->stack[fs_idx].regs);
- }
- re_free (fs->stack);
- }
- return REG_NOERROR;
-}
-
-static void
-update_regs (dfa, pmatch, cur_node, cur_idx, nmatch)
- re_dfa_t *dfa;
- regmatch_t *pmatch;
- int cur_node, cur_idx, nmatch;
-{
- int type = dfa->nodes[cur_node].type;
- int reg_num;
- if (type != OP_OPEN_SUBEXP && type != OP_CLOSE_SUBEXP)
- return;
- reg_num = dfa->nodes[cur_node].opr.idx + 1;
- if (reg_num >= nmatch)
- return;
- if (type == OP_OPEN_SUBEXP)
- {
- /* We are at the first node of this sub expression. */
- pmatch[reg_num].rm_so = cur_idx;
- pmatch[reg_num].rm_eo = -1;
- }
- else if (type == OP_CLOSE_SUBEXP)
- /* We are at the first node of this sub expression. */
- pmatch[reg_num].rm_eo = cur_idx;
-}
-
-#define NUMBER_OF_STATE 1
-
-/* This function checks the STATE_LOG from the SCTX->last_str_idx to 0
- and sift the nodes in each states according to the following rules.
- Updated state_log will be wrote to STATE_LOG.
-
- Rules: We throw away the Node `a' in the STATE_LOG[STR_IDX] if...
- 1. When STR_IDX == MATCH_LAST(the last index in the state_log):
- If `a' isn't the LAST_NODE and `a' can't epsilon transit to
- the LAST_NODE, we throw away the node `a'.
- 2. When 0 <= STR_IDX < MATCH_LAST and `a' accepts
- string `s' and transit to `b':
- i. If 'b' isn't in the STATE_LOG[STR_IDX+strlen('s')], we throw
- away the node `a'.
- ii. If 'b' is in the STATE_LOG[STR_IDX+strlen('s')] but 'b' is
- throwed away, we throw away the node `a'.
- 3. When 0 <= STR_IDX < n and 'a' epsilon transit to 'b':
- i. If 'b' isn't in the STATE_LOG[STR_IDX], we throw away the
- node `a'.
- ii. If 'b' is in the STATE_LOG[STR_IDX] but 'b' is throwed away,
- we throw away the node `a'. */
-
-#define STATE_NODE_CONTAINS(state,node) \
- ((state) != NULL && re_node_set_contains (&(state)->nodes, node))
-
-static reg_errcode_t
-sift_states_backward (preg, mctx, sctx)
- const regex_t *preg;
- re_match_context_t *mctx;
- re_sift_context_t *sctx;
-{
- reg_errcode_t err;
- re_dfa_t *dfa = (re_dfa_t *)preg->buffer;
- int null_cnt = 0;
- int str_idx = sctx->last_str_idx;
- re_node_set cur_dest;
- re_node_set *cur_src; /* Points the state_log[str_idx]->nodes */
-
-#ifdef DEBUG
- assert (mctx->state_log != NULL && mctx->state_log[str_idx] != NULL);
-#endif
- cur_src = &mctx->state_log[str_idx]->nodes;
-
- /* Build sifted state_log[str_idx]. It has the nodes which can epsilon
- transit to the last_node and the last_node itself. */
- err = re_node_set_init_1 (&cur_dest, sctx->last_node);
- if (BE (err != REG_NOERROR, 0))
- return err;
- err = update_cur_sifted_state (preg, mctx, sctx, str_idx, &cur_dest);
- if (BE (err != REG_NOERROR, 0))
- goto free_return;
-
- /* Then check each states in the state_log. */
- while (str_idx > 0)
- {
- int i, ret;
- /* Update counters. */
- null_cnt = (sctx->sifted_states[str_idx] == NULL) ? null_cnt + 1 : 0;
- if (null_cnt > mctx->max_mb_elem_len)
- {
- memset (sctx->sifted_states, '\0',
- sizeof (re_dfastate_t *) * str_idx);
- re_node_set_free (&cur_dest);
- return REG_NOERROR;
- }
- re_node_set_empty (&cur_dest);
- --str_idx;
- cur_src = ((mctx->state_log[str_idx] == NULL) ? &empty_set
- : &mctx->state_log[str_idx]->nodes);
-
- /* Then build the next sifted state.
- We build the next sifted state on `cur_dest', and update
- `sifted_states[str_idx]' with `cur_dest'.
- Note:
- `cur_dest' is the sifted state from `state_log[str_idx + 1]'.
- `cur_src' points the node_set of the old `state_log[str_idx]'. */
- for (i = 0; i < cur_src->nelem; i++)
- {
- int prev_node = cur_src->elems[i];
- int naccepted = 0;
- re_token_type_t type = dfa->nodes[prev_node].type;
-
- if (IS_EPSILON_NODE(type))
- continue;
-#ifdef RE_ENABLE_I18N
- /* If the node may accept `multi byte'. */
- if (ACCEPT_MB_NODE (type))
- naccepted = sift_states_iter_mb (preg, mctx, sctx, prev_node,
- str_idx, sctx->last_str_idx);
-
-#endif /* RE_ENABLE_I18N */
- /* We don't check backreferences here.
- See update_cur_sifted_state(). */
-
- if (!naccepted
- && check_node_accept (preg, dfa->nodes + prev_node, mctx,
- str_idx)
- && STATE_NODE_CONTAINS (sctx->sifted_states[str_idx + 1],
- dfa->nexts[prev_node]))
- naccepted = 1;
-
- if (naccepted == 0)
- continue;
-
- if (sctx->limits.nelem)
- {
- int to_idx = str_idx + naccepted;
- if (check_dst_limits (dfa, &sctx->limits, mctx,
- dfa->nexts[prev_node], to_idx,
- prev_node, str_idx))
- continue;
- }
- ret = re_node_set_insert (&cur_dest, prev_node);
- if (BE (ret == -1, 0))
- {
- err = REG_ESPACE;
- goto free_return;
- }
- }
-
- /* Add all the nodes which satisfy the following conditions:
- - It can epsilon transit to a node in CUR_DEST.
- - It is in CUR_SRC.
- And update state_log. */
- err = update_cur_sifted_state (preg, mctx, sctx, str_idx, &cur_dest);
- if (BE (err != REG_NOERROR, 0))
- goto free_return;
- }
- err = REG_NOERROR;
- free_return:
- re_node_set_free (&cur_dest);
- return err;
-}
-
-/* Helper functions. */
-
-static inline reg_errcode_t
-clean_state_log_if_need (mctx, next_state_log_idx)
- re_match_context_t *mctx;
- int next_state_log_idx;
-{
- int top = mctx->state_log_top;
-
- if (next_state_log_idx >= mctx->input->bufs_len
- || (next_state_log_idx >= mctx->input->valid_len
- && mctx->input->valid_len < mctx->input->len))
- {
- reg_errcode_t err;
- err = extend_buffers (mctx);
- if (BE (err != REG_NOERROR, 0))
- return err;
- }
-
- if (top < next_state_log_idx)
- {
- memset (mctx->state_log + top + 1, '\0',
- sizeof (re_dfastate_t *) * (next_state_log_idx - top));
- mctx->state_log_top = next_state_log_idx;
- }
- return REG_NOERROR;
-}
-
-static reg_errcode_t
-merge_state_array (dfa, dst, src, num)
- re_dfa_t *dfa;
- re_dfastate_t **dst;
- re_dfastate_t **src;
- int num;
-{
- int st_idx;
- reg_errcode_t err;
- for (st_idx = 0; st_idx < num; ++st_idx)
- {
- if (dst[st_idx] == NULL)
- dst[st_idx] = src[st_idx];
- else if (src[st_idx] != NULL)
- {
- re_node_set merged_set;
- err = re_node_set_init_union (&merged_set, &dst[st_idx]->nodes,
- &src[st_idx]->nodes);
- if (BE (err != REG_NOERROR, 0))
- return err;
- dst[st_idx] = re_acquire_state (&err, dfa, &merged_set);
- re_node_set_free (&merged_set);
- if (BE (err != REG_NOERROR, 0))
- return err;
- }
- }
- return REG_NOERROR;
-}
-
-static reg_errcode_t
-update_cur_sifted_state (preg, mctx, sctx, str_idx, dest_nodes)
- const regex_t *preg;
- re_match_context_t *mctx;
- re_sift_context_t *sctx;
- int str_idx;
- re_node_set *dest_nodes;
-{
- reg_errcode_t err;
- re_dfa_t *dfa = (re_dfa_t *)preg->buffer;
- const re_node_set *candidates;
- candidates = ((mctx->state_log[str_idx] == NULL) ? &empty_set
- : &mctx->state_log[str_idx]->nodes);
-
- /* At first, add the nodes which can epsilon transit to a node in
- DEST_NODE. */
- if (dest_nodes->nelem)
- {
- err = add_epsilon_src_nodes (dfa, dest_nodes, candidates);
- if (BE (err != REG_NOERROR, 0))
- return err;
- }
-
- /* Then, check the limitations in the current sift_context. */
- if (dest_nodes->nelem && sctx->limits.nelem)
- {
- err = check_subexp_limits (dfa, dest_nodes, candidates, &sctx->limits,
- mctx->bkref_ents, str_idx);
- if (BE (err != REG_NOERROR, 0))
- return err;
- }
-
- /* Update state_log. */
- sctx->sifted_states[str_idx] = re_acquire_state (&err, dfa, dest_nodes);
- if (BE (sctx->sifted_states[str_idx] == NULL && err != REG_NOERROR, 0))
- return err;
-
- if ((mctx->state_log[str_idx] != NULL
- && mctx->state_log[str_idx]->has_backref))
- {
- err = sift_states_bkref (preg, mctx, sctx, str_idx, dest_nodes);
- if (BE (err != REG_NOERROR, 0))
- return err;
- }
- return REG_NOERROR;
-}
-
-static reg_errcode_t
-add_epsilon_src_nodes (dfa, dest_nodes, candidates)
- re_dfa_t *dfa;
- re_node_set *dest_nodes;
- const re_node_set *candidates;
-{
- reg_errcode_t err;
- int src_idx;
- re_node_set src_copy;
-
- err = re_node_set_init_copy (&src_copy, dest_nodes);
- if (BE (err != REG_NOERROR, 0))
- return err;
- for (src_idx = 0; src_idx < src_copy.nelem; ++src_idx)
- {
- err = re_node_set_add_intersect (dest_nodes, candidates,
- dfa->inveclosures
- + src_copy.elems[src_idx]);
- if (BE (err != REG_NOERROR, 0))
- {
- re_node_set_free (&src_copy);
- return err;
- }
- }
- re_node_set_free (&src_copy);
- return REG_NOERROR;
-}
-
-static reg_errcode_t
-sub_epsilon_src_nodes (dfa, node, dest_nodes, candidates)
- re_dfa_t *dfa;
- int node;
- re_node_set *dest_nodes;
- const re_node_set *candidates;
-{
- int ecl_idx;
- reg_errcode_t err;
- re_node_set *inv_eclosure = dfa->inveclosures + node;
- re_node_set except_nodes;
- re_node_set_init_empty (&except_nodes);
- for (ecl_idx = 0; ecl_idx < inv_eclosure->nelem; ++ecl_idx)
- {
- int cur_node = inv_eclosure->elems[ecl_idx];
- if (cur_node == node)
- continue;
- if (IS_EPSILON_NODE (dfa->nodes[cur_node].type))
- {
- int edst1 = dfa->edests[cur_node].elems[0];
- int edst2 = ((dfa->edests[cur_node].nelem > 1)
- ? dfa->edests[cur_node].elems[1] : -1);
- if ((!re_node_set_contains (inv_eclosure, edst1)
- && re_node_set_contains (dest_nodes, edst1))
- || (edst2 > 0
- && !re_node_set_contains (inv_eclosure, edst2)
- && re_node_set_contains (dest_nodes, edst2)))
- {
- err = re_node_set_add_intersect (&except_nodes, candidates,
- dfa->inveclosures + cur_node);
- if (BE (err != REG_NOERROR, 0))
- {
- re_node_set_free (&except_nodes);
- return err;
- }
- }
- }
- }
- for (ecl_idx = 0; ecl_idx < inv_eclosure->nelem; ++ecl_idx)
- {
- int cur_node = inv_eclosure->elems[ecl_idx];
- if (!re_node_set_contains (&except_nodes, cur_node))
- {
- int idx = re_node_set_contains (dest_nodes, cur_node) - 1;
- re_node_set_remove_at (dest_nodes, idx);
- }
- }
- re_node_set_free (&except_nodes);
- return REG_NOERROR;
-}
-
-static int
-check_dst_limits (dfa, limits, mctx, dst_node, dst_idx, src_node, src_idx)
- re_dfa_t *dfa;
- re_node_set *limits;
- re_match_context_t *mctx;
- int dst_node, dst_idx, src_node, src_idx;
-{
- int lim_idx, src_pos, dst_pos;
-
- for (lim_idx = 0; lim_idx < limits->nelem; ++lim_idx)
- {
- int subexp_idx;
- struct re_backref_cache_entry *ent;
- ent = mctx->bkref_ents + limits->elems[lim_idx];
- subexp_idx = dfa->nodes[ent->node].opr.idx - 1;
-
- dst_pos = check_dst_limits_calc_pos (dfa, mctx, limits->elems[lim_idx],
- dfa->eclosures + dst_node,
- subexp_idx, dst_node, dst_idx);
- src_pos = check_dst_limits_calc_pos (dfa, mctx, limits->elems[lim_idx],
- dfa->eclosures + src_node,
- subexp_idx, src_node, src_idx);
-
- /* In case of:
- <src> <dst> ( <subexp> )
- ( <subexp> ) <src> <dst>
- ( <subexp1> <src> <subexp2> <dst> <subexp3> ) */
- if (src_pos == dst_pos)
- continue; /* This is unrelated limitation. */
- else
- return 1;
- }
- return 0;
-}
-
-static int
-check_dst_limits_calc_pos (dfa, mctx, limit, eclosures, subexp_idx, node,
- str_idx)
- re_dfa_t *dfa;
- re_match_context_t *mctx;
- re_node_set *eclosures;
- int limit, subexp_idx, node, str_idx;
-{
- struct re_backref_cache_entry *lim = mctx->bkref_ents + limit;
- int pos = (str_idx < lim->subexp_from ? -1
- : (lim->subexp_to < str_idx ? 1 : 0));
- if (pos == 0
- && (str_idx == lim->subexp_from || str_idx == lim->subexp_to))
- {
- int node_idx;
- for (node_idx = 0; node_idx < eclosures->nelem; ++node_idx)
- {
- int node = eclosures->elems[node_idx];
- re_token_type_t type= dfa->nodes[node].type;
- if (type == OP_BACK_REF)
- {
- int bi = search_cur_bkref_entry (mctx, str_idx);
- for (; bi < mctx->nbkref_ents; ++bi)
- {
- struct re_backref_cache_entry *ent = mctx->bkref_ents + bi;
- if (ent->str_idx > str_idx)
- break;
- if (ent->node == node && ent->subexp_from == ent->subexp_to)
- {
- int cpos, dst;
- dst = dfa->edests[node].elems[0];
- cpos = check_dst_limits_calc_pos (dfa, mctx, limit,
- dfa->eclosures + dst,
- subexp_idx, dst,
- str_idx);
- if ((str_idx == lim->subexp_from && cpos == -1)
- || (str_idx == lim->subexp_to && cpos == 0))
- return cpos;
- }
- }
- }
- if (type == OP_OPEN_SUBEXP && subexp_idx == dfa->nodes[node].opr.idx
- && str_idx == lim->subexp_from)
- {
- pos = -1;
- break;
- }
- if (type == OP_CLOSE_SUBEXP && subexp_idx == dfa->nodes[node].opr.idx
- && str_idx == lim->subexp_to)
- break;
- }
- if (node_idx == eclosures->nelem && str_idx == lim->subexp_to)
- pos = 1;
- }
- return pos;
-}
-
-/* Check the limitations of sub expressions LIMITS, and remove the nodes
- which are against limitations from DEST_NODES. */
-
-static reg_errcode_t
-check_subexp_limits (dfa, dest_nodes, candidates, limits, bkref_ents, str_idx)
- re_dfa_t *dfa;
- re_node_set *dest_nodes;
- const re_node_set *candidates;
- re_node_set *limits;
- struct re_backref_cache_entry *bkref_ents;
- int str_idx;
-{
- reg_errcode_t err;
- int node_idx, lim_idx;
-
- for (lim_idx = 0; lim_idx < limits->nelem; ++lim_idx)
- {
- int subexp_idx;
- struct re_backref_cache_entry *ent;
- ent = bkref_ents + limits->elems[lim_idx];
-
- if (str_idx <= ent->subexp_from || ent->str_idx < str_idx)
- continue; /* This is unrelated limitation. */
-
- subexp_idx = dfa->nodes[ent->node].opr.idx - 1;
- if (ent->subexp_to == str_idx)
- {
- int ops_node = -1;
- int cls_node = -1;
- for (node_idx = 0; node_idx < dest_nodes->nelem; ++node_idx)
- {
- int node = dest_nodes->elems[node_idx];
- re_token_type_t type= dfa->nodes[node].type;
- if (type == OP_OPEN_SUBEXP
- && subexp_idx == dfa->nodes[node].opr.idx)
- ops_node = node;
- else if (type == OP_CLOSE_SUBEXP
- && subexp_idx == dfa->nodes[node].opr.idx)
- cls_node = node;
- }
-
- /* Check the limitation of the open subexpression. */
- /* Note that (ent->subexp_to = str_idx != ent->subexp_from). */
- if (ops_node >= 0)
- {
- err = sub_epsilon_src_nodes(dfa, ops_node, dest_nodes,
- candidates);
- if (BE (err != REG_NOERROR, 0))
- return err;
- }
- /* Check the limitation of the close subexpression. */
- for (node_idx = 0; node_idx < dest_nodes->nelem; ++node_idx)
- {
- int node = dest_nodes->elems[node_idx];
- if (!re_node_set_contains (dfa->inveclosures + node, cls_node)
- && !re_node_set_contains (dfa->eclosures + node, cls_node))
- {
- /* It is against this limitation.
- Remove it form the current sifted state. */
- err = sub_epsilon_src_nodes(dfa, node, dest_nodes,
- candidates);
- if (BE (err != REG_NOERROR, 0))
- return err;
- --node_idx;
- }
- }
- }
- else /* (ent->subexp_to != str_idx) */
- {
- for (node_idx = 0; node_idx < dest_nodes->nelem; ++node_idx)
- {
- int node = dest_nodes->elems[node_idx];
- re_token_type_t type= dfa->nodes[node].type;
- if (type == OP_CLOSE_SUBEXP || type == OP_OPEN_SUBEXP)
- {
- if (subexp_idx != dfa->nodes[node].opr.idx)
- continue;
- if ((type == OP_CLOSE_SUBEXP && ent->subexp_to != str_idx)
- || (type == OP_OPEN_SUBEXP))
- {
- /* It is against this limitation.
- Remove it form the current sifted state. */
- err = sub_epsilon_src_nodes(dfa, node, dest_nodes,
- candidates);
- if (BE (err != REG_NOERROR, 0))
- return err;
- }
- }
- }
- }
- }
- return REG_NOERROR;
-}
-
-static reg_errcode_t
-sift_states_bkref (preg, mctx, sctx, str_idx, dest_nodes)
- const regex_t *preg;
- re_match_context_t *mctx;
- re_sift_context_t *sctx;
- int str_idx;
- re_node_set *dest_nodes;
-{
- reg_errcode_t err;
- re_dfa_t *dfa = (re_dfa_t *)preg->buffer;
- int node_idx, node;
- re_sift_context_t local_sctx;
- const re_node_set *candidates;
- candidates = ((mctx->state_log[str_idx] == NULL) ? &empty_set
- : &mctx->state_log[str_idx]->nodes);
- local_sctx.sifted_states = NULL; /* Mark that it hasn't been initialized. */
-
- for (node_idx = 0; node_idx < candidates->nelem; ++node_idx)
- {
- int cur_bkref_idx = re_string_cur_idx (mctx->input);
- re_token_type_t type;
- node = candidates->elems[node_idx];
- type = dfa->nodes[node].type;
- if (node == sctx->cur_bkref && str_idx == cur_bkref_idx)
- continue;
- /* Avoid infinite loop for the REs like "()\1+". */
- if (node == sctx->last_node && str_idx == sctx->last_str_idx)
- continue;
- if (type == OP_BACK_REF)
- {
- int enabled_idx = search_cur_bkref_entry (mctx, str_idx);
- for (; enabled_idx < mctx->nbkref_ents; ++enabled_idx)
- {
- int disabled_idx, subexp_len, to_idx, dst_node;
- struct re_backref_cache_entry *entry;
- entry = mctx->bkref_ents + enabled_idx;
- if (entry->str_idx > str_idx)
- break;
- if (entry->node != node)
- continue;
- subexp_len = entry->subexp_to - entry->subexp_from;
- to_idx = str_idx + subexp_len;
- dst_node = (subexp_len ? dfa->nexts[node]
- : dfa->edests[node].elems[0]);
-
- if (to_idx > sctx->last_str_idx
- || sctx->sifted_states[to_idx] == NULL
- || !STATE_NODE_CONTAINS (sctx->sifted_states[to_idx],
- dst_node)
- || check_dst_limits (dfa, &sctx->limits, mctx, node,
- str_idx, dst_node, to_idx))
- continue;
- {
- re_dfastate_t *cur_state;
- entry->flag = 0;
- for (disabled_idx = enabled_idx + 1;
- disabled_idx < mctx->nbkref_ents; ++disabled_idx)
- {
- struct re_backref_cache_entry *entry2;
- entry2 = mctx->bkref_ents + disabled_idx;
- if (entry2->str_idx > str_idx)
- break;
- entry2->flag = (entry2->node == node) ? 1 : entry2->flag;
- }
-
- if (local_sctx.sifted_states == NULL)
- {
- local_sctx = *sctx;
- err = re_node_set_init_copy (&local_sctx.limits,
- &sctx->limits);
- if (BE (err != REG_NOERROR, 0))
- goto free_return;
- }
- local_sctx.last_node = node;
- local_sctx.last_str_idx = str_idx;
- err = re_node_set_insert (&local_sctx.limits, enabled_idx);
- if (BE (err < 0, 0))
- {
- err = REG_ESPACE;
- goto free_return;
- }
- cur_state = local_sctx.sifted_states[str_idx];
- err = sift_states_backward (preg, mctx, &local_sctx);
- if (BE (err != REG_NOERROR, 0))
- goto free_return;
- if (sctx->limited_states != NULL)
- {
- err = merge_state_array (dfa, sctx->limited_states,
- local_sctx.sifted_states,
- str_idx + 1);
- if (BE (err != REG_NOERROR, 0))
- goto free_return;
- }
- local_sctx.sifted_states[str_idx] = cur_state;
- re_node_set_remove (&local_sctx.limits, enabled_idx);
- /* We must not use the variable entry here, since
- mctx->bkref_ents might be realloced. */
- mctx->bkref_ents[enabled_idx].flag = 1;
- }
- }
- enabled_idx = search_cur_bkref_entry (mctx, str_idx);
- for (; enabled_idx < mctx->nbkref_ents; ++enabled_idx)
- {
- struct re_backref_cache_entry *entry;
- entry = mctx->bkref_ents + enabled_idx;
- if (entry->str_idx > str_idx)
- break;
- if (entry->node == node)
- entry->flag = 0;
- }
- }
- }
- err = REG_NOERROR;
- free_return:
- if (local_sctx.sifted_states != NULL)
- {
- re_node_set_free (&local_sctx.limits);
- }
-
- return err;
-}
-
-
-#ifdef RE_ENABLE_I18N
-static int
-sift_states_iter_mb (preg, mctx, sctx, node_idx, str_idx, max_str_idx)
- const regex_t *preg;
- const re_match_context_t *mctx;
- re_sift_context_t *sctx;
- int node_idx, str_idx, max_str_idx;
-{
- re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- int naccepted;
- /* Check the node can accept `multi byte'. */
- naccepted = check_node_accept_bytes (preg, node_idx, mctx->input, str_idx);
- if (naccepted > 0 && str_idx + naccepted <= max_str_idx &&
- !STATE_NODE_CONTAINS (sctx->sifted_states[str_idx + naccepted],
- dfa->nexts[node_idx]))
- /* The node can't accept the `multi byte', or the
- destination was already throwed away, then the node
- could't accept the current input `multi byte'. */
- naccepted = 0;
- /* Otherwise, it is sure that the node could accept
- `naccepted' bytes input. */
- return naccepted;
-}
-#endif /* RE_ENABLE_I18N */
-
-\f
-/* Functions for state transition. */
-
-/* Return the next state to which the current state STATE will transit by
- accepting the current input byte, and update STATE_LOG if necessary.
- If STATE can accept a multibyte char/collating element/back reference
- update the destination of STATE_LOG. */
-
-static re_dfastate_t *
-transit_state (err, preg, mctx, state, fl_search)
- reg_errcode_t *err;
- const regex_t *preg;
- re_match_context_t *mctx;
- re_dfastate_t *state;
- int fl_search;
-{
- re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- re_dfastate_t **trtable, *next_state;
- unsigned char ch;
- int cur_idx;
-
- if (re_string_cur_idx (mctx->input) + 1 >= mctx->input->bufs_len
- || (re_string_cur_idx (mctx->input) + 1 >= mctx->input->valid_len
- && mctx->input->valid_len < mctx->input->len))
- {
- *err = extend_buffers (mctx);
- if (BE (*err != REG_NOERROR, 0))
- return NULL;
- }
-
- *err = REG_NOERROR;
- if (state == NULL)
- {
- next_state = state;
- re_string_skip_bytes (mctx->input, 1);
- }
- else
- {
-#ifdef RE_ENABLE_I18N
- /* If the current state can accept multibyte. */
- if (state->accept_mb)
- {
- *err = transit_state_mb (preg, state, mctx);
- if (BE (*err != REG_NOERROR, 0))
- return NULL;
- }
-#endif /* RE_ENABLE_I18N */
-
- /* Then decide the next state with the single byte. */
- if (1)
- {
- /* Use transition table */
- ch = re_string_fetch_byte (mctx->input);
- trtable = fl_search ? state->trtable_search : state->trtable;
- if (trtable == NULL)
- {
- trtable = build_trtable (preg, state, fl_search);
- if (fl_search)
- state->trtable_search = trtable;
- else
- state->trtable = trtable;
- }
- next_state = trtable[ch];
- }
- else
- {
- /* don't use transition table */
- next_state = transit_state_sb (err, preg, state, fl_search, mctx);
- if (BE (next_state == NULL && err != REG_NOERROR, 0))
- return NULL;
- }
- }
-
- cur_idx = re_string_cur_idx (mctx->input);
- /* Update the state_log if we need. */
- if (mctx->state_log != NULL)
- {
- if (cur_idx > mctx->state_log_top)
- {
- mctx->state_log[cur_idx] = next_state;
- mctx->state_log_top = cur_idx;
- }
- else if (mctx->state_log[cur_idx] == 0)
- {
- mctx->state_log[cur_idx] = next_state;
- }
- else
- {
- re_dfastate_t *pstate;
- unsigned int context;
- re_node_set next_nodes, *log_nodes, *table_nodes = NULL;
- /* If (state_log[cur_idx] != 0), it implies that cur_idx is
- the destination of a multibyte char/collating element/
- back reference. Then the next state is the union set of
- these destinations and the results of the transition table. */
- pstate = mctx->state_log[cur_idx];
- log_nodes = pstate->entrance_nodes;
- if (next_state != NULL)
- {
- table_nodes = next_state->entrance_nodes;
- *err = re_node_set_init_union (&next_nodes, table_nodes,
- log_nodes);
- if (BE (*err != REG_NOERROR, 0))
- return NULL;
- }
- else
- next_nodes = *log_nodes;
- /* Note: We already add the nodes of the initial state,
- then we don't need to add them here. */
-
- context = re_string_context_at (mctx->input,
- re_string_cur_idx (mctx->input) - 1,
- mctx->eflags, preg->newline_anchor);
- next_state = mctx->state_log[cur_idx]
- = re_acquire_state_context (err, dfa, &next_nodes, context);
- /* We don't need to check errors here, since the return value of
- this function is next_state and ERR is already set. */
-
- if (table_nodes != NULL)
- re_node_set_free (&next_nodes);
- }
- }
-
- /* Check OP_OPEN_SUBEXP in the current state in case that we use them
- later. We must check them here, since the back references in the
- next state might use them. */
- if (dfa->nbackref && next_state/* && fl_process_bkref */)
- {
- *err = check_subexp_matching_top (dfa, mctx, &next_state->nodes,
- cur_idx);
- if (BE (*err != REG_NOERROR, 0))
- return NULL;
- }
-
- /* If the next state has back references. */
- if (next_state != NULL && next_state->has_backref)
- {
- *err = transit_state_bkref (preg, &next_state->nodes, mctx);
- if (BE (*err != REG_NOERROR, 0))
- return NULL;
- next_state = mctx->state_log[cur_idx];
- }
- return next_state;
-}
-
-/* Helper functions for transit_state. */
-
-/* From the node set CUR_NODES, pick up the nodes whose types are
- OP_OPEN_SUBEXP and which have corresponding back references in the regular
- expression. And register them to use them later for evaluating the
- correspoding back references. */
-
-static reg_errcode_t
-check_subexp_matching_top (dfa, mctx, cur_nodes, str_idx)
- re_dfa_t *dfa;
- re_match_context_t *mctx;
- re_node_set *cur_nodes;
- int str_idx;
-{
- int node_idx;
- reg_errcode_t err;
-
- /* TODO: This isn't efficient.
- Because there might be more than one nodes whose types are
- OP_OPEN_SUBEXP and whose index is SUBEXP_IDX, we must check all
- nodes.
- E.g. RE: (a){2} */
- for (node_idx = 0; node_idx < cur_nodes->nelem; ++node_idx)
- {
- int node = cur_nodes->elems[node_idx];
- if (dfa->nodes[node].type == OP_OPEN_SUBEXP
- && dfa->used_bkref_map & (1 << dfa->nodes[node].opr.idx))
- {
- err = match_ctx_add_subtop (mctx, node, str_idx);
- if (BE (err != REG_NOERROR, 0))
- return err;
- }
- }
- return REG_NOERROR;
-}
-
-/* Return the next state to which the current state STATE will transit by
- accepting the current input byte. */
-
-static re_dfastate_t *
-transit_state_sb (err, preg, state, fl_search, mctx)
- reg_errcode_t *err;
- const regex_t *preg;
- re_dfastate_t *state;
- int fl_search;
- re_match_context_t *mctx;
-{
- re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- re_node_set next_nodes;
- re_dfastate_t *next_state;
- int node_cnt, cur_str_idx = re_string_cur_idx (mctx->input);
- unsigned int context;
-
- *err = re_node_set_alloc (&next_nodes, state->nodes.nelem + 1);
- if (BE (*err != REG_NOERROR, 0))
- return NULL;
- for (node_cnt = 0; node_cnt < state->nodes.nelem; ++node_cnt)
- {
- int cur_node = state->nodes.elems[node_cnt];
- if (check_node_accept (preg, dfa->nodes + cur_node, mctx, cur_str_idx))
- {
- *err = re_node_set_merge (&next_nodes,
- dfa->eclosures + dfa->nexts[cur_node]);
- if (BE (*err != REG_NOERROR, 0))
- {
- re_node_set_free (&next_nodes);
- return NULL;
- }
- }
- }
- if (fl_search)
- {
-#ifdef RE_ENABLE_I18N
- int not_initial = 0;
- if (MB_CUR_MAX > 1)
- for (node_cnt = 0; node_cnt < next_nodes.nelem; ++node_cnt)
- if (dfa->nodes[next_nodes.elems[node_cnt]].type == CHARACTER)
- {
- not_initial = dfa->nodes[next_nodes.elems[node_cnt]].mb_partial;
- break;
- }
- if (!not_initial)
-#endif
- {
- *err = re_node_set_merge (&next_nodes,
- dfa->init_state->entrance_nodes);
- if (BE (*err != REG_NOERROR, 0))
- {
- re_node_set_free (&next_nodes);
- return NULL;
- }
- }
- }
- context = re_string_context_at (mctx->input, cur_str_idx, mctx->eflags,
- preg->newline_anchor);
- next_state = re_acquire_state_context (err, dfa, &next_nodes, context);
- /* We don't need to check errors here, since the return value of
- this function is next_state and ERR is already set. */
-
- re_node_set_free (&next_nodes);
- re_string_skip_bytes (mctx->input, 1);
- return next_state;
-}
-
-#ifdef RE_ENABLE_I18N
-static reg_errcode_t
-transit_state_mb (preg, pstate, mctx)
- const regex_t *preg;
- re_dfastate_t *pstate;
- re_match_context_t *mctx;
-{
- reg_errcode_t err;
- re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- int i;
-
- for (i = 0; i < pstate->nodes.nelem; ++i)
- {
- re_node_set dest_nodes, *new_nodes;
- int cur_node_idx = pstate->nodes.elems[i];
- int naccepted = 0, dest_idx;
- unsigned int context;
- re_dfastate_t *dest_state;
-
- if (dfa->nodes[cur_node_idx].constraint)
- {
- context = re_string_context_at (mctx->input,
- re_string_cur_idx (mctx->input),
- mctx->eflags, preg->newline_anchor);
- if (NOT_SATISFY_NEXT_CONSTRAINT (dfa->nodes[cur_node_idx].constraint,
- context))
- continue;
- }
-
- /* How many bytes the node can accepts? */
- if (ACCEPT_MB_NODE (dfa->nodes[cur_node_idx].type))
- naccepted = check_node_accept_bytes (preg, cur_node_idx, mctx->input,
- re_string_cur_idx (mctx->input));
- if (naccepted == 0)
- continue;
-
- /* The node can accepts `naccepted' bytes. */
- dest_idx = re_string_cur_idx (mctx->input) + naccepted;
- mctx->max_mb_elem_len = ((mctx->max_mb_elem_len < naccepted) ? naccepted
- : mctx->max_mb_elem_len);
- err = clean_state_log_if_need (mctx, dest_idx);
- if (BE (err != REG_NOERROR, 0))
- return err;
-#ifdef DEBUG
- assert (dfa->nexts[cur_node_idx] != -1);
-#endif
- /* `cur_node_idx' may point the entity of the OP_CONTEXT_NODE,
- then we use pstate->nodes.elems[i] instead. */
- new_nodes = dfa->eclosures + dfa->nexts[pstate->nodes.elems[i]];
-
- dest_state = mctx->state_log[dest_idx];
- if (dest_state == NULL)
- dest_nodes = *new_nodes;
- else
- {
- err = re_node_set_init_union (&dest_nodes,
- dest_state->entrance_nodes, new_nodes);
- if (BE (err != REG_NOERROR, 0))
- return err;
- }
- context = re_string_context_at (mctx->input, dest_idx - 1, mctx->eflags,
- preg->newline_anchor);
- mctx->state_log[dest_idx]
- = re_acquire_state_context (&err, dfa, &dest_nodes, context);
- if (dest_state != NULL)
- re_node_set_free (&dest_nodes);
- if (BE (mctx->state_log[dest_idx] == NULL && err != REG_NOERROR, 0))
- return err;
- }
- return REG_NOERROR;
-}
-#endif /* RE_ENABLE_I18N */
-
-static reg_errcode_t
-transit_state_bkref (preg, nodes, mctx)
- const regex_t *preg;
- re_node_set *nodes;
- re_match_context_t *mctx;
-{
- reg_errcode_t err;
- re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- int i;
- int cur_str_idx = re_string_cur_idx (mctx->input);
-
- for (i = 0; i < nodes->nelem; ++i)
- {
- int dest_str_idx, prev_nelem, bkc_idx;
- int node_idx = nodes->elems[i];
- unsigned int context;
- re_token_t *node = dfa->nodes + node_idx;
- re_node_set *new_dest_nodes;
-
- /* Check whether `node' is a backreference or not. */
- if (node->type != OP_BACK_REF)
- continue;
-
- if (node->constraint)
- {
- context = re_string_context_at (mctx->input, cur_str_idx,
- mctx->eflags, preg->newline_anchor);
- if (NOT_SATISFY_NEXT_CONSTRAINT (node->constraint, context))
- continue;
- }
-
- /* `node' is a backreference.
- Check the substring which the substring matched. */
- bkc_idx = mctx->nbkref_ents;
- err = get_subexp (preg, mctx, node_idx, cur_str_idx);
- if (BE (err != REG_NOERROR, 0))
- goto free_return;
-
- /* And add the epsilon closures (which is `new_dest_nodes') of
- the backreference to appropriate state_log. */
-#ifdef DEBUG
- assert (dfa->nexts[node_idx] != -1);
-#endif
- for (; bkc_idx < mctx->nbkref_ents; ++bkc_idx)
- {
- int subexp_len;
- re_dfastate_t *dest_state;
- struct re_backref_cache_entry *bkref_ent;
- bkref_ent = mctx->bkref_ents + bkc_idx;
- if (bkref_ent->node != node_idx || bkref_ent->str_idx != cur_str_idx)
- continue;
- subexp_len = bkref_ent->subexp_to - bkref_ent->subexp_from;
- new_dest_nodes = (subexp_len == 0
- ? dfa->eclosures + dfa->edests[node_idx].elems[0]
- : dfa->eclosures + dfa->nexts[node_idx]);
- dest_str_idx = (cur_str_idx + bkref_ent->subexp_to
- - bkref_ent->subexp_from);
- context = re_string_context_at (mctx->input, dest_str_idx - 1,
- mctx->eflags, preg->newline_anchor);
- dest_state = mctx->state_log[dest_str_idx];
- prev_nelem = ((mctx->state_log[cur_str_idx] == NULL) ? 0
- : mctx->state_log[cur_str_idx]->nodes.nelem);
- /* Add `new_dest_node' to state_log. */
- if (dest_state == NULL)
- {
- mctx->state_log[dest_str_idx]
- = re_acquire_state_context (&err, dfa, new_dest_nodes,
- context);
- if (BE (mctx->state_log[dest_str_idx] == NULL
- && err != REG_NOERROR, 0))
- goto free_return;
- }
- else
- {
- re_node_set dest_nodes;
- err = re_node_set_init_union (&dest_nodes,
- dest_state->entrance_nodes,
- new_dest_nodes);
- if (BE (err != REG_NOERROR, 0))
- {
- re_node_set_free (&dest_nodes);
- goto free_return;
- }
- mctx->state_log[dest_str_idx]
- = re_acquire_state_context (&err, dfa, &dest_nodes, context);
- re_node_set_free (&dest_nodes);
- if (BE (mctx->state_log[dest_str_idx] == NULL
- && err != REG_NOERROR, 0))
- goto free_return;
- }
- /* We need to check recursively if the backreference can epsilon
- transit. */
- if (subexp_len == 0
- && mctx->state_log[cur_str_idx]->nodes.nelem > prev_nelem)
- {
- err = check_subexp_matching_top (dfa, mctx, new_dest_nodes,
- cur_str_idx);
- if (BE (err != REG_NOERROR, 0))
- goto free_return;
- err = transit_state_bkref (preg, new_dest_nodes, mctx);
- if (BE (err != REG_NOERROR, 0))
- goto free_return;
- }
- }
- }
- err = REG_NOERROR;
- free_return:
- return err;
-}
-
-/* Enumerate all the candidates which the backreference BKREF_NODE can match
- at BKREF_STR_IDX, and register them by match_ctx_add_entry().
- Note that we might collect inappropriate candidates here.
- However, the cost of checking them strictly here is too high, then we
- delay these checking for prune_impossible_nodes(). */
-
-static reg_errcode_t
-get_subexp (preg, mctx, bkref_node, bkref_str_idx)
- const regex_t *preg;
- re_match_context_t *mctx;
- int bkref_node, bkref_str_idx;
-{
- int subexp_num, sub_top_idx;
- re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- char *buf = (char *) re_string_get_buffer (mctx->input);
- /* Return if we have already checked BKREF_NODE at BKREF_STR_IDX. */
- int cache_idx = search_cur_bkref_entry (mctx, bkref_str_idx);
- for (; cache_idx < mctx->nbkref_ents; ++cache_idx)
- {
- struct re_backref_cache_entry *entry = mctx->bkref_ents + cache_idx;
- if (entry->str_idx > bkref_str_idx)
- break;
- if (entry->node == bkref_node)
- return REG_NOERROR; /* We already checked it. */
- }
- subexp_num = dfa->nodes[bkref_node].opr.idx - 1;
-
- /* For each sub expression */
- for (sub_top_idx = 0; sub_top_idx < mctx->nsub_tops; ++sub_top_idx)
- {
- reg_errcode_t err;
- re_sub_match_top_t *sub_top = mctx->sub_tops[sub_top_idx];
- re_sub_match_last_t *sub_last;
- int sub_last_idx, sl_str;
- char *bkref_str;
-
- if (dfa->nodes[sub_top->node].opr.idx != subexp_num)
- continue; /* It isn't related. */
-
- sl_str = sub_top->str_idx;
- bkref_str = buf + bkref_str_idx;
- /* At first, check the last node of sub expressions we already
- evaluated. */
- for (sub_last_idx = 0; sub_last_idx < sub_top->nlasts; ++sub_last_idx)
- {
- int sl_str_diff;
- sub_last = sub_top->lasts[sub_last_idx];
- sl_str_diff = sub_last->str_idx - sl_str;
- /* The matched string by the sub expression match with the substring
- at the back reference? */
- if (sl_str_diff > 0
- && memcmp (bkref_str, buf + sl_str, sl_str_diff) != 0)
- break; /* We don't need to search this sub expression any more. */
- bkref_str += sl_str_diff;
- sl_str += sl_str_diff;
- err = get_subexp_sub (preg, mctx, sub_top, sub_last, bkref_node,
- bkref_str_idx);
- if (err == REG_NOMATCH)
- continue;
- if (BE (err != REG_NOERROR, 0))
- return err;
- }
- if (sub_last_idx < sub_top->nlasts)
- continue;
- if (sub_last_idx > 0)
- ++sl_str;
- /* Then, search for the other last nodes of the sub expression. */
- for (; sl_str <= bkref_str_idx; ++sl_str)
- {
- int cls_node, sl_str_off;
- re_node_set *nodes;
- sl_str_off = sl_str - sub_top->str_idx;
- /* The matched string by the sub expression match with the substring
- at the back reference? */
- if (sl_str_off > 0
- && memcmp (bkref_str++, buf + sl_str - 1, 1) != 0)
- break; /* We don't need to search this sub expression any more. */
- if (mctx->state_log[sl_str] == NULL)
- continue;
- /* Does this state have a ')' of the sub expression? */
- nodes = &mctx->state_log[sl_str]->nodes;
- cls_node = find_subexp_node (dfa, nodes, subexp_num, 0);
- if (cls_node == -1)
- continue; /* No. */
- if (sub_top->path == NULL)
- {
- sub_top->path = calloc (sizeof (state_array_t),
- sl_str - sub_top->str_idx + 1);
- if (sub_top->path == NULL)
- return REG_ESPACE;
- }
- /* Can the OP_OPEN_SUBEXP node arrive the OP_CLOSE_SUBEXP node
- in the current context? */
- err = check_arrival (preg, mctx, sub_top->path, sub_top->node,
- sub_top->str_idx, cls_node, sl_str, 0);
- if (err == REG_NOMATCH)
- continue;
- if (BE (err != REG_NOERROR, 0))
- return err;
- sub_last = match_ctx_add_sublast (sub_top, cls_node, sl_str);
- if (BE (sub_last == NULL, 0))
- return REG_ESPACE;
- err = get_subexp_sub (preg, mctx, sub_top, sub_last, bkref_node,
- bkref_str_idx);
- if (err == REG_NOMATCH)
- continue;
- }
- }
- return REG_NOERROR;
-}
-
-/* Helper functions for get_subexp(). */
-
-/* Check SUB_LAST can arrive to the back reference BKREF_NODE at BKREF_STR.
- If it can arrive, register the sub expression expressed with SUB_TOP
- and SUB_LAST. */
-
-static reg_errcode_t
-get_subexp_sub (preg, mctx, sub_top, sub_last, bkref_node, bkref_str)
- const regex_t *preg;
- re_match_context_t *mctx;
- re_sub_match_top_t *sub_top;
- re_sub_match_last_t *sub_last;
- int bkref_node, bkref_str;
-{
- reg_errcode_t err;
- int to_idx;
- /* Can the subexpression arrive the back reference? */
- err = check_arrival (preg, mctx, &sub_last->path, sub_last->node,
- sub_last->str_idx, bkref_node, bkref_str, 1);
- if (err != REG_NOERROR)
- return err;
- err = match_ctx_add_entry (mctx, bkref_node, bkref_str, sub_top->str_idx,
- sub_last->str_idx);
- if (BE (err != REG_NOERROR, 0))
- return err;
- to_idx = bkref_str + sub_last->str_idx - sub_top->str_idx;
- clean_state_log_if_need (mctx, to_idx);
- return REG_NOERROR;
-}
-
-/* Find the first node which is '(' or ')' and whose index is SUBEXP_IDX.
- Search '(' if FL_OPEN, or search ')' otherwise.
- TODO: This function isn't efficient...
- Because there might be more than one nodes whose types are
- OP_OPEN_SUBEXP and whose index is SUBEXP_IDX, we must check all
- nodes.
- E.g. RE: (a){2} */
-
-static int
-find_subexp_node (dfa, nodes, subexp_idx, fl_open)
- re_dfa_t *dfa;
- re_node_set *nodes;
- int subexp_idx, fl_open;
-{
- int cls_idx;
- for (cls_idx = 0; cls_idx < nodes->nelem; ++cls_idx)
- {
- int cls_node = nodes->elems[cls_idx];
- re_token_t *node = dfa->nodes + cls_node;
- if (((fl_open && node->type == OP_OPEN_SUBEXP)
- || (!fl_open && node->type == OP_CLOSE_SUBEXP))
- && node->opr.idx == subexp_idx)
- return cls_node;
- }
- return -1;
-}
-
-/* Check whether the node TOP_NODE at TOP_STR can arrive to the node
- LAST_NODE at LAST_STR. We record the path onto PATH since it will be
- heavily reused.
- Return REG_NOERROR if it can arrive, or REG_NOMATCH otherwise. */
-
-static reg_errcode_t
-check_arrival (preg, mctx, path, top_node, top_str, last_node, last_str,
- fl_open)
- const regex_t *preg;
- re_match_context_t *mctx;
- state_array_t *path;
- int top_node, top_str, last_node, last_str, fl_open;
-{
- re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- reg_errcode_t err;
- int subexp_num, backup_cur_idx, str_idx, null_cnt;
- re_dfastate_t *cur_state = NULL;
- re_node_set *cur_nodes, next_nodes;
- re_dfastate_t **backup_state_log;
- unsigned int context;
-
- subexp_num = dfa->nodes[top_node].opr.idx;
- /* Extend the buffer if we need. */
- if (path->alloc < last_str + mctx->max_mb_elem_len + 1)
- {
- re_dfastate_t **new_array;
- int old_alloc = path->alloc;
- path->alloc += last_str + mctx->max_mb_elem_len + 1;
- new_array = re_realloc (path->array, re_dfastate_t *, path->alloc);
- if (new_array == NULL)
- return REG_ESPACE;
- path->array = new_array;
- memset (new_array + old_alloc, '\0',
- sizeof (re_dfastate_t *) * (path->alloc - old_alloc));
- }
-
- str_idx = path->next_idx == 0 ? top_str : path->next_idx;
-
- /* Temporary modify MCTX. */
- backup_state_log = mctx->state_log;
- backup_cur_idx = mctx->input->cur_idx;
- mctx->state_log = path->array;
- mctx->input->cur_idx = str_idx;
-
- /* Setup initial node set. */
- context = re_string_context_at (mctx->input, str_idx - 1, mctx->eflags,
- preg->newline_anchor);
- if (str_idx == top_str)
- {
- err = re_node_set_init_1 (&next_nodes, top_node);
- if (BE (err != REG_NOERROR, 0))
- return err;
- err = check_arrival_expand_ecl (dfa, &next_nodes, subexp_num, fl_open);
- if (BE (err != REG_NOERROR, 0))
- {
- re_node_set_free (&next_nodes);
- return err;
- }
- }
- else
- {
- cur_state = mctx->state_log[str_idx];
- if (cur_state && cur_state->has_backref)
- {
- err = re_node_set_init_copy (&next_nodes, &cur_state->nodes);
- if (BE ( err != REG_NOERROR, 0))
- return err;
- }
- else
- re_node_set_init_empty (&next_nodes);
- }
- if (str_idx == top_str || (cur_state && cur_state->has_backref))
- {
- if (next_nodes.nelem)
- {
- err = expand_bkref_cache (preg, mctx, &next_nodes, str_idx, last_str,
- subexp_num, fl_open);
- if (BE ( err != REG_NOERROR, 0))
- {
- re_node_set_free (&next_nodes);
- return err;
- }
- }
- cur_state = re_acquire_state_context (&err, dfa, &next_nodes, context);
- if (BE (cur_state == NULL && err != REG_NOERROR, 0))
- {
- re_node_set_free (&next_nodes);
- return err;
- }
- mctx->state_log[str_idx] = cur_state;
- }
-
- for (null_cnt = 0; str_idx < last_str && null_cnt <= mctx->max_mb_elem_len;)
- {
- re_node_set_empty (&next_nodes);
- if (mctx->state_log[str_idx + 1])
- {
- err = re_node_set_merge (&next_nodes,
- &mctx->state_log[str_idx + 1]->nodes);
- if (BE (err != REG_NOERROR, 0))
- {
- re_node_set_free (&next_nodes);
- return err;
- }
- }
- if (cur_state)
- {
- err = check_arrival_add_next_nodes(preg, dfa, mctx, str_idx,
- &cur_state->nodes, &next_nodes);
- if (BE (err != REG_NOERROR, 0))
- {
- re_node_set_free (&next_nodes);
- return err;
- }
- }
- ++str_idx;
- if (next_nodes.nelem)
- {
- err = check_arrival_expand_ecl (dfa, &next_nodes, subexp_num,
- fl_open);
- if (BE (err != REG_NOERROR, 0))
- {
- re_node_set_free (&next_nodes);
- return err;
- }
- err = expand_bkref_cache (preg, mctx, &next_nodes, str_idx, last_str,
- subexp_num, fl_open);
- if (BE ( err != REG_NOERROR, 0))
- {
- re_node_set_free (&next_nodes);
- return err;
- }
- }
- context = re_string_context_at (mctx->input, str_idx - 1, mctx->eflags,
- preg->newline_anchor);
- cur_state = re_acquire_state_context (&err, dfa, &next_nodes, context);
- if (BE (cur_state == NULL && err != REG_NOERROR, 0))
- {
- re_node_set_free (&next_nodes);
- return err;
- }
- mctx->state_log[str_idx] = cur_state;
- null_cnt = cur_state == NULL ? null_cnt + 1 : 0;
- }
- re_node_set_free (&next_nodes);
- cur_nodes = (mctx->state_log[last_str] == NULL ? NULL
- : &mctx->state_log[last_str]->nodes);
- path->next_idx = str_idx;
-
- /* Fix MCTX. */
- mctx->state_log = backup_state_log;
- mctx->input->cur_idx = backup_cur_idx;
-
- if (cur_nodes == NULL)
- return REG_NOMATCH;
- /* Then check the current node set has the node LAST_NODE. */
- return (re_node_set_contains (cur_nodes, last_node)
- || re_node_set_contains (cur_nodes, last_node) ? REG_NOERROR
- : REG_NOMATCH);
-}
-
-/* Helper functions for check_arrival. */
-
-/* Calculate the destination nodes of CUR_NODES at STR_IDX, and append them
- to NEXT_NODES.
- TODO: This function is similar to the functions transit_state*(),
- however this function has many additional works.
- Can't we unify them? */
-
-static reg_errcode_t
-check_arrival_add_next_nodes (preg, dfa, mctx, str_idx, cur_nodes, next_nodes)
- const regex_t *preg;
- re_dfa_t *dfa;
- re_match_context_t *mctx;
- int str_idx;
- re_node_set *cur_nodes, *next_nodes;
-{
- int cur_idx;
- reg_errcode_t err;
- re_node_set union_set;
- re_node_set_init_empty (&union_set);
- for (cur_idx = 0; cur_idx < cur_nodes->nelem; ++cur_idx)
- {
- int naccepted = 0;
- int cur_node = cur_nodes->elems[cur_idx];
- re_token_type_t type = dfa->nodes[cur_node].type;
- if (IS_EPSILON_NODE(type))
- continue;
-#ifdef RE_ENABLE_I18N
- /* If the node may accept `multi byte'. */
- if (ACCEPT_MB_NODE (type))
- {
- naccepted = check_node_accept_bytes (preg, cur_node, mctx->input,
- str_idx);
- if (naccepted > 1)
- {
- re_dfastate_t *dest_state;
- int next_node = dfa->nexts[cur_node];
- int next_idx = str_idx + naccepted;
- dest_state = mctx->state_log[next_idx];
- re_node_set_empty (&union_set);
- if (dest_state)
- {
- err = re_node_set_merge (&union_set, &dest_state->nodes);
- if (BE (err != REG_NOERROR, 0))
- {
- re_node_set_free (&union_set);
- return err;
- }
- err = re_node_set_insert (&union_set, next_node);
- if (BE (err < 0, 0))
- {
- re_node_set_free (&union_set);
- return REG_ESPACE;
- }
- }
- else
- {
- err = re_node_set_insert (&union_set, next_node);
- if (BE (err < 0, 0))
- {
- re_node_set_free (&union_set);
- return REG_ESPACE;
- }
- }
- mctx->state_log[next_idx] = re_acquire_state (&err, dfa,
- &union_set);
- if (BE (mctx->state_log[next_idx] == NULL
- && err != REG_NOERROR, 0))
- {
- re_node_set_free (&union_set);
- return err;
- }
- }
- }
-#endif /* RE_ENABLE_I18N */
- if (naccepted
- || check_node_accept (preg, dfa->nodes + cur_node, mctx,
- str_idx))
- {
- err = re_node_set_insert (next_nodes, dfa->nexts[cur_node]);
- if (BE (err < 0, 0))
- {
- re_node_set_free (&union_set);
- return REG_ESPACE;
- }
- }
- }
- re_node_set_free (&union_set);
- return REG_NOERROR;
-}
-
-/* For all the nodes in CUR_NODES, add the epsilon closures of them to
- CUR_NODES, however exclude the nodes which are:
- - inside the sub expression whose number is EX_SUBEXP, if FL_OPEN.
- - out of the sub expression whose number is EX_SUBEXP, if !FL_OPEN.
-*/
-
-static reg_errcode_t
-check_arrival_expand_ecl (dfa, cur_nodes, ex_subexp, fl_open)
- re_dfa_t *dfa;
- re_node_set *cur_nodes;
- int ex_subexp, fl_open;
-{
- reg_errcode_t err;
- int idx, outside_node;
- re_node_set new_nodes;
-#ifdef DEBUG
- assert (cur_nodes->nelem);
-#endif
- err = re_node_set_alloc (&new_nodes, cur_nodes->nelem);
- if (BE (err != REG_NOERROR, 0))
- return err;
- /* Create a new node set NEW_NODES with the nodes which are epsilon
- closures of the node in CUR_NODES. */
-
- for (idx = 0; idx < cur_nodes->nelem; ++idx)
- {
- int cur_node = cur_nodes->elems[idx];
- re_node_set *eclosure = dfa->eclosures + cur_node;
- outside_node = find_subexp_node (dfa, eclosure, ex_subexp, fl_open);
- if (outside_node == -1)
- {
- /* There are no problematic nodes, just merge them. */
- err = re_node_set_merge (&new_nodes, eclosure);
- if (BE (err != REG_NOERROR, 0))
- {
- re_node_set_free (&new_nodes);
- return err;
- }
- }
- else
- {
- /* There are problematic nodes, re-calculate incrementally. */
- err = check_arrival_expand_ecl_sub (dfa, &new_nodes, cur_node,
- ex_subexp, fl_open);
- if (BE (err != REG_NOERROR, 0))
- {
- re_node_set_free (&new_nodes);
- return err;
- }
- }
- }
- re_node_set_free (cur_nodes);
- *cur_nodes = new_nodes;
- return REG_NOERROR;
-}
-
-/* Helper function for check_arrival_expand_ecl.
- Check incrementally the epsilon closure of TARGET, and if it isn't
- problematic append it to DST_NODES. */
-
-static reg_errcode_t
-check_arrival_expand_ecl_sub (dfa, dst_nodes, target, ex_subexp, fl_open)
- re_dfa_t *dfa;
- int target, ex_subexp, fl_open;
- re_node_set *dst_nodes;
-{
- int cur_node, type;
- for (cur_node = target; !re_node_set_contains (dst_nodes, cur_node);)
- {
- int err;
- type = dfa->nodes[cur_node].type;
-
- if (((type == OP_OPEN_SUBEXP && fl_open)
- || (type == OP_CLOSE_SUBEXP && !fl_open))
- && dfa->nodes[cur_node].opr.idx == ex_subexp)
- {
- if (!fl_open)
- {
- err = re_node_set_insert (dst_nodes, cur_node);
- if (BE (err == -1, 0))
- return REG_ESPACE;
- }
- break;
- }
- err = re_node_set_insert (dst_nodes, cur_node);
- if (BE (err == -1, 0))
- return REG_ESPACE;
- if (dfa->edests[cur_node].nelem == 0)
- break;
- if (dfa->edests[cur_node].nelem == 2)
- {
- err = check_arrival_expand_ecl_sub (dfa, dst_nodes,
- dfa->edests[cur_node].elems[1],
- ex_subexp, fl_open);
- if (BE (err != REG_NOERROR, 0))
- return err;
- }
- cur_node = dfa->edests[cur_node].elems[0];
- }
- return REG_NOERROR;
-}
-
-
-/* For all the back references in the current state, calculate the
- destination of the back references by the appropriate entry
- in MCTX->BKREF_ENTS. */
-
-static reg_errcode_t
-expand_bkref_cache (preg, mctx, cur_nodes, cur_str, last_str, subexp_num,
- fl_open)
- const regex_t *preg;
- re_match_context_t *mctx;
- int cur_str, last_str, subexp_num, fl_open;
- re_node_set *cur_nodes;
-{
- reg_errcode_t err;
- re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- int cache_idx, cache_idx_start;
- /* The current state. */
-
- cache_idx_start = search_cur_bkref_entry (mctx, cur_str);
- for (cache_idx = cache_idx_start; cache_idx < mctx->nbkref_ents; ++cache_idx)
- {
- int to_idx, next_node;
- struct re_backref_cache_entry *ent = mctx->bkref_ents + cache_idx;
- if (ent->str_idx > cur_str)
- break;
- /* Is this entry ENT is appropriate? */
- if (!re_node_set_contains (cur_nodes, ent->node))
- continue; /* No. */
-
- to_idx = cur_str + ent->subexp_to - ent->subexp_from;
- /* Calculate the destination of the back reference, and append it
- to MCTX->STATE_LOG. */
- if (to_idx == cur_str)
- {
- /* The backreference did epsilon transit, we must re-check all the
- node in the current state. */
- re_node_set new_dests;
- reg_errcode_t err2, err3;
- next_node = dfa->edests[ent->node].elems[0];
- if (re_node_set_contains (cur_nodes, next_node))
- continue;
- err = re_node_set_init_1 (&new_dests, next_node);
- err2 = check_arrival_expand_ecl (dfa, &new_dests, subexp_num,
- fl_open);
- err3 = re_node_set_merge (cur_nodes, &new_dests);
- re_node_set_free (&new_dests);
- if (BE (err != REG_NOERROR || err2 != REG_NOERROR
- || err3 != REG_NOERROR, 0))
- {
- err = (err != REG_NOERROR ? err
- : (err2 != REG_NOERROR ? err2 : err3));
- return err;
- }
- /* TODO: It is still inefficient... */
- cache_idx = cache_idx_start - 1;
- continue;
- }
- else
- {
- re_node_set union_set;
- next_node = dfa->nexts[ent->node];
- if (mctx->state_log[to_idx])
- {
- int ret;
- if (re_node_set_contains (&mctx->state_log[to_idx]->nodes,
- next_node))
- continue;
- err = re_node_set_init_copy (&union_set,
- &mctx->state_log[to_idx]->nodes);
- ret = re_node_set_insert (&union_set, next_node);
- if (BE (err != REG_NOERROR || ret < 0, 0))
- {
- re_node_set_free (&union_set);
- err = err != REG_NOERROR ? err : REG_ESPACE;
- return err;
- }
- }
- else
- {
- err = re_node_set_init_1 (&union_set, next_node);
- if (BE (err != REG_NOERROR, 0))
- return err;
- }
- mctx->state_log[to_idx] = re_acquire_state (&err, dfa, &union_set);
- re_node_set_free (&union_set);
- if (BE (mctx->state_log[to_idx] == NULL
- && err != REG_NOERROR, 0))
- return err;
- }
- }
- return REG_NOERROR;
-}
-
-/* Build transition table for the state.
- Return the new table if succeeded, otherwise return NULL. */
-
-static re_dfastate_t **
-build_trtable (preg, state, fl_search)
- const regex_t *preg;
- const re_dfastate_t *state;
- int fl_search;
-{
- reg_errcode_t err;
- re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- int i, j, k, ch;
- int dests_node_malloced = 0, dest_states_malloced = 0;
- int ndests; /* Number of the destination states from `state'. */
- re_dfastate_t **trtable;
- re_dfastate_t **dest_states = NULL, **dest_states_word, **dest_states_nl;
- re_node_set follows, *dests_node;
- bitset *dests_ch;
- bitset acceptable;
-
- /* We build DFA states which corresponds to the destination nodes
- from `state'. `dests_node[i]' represents the nodes which i-th
- destination state contains, and `dests_ch[i]' represents the
- characters which i-th destination state accepts. */
-#ifdef _LIBC
- if (__libc_use_alloca ((sizeof (re_node_set) + sizeof (bitset)) * SBC_MAX))
- dests_node = (re_node_set *)
- alloca ((sizeof (re_node_set) + sizeof (bitset)) * SBC_MAX);
- else
-#endif
- {
- dests_node = (re_node_set *)
- malloc ((sizeof (re_node_set) + sizeof (bitset)) * SBC_MAX);
- if (BE (dests_node == NULL, 0))
- return NULL;
- dests_node_malloced = 1;
- }
- dests_ch = (bitset *) (dests_node + SBC_MAX);
-
- /* Initialize transiton table. */
- trtable = (re_dfastate_t **) calloc (sizeof (re_dfastate_t *), SBC_MAX);
- if (BE (trtable == NULL, 0))
- {
- if (dests_node_malloced)
- free (dests_node);
- return NULL;
- }
-
- /* At first, group all nodes belonging to `state' into several
- destinations. */
- ndests = group_nodes_into_DFAstates (preg, state, dests_node, dests_ch);
- if (BE (ndests <= 0, 0))
- {
- if (dests_node_malloced)
- free (dests_node);
- /* Return NULL in case of an error, trtable otherwise. */
- if (ndests == 0)
- return trtable;
- free (trtable);
- return NULL;
- }
-
- err = re_node_set_alloc (&follows, ndests + 1);
- if (BE (err != REG_NOERROR, 0))
- goto out_free;
-
-#ifdef _LIBC
- if (__libc_use_alloca ((sizeof (re_node_set) + sizeof (bitset)) * SBC_MAX
- + ndests * 3 * sizeof (re_dfastate_t *)))
- dest_states = (re_dfastate_t **)
- alloca (ndests * 3 * sizeof (re_dfastate_t *));
- else
-#endif
- {
- dest_states = (re_dfastate_t **)
- malloc (ndests * 3 * sizeof (re_dfastate_t *));
- if (BE (dest_states == NULL, 0))
- {
-out_free:
- if (dest_states_malloced)
- free (dest_states);
- re_node_set_free (&follows);
- for (i = 0; i < ndests; ++i)
- re_node_set_free (dests_node + i);
- free (trtable);
- if (dests_node_malloced)
- free (dests_node);
- return NULL;
- }
- dest_states_malloced = 1;
- }
- dest_states_word = dest_states + ndests;
- dest_states_nl = dest_states_word + ndests;
- bitset_empty (acceptable);
-
- /* Then build the states for all destinations. */
- for (i = 0; i < ndests; ++i)
- {
- int next_node;
- re_node_set_empty (&follows);
- /* Merge the follows of this destination states. */
- for (j = 0; j < dests_node[i].nelem; ++j)
- {
- next_node = dfa->nexts[dests_node[i].elems[j]];
- if (next_node != -1)
- {
- err = re_node_set_merge (&follows, dfa->eclosures + next_node);
- if (BE (err != REG_NOERROR, 0))
- goto out_free;
- }
- }
- /* If search flag is set, merge the initial state. */
- if (fl_search)
- {
-#ifdef RE_ENABLE_I18N
- int not_initial = 0;
- for (j = 0; j < follows.nelem; ++j)
- if (dfa->nodes[follows.elems[j]].type == CHARACTER)
- {
- not_initial = dfa->nodes[follows.elems[j]].mb_partial;
- break;
- }
- if (!not_initial)
-#endif
- {
- err = re_node_set_merge (&follows,
- dfa->init_state->entrance_nodes);
- if (BE (err != REG_NOERROR, 0))
- goto out_free;
- }
- }
- dest_states[i] = re_acquire_state_context (&err, dfa, &follows, 0);
- if (BE (dest_states[i] == NULL && err != REG_NOERROR, 0))
- goto out_free;
- /* If the new state has context constraint,
- build appropriate states for these contexts. */
- if (dest_states[i]->has_constraint)
- {
- dest_states_word[i] = re_acquire_state_context (&err, dfa, &follows,
- CONTEXT_WORD);
- if (BE (dest_states_word[i] == NULL && err != REG_NOERROR, 0))
- goto out_free;
- dest_states_nl[i] = re_acquire_state_context (&err, dfa, &follows,
- CONTEXT_NEWLINE);
- if (BE (dest_states_nl[i] == NULL && err != REG_NOERROR, 0))
- goto out_free;
- }
- else
- {
- dest_states_word[i] = dest_states[i];
- dest_states_nl[i] = dest_states[i];
- }
- bitset_merge (acceptable, dests_ch[i]);
- }
-
- /* Update the transition table. */
- /* For all characters ch...: */
- for (i = 0, ch = 0; i < BITSET_UINTS; ++i)
- for (j = 0; j < UINT_BITS; ++j, ++ch)
- if ((acceptable[i] >> j) & 1)
- {
- /* The current state accepts the character ch. */
- if (IS_WORD_CHAR (ch))
- {
- for (k = 0; k < ndests; ++k)
- if ((dests_ch[k][i] >> j) & 1)
- {
- /* k-th destination accepts the word character ch. */
- trtable[ch] = dest_states_word[k];
- /* There must be only one destination which accepts
- character ch. See group_nodes_into_DFAstates. */
- break;
- }
- }
- else /* not WORD_CHAR */
- {
- for (k = 0; k < ndests; ++k)
- if ((dests_ch[k][i] >> j) & 1)
- {
- /* k-th destination accepts the non-word character ch. */
- trtable[ch] = dest_states[k];
- /* There must be only one destination which accepts
- character ch. See group_nodes_into_DFAstates. */
- break;
- }
- }
- }
- /* new line */
- if (bitset_contain (acceptable, NEWLINE_CHAR))
- {
- /* The current state accepts newline character. */
- for (k = 0; k < ndests; ++k)
- if (bitset_contain (dests_ch[k], NEWLINE_CHAR))
- {
- /* k-th destination accepts newline character. */
- trtable[NEWLINE_CHAR] = dest_states_nl[k];
- /* There must be only one destination which accepts
- newline. See group_nodes_into_DFAstates. */
- break;
- }
- }
-
- if (dest_states_malloced)
- free (dest_states);
-
- re_node_set_free (&follows);
- for (i = 0; i < ndests; ++i)
- re_node_set_free (dests_node + i);
-
- if (dests_node_malloced)
- free (dests_node);
-
- return trtable;
-}
-
-/* Group all nodes belonging to STATE into several destinations.
- Then for all destinations, set the nodes belonging to the destination
- to DESTS_NODE[i] and set the characters accepted by the destination
- to DEST_CH[i]. This function return the number of destinations. */
-
-static int
-group_nodes_into_DFAstates (preg, state, dests_node, dests_ch)
- const regex_t *preg;
- const re_dfastate_t *state;
- re_node_set *dests_node;
- bitset *dests_ch;
-{
- reg_errcode_t err;
- const re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- int i, j, k;
- int ndests; /* Number of the destinations from `state'. */
- bitset accepts; /* Characters a node can accept. */
- const re_node_set *cur_nodes = &state->nodes;
- bitset_empty (accepts);
- ndests = 0;
-
- /* For all the nodes belonging to `state', */
- for (i = 0; i < cur_nodes->nelem; ++i)
- {
- re_token_t *node = &dfa->nodes[cur_nodes->elems[i]];
- re_token_type_t type = node->type;
- unsigned int constraint = node->constraint;
-
- /* Enumerate all single byte character this node can accept. */
- if (type == CHARACTER)
- bitset_set (accepts, node->opr.c);
- else if (type == SIMPLE_BRACKET)
- {
- bitset_merge (accepts, node->opr.sbcset);
- }
- else if (type == OP_PERIOD)
- {
- bitset_set_all (accepts);
- if (!(preg->syntax & RE_DOT_NEWLINE))
- bitset_clear (accepts, '\n');
- if (preg->syntax & RE_DOT_NOT_NULL)
- bitset_clear (accepts, '\0');
- }
- else
- continue;
-
- /* Check the `accepts' and sift the characters which are not
- match it the context. */
- if (constraint)
- {
- if (constraint & NEXT_WORD_CONSTRAINT)
- for (j = 0; j < BITSET_UINTS; ++j)
- accepts[j] &= dfa->word_char[j];
- if (constraint & NEXT_NOTWORD_CONSTRAINT)
- for (j = 0; j < BITSET_UINTS; ++j)
- accepts[j] &= ~dfa->word_char[j];
- if (constraint & NEXT_NEWLINE_CONSTRAINT)
- {
- int accepts_newline = bitset_contain (accepts, NEWLINE_CHAR);
- bitset_empty (accepts);
- if (accepts_newline)
- bitset_set (accepts, NEWLINE_CHAR);
- else
- continue;
- }
- }
-
- /* Then divide `accepts' into DFA states, or create a new
- state. */
- for (j = 0; j < ndests; ++j)
- {
- bitset intersec; /* Intersection sets, see below. */
- bitset remains;
- /* Flags, see below. */
- int has_intersec, not_subset, not_consumed;
-
- /* Optimization, skip if this state doesn't accept the character. */
- if (type == CHARACTER && !bitset_contain (dests_ch[j], node->opr.c))
- continue;
-
- /* Enumerate the intersection set of this state and `accepts'. */
- has_intersec = 0;
- for (k = 0; k < BITSET_UINTS; ++k)
- has_intersec |= intersec[k] = accepts[k] & dests_ch[j][k];
- /* And skip if the intersection set is empty. */
- if (!has_intersec)
- continue;
-
- /* Then check if this state is a subset of `accepts'. */
- not_subset = not_consumed = 0;
- for (k = 0; k < BITSET_UINTS; ++k)
- {
- not_subset |= remains[k] = ~accepts[k] & dests_ch[j][k];
- not_consumed |= accepts[k] = accepts[k] & ~dests_ch[j][k];
- }
-
- /* If this state isn't a subset of `accepts', create a
- new group state, which has the `remains'. */
- if (not_subset)
- {
- bitset_copy (dests_ch[ndests], remains);
- bitset_copy (dests_ch[j], intersec);
- err = re_node_set_init_copy (dests_node + ndests, &dests_node[j]);
- if (BE (err != REG_NOERROR, 0))
- goto error_return;
- ++ndests;
- }
-
- /* Put the position in the current group. */
- err = re_node_set_insert (&dests_node[j], cur_nodes->elems[i]);
- if (BE (err < 0, 0))
- goto error_return;
-
- /* If all characters are consumed, go to next node. */
- if (!not_consumed)
- break;
- }
- /* Some characters remain, create a new group. */
- if (j == ndests)
- {
- bitset_copy (dests_ch[ndests], accepts);
- err = re_node_set_init_1 (dests_node + ndests, cur_nodes->elems[i]);
- if (BE (err != REG_NOERROR, 0))
- goto error_return;
- ++ndests;
- bitset_empty (accepts);
- }
- }
- return ndests;
- error_return:
- for (j = 0; j < ndests; ++j)
- re_node_set_free (dests_node + j);
- return -1;
-}
-
-#ifdef RE_ENABLE_I18N
-/* Check how many bytes the node `dfa->nodes[node_idx]' accepts.
- Return the number of the bytes the node accepts.
- STR_IDX is the current index of the input string.
-
- This function handles the nodes which can accept one character, or
- one collating element like '.', '[a-z]', opposite to the other nodes
- can only accept one byte. */
-
-static int
-check_node_accept_bytes (preg, node_idx, input, str_idx)
- const regex_t *preg;
- int node_idx, str_idx;
- const re_string_t *input;
-{
- const re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
- const re_token_t *node = dfa->nodes + node_idx;
- int elem_len = re_string_elem_size_at (input, str_idx);
- int char_len = re_string_char_size_at (input, str_idx);
- int i;
-# ifdef _LIBC
- int j;
- uint32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES);
-# endif /* _LIBC */
- if (elem_len <= 1 && char_len <= 1)
- return 0;
- if (node->type == OP_PERIOD)
- {
- /* '.' accepts any one character except the following two cases. */
- if ((!(preg->syntax & RE_DOT_NEWLINE) &&
- re_string_byte_at (input, str_idx) == '\n') ||
- ((preg->syntax & RE_DOT_NOT_NULL) &&
- re_string_byte_at (input, str_idx) == '\0'))
- return 0;
- return char_len;
- }
- else if (node->type == COMPLEX_BRACKET)
- {
- const re_charset_t *cset = node->opr.mbcset;
-# ifdef _LIBC
- const unsigned char *pin = ((char *) re_string_get_buffer (input)
- + str_idx);
-# endif /* _LIBC */
- int match_len = 0;
- wchar_t wc = ((cset->nranges || cset->nchar_classes || cset->nmbchars)
- ? re_string_wchar_at (input, str_idx) : 0);
-
- /* match with multibyte character? */
- for (i = 0; i < cset->nmbchars; ++i)
- if (wc == cset->mbchars[i])
- {
- match_len = char_len;
- goto check_node_accept_bytes_match;
- }
- /* match with character_class? */
- for (i = 0; i < cset->nchar_classes; ++i)
- {
- wctype_t wt = cset->char_classes[i];
- if (__iswctype (wc, wt))
- {
- match_len = char_len;
- goto check_node_accept_bytes_match;
- }
- }
-
-# ifdef _LIBC
- if (nrules != 0)
- {
- unsigned int in_collseq = 0;
- const int32_t *table, *indirect;
- const unsigned char *weights, *extra;
- const char *collseqwc;
- int32_t idx;
- /* This #include defines a local function! */
-# include <locale/weight.h>
-
- /* match with collating_symbol? */
- if (cset->ncoll_syms)
- extra = (const unsigned char *)
- _NL_CURRENT (LC_COLLATE, _NL_COLLATE_SYMB_EXTRAMB);
- for (i = 0; i < cset->ncoll_syms; ++i)
- {
- const unsigned char *coll_sym = extra + cset->coll_syms[i];
- /* Compare the length of input collating element and
- the length of current collating element. */
- if (*coll_sym != elem_len)
- continue;
- /* Compare each bytes. */
- for (j = 0; j < *coll_sym; j++)
- if (pin[j] != coll_sym[1 + j])
- break;
- if (j == *coll_sym)
- {
- /* Match if every bytes is equal. */
- match_len = j;
- goto check_node_accept_bytes_match;
- }
- }
-
- if (cset->nranges)
- {
- if (elem_len <= char_len)
- {
- collseqwc = _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQWC);
- in_collseq = collseq_table_lookup (collseqwc, wc);
- }
- else
- in_collseq = find_collation_sequence_value (pin, elem_len);
- }
- /* match with range expression? */
- for (i = 0; i < cset->nranges; ++i)
- if (cset->range_starts[i] <= in_collseq
- && in_collseq <= cset->range_ends[i])
- {
- match_len = elem_len;
- goto check_node_accept_bytes_match;
- }
-
- /* match with equivalence_class? */
- if (cset->nequiv_classes)
- {
- const unsigned char *cp = pin;
- table = (const int32_t *)
- _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB);
- weights = (const unsigned char *)
- _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTMB);
- extra = (const unsigned char *)
- _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB);
- indirect = (const int32_t *)
- _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTMB);
- idx = findidx (&cp);
- if (idx > 0)
- for (i = 0; i < cset->nequiv_classes; ++i)
- {
- int32_t equiv_class_idx = cset->equiv_classes[i];
- size_t weight_len = weights[idx];
- if (weight_len == weights[equiv_class_idx])
- {
- int cnt = 0;
- while (cnt <= weight_len
- && (weights[equiv_class_idx + 1 + cnt]
- == weights[idx + 1 + cnt]))
- ++cnt;
- if (cnt > weight_len)
- {
- match_len = elem_len;
- goto check_node_accept_bytes_match;
- }
- }
- }
- }
- }
- else
-# endif /* _LIBC */
- {
- /* match with range expression? */
-#if __GNUC__ >= 2
- wchar_t cmp_buf[] = {L'\0', L'\0', wc, L'\0', L'\0', L'\0'};
-#else
- wchar_t cmp_buf[] = {L'\0', L'\0', L'\0', L'\0', L'\0', L'\0'};
- cmp_buf[2] = wc;
-#endif
- for (i = 0; i < cset->nranges; ++i)
- {
- cmp_buf[0] = cset->range_starts[i];
- cmp_buf[4] = cset->range_ends[i];
- if (wcscoll (cmp_buf, cmp_buf + 2) <= 0
- && wcscoll (cmp_buf + 2, cmp_buf + 4) <= 0)
- {
- match_len = char_len;
- goto check_node_accept_bytes_match;
- }
- }
- }
- check_node_accept_bytes_match:
- if (!cset->non_match)
- return match_len;
- else
- {
- if (match_len > 0)
- return 0;
- else
- return (elem_len > char_len) ? elem_len : char_len;
- }
- }
- return 0;
-}
-
-# ifdef _LIBC
-static unsigned int
-find_collation_sequence_value (mbs, mbs_len)
- const unsigned char *mbs;
- size_t mbs_len;
-{
- uint32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES);
- if (nrules == 0)
- {
- if (mbs_len == 1)
- {
- /* No valid character. Match it as a single byte character. */
- const unsigned char *collseq = (const unsigned char *)
- _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQMB);
- return collseq[mbs[0]];
- }
- return UINT_MAX;
- }
- else
- {
- int32_t idx;
- const unsigned char *extra = (const unsigned char *)
- _NL_CURRENT (LC_COLLATE, _NL_COLLATE_SYMB_EXTRAMB);
-
- for (idx = 0; ;)
- {
- int mbs_cnt, found = 0;
- int32_t elem_mbs_len;
- /* Skip the name of collating element name. */
- idx = idx + extra[idx] + 1;
- elem_mbs_len = extra[idx++];
- if (mbs_len == elem_mbs_len)
- {
- for (mbs_cnt = 0; mbs_cnt < elem_mbs_len; ++mbs_cnt)
- if (extra[idx + mbs_cnt] != mbs[mbs_cnt])
- break;
- if (mbs_cnt == elem_mbs_len)
- /* Found the entry. */
- found = 1;
- }
- /* Skip the byte sequence of the collating element. */
- idx += elem_mbs_len;
- /* Adjust for the alignment. */
- idx = (idx + 3) & ~3;
- /* Skip the collation sequence value. */
- idx += sizeof (uint32_t);
- /* Skip the wide char sequence of the collating element. */
- idx = idx + sizeof (uint32_t) * (extra[idx] + 1);
- /* If we found the entry, return the sequence value. */
- if (found)
- return *(uint32_t *) (extra + idx);
- /* Skip the collation sequence value. */
- idx += sizeof (uint32_t);
- }
- }
-}
-# endif /* _LIBC */
-#endif /* RE_ENABLE_I18N */
-
-/* Check whether the node accepts the byte which is IDX-th
- byte of the INPUT. */
-
-static int
-check_node_accept (preg, node, mctx, idx)
- const regex_t *preg;
- const re_token_t *node;
- const re_match_context_t *mctx;
- int idx;
-{
- unsigned char ch;
- if (node->constraint)
- {
- /* The node has constraints. Check whether the current context
- satisfies the constraints. */
- unsigned int context = re_string_context_at (mctx->input, idx,
- mctx->eflags,
- preg->newline_anchor);
- if (NOT_SATISFY_NEXT_CONSTRAINT (node->constraint, context))
- return 0;
- }
- ch = re_string_byte_at (mctx->input, idx);
- if (node->type == CHARACTER)
- return node->opr.c == ch;
- else if (node->type == SIMPLE_BRACKET)
- return bitset_contain (node->opr.sbcset, ch);
- else if (node->type == OP_PERIOD)
- return !((ch == '\n' && !(preg->syntax & RE_DOT_NEWLINE))
- || (ch == '\0' && (preg->syntax & RE_DOT_NOT_NULL)));
- else
- return 0;
-}
-
-/* Extend the buffers, if the buffers have run out. */
-
-static reg_errcode_t
-extend_buffers (mctx)
- re_match_context_t *mctx;
-{
- reg_errcode_t ret;
- re_string_t *pstr = mctx->input;
-
- /* Double the lengthes of the buffers. */
- ret = re_string_realloc_buffers (pstr, pstr->bufs_len * 2);
- if (BE (ret != REG_NOERROR, 0))
- return ret;
-
- if (mctx->state_log != NULL)
- {
- /* And double the length of state_log. */
- re_dfastate_t **new_array;
- new_array = re_realloc (mctx->state_log, re_dfastate_t *,
- pstr->bufs_len * 2);
- if (BE (new_array == NULL, 0))
- return REG_ESPACE;
- mctx->state_log = new_array;
- }
-
- /* Then reconstruct the buffers. */
- if (pstr->icase)
- {
-#ifdef RE_ENABLE_I18N
- if (MB_CUR_MAX > 1)
- build_wcs_upper_buffer (pstr);
- else
-#endif /* RE_ENABLE_I18N */
- build_upper_buffer (pstr);
- }
- else
- {
-#ifdef RE_ENABLE_I18N
- if (MB_CUR_MAX > 1)
- build_wcs_buffer (pstr);
- else
-#endif /* RE_ENABLE_I18N */
- {
- if (pstr->trans != NULL)
- re_string_translate_buffer (pstr);
- else
- pstr->valid_len = pstr->bufs_len;
- }
- }
- return REG_NOERROR;
-}
-
-\f
-/* Functions for matching context. */
-
-/* Initialize MCTX. */
-
-static reg_errcode_t
-match_ctx_init (mctx, eflags, input, n)
- re_match_context_t *mctx;
- int eflags, n;
- re_string_t *input;
-{
- mctx->eflags = eflags;
- mctx->input = input;
- mctx->match_last = -1;
- if (n > 0)
- {
- mctx->bkref_ents = re_malloc (struct re_backref_cache_entry, n);
- mctx->sub_tops = re_malloc (re_sub_match_top_t *, n);
- if (BE (mctx->bkref_ents == NULL || mctx->sub_tops == NULL, 0))
- return REG_ESPACE;
- }
- else
- mctx->bkref_ents = NULL;
- mctx->nbkref_ents = 0;
- mctx->abkref_ents = n;
- mctx->max_mb_elem_len = 1;
- mctx->nsub_tops = 0;
- mctx->asub_tops = n;
- return REG_NOERROR;
-}
-
-/* Clean the entries which depend on the current input in MCTX.
- This function must be invoked when the matcher changes the start index
- of the input, or changes the input string. */
-
-static void
-match_ctx_clean (mctx)
- re_match_context_t *mctx;
-{
- match_ctx_free_subtops (mctx);
- mctx->nsub_tops = 0;
- mctx->nbkref_ents = 0;
-}
-
-/* Free all the memory associated with MCTX. */
-
-static void
-match_ctx_free (mctx)
- re_match_context_t *mctx;
-{
- match_ctx_free_subtops (mctx);
- re_free (mctx->sub_tops);
- re_free (mctx->bkref_ents);
-}
-
-/* Free all the memory associated with MCTX->SUB_TOPS. */
-
-static void
-match_ctx_free_subtops (mctx)
- re_match_context_t *mctx;
-{
- int st_idx;
- for (st_idx = 0; st_idx < mctx->nsub_tops; ++st_idx)
- {
- int sl_idx;
- re_sub_match_top_t *top = mctx->sub_tops[st_idx];
- for (sl_idx = 0; sl_idx < top->nlasts; ++sl_idx)
- {
- re_sub_match_last_t *last = top->lasts[sl_idx];
- re_free (last->path.array);
- re_free (last);
- }
- re_free (top->lasts);
- if (top->path)
- {
- re_free (top->path->array);
- re_free (top->path);
- }
- free (top);
- }
-}
-
-/* Add a new backreference entry to MCTX.
- Note that we assume that caller never call this function with duplicate
- entry, and call with STR_IDX which isn't smaller than any existing entry.
-*/
-
-static reg_errcode_t
-match_ctx_add_entry (mctx, node, str_idx, from, to)
- re_match_context_t *mctx;
- int node, str_idx, from, to;
-{
- if (mctx->nbkref_ents >= mctx->abkref_ents)
- {
- struct re_backref_cache_entry* new_entry;
- new_entry = re_realloc (mctx->bkref_ents, struct re_backref_cache_entry,
- mctx->abkref_ents * 2);
- if (BE (new_entry == NULL, 0))
- {
- re_free (mctx->bkref_ents);
- return REG_ESPACE;
- }
- mctx->bkref_ents = new_entry;
- memset (mctx->bkref_ents + mctx->nbkref_ents, '\0',
- sizeof (struct re_backref_cache_entry) * mctx->abkref_ents);
- mctx->abkref_ents *= 2;
- }
- mctx->bkref_ents[mctx->nbkref_ents].node = node;
- mctx->bkref_ents[mctx->nbkref_ents].str_idx = str_idx;
- mctx->bkref_ents[mctx->nbkref_ents].subexp_from = from;
- mctx->bkref_ents[mctx->nbkref_ents].subexp_to = to;
- mctx->bkref_ents[mctx->nbkref_ents++].flag = 0;
- if (mctx->max_mb_elem_len < to - from)
- mctx->max_mb_elem_len = to - from;
- return REG_NOERROR;
-}
-
-/* Search for the first entry which has the same str_idx.
- Note that MCTX->BKREF_ENTS is already sorted by MCTX->STR_IDX. */
-
-static int
-search_cur_bkref_entry (mctx, str_idx)
- re_match_context_t *mctx;
- int str_idx;
-{
- int left, right, mid;
- right = mctx->nbkref_ents;
- for (left = 0; left < right;)
- {
- mid = (left + right) / 2;
- if (mctx->bkref_ents[mid].str_idx < str_idx)
- left = mid + 1;
- else
- right = mid;
- }
- return left;
-}
-
-static void
-match_ctx_clear_flag (mctx)
- re_match_context_t *mctx;
-{
- int i;
- for (i = 0; i < mctx->nbkref_ents; ++i)
- {
- mctx->bkref_ents[i].flag = 0;
- }
-}
-
-/* Register the node NODE, whose type is OP_OPEN_SUBEXP, and which matches
- at STR_IDX. */
-
-static reg_errcode_t
-match_ctx_add_subtop (mctx, node, str_idx)
- re_match_context_t *mctx;
- int node, str_idx;
-{
-#ifdef DEBUG
- assert (mctx->sub_tops != NULL);
- assert (mctx->asub_tops > 0);
-#endif
- if (mctx->nsub_tops == mctx->asub_tops)
- {
- re_sub_match_top_t **new_array;
- mctx->asub_tops *= 2;
- new_array = re_realloc (mctx->sub_tops, re_sub_match_top_t *,
- mctx->asub_tops);
- if (BE (new_array == NULL, 0))
- return REG_ESPACE;
- mctx->sub_tops = new_array;
- }
- mctx->sub_tops[mctx->nsub_tops] = calloc (1, sizeof (re_sub_match_top_t));
- if (mctx->sub_tops[mctx->nsub_tops] == NULL)
- return REG_ESPACE;
- mctx->sub_tops[mctx->nsub_tops]->node = node;
- mctx->sub_tops[mctx->nsub_tops++]->str_idx = str_idx;
- return REG_NOERROR;
-}
-
-/* Register the node NODE, whose type is OP_CLOSE_SUBEXP, and which matches
- at STR_IDX, whose corresponding OP_OPEN_SUBEXP is SUB_TOP. */
-
-static re_sub_match_last_t *
-match_ctx_add_sublast (subtop, node, str_idx)
- re_sub_match_top_t *subtop;
- int node, str_idx;
-{
- re_sub_match_last_t *new_entry;
- if (subtop->nlasts == subtop->alasts)
- {
- re_sub_match_last_t **new_array;
- subtop->alasts = 2 * subtop->alasts + 1;
- new_array = re_realloc (subtop->lasts, re_sub_match_last_t *,
- subtop->alasts);
- if (BE (new_array == NULL, 0))
- return NULL;
- subtop->lasts = new_array;
- }
- new_entry = calloc (1, sizeof (re_sub_match_last_t));
- if (BE (new_entry == NULL, 0))
- return NULL;
- subtop->lasts[subtop->nlasts] = new_entry;
- new_entry->node = node;
- new_entry->str_idx = str_idx;
- ++subtop->nlasts;
- return new_entry;
-}
-
-static void
-sift_ctx_init (sctx, sifted_sts, limited_sts, last_node, last_str_idx,
- check_subexp)
- re_sift_context_t *sctx;
- re_dfastate_t **sifted_sts, **limited_sts;
- int last_node, last_str_idx, check_subexp;
-{
- sctx->sifted_states = sifted_sts;
- sctx->limited_states = limited_sts;
- sctx->last_node = last_node;
- sctx->last_str_idx = last_str_idx;
- sctx->check_subexp = check_subexp;
- sctx->cur_bkref = -1;
- sctx->cls_subexp_idx = -1;
- re_node_set_init_empty (&sctx->limits);
-}
--- /dev/null
+/* Extended regular expression matching and search library.
+ Copyright (C) 2002, 2003 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+ Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, write to the Free
+ Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ 02111-1307 USA. */
+
+static reg_errcode_t re_compile_internal (regex_t *preg, const char * pattern,
+ int length, reg_syntax_t syntax);
+static void re_compile_fastmap_iter (regex_t *bufp,
+ const re_dfastate_t *init_state,
+ char *fastmap);
+static reg_errcode_t init_dfa (re_dfa_t *dfa, int pat_len);
+static reg_errcode_t init_word_char (re_dfa_t *dfa);
+#ifdef RE_ENABLE_I18N
+static void free_charset (re_charset_t *cset);
+#endif /* RE_ENABLE_I18N */
+static void free_workarea_compile (regex_t *preg);
+static reg_errcode_t create_initial_state (re_dfa_t *dfa);
+static reg_errcode_t analyze (re_dfa_t *dfa);
+static reg_errcode_t analyze_tree (re_dfa_t *dfa, bin_tree_t *node);
+static void calc_first (re_dfa_t *dfa, bin_tree_t *node);
+static void calc_next (re_dfa_t *dfa, bin_tree_t *node);
+static void calc_epsdest (re_dfa_t *dfa, bin_tree_t *node);
+static reg_errcode_t duplicate_node_closure (re_dfa_t *dfa, int top_org_node,
+ int top_clone_node, int root_node,
+ unsigned int constraint);
+static reg_errcode_t duplicate_node (int *new_idx, re_dfa_t *dfa, int org_idx,
+ unsigned int constraint);
+static int search_duplicated_node (re_dfa_t *dfa, int org_node,
+ unsigned int constraint);
+static reg_errcode_t calc_eclosure (re_dfa_t *dfa);
+static reg_errcode_t calc_eclosure_iter (re_node_set *new_set, re_dfa_t *dfa,
+ int node, int root);
+static void calc_inveclosure (re_dfa_t *dfa);
+static int fetch_number (re_string_t *input, re_token_t *token,
+ reg_syntax_t syntax);
+static re_token_t fetch_token (re_string_t *input, reg_syntax_t syntax);
+static int peek_token (re_token_t *token, re_string_t *input,
+ reg_syntax_t syntax);
+static int peek_token_bracket (re_token_t *token, re_string_t *input,
+ reg_syntax_t syntax);
+static bin_tree_t *parse (re_string_t *regexp, regex_t *preg,
+ reg_syntax_t syntax, reg_errcode_t *err);
+static bin_tree_t *parse_reg_exp (re_string_t *regexp, regex_t *preg,
+ re_token_t *token, reg_syntax_t syntax,
+ int nest, reg_errcode_t *err);
+static bin_tree_t *parse_branch (re_string_t *regexp, regex_t *preg,
+ re_token_t *token, reg_syntax_t syntax,
+ int nest, reg_errcode_t *err);
+static bin_tree_t *parse_expression (re_string_t *regexp, regex_t *preg,
+ re_token_t *token, reg_syntax_t syntax,
+ int nest, reg_errcode_t *err);
+static bin_tree_t *parse_sub_exp (re_string_t *regexp, regex_t *preg,
+ re_token_t *token, reg_syntax_t syntax,
+ int nest, reg_errcode_t *err);
+static bin_tree_t *parse_dup_op (bin_tree_t *dup_elem, re_string_t *regexp,
+ re_dfa_t *dfa, re_token_t *token,
+ reg_syntax_t syntax, reg_errcode_t *err);
+static bin_tree_t *parse_bracket_exp (re_string_t *regexp, re_dfa_t *dfa,
+ re_token_t *token, reg_syntax_t syntax,
+ reg_errcode_t *err);
+static reg_errcode_t parse_bracket_element (bracket_elem_t *elem,
+ re_string_t *regexp,
+ re_token_t *token, int token_len,
+ re_dfa_t *dfa,
+ reg_syntax_t syntax);
+static reg_errcode_t parse_bracket_symbol (bracket_elem_t *elem,
+ re_string_t *regexp,
+ re_token_t *token);
+#ifndef _LIBC
+# ifdef RE_ENABLE_I18N
+static reg_errcode_t build_range_exp (re_bitset_ptr_t sbcset,
+ re_charset_t *mbcset, int *range_alloc,
+ bracket_elem_t *start_elem,
+ bracket_elem_t *end_elem);
+static reg_errcode_t build_collating_symbol (re_bitset_ptr_t sbcset,
+ re_charset_t *mbcset,
+ int *coll_sym_alloc,
+ const unsigned char *name);
+# else /* not RE_ENABLE_I18N */
+static reg_errcode_t build_range_exp (re_bitset_ptr_t sbcset,
+ bracket_elem_t *start_elem,
+ bracket_elem_t *end_elem);
+static reg_errcode_t build_collating_symbol (re_bitset_ptr_t sbcset,
+ const unsigned char *name);
+# endif /* not RE_ENABLE_I18N */
+#endif /* not _LIBC */
+#ifdef RE_ENABLE_I18N
+static reg_errcode_t build_equiv_class (re_bitset_ptr_t sbcset,
+ re_charset_t *mbcset,
+ int *equiv_class_alloc,
+ const unsigned char *name);
+static reg_errcode_t build_charclass (re_bitset_ptr_t sbcset,
+ re_charset_t *mbcset,
+ int *char_class_alloc,
+ const unsigned char *class_name,
+ reg_syntax_t syntax);
+#else /* not RE_ENABLE_I18N */
+static reg_errcode_t build_equiv_class (re_bitset_ptr_t sbcset,
+ const unsigned char *name);
+static reg_errcode_t build_charclass (re_bitset_ptr_t sbcset,
+ const unsigned char *class_name,
+ reg_syntax_t syntax);
+#endif /* not RE_ENABLE_I18N */
+static bin_tree_t *build_word_op (re_dfa_t *dfa, int not, reg_errcode_t *err);
+static void free_bin_tree (bin_tree_t *tree);
+static bin_tree_t *create_tree (bin_tree_t *left, bin_tree_t *right,
+ re_token_type_t type, int index);
+static bin_tree_t *duplicate_tree (const bin_tree_t *src, re_dfa_t *dfa);
+\f
+/* This table gives an error message for each of the error codes listed
+ in regex.h. Obviously the order here has to be same as there.
+ POSIX doesn't require that we do anything for REG_NOERROR,
+ but why not be nice? */
+
+const char __re_error_msgid[] attribute_hidden =
+ {
+#define REG_NOERROR_IDX 0
+ gettext_noop ("Success") /* REG_NOERROR */
+ "\0"
+#define REG_NOMATCH_IDX (REG_NOERROR_IDX + sizeof "Success")
+ gettext_noop ("No match") /* REG_NOMATCH */
+ "\0"
+#define REG_BADPAT_IDX (REG_NOMATCH_IDX + sizeof "No match")
+ gettext_noop ("Invalid regular expression") /* REG_BADPAT */
+ "\0"
+#define REG_ECOLLATE_IDX (REG_BADPAT_IDX + sizeof "Invalid regular expression")
+ gettext_noop ("Invalid collation character") /* REG_ECOLLATE */
+ "\0"
+#define REG_ECTYPE_IDX (REG_ECOLLATE_IDX + sizeof "Invalid collation character")
+ gettext_noop ("Invalid character class name") /* REG_ECTYPE */
+ "\0"
+#define REG_EESCAPE_IDX (REG_ECTYPE_IDX + sizeof "Invalid character class name")
+ gettext_noop ("Trailing backslash") /* REG_EESCAPE */
+ "\0"
+#define REG_ESUBREG_IDX (REG_EESCAPE_IDX + sizeof "Trailing backslash")
+ gettext_noop ("Invalid back reference") /* REG_ESUBREG */
+ "\0"
+#define REG_EBRACK_IDX (REG_ESUBREG_IDX + sizeof "Invalid back reference")
+ gettext_noop ("Unmatched [ or [^") /* REG_EBRACK */
+ "\0"
+#define REG_EPAREN_IDX (REG_EBRACK_IDX + sizeof "Unmatched [ or [^")
+ gettext_noop ("Unmatched ( or \\(") /* REG_EPAREN */
+ "\0"
+#define REG_EBRACE_IDX (REG_EPAREN_IDX + sizeof "Unmatched ( or \\(")
+ gettext_noop ("Unmatched \\{") /* REG_EBRACE */
+ "\0"
+#define REG_BADBR_IDX (REG_EBRACE_IDX + sizeof "Unmatched \\{")
+ gettext_noop ("Invalid content of \\{\\}") /* REG_BADBR */
+ "\0"
+#define REG_ERANGE_IDX (REG_BADBR_IDX + sizeof "Invalid content of \\{\\}")
+ gettext_noop ("Invalid range end") /* REG_ERANGE */
+ "\0"
+#define REG_ESPACE_IDX (REG_ERANGE_IDX + sizeof "Invalid range end")
+ gettext_noop ("Memory exhausted") /* REG_ESPACE */
+ "\0"
+#define REG_BADRPT_IDX (REG_ESPACE_IDX + sizeof "Memory exhausted")
+ gettext_noop ("Invalid preceding regular expression") /* REG_BADRPT */
+ "\0"
+#define REG_EEND_IDX (REG_BADRPT_IDX + sizeof "Invalid preceding regular expression")
+ gettext_noop ("Premature end of regular expression") /* REG_EEND */
+ "\0"
+#define REG_ESIZE_IDX (REG_EEND_IDX + sizeof "Premature end of regular expression")
+ gettext_noop ("Regular expression too big") /* REG_ESIZE */
+ "\0"
+#define REG_ERPAREN_IDX (REG_ESIZE_IDX + sizeof "Regular expression too big")
+ gettext_noop ("Unmatched ) or \\)") /* REG_ERPAREN */
+ };
+
+const size_t __re_error_msgid_idx[] attribute_hidden =
+ {
+ REG_NOERROR_IDX,
+ REG_NOMATCH_IDX,
+ REG_BADPAT_IDX,
+ REG_ECOLLATE_IDX,
+ REG_ECTYPE_IDX,
+ REG_EESCAPE_IDX,
+ REG_ESUBREG_IDX,
+ REG_EBRACK_IDX,
+ REG_EPAREN_IDX,
+ REG_EBRACE_IDX,
+ REG_BADBR_IDX,
+ REG_ERANGE_IDX,
+ REG_ESPACE_IDX,
+ REG_BADRPT_IDX,
+ REG_EEND_IDX,
+ REG_ESIZE_IDX,
+ REG_ERPAREN_IDX
+ };
+\f
+/* Entry points for GNU code. */
+
+/* re_compile_pattern is the GNU regular expression compiler: it
+ compiles PATTERN (of length LENGTH) and puts the result in BUFP.
+ Returns 0 if the pattern was valid, otherwise an error string.
+
+ Assumes the `allocated' (and perhaps `buffer') and `translate' fields
+ are set in BUFP on entry. */
+
+const char *
+re_compile_pattern (pattern, length, bufp)
+ const char *pattern;
+ size_t length;
+ struct re_pattern_buffer *bufp;
+{
+ reg_errcode_t ret;
+
+ /* And GNU code determines whether or not to get register information
+ by passing null for the REGS argument to re_match, etc., not by
+ setting no_sub. */
+ bufp->no_sub = 0;
+
+ /* Match anchors at newline. */
+ bufp->newline_anchor = 1;
+
+ ret = re_compile_internal (bufp, pattern, length, re_syntax_options);
+
+ if (!ret)
+ return NULL;
+ return gettext (__re_error_msgid + __re_error_msgid_idx[(int) ret]);
+}
+#ifdef _LIBC
+weak_alias (__re_compile_pattern, re_compile_pattern)
+#endif
+
+/* Set by `re_set_syntax' to the current regexp syntax to recognize. Can
+ also be assigned to arbitrarily: each pattern buffer stores its own
+ syntax, so it can be changed between regex compilations. */
+/* This has no initializer because initialized variables in Emacs
+ become read-only after dumping. */
+reg_syntax_t re_syntax_options;
+
+
+/* Specify the precise syntax of regexps for compilation. This provides
+ for compatibility for various utilities which historically have
+ different, incompatible syntaxes.
+
+ The argument SYNTAX is a bit mask comprised of the various bits
+ defined in regex.h. We return the old syntax. */
+
+reg_syntax_t
+re_set_syntax (syntax)
+ reg_syntax_t syntax;
+{
+ reg_syntax_t ret = re_syntax_options;
+
+ re_syntax_options = syntax;
+ return ret;
+}
+#ifdef _LIBC
+weak_alias (__re_set_syntax, re_set_syntax)
+#endif
+
+int
+re_compile_fastmap (bufp)
+ struct re_pattern_buffer *bufp;
+{
+ re_dfa_t *dfa = (re_dfa_t *) bufp->buffer;
+ char *fastmap = bufp->fastmap;
+
+ memset (fastmap, '\0', sizeof (char) * SBC_MAX);
+ re_compile_fastmap_iter (bufp, dfa->init_state, fastmap);
+ if (dfa->init_state != dfa->init_state_word)
+ re_compile_fastmap_iter (bufp, dfa->init_state_word, fastmap);
+ if (dfa->init_state != dfa->init_state_nl)
+ re_compile_fastmap_iter (bufp, dfa->init_state_nl, fastmap);
+ if (dfa->init_state != dfa->init_state_begbuf)
+ re_compile_fastmap_iter (bufp, dfa->init_state_begbuf, fastmap);
+ bufp->fastmap_accurate = 1;
+ return 0;
+}
+#ifdef _LIBC
+weak_alias (__re_compile_fastmap, re_compile_fastmap)
+#endif
+
+static inline void
+re_set_fastmap (char *fastmap, int icase, int ch)
+{
+ fastmap[ch] = 1;
+ if (icase)
+ fastmap[tolower (ch)] = 1;
+}
+
+/* Helper function for re_compile_fastmap.
+ Compile fastmap for the initial_state INIT_STATE. */
+
+static void
+re_compile_fastmap_iter (bufp, init_state, fastmap)
+ regex_t *bufp;
+ const re_dfastate_t *init_state;
+ char *fastmap;
+{
+ re_dfa_t *dfa = (re_dfa_t *) bufp->buffer;
+ int node_cnt;
+ int icase = (MB_CUR_MAX == 1 && (bufp->syntax & RE_ICASE));
+ for (node_cnt = 0; node_cnt < init_state->nodes.nelem; ++node_cnt)
+ {
+ int node = init_state->nodes.elems[node_cnt];
+ re_token_type_t type = dfa->nodes[node].type;
+
+ if (type == CHARACTER)
+ re_set_fastmap (fastmap, icase, dfa->nodes[node].opr.c);
+ else if (type == SIMPLE_BRACKET)
+ {
+ int i, j, ch;
+ for (i = 0, ch = 0; i < BITSET_UINTS; ++i)
+ for (j = 0; j < UINT_BITS; ++j, ++ch)
+ if (dfa->nodes[node].opr.sbcset[i] & (1 << j))
+ re_set_fastmap (fastmap, icase, ch);
+ }
+#ifdef RE_ENABLE_I18N
+ else if (type == COMPLEX_BRACKET)
+ {
+ int i;
+ re_charset_t *cset = dfa->nodes[node].opr.mbcset;
+ if (cset->non_match || cset->ncoll_syms || cset->nequiv_classes
+ || cset->nranges || cset->nchar_classes)
+ {
+# ifdef _LIBC
+ if (_NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES) != 0)
+ {
+ /* In this case we want to catch the bytes which are
+ the first byte of any collation elements.
+ e.g. In da_DK, we want to catch 'a' since "aa"
+ is a valid collation element, and don't catch
+ 'b' since 'b' is the only collation element
+ which starts from 'b'. */
+ int j, ch;
+ const int32_t *table = (const int32_t *)
+ _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB);
+ for (i = 0, ch = 0; i < BITSET_UINTS; ++i)
+ for (j = 0; j < UINT_BITS; ++j, ++ch)
+ if (table[ch] < 0)
+ re_set_fastmap (fastmap, icase, ch);
+ }
+# else
+ if (MB_CUR_MAX > 1)
+ for (i = 0; i < SBC_MAX; ++i)
+ if (__btowc (i) == WEOF)
+ re_set_fastmap (fastmap, icase, i);
+# endif /* not _LIBC */
+ }
+ for (i = 0; i < cset->nmbchars; ++i)
+ {
+ char buf[256];
+ mbstate_t state;
+ memset (&state, '\0', sizeof (state));
+ __wcrtomb (buf, cset->mbchars[i], &state);
+ re_set_fastmap (fastmap, icase, *(unsigned char *) buf);
+ }
+ }
+#endif /* RE_ENABLE_I18N */
+ else if (type == END_OF_RE || type == OP_PERIOD)
+ {
+ memset (fastmap, '\1', sizeof (char) * SBC_MAX);
+ if (type == END_OF_RE)
+ bufp->can_be_null = 1;
+ return;
+ }
+ }
+}
+\f
+/* Entry point for POSIX code. */
+/* regcomp takes a regular expression as a string and compiles it.
+
+ PREG is a regex_t *. We do not expect any fields to be initialized,
+ since POSIX says we shouldn't. Thus, we set
+
+ `buffer' to the compiled pattern;
+ `used' to the length of the compiled pattern;
+ `syntax' to RE_SYNTAX_POSIX_EXTENDED if the
+ REG_EXTENDED bit in CFLAGS is set; otherwise, to
+ RE_SYNTAX_POSIX_BASIC;
+ `newline_anchor' to REG_NEWLINE being set in CFLAGS;
+ `fastmap' to an allocated space for the fastmap;
+ `fastmap_accurate' to zero;
+ `re_nsub' to the number of subexpressions in PATTERN.
+
+ PATTERN is the address of the pattern string.
+
+ CFLAGS is a series of bits which affect compilation.
+
+ If REG_EXTENDED is set, we use POSIX extended syntax; otherwise, we
+ use POSIX basic syntax.
+
+ If REG_NEWLINE is set, then . and [^...] don't match newline.
+ Also, regexec will try a match beginning after every newline.
+
+ If REG_ICASE is set, then we considers upper- and lowercase
+ versions of letters to be equivalent when matching.
+
+ If REG_NOSUB is set, then when PREG is passed to regexec, that
+ routine will report only success or failure, and nothing about the
+ registers.
+
+ It returns 0 if it succeeds, nonzero if it doesn't. (See regex.h for
+ the return codes and their meanings.) */
+
+int
+regcomp (preg, pattern, cflags)
+ regex_t *__restrict preg;
+ const char *__restrict pattern;
+ int cflags;
+{
+ reg_errcode_t ret;
+ reg_syntax_t syntax = ((cflags & REG_EXTENDED) ? RE_SYNTAX_POSIX_EXTENDED
+ : RE_SYNTAX_POSIX_BASIC);
+
+ preg->buffer = NULL;
+ preg->allocated = 0;
+ preg->used = 0;
+
+ /* Try to allocate space for the fastmap. */
+ preg->fastmap = re_malloc (char, SBC_MAX);
+ if (BE (preg->fastmap == NULL, 0))
+ return REG_ESPACE;
+
+ syntax |= (cflags & REG_ICASE) ? RE_ICASE : 0;
+
+ /* If REG_NEWLINE is set, newlines are treated differently. */
+ if (cflags & REG_NEWLINE)
+ { /* REG_NEWLINE implies neither . nor [^...] match newline. */
+ syntax &= ~RE_DOT_NEWLINE;
+ syntax |= RE_HAT_LISTS_NOT_NEWLINE;
+ /* It also changes the matching behavior. */
+ preg->newline_anchor = 1;
+ }
+ else
+ preg->newline_anchor = 0;
+ preg->no_sub = !!(cflags & REG_NOSUB);
+ preg->translate = NULL;
+
+ ret = re_compile_internal (preg, pattern, strlen (pattern), syntax);
+
+ /* POSIX doesn't distinguish between an unmatched open-group and an
+ unmatched close-group: both are REG_EPAREN. */
+ if (ret == REG_ERPAREN)
+ ret = REG_EPAREN;
+
+ /* We have already checked preg->fastmap != NULL. */
+ if (BE (ret == REG_NOERROR, 1))
+ /* Compute the fastmap now, since regexec cannot modify the pattern
+ buffer. This function nevers fails in this implementation. */
+ (void) re_compile_fastmap (preg);
+ else
+ {
+ /* Some error occurred while compiling the expression. */
+ re_free (preg->fastmap);
+ preg->fastmap = NULL;
+ }
+
+ return (int) ret;
+}
+#ifdef _LIBC
+weak_alias (__regcomp, regcomp)
+#endif
+
+/* Returns a message corresponding to an error code, ERRCODE, returned
+ from either regcomp or regexec. We don't use PREG here. */
+
+size_t
+regerror (errcode, preg, errbuf, errbuf_size)
+ int errcode;
+ const regex_t *preg;
+ char *errbuf;
+ size_t errbuf_size;
+{
+ const char *msg;
+ size_t msg_size;
+
+ if (BE (errcode < 0
+ || errcode >= (int) (sizeof (__re_error_msgid_idx)
+ / sizeof (__re_error_msgid_idx[0])), 0))
+ /* Only error codes returned by the rest of the code should be passed
+ to this routine. If we are given anything else, or if other regex
+ code generates an invalid error code, then the program has a bug.
+ Dump core so we can fix it. */
+ abort ();
+
+ msg = gettext (__re_error_msgid + __re_error_msgid_idx[errcode]);
+
+ msg_size = strlen (msg) + 1; /* Includes the null. */
+
+ if (BE (errbuf_size != 0, 1))
+ {
+ if (BE (msg_size > errbuf_size, 0))
+ {
+#if defined HAVE_MEMPCPY || defined _LIBC
+ *((char *) __mempcpy (errbuf, msg, errbuf_size - 1)) = '\0';
+#else
+ memcpy (errbuf, msg, errbuf_size - 1);
+ errbuf[errbuf_size - 1] = 0;
+#endif
+ }
+ else
+ memcpy (errbuf, msg, msg_size);
+ }
+
+ return msg_size;
+}
+#ifdef _LIBC
+weak_alias (__regerror, regerror)
+#endif
+
+
+static void
+free_dfa_content (re_dfa_t *dfa)
+{
+ int i, j;
+
+ re_free (dfa->subexps);
+
+ for (i = 0; i < dfa->nodes_len; ++i)
+ {
+ re_token_t *node = dfa->nodes + i;
+#ifdef RE_ENABLE_I18N
+ if (node->type == COMPLEX_BRACKET && node->duplicated == 0)
+ free_charset (node->opr.mbcset);
+ else
+#endif /* RE_ENABLE_I18N */
+ if (node->type == SIMPLE_BRACKET && node->duplicated == 0)
+ re_free (node->opr.sbcset);
+ }
+ re_free (dfa->nexts);
+ for (i = 0; i < dfa->nodes_len; ++i)
+ {
+ if (dfa->eclosures != NULL)
+ re_node_set_free (dfa->eclosures + i);
+ if (dfa->inveclosures != NULL)
+ re_node_set_free (dfa->inveclosures + i);
+ if (dfa->edests != NULL)
+ re_node_set_free (dfa->edests + i);
+ }
+ re_free (dfa->edests);
+ re_free (dfa->eclosures);
+ re_free (dfa->inveclosures);
+ re_free (dfa->nodes);
+
+ for (i = 0; i <= dfa->state_hash_mask; ++i)
+ {
+ struct re_state_table_entry *entry = dfa->state_table + i;
+ for (j = 0; j < entry->num; ++j)
+ {
+ re_dfastate_t *state = entry->array[j];
+ free_state (state);
+ }
+ re_free (entry->array);
+ }
+ re_free (dfa->state_table);
+
+ if (dfa->word_char != NULL)
+ re_free (dfa->word_char);
+#ifdef DEBUG
+ re_free (dfa->re_str);
+#endif
+
+ re_free (dfa);
+}
+
+
+/* Free dynamically allocated space used by PREG. */
+
+void
+regfree (preg)
+ regex_t *preg;
+{
+ re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ if (BE (dfa != NULL, 1))
+ free_dfa_content (dfa);
+
+ re_free (preg->fastmap);
+}
+#ifdef _LIBC
+weak_alias (__regfree, regfree)
+#endif
+\f
+/* Entry points compatible with 4.2 BSD regex library. We don't define
+ them unless specifically requested. */
+
+#if defined _REGEX_RE_COMP || defined _LIBC
+
+/* BSD has one and only one pattern buffer. */
+static struct re_pattern_buffer re_comp_buf;
+
+char *
+# ifdef _LIBC
+/* Make these definitions weak in libc, so POSIX programs can redefine
+ these names if they don't use our functions, and still use
+ regcomp/regexec above without link errors. */
+weak_function
+# endif
+re_comp (s)
+ const char *s;
+{
+ reg_errcode_t ret;
+ char *fastmap;
+
+ if (!s)
+ {
+ if (!re_comp_buf.buffer)
+ return gettext ("No previous regular expression");
+ return 0;
+ }
+
+ if (re_comp_buf.buffer)
+ {
+ fastmap = re_comp_buf.fastmap;
+ re_comp_buf.fastmap = NULL;
+ __regfree (&re_comp_buf);
+ memset (&re_comp_buf, '\0', sizeof (re_comp_buf));
+ re_comp_buf.fastmap = fastmap;
+ }
+
+ if (re_comp_buf.fastmap == NULL)
+ {
+ re_comp_buf.fastmap = (char *) malloc (SBC_MAX);
+ if (re_comp_buf.fastmap == NULL)
+ return (char *) gettext (__re_error_msgid
+ + __re_error_msgid_idx[(int) REG_ESPACE]);
+ }
+
+ /* Since `re_exec' always passes NULL for the `regs' argument, we
+ don't need to initialize the pattern buffer fields which affect it. */
+
+ /* Match anchors at newlines. */
+ re_comp_buf.newline_anchor = 1;
+
+ ret = re_compile_internal (&re_comp_buf, s, strlen (s), re_syntax_options);
+
+ if (!ret)
+ return NULL;
+
+ /* Yes, we're discarding `const' here if !HAVE_LIBINTL. */
+ return (char *) gettext (__re_error_msgid + __re_error_msgid_idx[(int) ret]);
+}
+
+#ifdef _LIBC
+libc_freeres_fn (free_mem)
+{
+ __regfree (&re_comp_buf);
+}
+#endif
+
+#endif /* _REGEX_RE_COMP */
+\f
+/* Internal entry point.
+ Compile the regular expression PATTERN, whose length is LENGTH.
+ SYNTAX indicate regular expression's syntax. */
+
+static reg_errcode_t
+re_compile_internal (preg, pattern, length, syntax)
+ regex_t *preg;
+ const char * pattern;
+ int length;
+ reg_syntax_t syntax;
+{
+ reg_errcode_t err = REG_NOERROR;
+ re_dfa_t *dfa;
+ re_string_t regexp;
+
+ /* Initialize the pattern buffer. */
+ preg->fastmap_accurate = 0;
+ preg->syntax = syntax;
+ preg->not_bol = preg->not_eol = 0;
+ preg->used = 0;
+ preg->re_nsub = 0;
+ preg->can_be_null = 0;
+ preg->regs_allocated = REGS_UNALLOCATED;
+
+ /* Initialize the dfa. */
+ dfa = (re_dfa_t *) preg->buffer;
+ if (preg->allocated < sizeof (re_dfa_t))
+ {
+ /* If zero allocated, but buffer is non-null, try to realloc
+ enough space. This loses if buffer's address is bogus, but
+ that is the user's responsibility. If ->buffer is NULL this
+ is a simple allocation. */
+ dfa = re_realloc (preg->buffer, re_dfa_t, 1);
+ if (dfa == NULL)
+ return REG_ESPACE;
+ preg->allocated = sizeof (re_dfa_t);
+ }
+ preg->buffer = (unsigned char *) dfa;
+ preg->used = sizeof (re_dfa_t);
+
+ err = init_dfa (dfa, length);
+ if (BE (err != REG_NOERROR, 0))
+ {
+ re_free (dfa);
+ preg->buffer = NULL;
+ preg->allocated = 0;
+ return err;
+ }
+#ifdef DEBUG
+ dfa->re_str = re_malloc (char, length + 1);
+ strncpy (dfa->re_str, pattern, length + 1);
+#endif
+
+ err = re_string_construct (®exp, pattern, length, preg->translate,
+ syntax & RE_ICASE);
+ if (BE (err != REG_NOERROR, 0))
+ {
+ re_free (dfa);
+ preg->buffer = NULL;
+ preg->allocated = 0;
+ return err;
+ }
+
+ /* Parse the regular expression, and build a structure tree. */
+ preg->re_nsub = 0;
+ dfa->str_tree = parse (®exp, preg, syntax, &err);
+ if (BE (dfa->str_tree == NULL, 0))
+ goto re_compile_internal_free_return;
+
+ /* Analyze the tree and collect information which is necessary to
+ create the dfa. */
+ err = analyze (dfa);
+ if (BE (err != REG_NOERROR, 0))
+ goto re_compile_internal_free_return;
+
+ /* Then create the initial state of the dfa. */
+ err = create_initial_state (dfa);
+
+ /* Release work areas. */
+ free_workarea_compile (preg);
+ re_string_destruct (®exp);
+
+ if (BE (err != REG_NOERROR, 0))
+ {
+ re_compile_internal_free_return:
+ free_dfa_content (dfa);
+ preg->buffer = NULL;
+ preg->allocated = 0;
+ }
+
+ return err;
+}
+
+/* Initialize DFA. We use the length of the regular expression PAT_LEN
+ as the initial length of some arrays. */
+
+static reg_errcode_t
+init_dfa (dfa, pat_len)
+ re_dfa_t *dfa;
+ int pat_len;
+{
+ int table_size;
+
+ memset (dfa, '\0', sizeof (re_dfa_t));
+
+ dfa->nodes_alloc = pat_len + 1;
+ dfa->nodes = re_malloc (re_token_t, dfa->nodes_alloc);
+
+ dfa->states_alloc = pat_len + 1;
+
+ /* table_size = 2 ^ ceil(log pat_len) */
+ for (table_size = 1; table_size > 0; table_size <<= 1)
+ if (table_size > pat_len)
+ break;
+
+ dfa->state_table = calloc (sizeof (struct re_state_table_entry), table_size);
+ dfa->state_hash_mask = table_size - 1;
+
+ dfa->subexps_alloc = 1;
+ dfa->subexps = re_malloc (re_subexp_t, dfa->subexps_alloc);
+ dfa->word_char = NULL;
+
+ if (BE (dfa->nodes == NULL || dfa->state_table == NULL
+ || dfa->subexps == NULL, 0))
+ {
+ /* We don't bother to free anything which was allocated. Very
+ soon the process will go down anyway. */
+ dfa->subexps = NULL;
+ dfa->state_table = NULL;
+ dfa->nodes = NULL;
+ return REG_ESPACE;
+ }
+ return REG_NOERROR;
+}
+
+/* Initialize WORD_CHAR table, which indicate which character is
+ "word". In this case "word" means that it is the word construction
+ character used by some operators like "\<", "\>", etc. */
+
+static reg_errcode_t
+init_word_char (dfa)
+ re_dfa_t *dfa;
+{
+ int i, j, ch;
+ dfa->word_char = (re_bitset_ptr_t) calloc (sizeof (bitset), 1);
+ if (BE (dfa->word_char == NULL, 0))
+ return REG_ESPACE;
+ for (i = 0, ch = 0; i < BITSET_UINTS; ++i)
+ for (j = 0; j < UINT_BITS; ++j, ++ch)
+ if (isalnum (ch) || ch == '_')
+ dfa->word_char[i] |= 1 << j;
+ return REG_NOERROR;
+}
+
+/* Free the work area which are only used while compiling. */
+
+static void
+free_workarea_compile (preg)
+ regex_t *preg;
+{
+ re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ free_bin_tree (dfa->str_tree);
+ dfa->str_tree = NULL;
+ re_free (dfa->org_indices);
+ dfa->org_indices = NULL;
+}
+
+/* Create initial states for all contexts. */
+
+static reg_errcode_t
+create_initial_state (dfa)
+ re_dfa_t *dfa;
+{
+ int first, i;
+ reg_errcode_t err;
+ re_node_set init_nodes;
+
+ /* Initial states have the epsilon closure of the node which is
+ the first node of the regular expression. */
+ first = dfa->str_tree->first;
+ dfa->init_node = first;
+ err = re_node_set_init_copy (&init_nodes, dfa->eclosures + first);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+
+ /* The back-references which are in initial states can epsilon transit,
+ since in this case all of the subexpressions can be null.
+ Then we add epsilon closures of the nodes which are the next nodes of
+ the back-references. */
+ if (dfa->nbackref > 0)
+ for (i = 0; i < init_nodes.nelem; ++i)
+ {
+ int node_idx = init_nodes.elems[i];
+ re_token_type_t type = dfa->nodes[node_idx].type;
+
+ int clexp_idx;
+ if (type != OP_BACK_REF)
+ continue;
+ for (clexp_idx = 0; clexp_idx < init_nodes.nelem; ++clexp_idx)
+ {
+ re_token_t *clexp_node;
+ clexp_node = dfa->nodes + init_nodes.elems[clexp_idx];
+ if (clexp_node->type == OP_CLOSE_SUBEXP
+ && clexp_node->opr.idx + 1 == dfa->nodes[node_idx].opr.idx)
+ break;
+ }
+ if (clexp_idx == init_nodes.nelem)
+ continue;
+
+ if (type == OP_BACK_REF)
+ {
+ int dest_idx = dfa->edests[node_idx].elems[0];
+ if (!re_node_set_contains (&init_nodes, dest_idx))
+ {
+ re_node_set_merge (&init_nodes, dfa->eclosures + dest_idx);
+ i = 0;
+ }
+ }
+ }
+
+ /* It must be the first time to invoke acquire_state. */
+ dfa->init_state = re_acquire_state_context (&err, dfa, &init_nodes, 0);
+ /* We don't check ERR here, since the initial state must not be NULL. */
+ if (BE (dfa->init_state == NULL, 0))
+ return err;
+ if (dfa->init_state->has_constraint)
+ {
+ dfa->init_state_word = re_acquire_state_context (&err, dfa, &init_nodes,
+ CONTEXT_WORD);
+ dfa->init_state_nl = re_acquire_state_context (&err, dfa, &init_nodes,
+ CONTEXT_NEWLINE);
+ dfa->init_state_begbuf = re_acquire_state_context (&err, dfa,
+ &init_nodes,
+ CONTEXT_NEWLINE
+ | CONTEXT_BEGBUF);
+ if (BE (dfa->init_state_word == NULL || dfa->init_state_nl == NULL
+ || dfa->init_state_begbuf == NULL, 0))
+ return err;
+ }
+ else
+ dfa->init_state_word = dfa->init_state_nl
+ = dfa->init_state_begbuf = dfa->init_state;
+
+ re_node_set_free (&init_nodes);
+ return REG_NOERROR;
+}
+\f
+/* Analyze the structure tree, and calculate "first", "next", "edest",
+ "eclosure", and "inveclosure". */
+
+static reg_errcode_t
+analyze (dfa)
+ re_dfa_t *dfa;
+{
+ int i;
+ reg_errcode_t ret;
+
+ /* Allocate arrays. */
+ dfa->nexts = re_malloc (int, dfa->nodes_alloc);
+ dfa->org_indices = re_malloc (int, dfa->nodes_alloc);
+ dfa->edests = re_malloc (re_node_set, dfa->nodes_alloc);
+ dfa->eclosures = re_malloc (re_node_set, dfa->nodes_alloc);
+ dfa->inveclosures = re_malloc (re_node_set, dfa->nodes_alloc);
+ if (BE (dfa->nexts == NULL || dfa->org_indices == NULL || dfa->edests == NULL
+ || dfa->eclosures == NULL || dfa->inveclosures == NULL, 0))
+ return REG_ESPACE;
+ /* Initialize them. */
+ for (i = 0; i < dfa->nodes_len; ++i)
+ {
+ dfa->nexts[i] = -1;
+ re_node_set_init_empty (dfa->edests + i);
+ re_node_set_init_empty (dfa->eclosures + i);
+ re_node_set_init_empty (dfa->inveclosures + i);
+ }
+
+ ret = analyze_tree (dfa, dfa->str_tree);
+ if (BE (ret == REG_NOERROR, 1))
+ {
+ ret = calc_eclosure (dfa);
+ if (ret == REG_NOERROR)
+ calc_inveclosure (dfa);
+ }
+ return ret;
+}
+
+/* Helper functions for analyze.
+ This function calculate "first", "next", and "edest" for the subtree
+ whose root is NODE. */
+
+static reg_errcode_t
+analyze_tree (dfa, node)
+ re_dfa_t *dfa;
+ bin_tree_t *node;
+{
+ reg_errcode_t ret;
+ if (node->first == -1)
+ calc_first (dfa, node);
+ if (node->next == -1)
+ calc_next (dfa, node);
+ if (node->eclosure.nelem == 0)
+ calc_epsdest (dfa, node);
+ /* Calculate "first" etc. for the left child. */
+ if (node->left != NULL)
+ {
+ ret = analyze_tree (dfa, node->left);
+ if (BE (ret != REG_NOERROR, 0))
+ return ret;
+ }
+ /* Calculate "first" etc. for the right child. */
+ if (node->right != NULL)
+ {
+ ret = analyze_tree (dfa, node->right);
+ if (BE (ret != REG_NOERROR, 0))
+ return ret;
+ }
+ return REG_NOERROR;
+}
+
+/* Calculate "first" for the node NODE. */
+static void
+calc_first (dfa, node)
+ re_dfa_t *dfa;
+ bin_tree_t *node;
+{
+ int idx, type;
+ idx = node->node_idx;
+ type = (node->type == 0) ? dfa->nodes[idx].type : node->type;
+
+ switch (type)
+ {
+#ifdef DEBUG
+ case OP_OPEN_BRACKET:
+ case OP_CLOSE_BRACKET:
+ case OP_OPEN_DUP_NUM:
+ case OP_CLOSE_DUP_NUM:
+ case OP_NON_MATCH_LIST:
+ case OP_OPEN_COLL_ELEM:
+ case OP_CLOSE_COLL_ELEM:
+ case OP_OPEN_EQUIV_CLASS:
+ case OP_CLOSE_EQUIV_CLASS:
+ case OP_OPEN_CHAR_CLASS:
+ case OP_CLOSE_CHAR_CLASS:
+ /* These must not be appeared here. */
+ assert (0);
+#endif
+ case END_OF_RE:
+ case CHARACTER:
+ case OP_PERIOD:
+ case OP_DUP_ASTERISK:
+ case OP_DUP_QUESTION:
+#ifdef RE_ENABLE_I18N
+ case COMPLEX_BRACKET:
+#endif /* RE_ENABLE_I18N */
+ case SIMPLE_BRACKET:
+ case OP_BACK_REF:
+ case ANCHOR:
+ case OP_OPEN_SUBEXP:
+ case OP_CLOSE_SUBEXP:
+ node->first = idx;
+ break;
+ case OP_DUP_PLUS:
+#ifdef DEBUG
+ assert (node->left != NULL);
+#endif
+ if (node->left->first == -1)
+ calc_first (dfa, node->left);
+ node->first = node->left->first;
+ break;
+ case OP_ALT:
+ node->first = idx;
+ break;
+ /* else fall through */
+ default:
+#ifdef DEBUG
+ assert (node->left != NULL);
+#endif
+ if (node->left->first == -1)
+ calc_first (dfa, node->left);
+ node->first = node->left->first;
+ break;
+ }
+}
+
+/* Calculate "next" for the node NODE. */
+
+static void
+calc_next (dfa, node)
+ re_dfa_t *dfa;
+ bin_tree_t *node;
+{
+ int idx, type;
+ bin_tree_t *parent = node->parent;
+ if (parent == NULL)
+ {
+ node->next = -1;
+ idx = node->node_idx;
+ if (node->type == 0)
+ dfa->nexts[idx] = node->next;
+ return;
+ }
+
+ idx = parent->node_idx;
+ type = (parent->type == 0) ? dfa->nodes[idx].type : parent->type;
+
+ switch (type)
+ {
+ case OP_DUP_ASTERISK:
+ case OP_DUP_PLUS:
+ node->next = idx;
+ break;
+ case CONCAT:
+ if (parent->left == node)
+ {
+ if (parent->right->first == -1)
+ calc_first (dfa, parent->right);
+ node->next = parent->right->first;
+ break;
+ }
+ /* else fall through */
+ default:
+ if (parent->next == -1)
+ calc_next (dfa, parent);
+ node->next = parent->next;
+ break;
+ }
+ idx = node->node_idx;
+ if (node->type == 0)
+ dfa->nexts[idx] = node->next;
+}
+
+/* Calculate "edest" for the node NODE. */
+
+static void
+calc_epsdest (dfa, node)
+ re_dfa_t *dfa;
+ bin_tree_t *node;
+{
+ int idx;
+ idx = node->node_idx;
+ if (node->type == 0)
+ {
+ if (dfa->nodes[idx].type == OP_DUP_ASTERISK
+ || dfa->nodes[idx].type == OP_DUP_PLUS
+ || dfa->nodes[idx].type == OP_DUP_QUESTION)
+ {
+ if (node->left->first == -1)
+ calc_first (dfa, node->left);
+ if (node->next == -1)
+ calc_next (dfa, node);
+ re_node_set_init_2 (dfa->edests + idx, node->left->first,
+ node->next);
+ }
+ else if (dfa->nodes[idx].type == OP_ALT)
+ {
+ int left, right;
+ if (node->left != NULL)
+ {
+ if (node->left->first == -1)
+ calc_first (dfa, node->left);
+ left = node->left->first;
+ }
+ else
+ {
+ if (node->next == -1)
+ calc_next (dfa, node);
+ left = node->next;
+ }
+ if (node->right != NULL)
+ {
+ if (node->right->first == -1)
+ calc_first (dfa, node->right);
+ right = node->right->first;
+ }
+ else
+ {
+ if (node->next == -1)
+ calc_next (dfa, node);
+ right = node->next;
+ }
+ re_node_set_init_2 (dfa->edests + idx, left, right);
+ }
+ else if (dfa->nodes[idx].type == ANCHOR
+ || dfa->nodes[idx].type == OP_OPEN_SUBEXP
+ || dfa->nodes[idx].type == OP_CLOSE_SUBEXP
+ || dfa->nodes[idx].type == OP_BACK_REF)
+ re_node_set_init_1 (dfa->edests + idx, node->next);
+ }
+}
+
+/* Duplicate the epsilon closure of the node ROOT_NODE.
+ Note that duplicated nodes have constraint INIT_CONSTRAINT in addition
+ to their own constraint. */
+
+static reg_errcode_t
+duplicate_node_closure (dfa, top_org_node, top_clone_node, root_node,
+ init_constraint)
+ re_dfa_t *dfa;
+ int top_org_node, top_clone_node, root_node;
+ unsigned int init_constraint;
+{
+ reg_errcode_t err;
+ int org_node, clone_node, ret;
+ unsigned int constraint = init_constraint;
+ for (org_node = top_org_node, clone_node = top_clone_node;;)
+ {
+ int org_dest, clone_dest;
+ if (dfa->nodes[org_node].type == OP_BACK_REF)
+ {
+ /* If the back reference epsilon-transit, its destination must
+ also have the constraint. Then duplicate the epsilon closure
+ of the destination of the back reference, and store it in
+ edests of the back reference. */
+ org_dest = dfa->nexts[org_node];
+ re_node_set_empty (dfa->edests + clone_node);
+ err = duplicate_node (&clone_dest, dfa, org_dest, constraint);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ dfa->nexts[clone_node] = dfa->nexts[org_node];
+ ret = re_node_set_insert (dfa->edests + clone_node, clone_dest);
+ if (BE (ret < 0, 0))
+ return REG_ESPACE;
+ }
+ else if (dfa->edests[org_node].nelem == 0)
+ {
+ /* In case of the node can't epsilon-transit, don't duplicate the
+ destination and store the original destination as the
+ destination of the node. */
+ dfa->nexts[clone_node] = dfa->nexts[org_node];
+ break;
+ }
+ else if (dfa->edests[org_node].nelem == 1)
+ {
+ /* In case of the node can epsilon-transit, and it has only one
+ destination. */
+ org_dest = dfa->edests[org_node].elems[0];
+ re_node_set_empty (dfa->edests + clone_node);
+ if (dfa->nodes[org_node].type == ANCHOR)
+ {
+ /* In case of the node has another constraint, append it. */
+ if (org_node == root_node && clone_node != org_node)
+ {
+ /* ...but if the node is root_node itself, it means the
+ epsilon closure have a loop, then tie it to the
+ destination of the root_node. */
+ ret = re_node_set_insert (dfa->edests + clone_node,
+ org_dest);
+ if (BE (ret < 0, 0))
+ return REG_ESPACE;
+ break;
+ }
+ constraint |= dfa->nodes[org_node].opr.ctx_type;
+ }
+ err = duplicate_node (&clone_dest, dfa, org_dest, constraint);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ ret = re_node_set_insert (dfa->edests + clone_node, clone_dest);
+ if (BE (ret < 0, 0))
+ return REG_ESPACE;
+ }
+ else /* dfa->edests[org_node].nelem == 2 */
+ {
+ /* In case of the node can epsilon-transit, and it has two
+ destinations. E.g. '|', '*', '+', '?'. */
+ org_dest = dfa->edests[org_node].elems[0];
+ re_node_set_empty (dfa->edests + clone_node);
+ /* Search for a duplicated node which satisfies the constraint. */
+ clone_dest = search_duplicated_node (dfa, org_dest, constraint);
+ if (clone_dest == -1)
+ {
+ /* There are no such a duplicated node, create a new one. */
+ err = duplicate_node (&clone_dest, dfa, org_dest, constraint);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ ret = re_node_set_insert (dfa->edests + clone_node, clone_dest);
+ if (BE (ret < 0, 0))
+ return REG_ESPACE;
+ err = duplicate_node_closure (dfa, org_dest, clone_dest,
+ root_node, constraint);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ }
+ else
+ {
+ /* There are a duplicated node which satisfy the constraint,
+ use it to avoid infinite loop. */
+ ret = re_node_set_insert (dfa->edests + clone_node, clone_dest);
+ if (BE (ret < 0, 0))
+ return REG_ESPACE;
+ }
+
+ org_dest = dfa->edests[org_node].elems[1];
+ err = duplicate_node (&clone_dest, dfa, org_dest, constraint);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ ret = re_node_set_insert (dfa->edests + clone_node, clone_dest);
+ if (BE (ret < 0, 0))
+ return REG_ESPACE;
+ }
+ org_node = org_dest;
+ clone_node = clone_dest;
+ }
+ return REG_NOERROR;
+}
+
+/* Search for a node which is duplicated from the node ORG_NODE, and
+ satisfies the constraint CONSTRAINT. */
+
+static int
+search_duplicated_node (dfa, org_node, constraint)
+ re_dfa_t *dfa;
+ int org_node;
+ unsigned int constraint;
+{
+ int idx;
+ for (idx = dfa->nodes_len - 1; dfa->nodes[idx].duplicated && idx > 0; --idx)
+ {
+ if (org_node == dfa->org_indices[idx]
+ && constraint == dfa->nodes[idx].constraint)
+ return idx; /* Found. */
+ }
+ return -1; /* Not found. */
+}
+
+/* Duplicate the node whose index is ORG_IDX and set the constraint CONSTRAINT.
+ The new index will be stored in NEW_IDX and return REG_NOERROR if succeeded,
+ otherwise return the error code. */
+
+static reg_errcode_t
+duplicate_node (new_idx, dfa, org_idx, constraint)
+ re_dfa_t *dfa;
+ int *new_idx, org_idx;
+ unsigned int constraint;
+{
+ re_token_t dup;
+ int dup_idx;
+
+ dup = dfa->nodes[org_idx];
+ dup_idx = re_dfa_add_node (dfa, dup, 1);
+ if (BE (dup_idx == -1, 0))
+ return REG_ESPACE;
+ dfa->nodes[dup_idx].constraint = constraint;
+ if (dfa->nodes[org_idx].type == ANCHOR)
+ dfa->nodes[dup_idx].constraint |= dfa->nodes[org_idx].opr.ctx_type;
+ dfa->nodes[dup_idx].duplicated = 1;
+ re_node_set_init_empty (dfa->edests + dup_idx);
+ re_node_set_init_empty (dfa->eclosures + dup_idx);
+ re_node_set_init_empty (dfa->inveclosures + dup_idx);
+
+ /* Store the index of the original node. */
+ dfa->org_indices[dup_idx] = org_idx;
+ *new_idx = dup_idx;
+ return REG_NOERROR;
+}
+
+static void
+calc_inveclosure (dfa)
+ re_dfa_t *dfa;
+{
+ int src, idx, dest;
+ for (src = 0; src < dfa->nodes_len; ++src)
+ {
+ for (idx = 0; idx < dfa->eclosures[src].nelem; ++idx)
+ {
+ dest = dfa->eclosures[src].elems[idx];
+ re_node_set_insert (dfa->inveclosures + dest, src);
+ }
+ }
+}
+
+/* Calculate "eclosure" for all the node in DFA. */
+
+static reg_errcode_t
+calc_eclosure (dfa)
+ re_dfa_t *dfa;
+{
+ int node_idx, incomplete;
+#ifdef DEBUG
+ assert (dfa->nodes_len > 0);
+#endif
+ incomplete = 0;
+ /* For each nodes, calculate epsilon closure. */
+ for (node_idx = 0; ; ++node_idx)
+ {
+ reg_errcode_t err;
+ re_node_set eclosure_elem;
+ if (node_idx == dfa->nodes_len)
+ {
+ if (!incomplete)
+ break;
+ incomplete = 0;
+ node_idx = 0;
+ }
+
+#ifdef DEBUG
+ assert (dfa->eclosures[node_idx].nelem != -1);
+#endif
+ /* If we have already calculated, skip it. */
+ if (dfa->eclosures[node_idx].nelem != 0)
+ continue;
+ /* Calculate epsilon closure of `node_idx'. */
+ err = calc_eclosure_iter (&eclosure_elem, dfa, node_idx, 1);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+
+ if (dfa->eclosures[node_idx].nelem == 0)
+ {
+ incomplete = 1;
+ re_node_set_free (&eclosure_elem);
+ }
+ }
+ return REG_NOERROR;
+}
+
+/* Calculate epsilon closure of NODE. */
+
+static reg_errcode_t
+calc_eclosure_iter (new_set, dfa, node, root)
+ re_node_set *new_set;
+ re_dfa_t *dfa;
+ int node, root;
+{
+ reg_errcode_t err;
+ unsigned int constraint;
+ int i, incomplete;
+ re_node_set eclosure;
+ incomplete = 0;
+ err = re_node_set_alloc (&eclosure, dfa->edests[node].nelem + 1);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+
+ /* This indicates that we are calculating this node now.
+ We reference this value to avoid infinite loop. */
+ dfa->eclosures[node].nelem = -1;
+
+ constraint = ((dfa->nodes[node].type == ANCHOR)
+ ? dfa->nodes[node].opr.ctx_type : 0);
+ /* If the current node has constraints, duplicate all nodes.
+ Since they must inherit the constraints. */
+ if (constraint && !dfa->nodes[dfa->edests[node].elems[0]].duplicated)
+ {
+ int org_node, cur_node;
+ org_node = cur_node = node;
+ err = duplicate_node_closure (dfa, node, node, node, constraint);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ }
+
+ /* Expand each epsilon destination nodes. */
+ if (IS_EPSILON_NODE(dfa->nodes[node].type))
+ for (i = 0; i < dfa->edests[node].nelem; ++i)
+ {
+ re_node_set eclosure_elem;
+ int edest = dfa->edests[node].elems[i];
+ /* If calculating the epsilon closure of `edest' is in progress,
+ return intermediate result. */
+ if (dfa->eclosures[edest].nelem == -1)
+ {
+ incomplete = 1;
+ continue;
+ }
+ /* If we haven't calculated the epsilon closure of `edest' yet,
+ calculate now. Otherwise use calculated epsilon closure. */
+ if (dfa->eclosures[edest].nelem == 0)
+ {
+ err = calc_eclosure_iter (&eclosure_elem, dfa, edest, 0);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ }
+ else
+ eclosure_elem = dfa->eclosures[edest];
+ /* Merge the epsilon closure of `edest'. */
+ re_node_set_merge (&eclosure, &eclosure_elem);
+ /* If the epsilon closure of `edest' is incomplete,
+ the epsilon closure of this node is also incomplete. */
+ if (dfa->eclosures[edest].nelem == 0)
+ {
+ incomplete = 1;
+ re_node_set_free (&eclosure_elem);
+ }
+ }
+
+ /* Epsilon closures include itself. */
+ re_node_set_insert (&eclosure, node);
+ if (incomplete && !root)
+ dfa->eclosures[node].nelem = 0;
+ else
+ dfa->eclosures[node] = eclosure;
+ *new_set = eclosure;
+ return REG_NOERROR;
+}
+\f
+/* Functions for token which are used in the parser. */
+
+/* Fetch a token from INPUT.
+ We must not use this function inside bracket expressions. */
+
+static re_token_t
+fetch_token (input, syntax)
+ re_string_t *input;
+ reg_syntax_t syntax;
+{
+ re_token_t token;
+ int consumed_byte;
+ consumed_byte = peek_token (&token, input, syntax);
+ re_string_skip_bytes (input, consumed_byte);
+ return token;
+}
+
+/* Peek a token from INPUT, and return the length of the token.
+ We must not use this function inside bracket expressions. */
+
+static int
+peek_token (token, input, syntax)
+ re_token_t *token;
+ re_string_t *input;
+ reg_syntax_t syntax;
+{
+ unsigned char c;
+
+ if (re_string_eoi (input))
+ {
+ token->type = END_OF_RE;
+ return 0;
+ }
+
+ c = re_string_peek_byte (input, 0);
+ token->opr.c = c;
+
+#ifdef RE_ENABLE_I18N
+ token->mb_partial = 0;
+ if (MB_CUR_MAX > 1 &&
+ !re_string_first_byte (input, re_string_cur_idx (input)))
+ {
+ token->type = CHARACTER;
+ token->mb_partial = 1;
+ return 1;
+ }
+#endif
+ if (c == '\\')
+ {
+ unsigned char c2;
+ if (re_string_cur_idx (input) + 1 >= re_string_length (input))
+ {
+ token->type = BACK_SLASH;
+ return 1;
+ }
+
+ c2 = re_string_peek_byte_case (input, 1);
+ token->opr.c = c2;
+ token->type = CHARACTER;
+ switch (c2)
+ {
+ case '|':
+ if (!(syntax & RE_LIMITED_OPS) && !(syntax & RE_NO_BK_VBAR))
+ token->type = OP_ALT;
+ break;
+ case '1': case '2': case '3': case '4': case '5':
+ case '6': case '7': case '8': case '9':
+ if (!(syntax & RE_NO_BK_REFS))
+ {
+ token->type = OP_BACK_REF;
+ token->opr.idx = c2 - '0';
+ }
+ break;
+ case '<':
+ if (!(syntax & RE_NO_GNU_OPS))
+ {
+ token->type = ANCHOR;
+ token->opr.idx = WORD_FIRST;
+ }
+ break;
+ case '>':
+ if (!(syntax & RE_NO_GNU_OPS))
+ {
+ token->type = ANCHOR;
+ token->opr.idx = WORD_LAST;
+ }
+ break;
+ case 'b':
+ if (!(syntax & RE_NO_GNU_OPS))
+ {
+ token->type = ANCHOR;
+ token->opr.idx = WORD_DELIM;
+ }
+ break;
+ case 'B':
+ if (!(syntax & RE_NO_GNU_OPS))
+ {
+ token->type = ANCHOR;
+ token->opr.idx = INSIDE_WORD;
+ }
+ break;
+ case 'w':
+ if (!(syntax & RE_NO_GNU_OPS))
+ token->type = OP_WORD;
+ break;
+ case 'W':
+ if (!(syntax & RE_NO_GNU_OPS))
+ token->type = OP_NOTWORD;
+ break;
+ case '`':
+ if (!(syntax & RE_NO_GNU_OPS))
+ {
+ token->type = ANCHOR;
+ token->opr.idx = BUF_FIRST;
+ }
+ break;
+ case '\'':
+ if (!(syntax & RE_NO_GNU_OPS))
+ {
+ token->type = ANCHOR;
+ token->opr.idx = BUF_LAST;
+ }
+ break;
+ case '(':
+ if (!(syntax & RE_NO_BK_PARENS))
+ token->type = OP_OPEN_SUBEXP;
+ break;
+ case ')':
+ if (!(syntax & RE_NO_BK_PARENS))
+ token->type = OP_CLOSE_SUBEXP;
+ break;
+ case '+':
+ if (!(syntax & RE_LIMITED_OPS) && (syntax & RE_BK_PLUS_QM))
+ token->type = OP_DUP_PLUS;
+ break;
+ case '?':
+ if (!(syntax & RE_LIMITED_OPS) && (syntax & RE_BK_PLUS_QM))
+ token->type = OP_DUP_QUESTION;
+ break;
+ case '{':
+ if ((syntax & RE_INTERVALS) && (!(syntax & RE_NO_BK_BRACES)))
+ token->type = OP_OPEN_DUP_NUM;
+ break;
+ case '}':
+ if ((syntax & RE_INTERVALS) && (!(syntax & RE_NO_BK_BRACES)))
+ token->type = OP_CLOSE_DUP_NUM;
+ break;
+ default:
+ break;
+ }
+ return 2;
+ }
+
+ token->type = CHARACTER;
+ switch (c)
+ {
+ case '\n':
+ if (syntax & RE_NEWLINE_ALT)
+ token->type = OP_ALT;
+ break;
+ case '|':
+ if (!(syntax & RE_LIMITED_OPS) && (syntax & RE_NO_BK_VBAR))
+ token->type = OP_ALT;
+ break;
+ case '*':
+ token->type = OP_DUP_ASTERISK;
+ break;
+ case '+':
+ if (!(syntax & RE_LIMITED_OPS) && !(syntax & RE_BK_PLUS_QM))
+ token->type = OP_DUP_PLUS;
+ break;
+ case '?':
+ if (!(syntax & RE_LIMITED_OPS) && !(syntax & RE_BK_PLUS_QM))
+ token->type = OP_DUP_QUESTION;
+ break;
+ case '{':
+ if ((syntax & RE_INTERVALS) && (syntax & RE_NO_BK_BRACES))
+ token->type = OP_OPEN_DUP_NUM;
+ break;
+ case '}':
+ if ((syntax & RE_INTERVALS) && (syntax & RE_NO_BK_BRACES))
+ token->type = OP_CLOSE_DUP_NUM;
+ break;
+ case '(':
+ if (syntax & RE_NO_BK_PARENS)
+ token->type = OP_OPEN_SUBEXP;
+ break;
+ case ')':
+ if (syntax & RE_NO_BK_PARENS)
+ token->type = OP_CLOSE_SUBEXP;
+ break;
+ case '[':
+ token->type = OP_OPEN_BRACKET;
+ break;
+ case '.':
+ token->type = OP_PERIOD;
+ break;
+ case '^':
+ if (!(syntax & RE_CONTEXT_INDEP_ANCHORS) &&
+ re_string_cur_idx (input) != 0)
+ {
+ char prev = re_string_peek_byte (input, -1);
+ if (prev != '|' && prev != '(' &&
+ (!(syntax & RE_NEWLINE_ALT) || prev != '\n'))
+ break;
+ }
+ token->type = ANCHOR;
+ token->opr.idx = LINE_FIRST;
+ break;
+ case '$':
+ if (!(syntax & RE_CONTEXT_INDEP_ANCHORS) &&
+ re_string_cur_idx (input) + 1 != re_string_length (input))
+ {
+ re_token_t next;
+ re_string_skip_bytes (input, 1);
+ peek_token (&next, input, syntax);
+ re_string_skip_bytes (input, -1);
+ if (next.type != OP_ALT && next.type != OP_CLOSE_SUBEXP)
+ break;
+ }
+ token->type = ANCHOR;
+ token->opr.idx = LINE_LAST;
+ break;
+ default:
+ break;
+ }
+ return 1;
+}
+
+/* Peek a token from INPUT, and return the length of the token.
+ We must not use this function out of bracket expressions. */
+
+static int
+peek_token_bracket (token, input, syntax)
+ re_token_t *token;
+ re_string_t *input;
+ reg_syntax_t syntax;
+{
+ unsigned char c;
+ if (re_string_eoi (input))
+ {
+ token->type = END_OF_RE;
+ return 0;
+ }
+ c = re_string_peek_byte (input, 0);
+ token->opr.c = c;
+
+#ifdef RE_ENABLE_I18N
+ if (MB_CUR_MAX > 1 &&
+ !re_string_first_byte (input, re_string_cur_idx (input)))
+ {
+ token->type = CHARACTER;
+ return 1;
+ }
+#endif /* RE_ENABLE_I18N */
+
+ if (c == '\\' && (syntax & RE_BACKSLASH_ESCAPE_IN_LISTS))
+ {
+ /* In this case, '\' escape a character. */
+ unsigned char c2;
+ re_string_skip_bytes (input, 1);
+ c2 = re_string_peek_byte (input, 0);
+ token->opr.c = c2;
+ token->type = CHARACTER;
+ return 1;
+ }
+ if (c == '[') /* '[' is a special char in a bracket exps. */
+ {
+ unsigned char c2;
+ int token_len;
+ c2 = re_string_peek_byte (input, 1);
+ token->opr.c = c2;
+ token_len = 2;
+ switch (c2)
+ {
+ case '.':
+ token->type = OP_OPEN_COLL_ELEM;
+ break;
+ case '=':
+ token->type = OP_OPEN_EQUIV_CLASS;
+ break;
+ case ':':
+ if (syntax & RE_CHAR_CLASSES)
+ {
+ token->type = OP_OPEN_CHAR_CLASS;
+ break;
+ }
+ /* else fall through. */
+ default:
+ token->type = CHARACTER;
+ token->opr.c = c;
+ token_len = 1;
+ break;
+ }
+ return token_len;
+ }
+ switch (c)
+ {
+ case '-':
+ token->type = OP_CHARSET_RANGE;
+ break;
+ case ']':
+ token->type = OP_CLOSE_BRACKET;
+ break;
+ case '^':
+ token->type = OP_NON_MATCH_LIST;
+ break;
+ default:
+ token->type = CHARACTER;
+ }
+ return 1;
+}
+\f
+/* Functions for parser. */
+
+/* Entry point of the parser.
+ Parse the regular expression REGEXP and return the structure tree.
+ If an error is occured, ERR is set by error code, and return NULL.
+ This function build the following tree, from regular expression <reg_exp>:
+ CAT
+ / \
+ / \
+ <reg_exp> EOR
+
+ CAT means concatenation.
+ EOR means end of regular expression. */
+
+static bin_tree_t *
+parse (regexp, preg, syntax, err)
+ re_string_t *regexp;
+ regex_t *preg;
+ reg_syntax_t syntax;
+ reg_errcode_t *err;
+{
+ re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ bin_tree_t *tree, *eor, *root;
+ re_token_t current_token;
+ int new_idx;
+ current_token = fetch_token (regexp, syntax);
+ tree = parse_reg_exp (regexp, preg, ¤t_token, syntax, 0, err);
+ if (BE (*err != REG_NOERROR && tree == NULL, 0))
+ return NULL;
+ new_idx = re_dfa_add_node (dfa, current_token, 0);
+ eor = create_tree (NULL, NULL, 0, new_idx);
+ if (tree != NULL)
+ root = create_tree (tree, eor, CONCAT, 0);
+ else
+ root = eor;
+ if (BE (new_idx == -1 || eor == NULL || root == NULL, 0))
+ {
+ *err = REG_ESPACE;
+ return NULL;
+ }
+ return root;
+}
+
+/* This function build the following tree, from regular expression
+ <branch1>|<branch2>:
+ ALT
+ / \
+ / \
+ <branch1> <branch2>
+
+ ALT means alternative, which represents the operator `|'. */
+
+static bin_tree_t *
+parse_reg_exp (regexp, preg, token, syntax, nest, err)
+ re_string_t *regexp;
+ regex_t *preg;
+ re_token_t *token;
+ reg_syntax_t syntax;
+ int nest;
+ reg_errcode_t *err;
+{
+ re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ bin_tree_t *tree, *branch = NULL;
+ int new_idx;
+ tree = parse_branch (regexp, preg, token, syntax, nest, err);
+ if (BE (*err != REG_NOERROR && tree == NULL, 0))
+ return NULL;
+
+ while (token->type == OP_ALT)
+ {
+ re_token_t alt_token = *token;
+ new_idx = re_dfa_add_node (dfa, alt_token, 0);
+ *token = fetch_token (regexp, syntax);
+ if (token->type != OP_ALT && token->type != END_OF_RE
+ && (nest == 0 || token->type != OP_CLOSE_SUBEXP))
+ {
+ branch = parse_branch (regexp, preg, token, syntax, nest, err);
+ if (BE (*err != REG_NOERROR && branch == NULL, 0))
+ {
+ free_bin_tree (tree);
+ return NULL;
+ }
+ }
+ else
+ branch = NULL;
+ tree = create_tree (tree, branch, 0, new_idx);
+ if (BE (new_idx == -1 || tree == NULL, 0))
+ {
+ *err = REG_ESPACE;
+ return NULL;
+ }
+ dfa->has_plural_match = 1;
+ }
+ return tree;
+}
+
+/* This function build the following tree, from regular expression
+ <exp1><exp2>:
+ CAT
+ / \
+ / \
+ <exp1> <exp2>
+
+ CAT means concatenation. */
+
+static bin_tree_t *
+parse_branch (regexp, preg, token, syntax, nest, err)
+ re_string_t *regexp;
+ regex_t *preg;
+ re_token_t *token;
+ reg_syntax_t syntax;
+ int nest;
+ reg_errcode_t *err;
+{
+ bin_tree_t *tree, *exp;
+ tree = parse_expression (regexp, preg, token, syntax, nest, err);
+ if (BE (*err != REG_NOERROR && tree == NULL, 0))
+ return NULL;
+
+ while (token->type != OP_ALT && token->type != END_OF_RE
+ && (nest == 0 || token->type != OP_CLOSE_SUBEXP))
+ {
+ exp = parse_expression (regexp, preg, token, syntax, nest, err);
+ if (BE (*err != REG_NOERROR && exp == NULL, 0))
+ {
+ free_bin_tree (tree);
+ return NULL;
+ }
+ if (tree != NULL && exp != NULL)
+ {
+ tree = create_tree (tree, exp, CONCAT, 0);
+ if (tree == NULL)
+ {
+ *err = REG_ESPACE;
+ return NULL;
+ }
+ }
+ else if (tree == NULL)
+ tree = exp;
+ /* Otherwise exp == NULL, we don't need to create new tree. */
+ }
+ return tree;
+}
+
+/* This function build the following tree, from regular expression a*:
+ *
+ |
+ a
+*/
+
+static bin_tree_t *
+parse_expression (regexp, preg, token, syntax, nest, err)
+ re_string_t *regexp;
+ regex_t *preg;
+ re_token_t *token;
+ reg_syntax_t syntax;
+ int nest;
+ reg_errcode_t *err;
+{
+ re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ bin_tree_t *tree;
+ int new_idx;
+ switch (token->type)
+ {
+ case CHARACTER:
+ new_idx = re_dfa_add_node (dfa, *token, 0);
+ tree = create_tree (NULL, NULL, 0, new_idx);
+ if (BE (new_idx == -1 || tree == NULL, 0))
+ {
+ *err = REG_ESPACE;
+ return NULL;
+ }
+#ifdef RE_ENABLE_I18N
+ if (MB_CUR_MAX > 1)
+ {
+ while (!re_string_eoi (regexp)
+ && !re_string_first_byte (regexp, re_string_cur_idx (regexp)))
+ {
+ bin_tree_t *mbc_remain;
+ *token = fetch_token (regexp, syntax);
+ new_idx = re_dfa_add_node (dfa, *token, 0);
+ mbc_remain = create_tree (NULL, NULL, 0, new_idx);
+ tree = create_tree (tree, mbc_remain, CONCAT, 0);
+ if (BE (new_idx == -1 || mbc_remain == NULL || tree == NULL, 0))
+ {
+ *err = REG_ESPACE;
+ return NULL;
+ }
+ }
+ }
+#endif
+ break;
+ case OP_OPEN_SUBEXP:
+ tree = parse_sub_exp (regexp, preg, token, syntax, nest + 1, err);
+ if (BE (*err != REG_NOERROR && tree == NULL, 0))
+ return NULL;
+ break;
+ case OP_OPEN_BRACKET:
+ tree = parse_bracket_exp (regexp, dfa, token, syntax, err);
+ if (BE (*err != REG_NOERROR && tree == NULL, 0))
+ return NULL;
+ break;
+ case OP_BACK_REF:
+ if (BE (preg->re_nsub < token->opr.idx
+ || dfa->subexps[token->opr.idx - 1].end == -1, 0))
+ {
+ *err = REG_ESUBREG;
+ return NULL;
+ }
+ dfa->used_bkref_map |= 1 << (token->opr.idx - 1);
+ new_idx = re_dfa_add_node (dfa, *token, 0);
+ tree = create_tree (NULL, NULL, 0, new_idx);
+ if (BE (new_idx == -1 || tree == NULL, 0))
+ {
+ *err = REG_ESPACE;
+ return NULL;
+ }
+ ++dfa->nbackref;
+ dfa->has_mb_node = 1;
+ break;
+ case OP_DUP_ASTERISK:
+ case OP_DUP_PLUS:
+ case OP_DUP_QUESTION:
+ case OP_OPEN_DUP_NUM:
+ if (syntax & RE_CONTEXT_INVALID_OPS)
+ {
+ *err = REG_BADRPT;
+ return NULL;
+ }
+ else if (syntax & RE_CONTEXT_INDEP_OPS)
+ {
+ *token = fetch_token (regexp, syntax);
+ return parse_expression (regexp, preg, token, syntax, nest, err);
+ }
+ /* else fall through */
+ case OP_CLOSE_SUBEXP:
+ if ((token->type == OP_CLOSE_SUBEXP) &&
+ !(syntax & RE_UNMATCHED_RIGHT_PAREN_ORD))
+ {
+ *err = REG_ERPAREN;
+ return NULL;
+ }
+ /* else fall through */
+ case OP_CLOSE_DUP_NUM:
+ /* We treat it as a normal character. */
+
+ /* Then we can these characters as normal characters. */
+ token->type = CHARACTER;
+ new_idx = re_dfa_add_node (dfa, *token, 0);
+ tree = create_tree (NULL, NULL, 0, new_idx);
+ if (BE (new_idx == -1 || tree == NULL, 0))
+ {
+ *err = REG_ESPACE;
+ return NULL;
+ }
+ break;
+ case ANCHOR:
+ if (dfa->word_char == NULL)
+ {
+ *err = init_word_char (dfa);
+ if (BE (*err != REG_NOERROR, 0))
+ return NULL;
+ }
+ if (token->opr.ctx_type == WORD_DELIM)
+ {
+ bin_tree_t *tree_first, *tree_last;
+ int idx_first, idx_last;
+ token->opr.ctx_type = WORD_FIRST;
+ idx_first = re_dfa_add_node (dfa, *token, 0);
+ tree_first = create_tree (NULL, NULL, 0, idx_first);
+ token->opr.ctx_type = WORD_LAST;
+ idx_last = re_dfa_add_node (dfa, *token, 0);
+ tree_last = create_tree (NULL, NULL, 0, idx_last);
+ token->type = OP_ALT;
+ new_idx = re_dfa_add_node (dfa, *token, 0);
+ tree = create_tree (tree_first, tree_last, 0, new_idx);
+ if (BE (idx_first == -1 || idx_last == -1 || new_idx == -1
+ || tree_first == NULL || tree_last == NULL
+ || tree == NULL, 0))
+ {
+ *err = REG_ESPACE;
+ return NULL;
+ }
+ }
+ else
+ {
+ new_idx = re_dfa_add_node (dfa, *token, 0);
+ tree = create_tree (NULL, NULL, 0, new_idx);
+ if (BE (new_idx == -1 || tree == NULL, 0))
+ {
+ *err = REG_ESPACE;
+ return NULL;
+ }
+ }
+ /* We must return here, since ANCHORs can't be followed
+ by repetition operators.
+ eg. RE"^*" is invalid or "<ANCHOR(^)><CHAR(*)>",
+ it must not be "<ANCHOR(^)><REPEAT(*)>". */
+ *token = fetch_token (regexp, syntax);
+ return tree;
+ case OP_PERIOD:
+ new_idx = re_dfa_add_node (dfa, *token, 0);
+ tree = create_tree (NULL, NULL, 0, new_idx);
+ if (BE (new_idx == -1 || tree == NULL, 0))
+ {
+ *err = REG_ESPACE;
+ return NULL;
+ }
+ if (MB_CUR_MAX > 1)
+ dfa->has_mb_node = 1;
+ break;
+ case OP_WORD:
+ tree = build_word_op (dfa, 0, err);
+ if (BE (*err != REG_NOERROR && tree == NULL, 0))
+ return NULL;
+ break;
+ case OP_NOTWORD:
+ tree = build_word_op (dfa, 1, err);
+ if (BE (*err != REG_NOERROR && tree == NULL, 0))
+ return NULL;
+ break;
+ case OP_ALT:
+ case END_OF_RE:
+ return NULL;
+ case BACK_SLASH:
+ *err = REG_EESCAPE;
+ return NULL;
+ default:
+ /* Must not happen? */
+#ifdef DEBUG
+ assert (0);
+#endif
+ return NULL;
+ }
+ *token = fetch_token (regexp, syntax);
+
+ while (token->type == OP_DUP_ASTERISK || token->type == OP_DUP_PLUS
+ || token->type == OP_DUP_QUESTION || token->type == OP_OPEN_DUP_NUM)
+ {
+ tree = parse_dup_op (tree, regexp, dfa, token, syntax, err);
+ if (BE (*err != REG_NOERROR && tree == NULL, 0))
+ return NULL;
+ dfa->has_plural_match = 1;
+ }
+
+ return tree;
+}
+
+/* This function build the following tree, from regular expression
+ (<reg_exp>):
+ SUBEXP
+ |
+ <reg_exp>
+*/
+
+static bin_tree_t *
+parse_sub_exp (regexp, preg, token, syntax, nest, err)
+ re_string_t *regexp;
+ regex_t *preg;
+ re_token_t *token;
+ reg_syntax_t syntax;
+ int nest;
+ reg_errcode_t *err;
+{
+ re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ bin_tree_t *tree, *left_par, *right_par;
+ size_t cur_nsub;
+ int new_idx;
+ cur_nsub = preg->re_nsub++;
+ if (dfa->subexps_alloc < preg->re_nsub)
+ {
+ re_subexp_t *new_array;
+ dfa->subexps_alloc *= 2;
+ new_array = re_realloc (dfa->subexps, re_subexp_t, dfa->subexps_alloc);
+ if (BE (new_array == NULL, 0))
+ {
+ dfa->subexps_alloc /= 2;
+ *err = REG_ESPACE;
+ return NULL;
+ }
+ dfa->subexps = new_array;
+ }
+ dfa->subexps[cur_nsub].start = dfa->nodes_len;
+ dfa->subexps[cur_nsub].end = -1;
+
+ new_idx = re_dfa_add_node (dfa, *token, 0);
+ left_par = create_tree (NULL, NULL, 0, new_idx);
+ if (BE (new_idx == -1 || left_par == NULL, 0))
+ {
+ *err = REG_ESPACE;
+ return NULL;
+ }
+ dfa->nodes[new_idx].opr.idx = cur_nsub;
+ *token = fetch_token (regexp, syntax);
+
+ /* The subexpression may be a null string. */
+ if (token->type == OP_CLOSE_SUBEXP)
+ tree = NULL;
+ else
+ {
+ tree = parse_reg_exp (regexp, preg, token, syntax, nest, err);
+ if (BE (*err != REG_NOERROR && tree == NULL, 0))
+ return NULL;
+ }
+ if (BE (token->type != OP_CLOSE_SUBEXP, 0))
+ {
+ free_bin_tree (tree);
+ *err = REG_BADPAT;
+ return NULL;
+ }
+ new_idx = re_dfa_add_node (dfa, *token, 0);
+ dfa->subexps[cur_nsub].end = dfa->nodes_len;
+ right_par = create_tree (NULL, NULL, 0, new_idx);
+ tree = ((tree == NULL) ? right_par
+ : create_tree (tree, right_par, CONCAT, 0));
+ tree = create_tree (left_par, tree, CONCAT, 0);
+ if (BE (new_idx == -1 || right_par == NULL || tree == NULL, 0))
+ {
+ *err = REG_ESPACE;
+ return NULL;
+ }
+ dfa->nodes[new_idx].opr.idx = cur_nsub;
+
+ return tree;
+}
+
+/* This function parse repetition operators like "*", "+", "{1,3}" etc. */
+
+static bin_tree_t *
+parse_dup_op (dup_elem, regexp, dfa, token, syntax, err)
+ bin_tree_t *dup_elem;
+ re_string_t *regexp;
+ re_dfa_t *dfa;
+ re_token_t *token;
+ reg_syntax_t syntax;
+ reg_errcode_t *err;
+{
+ re_token_t dup_token;
+ bin_tree_t *tree = dup_elem, *work_tree;
+ int new_idx, start_idx = re_string_cur_idx (regexp);
+ re_token_t start_token = *token;
+ if (token->type == OP_OPEN_DUP_NUM)
+ {
+ int i;
+ int end = 0;
+ int start = fetch_number (regexp, token, syntax);
+ bin_tree_t *elem;
+ if (start == -1)
+ {
+ if (token->type == CHARACTER && token->opr.c == ',')
+ start = 0; /* We treat "{,m}" as "{0,m}". */
+ else
+ {
+ *err = REG_BADBR; /* <re>{} is invalid. */
+ return NULL;
+ }
+ }
+ if (BE (start != -2, 1))
+ {
+ /* We treat "{n}" as "{n,n}". */
+ end = ((token->type == OP_CLOSE_DUP_NUM) ? start
+ : ((token->type == CHARACTER && token->opr.c == ',')
+ ? fetch_number (regexp, token, syntax) : -2));
+ }
+ if (BE (start == -2 || end == -2, 0))
+ {
+ /* Invalid sequence. */
+ if (token->type == OP_CLOSE_DUP_NUM)
+ goto parse_dup_op_invalid_interval;
+ else
+ goto parse_dup_op_ebrace;
+ }
+ if (BE (start == 0 && end == 0, 0))
+ {
+ /* We treat "<re>{0}" and "<re>{0,0}" as null string. */
+ *token = fetch_token (regexp, syntax);
+ free_bin_tree (dup_elem);
+ return NULL;
+ }
+
+ /* Extract "<re>{n,m}" to "<re><re>...<re><re>{0,<m-n>}". */
+ elem = tree;
+ for (i = 0; i < start; ++i)
+ if (i != 0)
+ {
+ work_tree = duplicate_tree (elem, dfa);
+ tree = create_tree (tree, work_tree, CONCAT, 0);
+ if (BE (work_tree == NULL || tree == NULL, 0))
+ goto parse_dup_op_espace;
+ }
+
+ if (end == -1)
+ {
+ /* We treat "<re>{0,}" as "<re>*". */
+ dup_token.type = OP_DUP_ASTERISK;
+ if (start > 0)
+ {
+ elem = duplicate_tree (elem, dfa);
+ new_idx = re_dfa_add_node (dfa, dup_token, 0);
+ work_tree = create_tree (elem, NULL, 0, new_idx);
+ tree = create_tree (tree, work_tree, CONCAT, 0);
+ if (BE (elem == NULL || new_idx == -1 || work_tree == NULL
+ || tree == NULL, 0))
+ goto parse_dup_op_espace;
+ }
+ else
+ {
+ new_idx = re_dfa_add_node (dfa, dup_token, 0);
+ tree = create_tree (elem, NULL, 0, new_idx);
+ if (BE (new_idx == -1 || tree == NULL, 0))
+ goto parse_dup_op_espace;
+ }
+ }
+ else if (end - start > 0)
+ {
+ /* Then extract "<re>{0,m}" to "<re>?<re>?...<re>?". */
+ dup_token.type = OP_DUP_QUESTION;
+ if (start > 0)
+ {
+ elem = duplicate_tree (elem, dfa);
+ new_idx = re_dfa_add_node (dfa, dup_token, 0);
+ elem = create_tree (elem, NULL, 0, new_idx);
+ tree = create_tree (tree, elem, CONCAT, 0);
+ if (BE (elem == NULL || new_idx == -1 || tree == NULL, 0))
+ goto parse_dup_op_espace;
+ }
+ else
+ {
+ new_idx = re_dfa_add_node (dfa, dup_token, 0);
+ tree = elem = create_tree (elem, NULL, 0, new_idx);
+ if (BE (new_idx == -1 || tree == NULL, 0))
+ goto parse_dup_op_espace;
+ }
+ for (i = 1; i < end - start; ++i)
+ {
+ work_tree = duplicate_tree (elem, dfa);
+ tree = create_tree (tree, work_tree, CONCAT, 0);
+ if (BE (work_tree == NULL || tree == NULL, 0))
+ {
+ *err = REG_ESPACE;
+ return NULL;
+ }
+ }
+ }
+ }
+ else
+ {
+ new_idx = re_dfa_add_node (dfa, *token, 0);
+ tree = create_tree (tree, NULL, 0, new_idx);
+ if (BE (new_idx == -1 || tree == NULL, 0))
+ {
+ *err = REG_ESPACE;
+ return NULL;
+ }
+ }
+ *token = fetch_token (regexp, syntax);
+ return tree;
+
+ parse_dup_op_espace:
+ free_bin_tree (tree);
+ *err = REG_ESPACE;
+ return NULL;
+
+ parse_dup_op_ebrace:
+ if (BE (!(syntax & RE_INVALID_INTERVAL_ORD), 0))
+ {
+ *err = REG_EBRACE;
+ return NULL;
+ }
+ goto parse_dup_op_rollback;
+ parse_dup_op_invalid_interval:
+ if (BE (!(syntax & RE_INVALID_INTERVAL_ORD), 0))
+ {
+ *err = REG_BADBR;
+ return NULL;
+ }
+ parse_dup_op_rollback:
+ re_string_set_index (regexp, start_idx);
+ *token = start_token;
+ token->type = CHARACTER;
+ return dup_elem;
+}
+
+/* Size of the names for collating symbol/equivalence_class/character_class.
+ I'm not sure, but maybe enough. */
+#define BRACKET_NAME_BUF_SIZE 32
+
+#ifndef _LIBC
+ /* Local function for parse_bracket_exp only used in case of NOT _LIBC.
+ Build the range expression which starts from START_ELEM, and ends
+ at END_ELEM. The result are written to MBCSET and SBCSET.
+ RANGE_ALLOC is the allocated size of mbcset->range_starts, and
+ mbcset->range_ends, is a pointer argument sinse we may
+ update it. */
+
+static reg_errcode_t
+# ifdef RE_ENABLE_I18N
+build_range_exp (sbcset, mbcset, range_alloc, start_elem, end_elem)
+ re_charset_t *mbcset;
+ int *range_alloc;
+# else /* not RE_ENABLE_I18N */
+build_range_exp (sbcset, start_elem, end_elem)
+# endif /* not RE_ENABLE_I18N */
+ re_bitset_ptr_t sbcset;
+ bracket_elem_t *start_elem, *end_elem;
+{
+ unsigned int start_ch, end_ch;
+ /* Equivalence Classes and Character Classes can't be a range start/end. */
+ if (BE (start_elem->type == EQUIV_CLASS || start_elem->type == CHAR_CLASS
+ || end_elem->type == EQUIV_CLASS || end_elem->type == CHAR_CLASS,
+ 0))
+ return REG_ERANGE;
+
+ /* We can handle no multi character collating elements without libc
+ support. */
+ if (BE ((start_elem->type == COLL_SYM
+ && strlen ((char *) start_elem->opr.name) > 1)
+ || (end_elem->type == COLL_SYM
+ && strlen ((char *) end_elem->opr.name) > 1), 0))
+ return REG_ECOLLATE;
+
+# ifdef RE_ENABLE_I18N
+ {
+ wchar_t wc, start_wc, end_wc;
+ wchar_t cmp_buf[6] = {L'\0', L'\0', L'\0', L'\0', L'\0', L'\0'};
+
+ start_ch = ((start_elem->type == SB_CHAR) ? start_elem->opr.ch
+ : ((start_elem->type == COLL_SYM) ? start_elem->opr.name[0]
+ : 0));
+ end_ch = ((end_elem->type == SB_CHAR) ? end_elem->opr.ch
+ : ((end_elem->type == COLL_SYM) ? end_elem->opr.name[0]
+ : 0));
+ start_wc = ((start_elem->type == SB_CHAR || start_elem->type == COLL_SYM)
+ ? __btowc (start_ch) : start_elem->opr.wch);
+ end_wc = ((end_elem->type == SB_CHAR || end_elem->type == COLL_SYM)
+ ? __btowc (end_ch) : end_elem->opr.wch);
+ cmp_buf[0] = start_wc;
+ cmp_buf[4] = end_wc;
+ if (wcscoll (cmp_buf, cmp_buf + 4) > 0)
+ return REG_ERANGE;
+
+ /* Check the space of the arrays. */
+ if (*range_alloc == mbcset->nranges)
+ {
+ /* There are not enough space, need realloc. */
+ wchar_t *new_array_start, *new_array_end;
+ int new_nranges;
+
+ /* +1 in case of mbcset->nranges is 0. */
+ new_nranges = 2 * mbcset->nranges + 1;
+ /* Use realloc since mbcset->range_starts and mbcset->range_ends
+ are NULL if *range_alloc == 0. */
+ new_array_start = re_realloc (mbcset->range_starts, wchar_t,
+ new_nranges);
+ new_array_end = re_realloc (mbcset->range_ends, wchar_t,
+ new_nranges);
+
+ if (BE (new_array_start == NULL || new_array_end == NULL, 0))
+ return REG_ESPACE;
+
+ mbcset->range_starts = new_array_start;
+ mbcset->range_ends = new_array_end;
+ *range_alloc = new_nranges;
+ }
+
+ mbcset->range_starts[mbcset->nranges] = start_wc;
+ mbcset->range_ends[mbcset->nranges++] = end_wc;
+
+ /* Build the table for single byte characters. */
+ for (wc = 0; wc <= SBC_MAX; ++wc)
+ {
+ cmp_buf[2] = wc;
+ if (wcscoll (cmp_buf, cmp_buf + 2) <= 0
+ && wcscoll (cmp_buf + 2, cmp_buf + 4) <= 0)
+ bitset_set (sbcset, wc);
+ }
+ }
+# else /* not RE_ENABLE_I18N */
+ {
+ unsigned int ch;
+ start_ch = ((start_elem->type == SB_CHAR ) ? start_elem->opr.ch
+ : ((start_elem->type == COLL_SYM) ? start_elem->opr.name[0]
+ : 0));
+ end_ch = ((end_elem->type == SB_CHAR ) ? end_elem->opr.ch
+ : ((end_elem->type == COLL_SYM) ? end_elem->opr.name[0]
+ : 0));
+ if (start_ch > end_ch)
+ return REG_ERANGE;
+ /* Build the table for single byte characters. */
+ for (ch = 0; ch <= SBC_MAX; ++ch)
+ if (start_ch <= ch && ch <= end_ch)
+ bitset_set (sbcset, ch);
+ }
+# endif /* not RE_ENABLE_I18N */
+ return REG_NOERROR;
+}
+#endif /* not _LIBC */
+
+#ifndef _LIBC
+/* Helper function for parse_bracket_exp only used in case of NOT _LIBC..
+ Build the collating element which is represented by NAME.
+ The result are written to MBCSET and SBCSET.
+ COLL_SYM_ALLOC is the allocated size of mbcset->coll_sym, is a
+ pointer argument since we may update it. */
+
+static reg_errcode_t
+# ifdef RE_ENABLE_I18N
+build_collating_symbol (sbcset, mbcset, coll_sym_alloc, name)
+ re_charset_t *mbcset;
+ int *coll_sym_alloc;
+# else /* not RE_ENABLE_I18N */
+build_collating_symbol (sbcset, name)
+# endif /* not RE_ENABLE_I18N */
+ re_bitset_ptr_t sbcset;
+ const unsigned char *name;
+{
+ size_t name_len = strlen ((const char *) name);
+ if (BE (name_len != 1, 0))
+ return REG_ECOLLATE;
+ else
+ {
+ bitset_set (sbcset, name[0]);
+ return REG_NOERROR;
+ }
+}
+#endif /* not _LIBC */
+
+/* This function parse bracket expression like "[abc]", "[a-c]",
+ "[[.a-a.]]" etc. */
+
+static bin_tree_t *
+parse_bracket_exp (regexp, dfa, token, syntax, err)
+ re_string_t *regexp;
+ re_dfa_t *dfa;
+ re_token_t *token;
+ reg_syntax_t syntax;
+ reg_errcode_t *err;
+{
+#ifdef _LIBC
+ const unsigned char *collseqmb;
+ const char *collseqwc;
+ uint32_t nrules;
+ int32_t table_size;
+ const int32_t *symb_table;
+ const unsigned char *extra;
+
+ /* Local function for parse_bracket_exp used in _LIBC environement.
+ Seek the collating symbol entry correspondings to NAME.
+ Return the index of the symbol in the SYMB_TABLE. */
+
+ static inline int32_t
+ seek_collating_symbol_entry (name, name_len)
+ const unsigned char *name;
+ size_t name_len;
+ {
+ int32_t hash = elem_hash ((const char *) name, name_len);
+ int32_t elem = hash % table_size;
+ int32_t second = hash % (table_size - 2);
+ while (symb_table[2 * elem] != 0)
+ {
+ /* First compare the hashing value. */
+ if (symb_table[2 * elem] == hash
+ /* Compare the length of the name. */
+ && name_len == extra[symb_table[2 * elem + 1]]
+ /* Compare the name. */
+ && memcmp (name, &extra[symb_table[2 * elem + 1] + 1],
+ name_len) == 0)
+ {
+ /* Yep, this is the entry. */
+ break;
+ }
+
+ /* Next entry. */
+ elem += second;
+ }
+ return elem;
+ }
+
+ /* Local function for parse_bracket_exp used in _LIBC environement.
+ Look up the collation sequence value of BR_ELEM.
+ Return the value if succeeded, UINT_MAX otherwise. */
+
+ static inline unsigned int
+ lookup_collation_sequence_value (br_elem)
+ bracket_elem_t *br_elem;
+ {
+ if (br_elem->type == SB_CHAR)
+ {
+ /*
+ if (MB_CUR_MAX == 1)
+ */
+ if (nrules == 0)
+ return collseqmb[br_elem->opr.ch];
+ else
+ {
+ wint_t wc = __btowc (br_elem->opr.ch);
+ return collseq_table_lookup (collseqwc, wc);
+ }
+ }
+ else if (br_elem->type == MB_CHAR)
+ {
+ return collseq_table_lookup (collseqwc, br_elem->opr.wch);
+ }
+ else if (br_elem->type == COLL_SYM)
+ {
+ size_t sym_name_len = strlen ((char *) br_elem->opr.name);
+ if (nrules != 0)
+ {
+ int32_t elem, idx;
+ elem = seek_collating_symbol_entry (br_elem->opr.name,
+ sym_name_len);
+ if (symb_table[2 * elem] != 0)
+ {
+ /* We found the entry. */
+ idx = symb_table[2 * elem + 1];
+ /* Skip the name of collating element name. */
+ idx += 1 + extra[idx];
+ /* Skip the byte sequence of the collating element. */
+ idx += 1 + extra[idx];
+ /* Adjust for the alignment. */
+ idx = (idx + 3) & ~3;
+ /* Skip the multibyte collation sequence value. */
+ idx += sizeof (unsigned int);
+ /* Skip the wide char sequence of the collating element. */
+ idx += sizeof (unsigned int) *
+ (1 + *(unsigned int *) (extra + idx));
+ /* Return the collation sequence value. */
+ return *(unsigned int *) (extra + idx);
+ }
+ else if (symb_table[2 * elem] == 0 && sym_name_len == 1)
+ {
+ /* No valid character. Match it as a single byte
+ character. */
+ return collseqmb[br_elem->opr.name[0]];
+ }
+ }
+ else if (sym_name_len == 1)
+ return collseqmb[br_elem->opr.name[0]];
+ }
+ return UINT_MAX;
+ }
+
+ /* Local function for parse_bracket_exp used in _LIBC environement.
+ Build the range expression which starts from START_ELEM, and ends
+ at END_ELEM. The result are written to MBCSET and SBCSET.
+ RANGE_ALLOC is the allocated size of mbcset->range_starts, and
+ mbcset->range_ends, is a pointer argument sinse we may
+ update it. */
+
+ static inline reg_errcode_t
+# ifdef RE_ENABLE_I18N
+ build_range_exp (sbcset, mbcset, range_alloc, start_elem, end_elem)
+ re_charset_t *mbcset;
+ int *range_alloc;
+# else /* not RE_ENABLE_I18N */
+ build_range_exp (sbcset, start_elem, end_elem)
+# endif /* not RE_ENABLE_I18N */
+ re_bitset_ptr_t sbcset;
+ bracket_elem_t *start_elem, *end_elem;
+ {
+ unsigned int ch;
+ uint32_t start_collseq;
+ uint32_t end_collseq;
+
+# ifdef RE_ENABLE_I18N
+ /* Check the space of the arrays. */
+ if (*range_alloc == mbcset->nranges)
+ {
+ /* There are not enough space, need realloc. */
+ uint32_t *new_array_start;
+ uint32_t *new_array_end;
+ int new_nranges;
+
+ /* +1 in case of mbcset->nranges is 0. */
+ new_nranges = 2 * mbcset->nranges + 1;
+ /* Use realloc since mbcset->range_starts and mbcset->range_ends
+ are NULL if *range_alloc == 0. */
+ new_array_start = re_realloc (mbcset->range_starts, uint32_t,
+ new_nranges);
+ new_array_end = re_realloc (mbcset->range_ends, uint32_t,
+ new_nranges);
+
+ if (BE (new_array_start == NULL || new_array_end == NULL, 0))
+ return REG_ESPACE;
+
+ mbcset->range_starts = new_array_start;
+ mbcset->range_ends = new_array_end;
+ *range_alloc = new_nranges;
+ }
+# endif /* RE_ENABLE_I18N */
+
+ /* Equivalence Classes and Character Classes can't be a range
+ start/end. */
+ if (BE (start_elem->type == EQUIV_CLASS || start_elem->type == CHAR_CLASS
+ || end_elem->type == EQUIV_CLASS || end_elem->type == CHAR_CLASS,
+ 0))
+ return REG_ERANGE;
+
+ start_collseq = lookup_collation_sequence_value (start_elem);
+ end_collseq = lookup_collation_sequence_value (end_elem);
+ /* Check start/end collation sequence values. */
+ if (BE (start_collseq == UINT_MAX || end_collseq == UINT_MAX, 0))
+ return REG_ECOLLATE;
+ if (BE ((syntax & RE_NO_EMPTY_RANGES) && start_collseq > end_collseq, 0))
+ return REG_ERANGE;
+
+# ifdef RE_ENABLE_I18N
+ /* Got valid collation sequence values, add them as a new entry. */
+ mbcset->range_starts[mbcset->nranges] = start_collseq;
+ mbcset->range_ends[mbcset->nranges++] = end_collseq;
+# endif /* RE_ENABLE_I18N */
+
+ /* Build the table for single byte characters. */
+ for (ch = 0; ch <= SBC_MAX; ch++)
+ {
+ uint32_t ch_collseq;
+ /*
+ if (MB_CUR_MAX == 1)
+ */
+ if (nrules == 0)
+ ch_collseq = collseqmb[ch];
+ else
+ ch_collseq = collseq_table_lookup (collseqwc, __btowc (ch));
+ if (start_collseq <= ch_collseq && ch_collseq <= end_collseq)
+ bitset_set (sbcset, ch);
+ }
+ return REG_NOERROR;
+ }
+
+ /* Local function for parse_bracket_exp used in _LIBC environement.
+ Build the collating element which is represented by NAME.
+ The result are written to MBCSET and SBCSET.
+ COLL_SYM_ALLOC is the allocated size of mbcset->coll_sym, is a
+ pointer argument sinse we may update it. */
+
+ static inline reg_errcode_t
+# ifdef RE_ENABLE_I18N
+ build_collating_symbol (sbcset, mbcset, coll_sym_alloc, name)
+ re_charset_t *mbcset;
+ int *coll_sym_alloc;
+# else /* not RE_ENABLE_I18N */
+ build_collating_symbol (sbcset, name)
+# endif /* not RE_ENABLE_I18N */
+ re_bitset_ptr_t sbcset;
+ const unsigned char *name;
+ {
+ int32_t elem, idx;
+ size_t name_len = strlen ((const char *) name);
+ if (nrules != 0)
+ {
+ elem = seek_collating_symbol_entry (name, name_len);
+ if (symb_table[2 * elem] != 0)
+ {
+ /* We found the entry. */
+ idx = symb_table[2 * elem + 1];
+ /* Skip the name of collating element name. */
+ idx += 1 + extra[idx];
+ }
+ else if (symb_table[2 * elem] == 0 && name_len == 1)
+ {
+ /* No valid character, treat it as a normal
+ character. */
+ bitset_set (sbcset, name[0]);
+ return REG_NOERROR;
+ }
+ else
+ return REG_ECOLLATE;
+
+# ifdef RE_ENABLE_I18N
+ /* Got valid collation sequence, add it as a new entry. */
+ /* Check the space of the arrays. */
+ if (*coll_sym_alloc == mbcset->ncoll_syms)
+ {
+ /* Not enough, realloc it. */
+ /* +1 in case of mbcset->ncoll_syms is 0. */
+ *coll_sym_alloc = 2 * mbcset->ncoll_syms + 1;
+ /* Use realloc since mbcset->coll_syms is NULL
+ if *alloc == 0. */
+ mbcset->coll_syms = re_realloc (mbcset->coll_syms, int32_t,
+ *coll_sym_alloc);
+ if (BE (mbcset->coll_syms == NULL, 0))
+ return REG_ESPACE;
+ }
+ mbcset->coll_syms[mbcset->ncoll_syms++] = idx;
+# endif /* RE_ENABLE_I18N */
+ return REG_NOERROR;
+ }
+ else
+ {
+ if (BE (name_len != 1, 0))
+ return REG_ECOLLATE;
+ else
+ {
+ bitset_set (sbcset, name[0]);
+ return REG_NOERROR;
+ }
+ }
+ }
+#endif
+
+ re_token_t br_token;
+ re_bitset_ptr_t sbcset;
+#ifdef RE_ENABLE_I18N
+ re_charset_t *mbcset;
+ int coll_sym_alloc = 0, range_alloc = 0, mbchar_alloc = 0;
+ int equiv_class_alloc = 0, char_class_alloc = 0;
+#else /* not RE_ENABLE_I18N */
+ int non_match = 0;
+#endif /* not RE_ENABLE_I18N */
+ bin_tree_t *work_tree;
+ int token_len, new_idx;
+#ifdef _LIBC
+ collseqmb = (const unsigned char *)
+ _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQMB);
+ nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES);
+ if (nrules)
+ {
+ /*
+ if (MB_CUR_MAX > 1)
+ */
+ collseqwc = _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQWC);
+ table_size = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_SYMB_HASH_SIZEMB);
+ symb_table = (const int32_t *) _NL_CURRENT (LC_COLLATE,
+ _NL_COLLATE_SYMB_TABLEMB);
+ extra = (const unsigned char *) _NL_CURRENT (LC_COLLATE,
+ _NL_COLLATE_SYMB_EXTRAMB);
+ }
+#endif
+ sbcset = (re_bitset_ptr_t) calloc (sizeof (unsigned int), BITSET_UINTS);
+#ifdef RE_ENABLE_I18N
+ mbcset = (re_charset_t *) calloc (sizeof (re_charset_t), 1);
+#endif /* RE_ENABLE_I18N */
+#ifdef RE_ENABLE_I18N
+ if (BE (sbcset == NULL || mbcset == NULL, 0))
+#else
+ if (BE (sbcset == NULL, 0))
+#endif /* RE_ENABLE_I18N */
+ {
+ *err = REG_ESPACE;
+ return NULL;
+ }
+
+ token_len = peek_token_bracket (token, regexp, syntax);
+ if (BE (token->type == END_OF_RE, 0))
+ {
+ *err = REG_BADPAT;
+ goto parse_bracket_exp_free_return;
+ }
+ if (token->type == OP_NON_MATCH_LIST)
+ {
+#ifdef RE_ENABLE_I18N
+ int i;
+ mbcset->non_match = 1;
+#else /* not RE_ENABLE_I18N */
+ non_match = 1;
+#endif /* not RE_ENABLE_I18N */
+ if (syntax & RE_HAT_LISTS_NOT_NEWLINE)
+ bitset_set (sbcset, '\0');
+ re_string_skip_bytes (regexp, token_len); /* Skip a token. */
+ token_len = peek_token_bracket (token, regexp, syntax);
+ if (BE (token->type == END_OF_RE, 0))
+ {
+ *err = REG_BADPAT;
+ goto parse_bracket_exp_free_return;
+ }
+#ifdef RE_ENABLE_I18N
+ if (MB_CUR_MAX > 1)
+ for (i = 0; i < SBC_MAX; ++i)
+ if (__btowc (i) == WEOF)
+ bitset_set (sbcset, i);
+#endif /* RE_ENABLE_I18N */
+ }
+
+ /* We treat the first ']' as a normal character. */
+ if (token->type == OP_CLOSE_BRACKET)
+ token->type = CHARACTER;
+
+ while (1)
+ {
+ bracket_elem_t start_elem, end_elem;
+ unsigned char start_name_buf[BRACKET_NAME_BUF_SIZE];
+ unsigned char end_name_buf[BRACKET_NAME_BUF_SIZE];
+ reg_errcode_t ret;
+ int token_len2 = 0, is_range_exp = 0;
+ re_token_t token2;
+
+ start_elem.opr.name = start_name_buf;
+ ret = parse_bracket_element (&start_elem, regexp, token, token_len, dfa,
+ syntax);
+ if (BE (ret != REG_NOERROR, 0))
+ {
+ *err = ret;
+ goto parse_bracket_exp_free_return;
+ }
+
+ token_len = peek_token_bracket (token, regexp, syntax);
+ if (BE (token->type == END_OF_RE, 0))
+ {
+ *err = REG_BADPAT;
+ goto parse_bracket_exp_free_return;
+ }
+ if (token->type == OP_CHARSET_RANGE)
+ {
+ re_string_skip_bytes (regexp, token_len); /* Skip '-'. */
+ token_len2 = peek_token_bracket (&token2, regexp, syntax);
+ if (BE (token->type == END_OF_RE, 0))
+ {
+ *err = REG_BADPAT;
+ goto parse_bracket_exp_free_return;
+ }
+ if (token2.type == OP_CLOSE_BRACKET)
+ {
+ /* We treat the last '-' as a normal character. */
+ re_string_skip_bytes (regexp, -token_len);
+ token->type = CHARACTER;
+ }
+ else
+ is_range_exp = 1;
+ }
+
+ if (is_range_exp == 1)
+ {
+ end_elem.opr.name = end_name_buf;
+ ret = parse_bracket_element (&end_elem, regexp, &token2, token_len2,
+ dfa, syntax);
+ if (BE (ret != REG_NOERROR, 0))
+ {
+ *err = ret;
+ goto parse_bracket_exp_free_return;
+ }
+
+ token_len = peek_token_bracket (token, regexp, syntax);
+ if (BE (token->type == END_OF_RE, 0))
+ {
+ *err = REG_BADPAT;
+ goto parse_bracket_exp_free_return;
+ }
+ *err = build_range_exp (sbcset,
+#ifdef RE_ENABLE_I18N
+ mbcset, &range_alloc,
+#endif /* RE_ENABLE_I18N */
+ &start_elem, &end_elem);
+ if (BE (*err != REG_NOERROR, 0))
+ goto parse_bracket_exp_free_return;
+ }
+ else
+ {
+ switch (start_elem.type)
+ {
+ case SB_CHAR:
+ bitset_set (sbcset, start_elem.opr.ch);
+ break;
+#ifdef RE_ENABLE_I18N
+ case MB_CHAR:
+ /* Check whether the array has enough space. */
+ if (mbchar_alloc == mbcset->nmbchars)
+ {
+ /* Not enough, realloc it. */
+ /* +1 in case of mbcset->nmbchars is 0. */
+ mbchar_alloc = 2 * mbcset->nmbchars + 1;
+ /* Use realloc since array is NULL if *alloc == 0. */
+ mbcset->mbchars = re_realloc (mbcset->mbchars, wchar_t,
+ mbchar_alloc);
+ if (BE (mbcset->mbchars == NULL, 0))
+ goto parse_bracket_exp_espace;
+ }
+ mbcset->mbchars[mbcset->nmbchars++] = start_elem.opr.wch;
+ break;
+#endif /* RE_ENABLE_I18N */
+ case EQUIV_CLASS:
+ *err = build_equiv_class (sbcset,
+#ifdef RE_ENABLE_I18N
+ mbcset, &equiv_class_alloc,
+#endif /* RE_ENABLE_I18N */
+ start_elem.opr.name);
+ if (BE (*err != REG_NOERROR, 0))
+ goto parse_bracket_exp_free_return;
+ break;
+ case COLL_SYM:
+ *err = build_collating_symbol (sbcset,
+#ifdef RE_ENABLE_I18N
+ mbcset, &coll_sym_alloc,
+#endif /* RE_ENABLE_I18N */
+ start_elem.opr.name);
+ if (BE (*err != REG_NOERROR, 0))
+ goto parse_bracket_exp_free_return;
+ break;
+ case CHAR_CLASS:
+ *err = build_charclass (sbcset,
+#ifdef RE_ENABLE_I18N
+ mbcset, &char_class_alloc,
+#endif /* RE_ENABLE_I18N */
+ start_elem.opr.name, syntax);
+ if (BE (*err != REG_NOERROR, 0))
+ goto parse_bracket_exp_free_return;
+ break;
+ default:
+ assert (0);
+ break;
+ }
+ }
+ if (token->type == OP_CLOSE_BRACKET)
+ break;
+ }
+
+ re_string_skip_bytes (regexp, token_len); /* Skip a token. */
+
+ /* If it is non-matching list. */
+#ifdef RE_ENABLE_I18N
+ if (mbcset->non_match)
+#else /* not RE_ENABLE_I18N */
+ if (non_match)
+#endif /* not RE_ENABLE_I18N */
+ bitset_not (sbcset);
+
+ /* Build a tree for simple bracket. */
+ br_token.type = SIMPLE_BRACKET;
+ br_token.opr.sbcset = sbcset;
+ new_idx = re_dfa_add_node (dfa, br_token, 0);
+ work_tree = create_tree (NULL, NULL, 0, new_idx);
+ if (BE (new_idx == -1 || work_tree == NULL, 0))
+ goto parse_bracket_exp_espace;
+
+#ifdef RE_ENABLE_I18N
+ if (mbcset->nmbchars || mbcset->ncoll_syms || mbcset->nequiv_classes
+ || mbcset->nranges || (MB_CUR_MAX > 1 && (mbcset->nchar_classes
+ || mbcset->non_match)))
+ {
+ re_token_t alt_token;
+ bin_tree_t *mbc_tree;
+ /* Build a tree for complex bracket. */
+ br_token.type = COMPLEX_BRACKET;
+ br_token.opr.mbcset = mbcset;
+ dfa->has_mb_node = 1;
+ new_idx = re_dfa_add_node (dfa, br_token, 0);
+ mbc_tree = create_tree (NULL, NULL, 0, new_idx);
+ if (BE (new_idx == -1 || mbc_tree == NULL, 0))
+ goto parse_bracket_exp_espace;
+ /* Then join them by ALT node. */
+ dfa->has_plural_match = 1;
+ alt_token.type = OP_ALT;
+ new_idx = re_dfa_add_node (dfa, alt_token, 0);
+ work_tree = create_tree (work_tree, mbc_tree, 0, new_idx);
+ if (BE (new_idx != -1 && mbc_tree != NULL, 1))
+ return work_tree;
+ }
+ else
+ {
+ free_charset (mbcset);
+ return work_tree;
+ }
+#else /* not RE_ENABLE_I18N */
+ return work_tree;
+#endif /* not RE_ENABLE_I18N */
+
+ parse_bracket_exp_espace:
+ *err = REG_ESPACE;
+ parse_bracket_exp_free_return:
+ re_free (sbcset);
+#ifdef RE_ENABLE_I18N
+ free_charset (mbcset);
+#endif /* RE_ENABLE_I18N */
+ return NULL;
+}
+
+/* Parse an element in the bracket expression. */
+
+static reg_errcode_t
+parse_bracket_element (elem, regexp, token, token_len, dfa, syntax)
+ bracket_elem_t *elem;
+ re_string_t *regexp;
+ re_token_t *token;
+ int token_len;
+ re_dfa_t *dfa;
+ reg_syntax_t syntax;
+{
+#ifdef RE_ENABLE_I18N
+ int cur_char_size;
+ cur_char_size = re_string_char_size_at (regexp, re_string_cur_idx (regexp));
+ if (cur_char_size > 1)
+ {
+ elem->type = MB_CHAR;
+ elem->opr.wch = re_string_wchar_at (regexp, re_string_cur_idx (regexp));
+ re_string_skip_bytes (regexp, cur_char_size);
+ return REG_NOERROR;
+ }
+#endif /* RE_ENABLE_I18N */
+ re_string_skip_bytes (regexp, token_len); /* Skip a token. */
+ if (token->type == OP_OPEN_COLL_ELEM || token->type == OP_OPEN_CHAR_CLASS
+ || token->type == OP_OPEN_EQUIV_CLASS)
+ return parse_bracket_symbol (elem, regexp, token);
+ elem->type = SB_CHAR;
+ elem->opr.ch = token->opr.c;
+ return REG_NOERROR;
+}
+
+/* Parse a bracket symbol in the bracket expression. Bracket symbols are
+ such as [:<character_class>:], [.<collating_element>.], and
+ [=<equivalent_class>=]. */
+
+static reg_errcode_t
+parse_bracket_symbol (elem, regexp, token)
+ bracket_elem_t *elem;
+ re_string_t *regexp;
+ re_token_t *token;
+{
+ unsigned char ch, delim = token->opr.c;
+ int i = 0;
+ for (;; ++i)
+ {
+ if (re_string_eoi(regexp) || i >= BRACKET_NAME_BUF_SIZE)
+ return REG_EBRACK;
+ if (token->type == OP_OPEN_CHAR_CLASS)
+ ch = re_string_fetch_byte_case (regexp);
+ else
+ ch = re_string_fetch_byte (regexp);
+ if (ch == delim && re_string_peek_byte (regexp, 0) == ']')
+ break;
+ elem->opr.name[i] = ch;
+ }
+ re_string_skip_bytes (regexp, 1);
+ elem->opr.name[i] = '\0';
+ switch (token->type)
+ {
+ case OP_OPEN_COLL_ELEM:
+ elem->type = COLL_SYM;
+ break;
+ case OP_OPEN_EQUIV_CLASS:
+ elem->type = EQUIV_CLASS;
+ break;
+ case OP_OPEN_CHAR_CLASS:
+ elem->type = CHAR_CLASS;
+ break;
+ default:
+ break;
+ }
+ return REG_NOERROR;
+}
+
+ /* Helper function for parse_bracket_exp.
+ Build the equivalence class which is represented by NAME.
+ The result are written to MBCSET and SBCSET.
+ EQUIV_CLASS_ALLOC is the allocated size of mbcset->equiv_classes,
+ is a pointer argument sinse we may update it. */
+
+static reg_errcode_t
+#ifdef RE_ENABLE_I18N
+build_equiv_class (sbcset, mbcset, equiv_class_alloc, name)
+ re_charset_t *mbcset;
+ int *equiv_class_alloc;
+#else /* not RE_ENABLE_I18N */
+build_equiv_class (sbcset, name)
+#endif /* not RE_ENABLE_I18N */
+ re_bitset_ptr_t sbcset;
+ const unsigned char *name;
+{
+#if defined _LIBC && defined RE_ENABLE_I18N
+ uint32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES);
+ if (nrules != 0)
+ {
+ const int32_t *table, *indirect;
+ const unsigned char *weights, *extra, *cp;
+ unsigned char char_buf[2];
+ int32_t idx1, idx2;
+ unsigned int ch;
+ size_t len;
+ /* This #include defines a local function! */
+# include <locale/weight.h>
+ /* Calculate the index for equivalence class. */
+ cp = name;
+ table = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB);
+ weights = (const unsigned char *) _NL_CURRENT (LC_COLLATE,
+ _NL_COLLATE_WEIGHTMB);
+ extra = (const unsigned char *) _NL_CURRENT (LC_COLLATE,
+ _NL_COLLATE_EXTRAMB);
+ indirect = (const int32_t *) _NL_CURRENT (LC_COLLATE,
+ _NL_COLLATE_INDIRECTMB);
+ idx1 = findidx (&cp);
+ if (BE (idx1 == 0 || cp < name + strlen ((const char *) name), 0))
+ /* This isn't a valid character. */
+ return REG_ECOLLATE;
+
+ /* Build single byte matcing table for this equivalence class. */
+ char_buf[1] = (unsigned char) '\0';
+ len = weights[idx1];
+ for (ch = 0; ch < SBC_MAX; ++ch)
+ {
+ char_buf[0] = ch;
+ cp = char_buf;
+ idx2 = findidx (&cp);
+/*
+ idx2 = table[ch];
+*/
+ if (idx2 == 0)
+ /* This isn't a valid character. */
+ continue;
+ if (len == weights[idx2])
+ {
+ int cnt = 0;
+ while (cnt <= len &&
+ weights[idx1 + 1 + cnt] == weights[idx2 + 1 + cnt])
+ ++cnt;
+
+ if (cnt > len)
+ bitset_set (sbcset, ch);
+ }
+ }
+ /* Check whether the array has enough space. */
+ if (*equiv_class_alloc == mbcset->nequiv_classes)
+ {
+ /* Not enough, realloc it. */
+ /* +1 in case of mbcset->nequiv_classes is 0. */
+ *equiv_class_alloc = 2 * mbcset->nequiv_classes + 1;
+ /* Use realloc since the array is NULL if *alloc == 0. */
+ mbcset->equiv_classes = re_realloc (mbcset->equiv_classes, int32_t,
+ *equiv_class_alloc);
+ if (BE (mbcset->equiv_classes == NULL, 0))
+ return REG_ESPACE;
+ }
+ mbcset->equiv_classes[mbcset->nequiv_classes++] = idx1;
+ }
+ else
+#endif /* _LIBC && RE_ENABLE_I18N */
+ {
+ if (BE (strlen ((const char *) name) != 1, 0))
+ return REG_ECOLLATE;
+ bitset_set (sbcset, *name);
+ }
+ return REG_NOERROR;
+}
+
+ /* Helper function for parse_bracket_exp.
+ Build the character class which is represented by NAME.
+ The result are written to MBCSET and SBCSET.
+ CHAR_CLASS_ALLOC is the allocated size of mbcset->char_classes,
+ is a pointer argument sinse we may update it. */
+
+static reg_errcode_t
+#ifdef RE_ENABLE_I18N
+build_charclass (sbcset, mbcset, char_class_alloc, class_name, syntax)
+ re_charset_t *mbcset;
+ int *char_class_alloc;
+#else /* not RE_ENABLE_I18N */
+build_charclass (sbcset, class_name, syntax)
+#endif /* not RE_ENABLE_I18N */
+ re_bitset_ptr_t sbcset;
+ const unsigned char *class_name;
+ reg_syntax_t syntax;
+{
+ int i;
+ const char *name = (const char *) class_name;
+
+ /* In case of REG_ICASE "upper" and "lower" match the both of
+ upper and lower cases. */
+ if ((syntax & RE_ICASE)
+ && (strcmp (name, "upper") == 0 || strcmp (name, "lower") == 0))
+ name = "alpha";
+
+#ifdef RE_ENABLE_I18N
+ /* Check the space of the arrays. */
+ if (*char_class_alloc == mbcset->nchar_classes)
+ {
+ /* Not enough, realloc it. */
+ /* +1 in case of mbcset->nchar_classes is 0. */
+ *char_class_alloc = 2 * mbcset->nchar_classes + 1;
+ /* Use realloc since array is NULL if *alloc == 0. */
+ mbcset->char_classes = re_realloc (mbcset->char_classes, wctype_t,
+ *char_class_alloc);
+ if (BE (mbcset->char_classes == NULL, 0))
+ return REG_ESPACE;
+ }
+ mbcset->char_classes[mbcset->nchar_classes++] = __wctype (name);
+#endif /* RE_ENABLE_I18N */
+
+#define BUILD_CHARCLASS_LOOP(ctype_func)\
+ for (i = 0; i < SBC_MAX; ++i) \
+ { \
+ if (ctype_func (i)) \
+ bitset_set (sbcset, i); \
+ }
+
+ if (strcmp (name, "alnum") == 0)
+ BUILD_CHARCLASS_LOOP (isalnum)
+ else if (strcmp (name, "cntrl") == 0)
+ BUILD_CHARCLASS_LOOP (iscntrl)
+ else if (strcmp (name, "lower") == 0)
+ BUILD_CHARCLASS_LOOP (islower)
+ else if (strcmp (name, "space") == 0)
+ BUILD_CHARCLASS_LOOP (isspace)
+ else if (strcmp (name, "alpha") == 0)
+ BUILD_CHARCLASS_LOOP (isalpha)
+ else if (strcmp (name, "digit") == 0)
+ BUILD_CHARCLASS_LOOP (isdigit)
+ else if (strcmp (name, "print") == 0)
+ BUILD_CHARCLASS_LOOP (isprint)
+ else if (strcmp (name, "upper") == 0)
+ BUILD_CHARCLASS_LOOP (isupper)
+ else if (strcmp (name, "blank") == 0)
+ BUILD_CHARCLASS_LOOP (isblank)
+ else if (strcmp (name, "graph") == 0)
+ BUILD_CHARCLASS_LOOP (isgraph)
+ else if (strcmp (name, "punct") == 0)
+ BUILD_CHARCLASS_LOOP (ispunct)
+ else if (strcmp (name, "xdigit") == 0)
+ BUILD_CHARCLASS_LOOP (isxdigit)
+ else
+ return REG_ECTYPE;
+
+ return REG_NOERROR;
+}
+
+static bin_tree_t *
+build_word_op (dfa, not, err)
+ re_dfa_t *dfa;
+ int not;
+ reg_errcode_t *err;
+{
+ re_bitset_ptr_t sbcset;
+#ifdef RE_ENABLE_I18N
+ re_charset_t *mbcset;
+ int alloc = 0;
+#else /* not RE_ENABLE_I18N */
+ int non_match = 0;
+#endif /* not RE_ENABLE_I18N */
+ reg_errcode_t ret;
+ re_token_t br_token;
+ bin_tree_t *tree;
+ int new_idx;
+
+ sbcset = (re_bitset_ptr_t) calloc (sizeof (unsigned int), BITSET_UINTS);
+#ifdef RE_ENABLE_I18N
+ mbcset = (re_charset_t *) calloc (sizeof (re_charset_t), 1);
+#endif /* RE_ENABLE_I18N */
+
+#ifdef RE_ENABLE_I18N
+ if (BE (sbcset == NULL || mbcset == NULL, 0))
+#else /* not RE_ENABLE_I18N */
+ if (BE (sbcset == NULL, 0))
+#endif /* not RE_ENABLE_I18N */
+ {
+ *err = REG_ESPACE;
+ return NULL;
+ }
+
+ if (not)
+ {
+#ifdef RE_ENABLE_I18N
+ int i;
+ /*
+ if (syntax & RE_HAT_LISTS_NOT_NEWLINE)
+ bitset_set(cset->sbcset, '\0');
+ */
+ mbcset->non_match = 1;
+ if (MB_CUR_MAX > 1)
+ for (i = 0; i < SBC_MAX; ++i)
+ if (__btowc (i) == WEOF)
+ bitset_set (sbcset, i);
+#else /* not RE_ENABLE_I18N */
+ non_match = 1;
+#endif /* not RE_ENABLE_I18N */
+ }
+
+ /* We don't care the syntax in this case. */
+ ret = build_charclass (sbcset,
+#ifdef RE_ENABLE_I18N
+ mbcset, &alloc,
+#endif /* RE_ENABLE_I18N */
+ (const unsigned char *) "alpha", 0);
+
+ if (BE (ret != REG_NOERROR, 0))
+ {
+ re_free (sbcset);
+#ifdef RE_ENABLE_I18N
+ free_charset (mbcset);
+#endif /* RE_ENABLE_I18N */
+ *err = ret;
+ return NULL;
+ }
+ /* \w match '_' also. */
+ bitset_set (sbcset, '_');
+
+ /* If it is non-matching list. */
+#ifdef RE_ENABLE_I18N
+ if (mbcset->non_match)
+#else /* not RE_ENABLE_I18N */
+ if (non_match)
+#endif /* not RE_ENABLE_I18N */
+ bitset_not (sbcset);
+
+ /* Build a tree for simple bracket. */
+ br_token.type = SIMPLE_BRACKET;
+ br_token.opr.sbcset = sbcset;
+ new_idx = re_dfa_add_node (dfa, br_token, 0);
+ tree = create_tree (NULL, NULL, 0, new_idx);
+ if (BE (new_idx == -1 || tree == NULL, 0))
+ goto build_word_op_espace;
+
+#ifdef RE_ENABLE_I18N
+ if (MB_CUR_MAX > 1)
+ {
+ re_token_t alt_token;
+ bin_tree_t *mbc_tree;
+ /* Build a tree for complex bracket. */
+ br_token.type = COMPLEX_BRACKET;
+ br_token.opr.mbcset = mbcset;
+ dfa->has_mb_node = 1;
+ new_idx = re_dfa_add_node (dfa, br_token, 0);
+ mbc_tree = create_tree (NULL, NULL, 0, new_idx);
+ if (BE (new_idx == -1 || mbc_tree == NULL, 0))
+ goto build_word_op_espace;
+ /* Then join them by ALT node. */
+ alt_token.type = OP_ALT;
+ new_idx = re_dfa_add_node (dfa, alt_token, 0);
+ tree = create_tree (tree, mbc_tree, 0, new_idx);
+ if (BE (new_idx != -1 && mbc_tree != NULL, 1))
+ return tree;
+ }
+ else
+ {
+ free_charset (mbcset);
+ return tree;
+ }
+#else /* not RE_ENABLE_I18N */
+ return tree;
+#endif /* not RE_ENABLE_I18N */
+
+ build_word_op_espace:
+ re_free (sbcset);
+#ifdef RE_ENABLE_I18N
+ free_charset (mbcset);
+#endif /* RE_ENABLE_I18N */
+ *err = REG_ESPACE;
+ return NULL;
+}
+
+/* This is intended for the expressions like "a{1,3}".
+ Fetch a number from `input', and return the number.
+ Return -1, if the number field is empty like "{,1}".
+ Return -2, If an error is occured. */
+
+static int
+fetch_number (input, token, syntax)
+ re_string_t *input;
+ re_token_t *token;
+ reg_syntax_t syntax;
+{
+ int num = -1;
+ unsigned char c;
+ while (1)
+ {
+ *token = fetch_token (input, syntax);
+ c = token->opr.c;
+ if (BE (token->type == END_OF_RE, 0))
+ return -2;
+ if (token->type == OP_CLOSE_DUP_NUM || c == ',')
+ break;
+ num = ((token->type != CHARACTER || c < '0' || '9' < c || num == -2)
+ ? -2 : ((num == -1) ? c - '0' : num * 10 + c - '0'));
+ num = (num > RE_DUP_MAX) ? -2 : num;
+ }
+ return num;
+}
+\f
+#ifdef RE_ENABLE_I18N
+static void
+free_charset (re_charset_t *cset)
+{
+ re_free (cset->mbchars);
+# ifdef _LIBC
+ re_free (cset->coll_syms);
+ re_free (cset->equiv_classes);
+ re_free (cset->range_starts);
+ re_free (cset->range_ends);
+# endif
+ re_free (cset->char_classes);
+ re_free (cset);
+}
+#endif /* RE_ENABLE_I18N */
+\f
+/* Functions for binary tree operation. */
+
+/* Create a node of tree.
+ Note: This function automatically free left and right if malloc fails. */
+
+static bin_tree_t *
+create_tree (left, right, type, index)
+ bin_tree_t *left;
+ bin_tree_t *right;
+ re_token_type_t type;
+ int index;
+{
+ bin_tree_t *tree;
+ tree = re_malloc (bin_tree_t, 1);
+ if (BE (tree == NULL, 0))
+ {
+ free_bin_tree (left);
+ free_bin_tree (right);
+ return NULL;
+ }
+ tree->parent = NULL;
+ tree->left = left;
+ tree->right = right;
+ tree->type = type;
+ tree->node_idx = index;
+ tree->first = -1;
+ tree->next = -1;
+ re_node_set_init_empty (&tree->eclosure);
+
+ if (left != NULL)
+ left->parent = tree;
+ if (right != NULL)
+ right->parent = tree;
+ return tree;
+}
+
+/* Free the sub tree pointed by TREE. */
+
+static void
+free_bin_tree (tree)
+ bin_tree_t *tree;
+{
+ if (tree == NULL)
+ return;
+ /*re_node_set_free (&tree->eclosure);*/
+ free_bin_tree (tree->left);
+ free_bin_tree (tree->right);
+ re_free (tree);
+}
+
+/* Duplicate the node SRC, and return new node. */
+
+static bin_tree_t *
+duplicate_tree (src, dfa)
+ const bin_tree_t *src;
+ re_dfa_t *dfa;
+{
+ bin_tree_t *left = NULL, *right = NULL, *new_tree;
+ int new_node_idx;
+ /* Since node indies must be according to Post-order of the tree,
+ we must duplicate the left at first. */
+ if (src->left != NULL)
+ {
+ left = duplicate_tree (src->left, dfa);
+ if (left == NULL)
+ return NULL;
+ }
+
+ /* Secondaly, duplicate the right. */
+ if (src->right != NULL)
+ {
+ right = duplicate_tree (src->right, dfa);
+ if (right == NULL)
+ {
+ free_bin_tree (left);
+ return NULL;
+ }
+ }
+
+ /* At last, duplicate itself. */
+ if (src->type == NON_TYPE)
+ {
+ new_node_idx = re_dfa_add_node (dfa, dfa->nodes[src->node_idx], 0);
+ dfa->nodes[new_node_idx].duplicated = 1;
+ if (BE (new_node_idx == -1, 0))
+ {
+ free_bin_tree (left);
+ free_bin_tree (right);
+ return NULL;
+ }
+ }
+ else
+ new_node_idx = src->type;
+
+ new_tree = create_tree (left, right, src->type, new_node_idx);
+ if (BE (new_tree == NULL, 0))
+ {
+ free_bin_tree (left);
+ free_bin_tree (right);
+ }
+ return new_tree;
+}
--- /dev/null
+/* Extended regular expression matching and search library.
+ Copyright (C) 2002, 2003 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+ Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, write to the Free
+ Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ 02111-1307 USA. */
+
+#ifdef _LIBC
+/* We have to keep the namespace clean. */
+# define regfree(preg) __regfree (preg)
+# define regexec(pr, st, nm, pm, ef) __regexec (pr, st, nm, pm, ef)
+# define regcomp(preg, pattern, cflags) __regcomp (preg, pattern, cflags)
+# define regerror(errcode, preg, errbuf, errbuf_size) \
+ __regerror(errcode, preg, errbuf, errbuf_size)
+# define re_set_registers(bu, re, nu, st, en) \
+ __re_set_registers (bu, re, nu, st, en)
+# define re_match_2(bufp, string1, size1, string2, size2, pos, regs, stop) \
+ __re_match_2 (bufp, string1, size1, string2, size2, pos, regs, stop)
+# define re_match(bufp, string, size, pos, regs) \
+ __re_match (bufp, string, size, pos, regs)
+# define re_search(bufp, string, size, startpos, range, regs) \
+ __re_search (bufp, string, size, startpos, range, regs)
+# define re_compile_pattern(pattern, length, bufp) \
+ __re_compile_pattern (pattern, length, bufp)
+# define re_set_syntax(syntax) __re_set_syntax (syntax)
+# define re_search_2(bufp, st1, s1, st2, s2, startpos, range, regs, stop) \
+ __re_search_2 (bufp, st1, s1, st2, s2, startpos, range, regs, stop)
+# define re_compile_fastmap(bufp) __re_compile_fastmap (bufp)
+#endif
+
+/* POSIX says that <sys/types.h> must be included (by the caller) before
+ <regex.h>. */
+#include <sys/types.h>
+#include <regex.h>
+#include "regex_internal.h"
+
+#include "regex_internal.c"
+#include "regcomp.c"
+#include "regexec.c"
+
+/* Binary backward compatibility. */
+#if _LIBC
+# include <shlib-compat.h>
+# if SHLIB_COMPAT (libc, GLIBC_2_0, GLIBC_2_3)
+link_warning (re_max_failures, "the 're_max_failures' variable is obsolete and will go away.")
+int re_max_failures = 2000;
+# endif
+#endif
--- /dev/null
+/* Definitions for data structures and routines for the regular
+ expression library.
+ Copyright (C) 1985,1989-93,1995-98,2000,2001,2002
+ Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, write to the Free
+ Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ 02111-1307 USA. */
+
+#ifndef _REGEX_H
+#define _REGEX_H 1
+
+/* Allow the use in C++ code. */
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* POSIX says that <sys/types.h> must be included (by the caller) before
+ <regex.h>. */
+
+#if !defined _POSIX_C_SOURCE && !defined _POSIX_SOURCE && (defined VMS || defined _MSC_VER)
+/* VMS doesn't have `size_t' in <sys/types.h>, even though POSIX says it
+ should be there. Same for Microsoft Visual C++ 6.0 */
+# include <stddef.h>
+#endif
+
+/* The following two types have to be signed and unsigned integer type
+ wide enough to hold a value of a pointer. For most ANSI compilers
+ ptrdiff_t and size_t should be likely OK. Still size of these two
+ types is 2 for Microsoft C. Ugh... */
+typedef long int s_reg_t;
+typedef unsigned long int active_reg_t;
+
+/* The following bits are used to determine the regexp syntax we
+ recognize. The set/not-set meanings are chosen so that Emacs syntax
+ remains the value 0. The bits are given in alphabetical order, and
+ the definitions shifted by one from the previous bit; thus, when we
+ add or remove a bit, only one other definition need change. */
+typedef unsigned long int reg_syntax_t;
+
+/* If this bit is not set, then \ inside a bracket expression is literal.
+ If set, then such a \ quotes the following character. */
+#define RE_BACKSLASH_ESCAPE_IN_LISTS ((unsigned long int) 1)
+
+/* If this bit is not set, then + and ? are operators, and \+ and \? are
+ literals.
+ If set, then \+ and \? are operators and + and ? are literals. */
+#define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1)
+
+/* If this bit is set, then character classes are supported. They are:
+ [:alpha:], [:upper:], [:lower:], [:digit:], [:alnum:], [:xdigit:],
+ [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:].
+ If not set, then character classes are not supported. */
+#define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1)
+
+/* If this bit is set, then ^ and $ are always anchors (outside bracket
+ expressions, of course).
+ If this bit is not set, then it depends:
+ ^ is an anchor if it is at the beginning of a regular
+ expression or after an open-group or an alternation operator;
+ $ is an anchor if it is at the end of a regular expression, or
+ before a close-group or an alternation operator.
+
+ This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because
+ POSIX draft 11.2 says that * etc. in leading positions is undefined.
+ We already implemented a previous draft which made those constructs
+ invalid, though, so we haven't changed the code back. */
+#define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1)
+
+/* If this bit is set, then special characters are always special
+ regardless of where they are in the pattern.
+ If this bit is not set, then special characters are special only in
+ some contexts; otherwise they are ordinary. Specifically,
+ * + ? and intervals are only special when not after the beginning,
+ open-group, or alternation operator. */
+#define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1)
+
+/* If this bit is set, then *, +, ?, and { cannot be first in an re or
+ immediately after an alternation or begin-group operator. */
+#define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1)
+
+/* If this bit is set, then . matches newline.
+ If not set, then it doesn't. */
+#define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1)
+
+/* If this bit is set, then . doesn't match NUL.
+ If not set, then it does. */
+#define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1)
+
+/* If this bit is set, nonmatching lists [^...] do not match newline.
+ If not set, they do. */
+#define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1)
+
+/* If this bit is set, either \{...\} or {...} defines an
+ interval, depending on RE_NO_BK_BRACES.
+ If not set, \{, \}, {, and } are literals. */
+#define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1)
+
+/* If this bit is set, +, ? and | aren't recognized as operators.
+ If not set, they are. */
+#define RE_LIMITED_OPS (RE_INTERVALS << 1)
+
+/* If this bit is set, newline is an alternation operator.
+ If not set, newline is literal. */
+#define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1)
+
+/* If this bit is set, then `{...}' defines an interval, and \{ and \}
+ are literals.
+ If not set, then `\{...\}' defines an interval. */
+#define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1)
+
+/* If this bit is set, (...) defines a group, and \( and \) are literals.
+ If not set, \(...\) defines a group, and ( and ) are literals. */
+#define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1)
+
+/* If this bit is set, then \<digit> matches <digit>.
+ If not set, then \<digit> is a back-reference. */
+#define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1)
+
+/* If this bit is set, then | is an alternation operator, and \| is literal.
+ If not set, then \| is an alternation operator, and | is literal. */
+#define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1)
+
+/* If this bit is set, then an ending range point collating higher
+ than the starting range point, as in [z-a], is invalid.
+ If not set, then when ending range point collates higher than the
+ starting range point, the range is ignored. */
+#define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1)
+
+/* If this bit is set, then an unmatched ) is ordinary.
+ If not set, then an unmatched ) is invalid. */
+#define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1)
+
+/* If this bit is set, succeed as soon as we match the whole pattern,
+ without further backtracking. */
+#define RE_NO_POSIX_BACKTRACKING (RE_UNMATCHED_RIGHT_PAREN_ORD << 1)
+
+/* If this bit is set, do not process the GNU regex operators.
+ If not set, then the GNU regex operators are recognized. */
+#define RE_NO_GNU_OPS (RE_NO_POSIX_BACKTRACKING << 1)
+
+/* If this bit is set, turn on internal regex debugging.
+ If not set, and debugging was on, turn it off.
+ This only works if regex.c is compiled -DDEBUG.
+ We define this bit always, so that all that's needed to turn on
+ debugging is to recompile regex.c; the calling code can always have
+ this bit set, and it won't affect anything in the normal case. */
+#define RE_DEBUG (RE_NO_GNU_OPS << 1)
+
+/* If this bit is set, a syntactically invalid interval is treated as
+ a string of ordinary characters. For example, the ERE 'a{1' is
+ treated as 'a\{1'. */
+#define RE_INVALID_INTERVAL_ORD (RE_DEBUG << 1)
+
+/* If this bit is set, then ignore case when matching.
+ If not set, then case is significant. */
+#define RE_ICASE (RE_INVALID_INTERVAL_ORD << 1)
+
+/* This global variable defines the particular regexp syntax to use (for
+ some interfaces). When a regexp is compiled, the syntax used is
+ stored in the pattern buffer, so changing this does not affect
+ already-compiled regexps. */
+extern reg_syntax_t re_syntax_options;
+\f
+/* Define combinations of the above bits for the standard possibilities.
+ (The [[[ comments delimit what gets put into the Texinfo file, so
+ don't delete them!) */
+/* [[[begin syntaxes]]] */
+#define RE_SYNTAX_EMACS 0
+
+#define RE_SYNTAX_AWK \
+ (RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL \
+ | RE_NO_BK_PARENS | RE_NO_BK_REFS \
+ | RE_NO_BK_VBAR | RE_NO_EMPTY_RANGES \
+ | RE_DOT_NEWLINE | RE_CONTEXT_INDEP_ANCHORS \
+ | RE_UNMATCHED_RIGHT_PAREN_ORD | RE_NO_GNU_OPS)
+
+#define RE_SYNTAX_GNU_AWK \
+ ((RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DEBUG) \
+ & ~(RE_DOT_NOT_NULL | RE_INTERVALS | RE_CONTEXT_INDEP_OPS \
+ | RE_CONTEXT_INVALID_OPS ))
+
+#define RE_SYNTAX_POSIX_AWK \
+ (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS \
+ | RE_INTERVALS | RE_NO_GNU_OPS)
+
+#define RE_SYNTAX_GREP \
+ (RE_BK_PLUS_QM | RE_CHAR_CLASSES \
+ | RE_HAT_LISTS_NOT_NEWLINE | RE_INTERVALS \
+ | RE_NEWLINE_ALT)
+
+#define RE_SYNTAX_EGREP \
+ (RE_CHAR_CLASSES | RE_CONTEXT_INDEP_ANCHORS \
+ | RE_CONTEXT_INDEP_OPS | RE_HAT_LISTS_NOT_NEWLINE \
+ | RE_NEWLINE_ALT | RE_NO_BK_PARENS \
+ | RE_NO_BK_VBAR)
+
+#define RE_SYNTAX_POSIX_EGREP \
+ (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES \
+ | RE_INVALID_INTERVAL_ORD)
+
+/* P1003.2/D11.2, section 4.20.7.1, lines 5078ff. */
+#define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC
+
+#define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC
+
+/* Syntax bits common to both basic and extended POSIX regex syntax. */
+#define _RE_SYNTAX_POSIX_COMMON \
+ (RE_CHAR_CLASSES | RE_DOT_NEWLINE | RE_DOT_NOT_NULL \
+ | RE_INTERVALS | RE_NO_EMPTY_RANGES)
+
+#define RE_SYNTAX_POSIX_BASIC \
+ (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM)
+
+/* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes
+ RE_LIMITED_OPS, i.e., \? \+ \| are not recognized. Actually, this
+ isn't minimal, since other operators, such as \`, aren't disabled. */
+#define RE_SYNTAX_POSIX_MINIMAL_BASIC \
+ (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS)
+
+#define RE_SYNTAX_POSIX_EXTENDED \
+ (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \
+ | RE_CONTEXT_INDEP_OPS | RE_NO_BK_BRACES \
+ | RE_NO_BK_PARENS | RE_NO_BK_VBAR \
+ | RE_CONTEXT_INVALID_OPS | RE_UNMATCHED_RIGHT_PAREN_ORD)
+
+/* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INDEP_OPS is
+ removed and RE_NO_BK_REFS is added. */
+#define RE_SYNTAX_POSIX_MINIMAL_EXTENDED \
+ (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \
+ | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES \
+ | RE_NO_BK_PARENS | RE_NO_BK_REFS \
+ | RE_NO_BK_VBAR | RE_UNMATCHED_RIGHT_PAREN_ORD)
+/* [[[end syntaxes]]] */
+\f
+/* Maximum number of duplicates an interval can allow. Some systems
+ (erroneously) define this in other header files, but we want our
+ value, so remove any previous define. */
+#ifdef RE_DUP_MAX
+# undef RE_DUP_MAX
+#endif
+/* If sizeof(int) == 2, then ((1 << 15) - 1) overflows. */
+#define RE_DUP_MAX (0x7fff)
+
+
+/* POSIX `cflags' bits (i.e., information for `regcomp'). */
+
+/* If this bit is set, then use extended regular expression syntax.
+ If not set, then use basic regular expression syntax. */
+#define REG_EXTENDED 1
+
+/* If this bit is set, then ignore case when matching.
+ If not set, then case is significant. */
+#define REG_ICASE (REG_EXTENDED << 1)
+
+/* If this bit is set, then anchors do not match at newline
+ characters in the string.
+ If not set, then anchors do match at newlines. */
+#define REG_NEWLINE (REG_ICASE << 1)
+
+/* If this bit is set, then report only success or fail in regexec.
+ If not set, then returns differ between not matching and errors. */
+#define REG_NOSUB (REG_NEWLINE << 1)
+
+
+/* POSIX `eflags' bits (i.e., information for regexec). */
+
+/* If this bit is set, then the beginning-of-line operator doesn't match
+ the beginning of the string (presumably because it's not the
+ beginning of a line).
+ If not set, then the beginning-of-line operator does match the
+ beginning of the string. */
+#define REG_NOTBOL 1
+
+/* Like REG_NOTBOL, except for the end-of-line. */
+#define REG_NOTEOL (1 << 1)
+
+
+/* If any error codes are removed, changed, or added, update the
+ `re_error_msg' table in regex.c. */
+typedef enum
+{
+#ifdef _XOPEN_SOURCE
+ REG_ENOSYS = -1, /* This will never happen for this implementation. */
+#endif
+
+ REG_NOERROR = 0, /* Success. */
+ REG_NOMATCH, /* Didn't find a match (for regexec). */
+
+ /* POSIX regcomp return error codes. (In the order listed in the
+ standard.) */
+ REG_BADPAT, /* Invalid pattern. */
+ REG_ECOLLATE, /* Not implemented. */
+ REG_ECTYPE, /* Invalid character class name. */
+ REG_EESCAPE, /* Trailing backslash. */
+ REG_ESUBREG, /* Invalid back reference. */
+ REG_EBRACK, /* Unmatched left bracket. */
+ REG_EPAREN, /* Parenthesis imbalance. */
+ REG_EBRACE, /* Unmatched \{. */
+ REG_BADBR, /* Invalid contents of \{\}. */
+ REG_ERANGE, /* Invalid range end. */
+ REG_ESPACE, /* Ran out of memory. */
+ REG_BADRPT, /* No preceding re for repetition op. */
+
+ /* Error codes we've added. */
+ REG_EEND, /* Premature end. */
+ REG_ESIZE, /* Compiled pattern bigger than 2^16 bytes. */
+ REG_ERPAREN /* Unmatched ) or \); not returned from regcomp. */
+} reg_errcode_t;
+\f
+/* This data structure represents a compiled pattern. Before calling
+ the pattern compiler, the fields `buffer', `allocated', `fastmap',
+ `translate', and `no_sub' can be set. After the pattern has been
+ compiled, the `re_nsub' field is available. All other fields are
+ private to the regex routines. */
+
+#ifndef RE_TRANSLATE_TYPE
+# define RE_TRANSLATE_TYPE char *
+#endif
+
+struct re_pattern_buffer
+{
+/* [[[begin pattern_buffer]]] */
+ /* Space that holds the compiled pattern. It is declared as
+ `unsigned char *' because its elements are
+ sometimes used as array indexes. */
+ unsigned char *buffer;
+
+ /* Number of bytes to which `buffer' points. */
+ unsigned long int allocated;
+
+ /* Number of bytes actually used in `buffer'. */
+ unsigned long int used;
+
+ /* Syntax setting with which the pattern was compiled. */
+ reg_syntax_t syntax;
+
+ /* Pointer to a fastmap, if any, otherwise zero. re_search uses
+ the fastmap, if there is one, to skip over impossible
+ starting points for matches. */
+ char *fastmap;
+
+ /* Either a translate table to apply to all characters before
+ comparing them, or zero for no translation. The translation
+ is applied to a pattern when it is compiled and to a string
+ when it is matched. */
+ RE_TRANSLATE_TYPE translate;
+
+ /* Number of subexpressions found by the compiler. */
+ size_t re_nsub;
+
+ /* Zero if this pattern cannot match the empty string, one else.
+ Well, in truth it's used only in `re_search_2', to see
+ whether or not we should use the fastmap, so we don't set
+ this absolutely perfectly; see `re_compile_fastmap' (the
+ `duplicate' case). */
+ unsigned can_be_null : 1;
+
+ /* If REGS_UNALLOCATED, allocate space in the `regs' structure
+ for `max (RE_NREGS, re_nsub + 1)' groups.
+ If REGS_REALLOCATE, reallocate space if necessary.
+ If REGS_FIXED, use what's there. */
+#define REGS_UNALLOCATED 0
+#define REGS_REALLOCATE 1
+#define REGS_FIXED 2
+ unsigned regs_allocated : 2;
+
+ /* Set to zero when `regex_compile' compiles a pattern; set to one
+ by `re_compile_fastmap' if it updates the fastmap. */
+ unsigned fastmap_accurate : 1;
+
+ /* If set, `re_match_2' does not return information about
+ subexpressions. */
+ unsigned no_sub : 1;
+
+ /* If set, a beginning-of-line anchor doesn't match at the
+ beginning of the string. */
+ unsigned not_bol : 1;
+
+ /* Similarly for an end-of-line anchor. */
+ unsigned not_eol : 1;
+
+ /* If true, an anchor at a newline matches. */
+ unsigned newline_anchor : 1;
+
+/* [[[end pattern_buffer]]] */
+};
+
+typedef struct re_pattern_buffer regex_t;
+\f
+/* Type for byte offsets within the string. POSIX mandates this. */
+typedef int regoff_t;
+
+
+/* This is the structure we store register match data in. See
+ regex.texinfo for a full description of what registers match. */
+struct re_registers
+{
+ unsigned num_regs;
+ regoff_t *start;
+ regoff_t *end;
+};
+
+
+/* If `regs_allocated' is REGS_UNALLOCATED in the pattern buffer,
+ `re_match_2' returns information about at least this many registers
+ the first time a `regs' structure is passed. */
+#ifndef RE_NREGS
+# define RE_NREGS 30
+#endif
+
+
+/* POSIX specification for registers. Aside from the different names than
+ `re_registers', POSIX uses an array of structures, instead of a
+ structure of arrays. */
+typedef struct
+{
+ regoff_t rm_so; /* Byte offset from string's start to substring's start. */
+ regoff_t rm_eo; /* Byte offset from string's start to substring's end. */
+} regmatch_t;
+\f
+/* Declarations for routines. */
+
+/* To avoid duplicating every routine declaration -- once with a
+ prototype (if we are ANSI), and once without (if we aren't) -- we
+ use the following macro to declare argument types. This
+ unfortunately clutters up the declarations a bit, but I think it's
+ worth it. */
+
+#if defined(__STDC__) || defined(__cplusplus)
+
+# define _RE_ARGS(args) args
+
+#else /* not __STDC__ */
+
+# define _RE_ARGS(args) ()
+
+#endif /* not __STDC__ */
+
+/* Sets the current default syntax to SYNTAX, and return the old syntax.
+ You can also simply assign to the `re_syntax_options' variable. */
+extern reg_syntax_t re_set_syntax _RE_ARGS ((reg_syntax_t syntax));
+
+/* Compile the regular expression PATTERN, with length LENGTH
+ and syntax given by the global `re_syntax_options', into the buffer
+ BUFFER. Return NULL if successful, and an error string if not. */
+extern const char *re_compile_pattern
+ _RE_ARGS ((const char *pattern, size_t length,
+ struct re_pattern_buffer *buffer));
+
+
+/* Compile a fastmap for the compiled pattern in BUFFER; used to
+ accelerate searches. Return 0 if successful and -2 if was an
+ internal error. */
+extern int re_compile_fastmap _RE_ARGS ((struct re_pattern_buffer *buffer));
+
+
+/* Search in the string STRING (with length LENGTH) for the pattern
+ compiled into BUFFER. Start searching at position START, for RANGE
+ characters. Return the starting position of the match, -1 for no
+ match, or -2 for an internal error. Also return register
+ information in REGS (if REGS and BUFFER->no_sub are nonzero). */
+extern int re_search
+ _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string,
+ int length, int start, int range, struct re_registers *regs));
+
+
+/* Like `re_search', but search in the concatenation of STRING1 and
+ STRING2. Also, stop searching at index START + STOP. */
+extern int re_search_2
+ _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1,
+ int length1, const char *string2, int length2,
+ int start, int range, struct re_registers *regs, int stop));
+
+
+/* Like `re_search', but return how many characters in STRING the regexp
+ in BUFFER matched, starting at position START. */
+extern int re_match
+ _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string,
+ int length, int start, struct re_registers *regs));
+
+
+/* Relates to `re_match' as `re_search_2' relates to `re_search'. */
+extern int re_match_2
+ _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1,
+ int length1, const char *string2, int length2,
+ int start, struct re_registers *regs, int stop));
+
+
+/* Set REGS to hold NUM_REGS registers, storing them in STARTS and
+ ENDS. Subsequent matches using BUFFER and REGS will use this memory
+ for recording register information. STARTS and ENDS must be
+ allocated with malloc, and must each be at least `NUM_REGS * sizeof
+ (regoff_t)' bytes long.
+
+ If NUM_REGS == 0, then subsequent matches should allocate their own
+ register data.
+
+ Unless this function is called, the first search or match using
+ PATTERN_BUFFER will allocate its own register data, without
+ freeing the old data. */
+extern void re_set_registers
+ _RE_ARGS ((struct re_pattern_buffer *buffer, struct re_registers *regs,
+ unsigned num_regs, regoff_t *starts, regoff_t *ends));
+
+#if defined _REGEX_RE_COMP || defined _LIBC
+# ifndef _CRAY
+/* 4.2 bsd compatibility. */
+extern char *re_comp _RE_ARGS ((const char *));
+extern int re_exec _RE_ARGS ((const char *));
+# endif
+#endif
+
+/* GCC 2.95 and later have "__restrict"; C99 compilers have
+ "restrict", and "configure" may have defined "restrict". */
+#ifndef __restrict
+# if ! (2 < __GNUC__ || (2 == __GNUC__ && 95 <= __GNUC_MINOR__))
+# if defined restrict || 199901L <= __STDC_VERSION__
+# define __restrict restrict
+# else
+# define __restrict
+# endif
+# endif
+#endif
+/* gcc 3.1 and up support the [restrict] syntax. */
+#ifndef __restrict_arr
+# if (__GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 1)) && !defined(__WIN32__)
+# define __restrict_arr __restrict
+# else
+# define __restrict_arr
+# endif
+#endif
+
+/* POSIX compatibility. */
+extern int regcomp _RE_ARGS ((regex_t *__restrict __preg,
+ const char *__restrict __pattern,
+ int __cflags));
+
+extern int regexec _RE_ARGS ((const regex_t *__restrict __preg,
+ const char *__restrict __string, size_t __nmatch,
+ regmatch_t __pmatch[__restrict_arr],
+ int __eflags));
+
+extern size_t regerror _RE_ARGS ((int __errcode, const regex_t *__preg,
+ char *__errbuf, size_t __errbuf_size));
+
+extern void regfree _RE_ARGS ((regex_t *__preg));
+
+
+#ifdef __cplusplus
+}
+#endif /* C++ */
+
+#endif /* regex.h */
+\f
+/*
+Local variables:
+make-backup-files: t
+version-control: t
+trim-versions-without-asking: nil
+End:
+*/
--- /dev/null
+/* Extended regular expression matching and search library.
+ Copyright (C) 2002, 2003 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+ Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, write to the Free
+ Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ 02111-1307 USA. */
+
+static void re_string_construct_common (const char *str, int len,
+ re_string_t *pstr,
+ RE_TRANSLATE_TYPE trans, int icase);
+#ifdef RE_ENABLE_I18N
+static int re_string_skip_chars (re_string_t *pstr, int new_raw_idx,
+ wint_t *last_wc);
+#endif /* RE_ENABLE_I18N */
+static re_dfastate_t *create_newstate_common (re_dfa_t *dfa,
+ const re_node_set *nodes,
+ unsigned int hash);
+static reg_errcode_t register_state (re_dfa_t *dfa, re_dfastate_t *newstate,
+ unsigned int hash);
+static re_dfastate_t *create_ci_newstate (re_dfa_t *dfa,
+ const re_node_set *nodes,
+ unsigned int hash);
+static re_dfastate_t *create_cd_newstate (re_dfa_t *dfa,
+ const re_node_set *nodes,
+ unsigned int context,
+ unsigned int hash);
+static unsigned int inline calc_state_hash (const re_node_set *nodes,
+ unsigned int context);
+\f
+/* Functions for string operation. */
+
+/* This function allocate the buffers. It is necessary to call
+ re_string_reconstruct before using the object. */
+
+static reg_errcode_t
+re_string_allocate (pstr, str, len, init_len, trans, icase)
+ re_string_t *pstr;
+ const char *str;
+ int len, init_len, icase;
+ RE_TRANSLATE_TYPE trans;
+{
+ reg_errcode_t ret;
+ int init_buf_len = (len + 1 < init_len) ? len + 1: init_len;
+ re_string_construct_common (str, len, pstr, trans, icase);
+ pstr->stop = pstr->len;
+
+ ret = re_string_realloc_buffers (pstr, init_buf_len);
+ if (BE (ret != REG_NOERROR, 0))
+ return ret;
+
+ pstr->mbs_case = (MBS_CASE_ALLOCATED (pstr) ? pstr->mbs_case
+ : (unsigned char *) str);
+ pstr->mbs = MBS_ALLOCATED (pstr) ? pstr->mbs : pstr->mbs_case;
+ pstr->valid_len = (MBS_CASE_ALLOCATED (pstr) || MBS_ALLOCATED (pstr)
+ || MB_CUR_MAX > 1) ? pstr->valid_len : len;
+ return REG_NOERROR;
+}
+
+/* This function allocate the buffers, and initialize them. */
+
+static reg_errcode_t
+re_string_construct (pstr, str, len, trans, icase)
+ re_string_t *pstr;
+ const char *str;
+ int len, icase;
+ RE_TRANSLATE_TYPE trans;
+{
+ reg_errcode_t ret;
+ re_string_construct_common (str, len, pstr, trans, icase);
+ pstr->stop = pstr->len;
+ /* Set 0 so that this function can initialize whole buffers. */
+ pstr->valid_len = 0;
+
+ if (len > 0)
+ {
+ ret = re_string_realloc_buffers (pstr, len + 1);
+ if (BE (ret != REG_NOERROR, 0))
+ return ret;
+ }
+ pstr->mbs_case = (MBS_CASE_ALLOCATED (pstr) ? pstr->mbs_case
+ : (unsigned char *) str);
+ pstr->mbs = MBS_ALLOCATED (pstr) ? pstr->mbs : pstr->mbs_case;
+
+ if (icase)
+ {
+#ifdef RE_ENABLE_I18N
+ if (MB_CUR_MAX > 1)
+ build_wcs_upper_buffer (pstr);
+ else
+#endif /* RE_ENABLE_I18N */
+ build_upper_buffer (pstr);
+ }
+ else
+ {
+#ifdef RE_ENABLE_I18N
+ if (MB_CUR_MAX > 1)
+ build_wcs_buffer (pstr);
+ else
+#endif /* RE_ENABLE_I18N */
+ {
+ if (trans != NULL)
+ re_string_translate_buffer (pstr);
+ else
+ pstr->valid_len = len;
+ }
+ }
+
+ /* Initialized whole buffers, then valid_len == bufs_len. */
+ pstr->valid_len = pstr->bufs_len;
+ return REG_NOERROR;
+}
+
+/* Helper functions for re_string_allocate, and re_string_construct. */
+
+static reg_errcode_t
+re_string_realloc_buffers (pstr, new_buf_len)
+ re_string_t *pstr;
+ int new_buf_len;
+{
+#ifdef RE_ENABLE_I18N
+ if (MB_CUR_MAX > 1)
+ {
+ wint_t *new_array = re_realloc (pstr->wcs, wint_t, new_buf_len);
+ if (BE (new_array == NULL, 0))
+ return REG_ESPACE;
+ pstr->wcs = new_array;
+ }
+#endif /* RE_ENABLE_I18N */
+ if (MBS_ALLOCATED (pstr))
+ {
+ unsigned char *new_array = re_realloc (pstr->mbs, unsigned char,
+ new_buf_len);
+ if (BE (new_array == NULL, 0))
+ return REG_ESPACE;
+ pstr->mbs = new_array;
+ }
+ if (MBS_CASE_ALLOCATED (pstr))
+ {
+ unsigned char *new_array = re_realloc (pstr->mbs_case, unsigned char,
+ new_buf_len);
+ if (BE (new_array == NULL, 0))
+ return REG_ESPACE;
+ pstr->mbs_case = new_array;
+ if (!MBS_ALLOCATED (pstr))
+ pstr->mbs = pstr->mbs_case;
+ }
+ pstr->bufs_len = new_buf_len;
+ return REG_NOERROR;
+}
+
+
+static void
+re_string_construct_common (str, len, pstr, trans, icase)
+ const char *str;
+ int len;
+ re_string_t *pstr;
+ RE_TRANSLATE_TYPE trans;
+ int icase;
+{
+ memset (pstr, '\0', sizeof (re_string_t));
+ pstr->raw_mbs = (const unsigned char *) str;
+ pstr->len = len;
+ pstr->trans = trans;
+ pstr->icase = icase ? 1 : 0;
+}
+
+#ifdef RE_ENABLE_I18N
+
+/* Build wide character buffer PSTR->WCS.
+ If the byte sequence of the string are:
+ <mb1>(0), <mb1>(1), <mb2>(0), <mb2>(1), <sb3>
+ Then wide character buffer will be:
+ <wc1> , WEOF , <wc2> , WEOF , <wc3>
+ We use WEOF for padding, they indicate that the position isn't
+ a first byte of a multibyte character.
+
+ Note that this function assumes PSTR->VALID_LEN elements are already
+ built and starts from PSTR->VALID_LEN. */
+
+static void
+build_wcs_buffer (pstr)
+ re_string_t *pstr;
+{
+ mbstate_t prev_st;
+ int byte_idx, end_idx, mbclen, remain_len;
+ /* Build the buffers from pstr->valid_len to either pstr->len or
+ pstr->bufs_len. */
+ end_idx = (pstr->bufs_len > pstr->len)? pstr->len : pstr->bufs_len;
+ for (byte_idx = pstr->valid_len; byte_idx < end_idx;)
+ {
+ wchar_t wc;
+ remain_len = end_idx - byte_idx;
+ prev_st = pstr->cur_state;
+ mbclen = mbrtowc (&wc, ((const char *) pstr->raw_mbs + pstr->raw_mbs_idx
+ + byte_idx), remain_len, &pstr->cur_state);
+ if (BE (mbclen == (size_t) -2, 0))
+ {
+ /* The buffer doesn't have enough space, finish to build. */
+ pstr->cur_state = prev_st;
+ break;
+ }
+ else if (BE (mbclen == (size_t) -1 || mbclen == 0, 0))
+ {
+ /* We treat these cases as a singlebyte character. */
+ mbclen = 1;
+ wc = (wchar_t) pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx];
+ pstr->cur_state = prev_st;
+ }
+
+ /* Apply the translateion if we need. */
+ if (pstr->trans != NULL && mbclen == 1)
+ {
+ int ch = pstr->trans[pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx]];
+ pstr->mbs_case[byte_idx] = ch;
+ }
+ /* Write wide character and padding. */
+ pstr->wcs[byte_idx++] = wc;
+ /* Write paddings. */
+ for (remain_len = byte_idx + mbclen - 1; byte_idx < remain_len ;)
+ pstr->wcs[byte_idx++] = WEOF;
+ }
+ pstr->valid_len = byte_idx;
+}
+
+/* Build wide character buffer PSTR->WCS like build_wcs_buffer,
+ but for REG_ICASE. */
+
+static void
+build_wcs_upper_buffer (pstr)
+ re_string_t *pstr;
+{
+ mbstate_t prev_st;
+ int byte_idx, end_idx, mbclen, remain_len;
+ /* Build the buffers from pstr->valid_len to either pstr->len or
+ pstr->bufs_len. */
+ end_idx = (pstr->bufs_len > pstr->len)? pstr->len : pstr->bufs_len;
+ for (byte_idx = pstr->valid_len; byte_idx < end_idx;)
+ {
+ wchar_t wc;
+ remain_len = end_idx - byte_idx;
+ prev_st = pstr->cur_state;
+ mbclen = mbrtowc (&wc, ((const char *) pstr->raw_mbs + pstr->raw_mbs_idx
+ + byte_idx), remain_len, &pstr->cur_state);
+ if (BE (mbclen == (size_t) -2, 0))
+ {
+ /* The buffer doesn't have enough space, finish to build. */
+ pstr->cur_state = prev_st;
+ break;
+ }
+ else if (mbclen == 1 || mbclen == (size_t) -1 || mbclen == 0)
+ {
+ /* In case of a singlebyte character. */
+ int ch = pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx];
+ /* Apply the translateion if we need. */
+ if (pstr->trans != NULL && mbclen == 1)
+ {
+ ch = pstr->trans[ch];
+ pstr->mbs_case[byte_idx] = ch;
+ }
+ pstr->wcs[byte_idx] = iswlower (wc) ? toupper (wc) : wc;
+ pstr->mbs[byte_idx++] = islower (ch) ? toupper (ch) : ch;
+ if (BE (mbclen == (size_t) -1, 0))
+ pstr->cur_state = prev_st;
+ }
+ else /* mbclen > 1 */
+ {
+ if (iswlower (wc))
+ wcrtomb ((char *) pstr->mbs + byte_idx, towupper (wc), &prev_st);
+ else
+ memcpy (pstr->mbs + byte_idx,
+ pstr->raw_mbs + pstr->raw_mbs_idx + byte_idx, mbclen);
+ pstr->wcs[byte_idx++] = iswlower (wc) ? toupper (wc) : wc;
+ /* Write paddings. */
+ for (remain_len = byte_idx + mbclen - 1; byte_idx < remain_len ;)
+ pstr->wcs[byte_idx++] = WEOF;
+ }
+ }
+ pstr->valid_len = byte_idx;
+}
+
+/* Skip characters until the index becomes greater than NEW_RAW_IDX.
+ Return the index. */
+
+static int
+re_string_skip_chars (pstr, new_raw_idx, last_wc)
+ re_string_t *pstr;
+ int new_raw_idx;
+ wint_t *last_wc;
+{
+ mbstate_t prev_st;
+ int rawbuf_idx, mbclen;
+ wchar_t wc = 0;
+
+ /* Skip the characters which are not necessary to check. */
+ for (rawbuf_idx = pstr->raw_mbs_idx + pstr->valid_len;
+ rawbuf_idx < new_raw_idx;)
+ {
+ int remain_len;
+ remain_len = pstr->len - rawbuf_idx;
+ prev_st = pstr->cur_state;
+ mbclen = mbrtowc (&wc, (const char *) pstr->raw_mbs + rawbuf_idx,
+ remain_len, &pstr->cur_state);
+ if (BE (mbclen == (size_t) -2 || mbclen == (size_t) -1 || mbclen == 0, 0))
+ {
+ /* We treat these cases as a singlebyte character. */
+ mbclen = 1;
+ pstr->cur_state = prev_st;
+ }
+ /* Then proceed the next character. */
+ rawbuf_idx += mbclen;
+ }
+ *last_wc = (wint_t) wc;
+ return rawbuf_idx;
+}
+#endif /* RE_ENABLE_I18N */
+
+/* Build the buffer PSTR->MBS, and apply the translation if we need.
+ This function is used in case of REG_ICASE. */
+
+static void
+build_upper_buffer (pstr)
+ re_string_t *pstr;
+{
+ int char_idx, end_idx;
+ end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len;
+
+ for (char_idx = pstr->valid_len; char_idx < end_idx; ++char_idx)
+ {
+ int ch = pstr->raw_mbs[pstr->raw_mbs_idx + char_idx];
+ if (pstr->trans != NULL)
+ {
+ ch = pstr->trans[ch];
+ pstr->mbs_case[char_idx] = ch;
+ }
+ if (islower (ch))
+ pstr->mbs[char_idx] = toupper (ch);
+ else
+ pstr->mbs[char_idx] = ch;
+ }
+ pstr->valid_len = char_idx;
+}
+
+/* Apply TRANS to the buffer in PSTR. */
+
+static void
+re_string_translate_buffer (pstr)
+ re_string_t *pstr;
+{
+ int buf_idx, end_idx;
+ end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len;
+
+ for (buf_idx = pstr->valid_len; buf_idx < end_idx; ++buf_idx)
+ {
+ int ch = pstr->raw_mbs[pstr->raw_mbs_idx + buf_idx];
+ pstr->mbs_case[buf_idx] = pstr->trans[ch];
+ }
+
+ pstr->valid_len = buf_idx;
+}
+
+/* This function re-construct the buffers.
+ Concretely, convert to wide character in case of MB_CUR_MAX > 1,
+ convert to upper case in case of REG_ICASE, apply translation. */
+
+static reg_errcode_t
+re_string_reconstruct (pstr, idx, eflags, newline)
+ re_string_t *pstr;
+ int idx, eflags, newline;
+{
+ int offset = idx - pstr->raw_mbs_idx;
+ if (offset < 0)
+ {
+ /* Reset buffer. */
+#ifdef RE_ENABLE_I18N
+ if (MB_CUR_MAX > 1)
+ memset (&pstr->cur_state, '\0', sizeof (mbstate_t));
+#endif /* RE_ENABLE_I18N */
+ pstr->len += pstr->raw_mbs_idx;
+ pstr->stop += pstr->raw_mbs_idx;
+ pstr->valid_len = pstr->raw_mbs_idx = 0;
+ pstr->tip_context = ((eflags & REG_NOTBOL) ? CONTEXT_BEGBUF
+ : CONTEXT_NEWLINE | CONTEXT_BEGBUF);
+ if (!MBS_CASE_ALLOCATED (pstr))
+ pstr->mbs_case = (unsigned char *) pstr->raw_mbs;
+ if (!MBS_ALLOCATED (pstr) && !MBS_CASE_ALLOCATED (pstr))
+ pstr->mbs = (unsigned char *) pstr->raw_mbs;
+ offset = idx;
+ }
+
+ if (offset != 0)
+ {
+ /* Are the characters which are already checked remain? */
+ if (offset < pstr->valid_len)
+ {
+ /* Yes, move them to the front of the buffer. */
+ pstr->tip_context = re_string_context_at (pstr, offset - 1, eflags,
+ newline);
+#ifdef RE_ENABLE_I18N
+ if (MB_CUR_MAX > 1)
+ memmove (pstr->wcs, pstr->wcs + offset,
+ (pstr->valid_len - offset) * sizeof (wint_t));
+#endif /* RE_ENABLE_I18N */
+ if (MBS_ALLOCATED (pstr))
+ memmove (pstr->mbs, pstr->mbs + offset,
+ pstr->valid_len - offset);
+ if (MBS_CASE_ALLOCATED (pstr))
+ memmove (pstr->mbs_case, pstr->mbs_case + offset,
+ pstr->valid_len - offset);
+ pstr->valid_len -= offset;
+#if DEBUG
+ assert (pstr->valid_len > 0);
+#endif
+ }
+ else
+ {
+ /* No, skip all characters until IDX. */
+ pstr->valid_len = 0;
+#ifdef RE_ENABLE_I18N
+ if (MB_CUR_MAX > 1)
+ {
+ int wcs_idx;
+ wint_t wc;
+ pstr->valid_len = re_string_skip_chars (pstr, idx, &wc) - idx;
+ for (wcs_idx = 0; wcs_idx < pstr->valid_len; ++wcs_idx)
+ pstr->wcs[wcs_idx] = WEOF;
+ if (pstr->trans && wc <= 0xff)
+ wc = pstr->trans[wc];
+ pstr->tip_context = (IS_WIDE_WORD_CHAR (wc) ? CONTEXT_WORD
+ : ((newline && IS_WIDE_NEWLINE (wc))
+ ? CONTEXT_NEWLINE : 0));
+ }
+ else
+#endif /* RE_ENABLE_I18N */
+ {
+ int c = pstr->raw_mbs[pstr->raw_mbs_idx + offset - 1];
+ if (pstr->trans)
+ c = pstr->trans[c];
+ pstr->tip_context = (IS_WORD_CHAR (c) ? CONTEXT_WORD
+ : ((newline && IS_NEWLINE (c))
+ ? CONTEXT_NEWLINE : 0));
+ }
+ }
+ if (!MBS_CASE_ALLOCATED (pstr))
+ {
+ pstr->mbs_case += offset;
+ /* In case of !MBS_ALLOCATED && !MBS_CASE_ALLOCATED. */
+ if (!MBS_ALLOCATED (pstr))
+ pstr->mbs += offset;
+ }
+ }
+ pstr->raw_mbs_idx = idx;
+ pstr->len -= offset;
+ pstr->stop -= offset;
+
+ /* Then build the buffers. */
+#ifdef RE_ENABLE_I18N
+ if (MB_CUR_MAX > 1)
+ {
+ if (pstr->icase)
+ build_wcs_upper_buffer (pstr);
+ else
+ build_wcs_buffer (pstr);
+ }
+ else
+#endif /* RE_ENABLE_I18N */
+ {
+ if (pstr->icase)
+ build_upper_buffer (pstr);
+ else if (pstr->trans != NULL)
+ re_string_translate_buffer (pstr);
+ }
+ pstr->cur_idx = 0;
+
+ return REG_NOERROR;
+}
+
+static void
+re_string_destruct (pstr)
+ re_string_t *pstr;
+{
+#ifdef RE_ENABLE_I18N
+ re_free (pstr->wcs);
+#endif /* RE_ENABLE_I18N */
+ if (MBS_ALLOCATED (pstr))
+ re_free (pstr->mbs);
+ if (MBS_CASE_ALLOCATED (pstr))
+ re_free (pstr->mbs_case);
+}
+
+/* Return the context at IDX in INPUT. */
+
+static unsigned int
+re_string_context_at (input, idx, eflags, newline_anchor)
+ const re_string_t *input;
+ int idx, eflags, newline_anchor;
+{
+ int c;
+ if (idx < 0 || idx == input->len)
+ {
+ if (idx < 0)
+ /* In this case, we use the value stored in input->tip_context,
+ since we can't know the character in input->mbs[-1] here. */
+ return input->tip_context;
+ else /* (idx == input->len) */
+ return ((eflags & REG_NOTEOL) ? CONTEXT_ENDBUF
+ : CONTEXT_NEWLINE | CONTEXT_ENDBUF);
+ }
+#ifdef RE_ENABLE_I18N
+ if (MB_CUR_MAX > 1)
+ {
+ wint_t wc;
+ int wc_idx = idx;
+ while(input->wcs[wc_idx] == WEOF)
+ {
+#ifdef DEBUG
+ /* It must not happen. */
+ assert (wc_idx >= 0);
+#endif
+ --wc_idx;
+ if (wc_idx < 0)
+ return input->tip_context;
+ }
+ wc = input->wcs[wc_idx];
+ if (IS_WIDE_WORD_CHAR (wc))
+ return CONTEXT_WORD;
+ return (newline_anchor && IS_WIDE_NEWLINE (wc)) ? CONTEXT_NEWLINE : 0;
+ }
+ else
+#endif
+ {
+ c = re_string_byte_at (input, idx);
+ if (IS_WORD_CHAR (c))
+ return CONTEXT_WORD;
+ return (newline_anchor && IS_NEWLINE (c)) ? CONTEXT_NEWLINE : 0;
+ }
+}
+\f
+/* Functions for set operation. */
+
+static reg_errcode_t
+re_node_set_alloc (set, size)
+ re_node_set *set;
+ int size;
+{
+ set->alloc = size;
+ set->nelem = 0;
+ set->elems = re_malloc (int, size);
+ if (BE (set->elems == NULL, 0))
+ return REG_ESPACE;
+ return REG_NOERROR;
+}
+
+static reg_errcode_t
+re_node_set_init_1 (set, elem)
+ re_node_set *set;
+ int elem;
+{
+ set->alloc = 1;
+ set->nelem = 1;
+ set->elems = re_malloc (int, 1);
+ if (BE (set->elems == NULL, 0))
+ {
+ set->alloc = set->nelem = 0;
+ return REG_ESPACE;
+ }
+ set->elems[0] = elem;
+ return REG_NOERROR;
+}
+
+static reg_errcode_t
+re_node_set_init_2 (set, elem1, elem2)
+ re_node_set *set;
+ int elem1, elem2;
+{
+ set->alloc = 2;
+ set->elems = re_malloc (int, 2);
+ if (BE (set->elems == NULL, 0))
+ return REG_ESPACE;
+ if (elem1 == elem2)
+ {
+ set->nelem = 1;
+ set->elems[0] = elem1;
+ }
+ else
+ {
+ set->nelem = 2;
+ if (elem1 < elem2)
+ {
+ set->elems[0] = elem1;
+ set->elems[1] = elem2;
+ }
+ else
+ {
+ set->elems[0] = elem2;
+ set->elems[1] = elem1;
+ }
+ }
+ return REG_NOERROR;
+}
+
+static reg_errcode_t
+re_node_set_init_copy (dest, src)
+ re_node_set *dest;
+ const re_node_set *src;
+{
+ dest->nelem = src->nelem;
+ if (src->nelem > 0)
+ {
+ dest->alloc = dest->nelem;
+ dest->elems = re_malloc (int, dest->alloc);
+ if (BE (dest->elems == NULL, 0))
+ {
+ dest->alloc = dest->nelem = 0;
+ return REG_ESPACE;
+ }
+ memcpy (dest->elems, src->elems, src->nelem * sizeof (int));
+ }
+ else
+ re_node_set_init_empty (dest);
+ return REG_NOERROR;
+}
+
+/* Calculate the intersection of the sets SRC1 and SRC2. And merge it to
+ DEST. Return value indicate the error code or REG_NOERROR if succeeded.
+ Note: We assume dest->elems is NULL, when dest->alloc is 0. */
+
+static reg_errcode_t
+re_node_set_add_intersect (dest, src1, src2)
+ re_node_set *dest;
+ const re_node_set *src1, *src2;
+{
+ int i1, i2, id;
+ if (src1->nelem > 0 && src2->nelem > 0)
+ {
+ if (src1->nelem + src2->nelem + dest->nelem > dest->alloc)
+ {
+ dest->alloc = src1->nelem + src2->nelem + dest->nelem;
+ dest->elems = re_realloc (dest->elems, int, dest->alloc);
+ if (BE (dest->elems == NULL, 0))
+ return REG_ESPACE;
+ }
+ }
+ else
+ return REG_NOERROR;
+
+ for (i1 = i2 = id = 0 ; i1 < src1->nelem && i2 < src2->nelem ;)
+ {
+ if (src1->elems[i1] > src2->elems[i2])
+ {
+ ++i2;
+ continue;
+ }
+ if (src1->elems[i1] == src2->elems[i2])
+ {
+ while (id < dest->nelem && dest->elems[id] < src2->elems[i2])
+ ++id;
+ if (id < dest->nelem && dest->elems[id] == src2->elems[i2])
+ ++id;
+ else
+ {
+ memmove (dest->elems + id + 1, dest->elems + id,
+ sizeof (int) * (dest->nelem - id));
+ dest->elems[id++] = src2->elems[i2++];
+ ++dest->nelem;
+ }
+ }
+ ++i1;
+ }
+ return REG_NOERROR;
+}
+
+/* Calculate the union set of the sets SRC1 and SRC2. And store it to
+ DEST. Return value indicate the error code or REG_NOERROR if succeeded. */
+
+static reg_errcode_t
+re_node_set_init_union (dest, src1, src2)
+ re_node_set *dest;
+ const re_node_set *src1, *src2;
+{
+ int i1, i2, id;
+ if (src1 != NULL && src1->nelem > 0 && src2 != NULL && src2->nelem > 0)
+ {
+ dest->alloc = src1->nelem + src2->nelem;
+ dest->elems = re_malloc (int, dest->alloc);
+ if (BE (dest->elems == NULL, 0))
+ return REG_ESPACE;
+ }
+ else
+ {
+ if (src1 != NULL && src1->nelem > 0)
+ return re_node_set_init_copy (dest, src1);
+ else if (src2 != NULL && src2->nelem > 0)
+ return re_node_set_init_copy (dest, src2);
+ else
+ re_node_set_init_empty (dest);
+ return REG_NOERROR;
+ }
+ for (i1 = i2 = id = 0 ; i1 < src1->nelem && i2 < src2->nelem ;)
+ {
+ if (src1->elems[i1] > src2->elems[i2])
+ {
+ dest->elems[id++] = src2->elems[i2++];
+ continue;
+ }
+ if (src1->elems[i1] == src2->elems[i2])
+ ++i2;
+ dest->elems[id++] = src1->elems[i1++];
+ }
+ if (i1 < src1->nelem)
+ {
+ memcpy (dest->elems + id, src1->elems + i1,
+ (src1->nelem - i1) * sizeof (int));
+ id += src1->nelem - i1;
+ }
+ else if (i2 < src2->nelem)
+ {
+ memcpy (dest->elems + id, src2->elems + i2,
+ (src2->nelem - i2) * sizeof (int));
+ id += src2->nelem - i2;
+ }
+ dest->nelem = id;
+ return REG_NOERROR;
+}
+
+/* Calculate the union set of the sets DEST and SRC. And store it to
+ DEST. Return value indicate the error code or REG_NOERROR if succeeded. */
+
+static reg_errcode_t
+re_node_set_merge (dest, src)
+ re_node_set *dest;
+ const re_node_set *src;
+{
+ int si, di;
+ if (src == NULL || src->nelem == 0)
+ return REG_NOERROR;
+ if (dest->alloc < src->nelem + dest->nelem)
+ {
+ int *new_buffer;
+ dest->alloc = 2 * (src->nelem + dest->alloc);
+ new_buffer = re_realloc (dest->elems, int, dest->alloc);
+ if (BE (new_buffer == NULL, 0))
+ return REG_ESPACE;
+ dest->elems = new_buffer;
+ }
+
+ for (si = 0, di = 0 ; si < src->nelem && di < dest->nelem ;)
+ {
+ int cp_from, ncp, mid, right, src_elem = src->elems[si];
+ /* Binary search the spot we will add the new element. */
+ right = dest->nelem;
+ while (di < right)
+ {
+ mid = (di + right) / 2;
+ if (dest->elems[mid] < src_elem)
+ di = mid + 1;
+ else
+ right = mid;
+ }
+ if (di >= dest->nelem)
+ break;
+
+ if (dest->elems[di] == src_elem)
+ {
+ /* Skip since, DEST already has the element. */
+ ++di;
+ ++si;
+ continue;
+ }
+
+ /* Skip the src elements which are less than dest->elems[di]. */
+ cp_from = si;
+ while (si < src->nelem && src->elems[si] < dest->elems[di])
+ ++si;
+ /* Copy these src elements. */
+ ncp = si - cp_from;
+ memmove (dest->elems + di + ncp, dest->elems + di,
+ sizeof (int) * (dest->nelem - di));
+ memcpy (dest->elems + di, src->elems + cp_from,
+ sizeof (int) * ncp);
+ /* Update counters. */
+ di += ncp;
+ dest->nelem += ncp;
+ }
+
+ /* Copy remaining src elements. */
+ if (si < src->nelem)
+ {
+ memcpy (dest->elems + di, src->elems + si,
+ sizeof (int) * (src->nelem - si));
+ dest->nelem += src->nelem - si;
+ }
+ return REG_NOERROR;
+}
+
+/* Insert the new element ELEM to the re_node_set* SET.
+ return 0 if SET already has ELEM,
+ return -1 if an error is occured, return 1 otherwise. */
+
+static int
+re_node_set_insert (set, elem)
+ re_node_set *set;
+ int elem;
+{
+ int idx, right, mid;
+ /* In case of the set is empty. */
+ if (set->elems == NULL || set->alloc == 0)
+ {
+ if (BE (re_node_set_init_1 (set, elem) == REG_NOERROR, 1))
+ return 1;
+ else
+ return -1;
+ }
+
+ /* Binary search the spot we will add the new element. */
+ idx = 0;
+ right = set->nelem;
+ while (idx < right)
+ {
+ mid = (idx + right) / 2;
+ if (set->elems[mid] < elem)
+ idx = mid + 1;
+ else
+ right = mid;
+ }
+
+ /* Realloc if we need. */
+ if (set->alloc < set->nelem + 1)
+ {
+ int *new_array;
+ set->alloc = set->alloc * 2;
+ new_array = re_malloc (int, set->alloc);
+ if (BE (new_array == NULL, 0))
+ return -1;
+ /* Copy the elements they are followed by the new element. */
+ if (idx > 0)
+ memcpy (new_array, set->elems, sizeof (int) * (idx));
+ /* Copy the elements which follows the new element. */
+ if (set->nelem - idx > 0)
+ memcpy (new_array + idx + 1, set->elems + idx,
+ sizeof (int) * (set->nelem - idx));
+ re_free (set->elems);
+ set->elems = new_array;
+ }
+ else
+ {
+ /* Move the elements which follows the new element. */
+ if (set->nelem - idx > 0)
+ memmove (set->elems + idx + 1, set->elems + idx,
+ sizeof (int) * (set->nelem - idx));
+ }
+ /* Insert the new element. */
+ set->elems[idx] = elem;
+ ++set->nelem;
+ return 1;
+}
+
+/* Compare two node sets SET1 and SET2.
+ return 1 if SET1 and SET2 are equivalent, retrun 0 otherwise. */
+
+static int
+re_node_set_compare (set1, set2)
+ const re_node_set *set1, *set2;
+{
+ int i;
+ if (set1 == NULL || set2 == NULL || set1->nelem != set2->nelem)
+ return 0;
+ for (i = 0 ; i < set1->nelem ; i++)
+ if (set1->elems[i] != set2->elems[i])
+ return 0;
+ return 1;
+}
+
+/* Return (idx + 1) if SET contains the element ELEM, return 0 otherwise. */
+
+static int
+re_node_set_contains (set, elem)
+ const re_node_set *set;
+ int elem;
+{
+ int idx, right, mid;
+ if (set->nelem <= 0)
+ return 0;
+
+ /* Binary search the element. */
+ idx = 0;
+ right = set->nelem - 1;
+ while (idx < right)
+ {
+ mid = (idx + right) / 2;
+ if (set->elems[mid] < elem)
+ idx = mid + 1;
+ else
+ right = mid;
+ }
+ return set->elems[idx] == elem ? idx + 1 : 0;
+}
+
+static void
+re_node_set_remove_at (set, idx)
+ re_node_set *set;
+ int idx;
+{
+ if (idx < 0 || idx >= set->nelem)
+ return;
+ if (idx < set->nelem - 1)
+ memmove (set->elems + idx, set->elems + idx + 1,
+ sizeof (int) * (set->nelem - idx - 1));
+ --set->nelem;
+}
+\f
+
+/* Add the token TOKEN to dfa->nodes, and return the index of the token.
+ Or return -1, if an error will be occured. */
+
+static int
+re_dfa_add_node (dfa, token, mode)
+ re_dfa_t *dfa;
+ re_token_t token;
+ int mode;
+{
+ if (dfa->nodes_len >= dfa->nodes_alloc)
+ {
+ re_token_t *new_array;
+ dfa->nodes_alloc *= 2;
+ new_array = re_realloc (dfa->nodes, re_token_t, dfa->nodes_alloc);
+ if (BE (new_array == NULL, 0))
+ return -1;
+ else
+ dfa->nodes = new_array;
+ if (mode)
+ {
+ int *new_nexts, *new_indices;
+ re_node_set *new_edests, *new_eclosures, *new_inveclosures;
+
+ new_nexts = re_realloc (dfa->nexts, int, dfa->nodes_alloc);
+ new_indices = re_realloc (dfa->org_indices, int, dfa->nodes_alloc);
+ new_edests = re_realloc (dfa->edests, re_node_set, dfa->nodes_alloc);
+ new_eclosures = re_realloc (dfa->eclosures, re_node_set,
+ dfa->nodes_alloc);
+ new_inveclosures = re_realloc (dfa->inveclosures, re_node_set,
+ dfa->nodes_alloc);
+ if (BE (new_nexts == NULL || new_indices == NULL
+ || new_edests == NULL || new_eclosures == NULL
+ || new_inveclosures == NULL, 0))
+ return -1;
+ dfa->nexts = new_nexts;
+ dfa->org_indices = new_indices;
+ dfa->edests = new_edests;
+ dfa->eclosures = new_eclosures;
+ dfa->inveclosures = new_inveclosures;
+ }
+ }
+ dfa->nodes[dfa->nodes_len] = token;
+ dfa->nodes[dfa->nodes_len].duplicated = 0;
+ dfa->nodes[dfa->nodes_len].constraint = 0;
+ return dfa->nodes_len++;
+}
+
+static unsigned int inline
+calc_state_hash (nodes, context)
+ const re_node_set *nodes;
+ unsigned int context;
+{
+ unsigned int hash = nodes->nelem + context;
+ int i;
+ for (i = 0 ; i < nodes->nelem ; i++)
+ hash += nodes->elems[i];
+ return hash;
+}
+
+/* Search for the state whose node_set is equivalent to NODES.
+ Return the pointer to the state, if we found it in the DFA.
+ Otherwise create the new one and return it. In case of an error
+ return NULL and set the error code in ERR.
+ Note: - We assume NULL as the invalid state, then it is possible that
+ return value is NULL and ERR is REG_NOERROR.
+ - We never return non-NULL value in case of any errors, it is for
+ optimization. */
+
+static re_dfastate_t*
+re_acquire_state (err, dfa, nodes)
+ reg_errcode_t *err;
+ re_dfa_t *dfa;
+ const re_node_set *nodes;
+{
+ unsigned int hash;
+ re_dfastate_t *new_state;
+ struct re_state_table_entry *spot;
+ int i;
+ if (BE (nodes->nelem == 0, 0))
+ {
+ *err = REG_NOERROR;
+ return NULL;
+ }
+ hash = calc_state_hash (nodes, 0);
+ spot = dfa->state_table + (hash & dfa->state_hash_mask);
+
+ for (i = 0 ; i < spot->num ; i++)
+ {
+ re_dfastate_t *state = spot->array[i];
+ if (hash != state->hash)
+ continue;
+ if (re_node_set_compare (&state->nodes, nodes))
+ return state;
+ }
+
+ /* There are no appropriate state in the dfa, create the new one. */
+ new_state = create_ci_newstate (dfa, nodes, hash);
+ if (BE (new_state != NULL, 1))
+ return new_state;
+ else
+ {
+ *err = REG_ESPACE;
+ return NULL;
+ }
+}
+
+/* Search for the state whose node_set is equivalent to NODES and
+ whose context is equivalent to CONTEXT.
+ Return the pointer to the state, if we found it in the DFA.
+ Otherwise create the new one and return it. In case of an error
+ return NULL and set the error code in ERR.
+ Note: - We assume NULL as the invalid state, then it is possible that
+ return value is NULL and ERR is REG_NOERROR.
+ - We never return non-NULL value in case of any errors, it is for
+ optimization. */
+
+static re_dfastate_t*
+re_acquire_state_context (err, dfa, nodes, context)
+ reg_errcode_t *err;
+ re_dfa_t *dfa;
+ const re_node_set *nodes;
+ unsigned int context;
+{
+ unsigned int hash;
+ re_dfastate_t *new_state;
+ struct re_state_table_entry *spot;
+ int i;
+ if (nodes->nelem == 0)
+ {
+ *err = REG_NOERROR;
+ return NULL;
+ }
+ hash = calc_state_hash (nodes, context);
+ spot = dfa->state_table + (hash & dfa->state_hash_mask);
+
+ for (i = 0 ; i < spot->num ; i++)
+ {
+ re_dfastate_t *state = spot->array[i];
+ if (hash != state->hash)
+ continue;
+ if (re_node_set_compare (state->entrance_nodes, nodes)
+ && state->context == context)
+ return state;
+ }
+ /* There are no appropriate state in `dfa', create the new one. */
+ new_state = create_cd_newstate (dfa, nodes, context, hash);
+ if (BE (new_state != NULL, 1))
+ return new_state;
+ else
+ {
+ *err = REG_ESPACE;
+ return NULL;
+ }
+}
+
+/* Allocate memory for DFA state and initialize common properties.
+ Return the new state if succeeded, otherwise return NULL. */
+
+static re_dfastate_t *
+create_newstate_common (dfa, nodes, hash)
+ re_dfa_t *dfa;
+ const re_node_set *nodes;
+ unsigned int hash;
+{
+ re_dfastate_t *newstate;
+ reg_errcode_t err;
+ newstate = (re_dfastate_t *) calloc (sizeof (re_dfastate_t), 1);
+ if (BE (newstate == NULL, 0))
+ return NULL;
+ err = re_node_set_init_copy (&newstate->nodes, nodes);
+ if (BE (err != REG_NOERROR, 0))
+ {
+ re_free (newstate);
+ return NULL;
+ }
+ newstate->trtable = NULL;
+ newstate->trtable_search = NULL;
+ newstate->hash = hash;
+ return newstate;
+}
+
+/* Store the new state NEWSTATE whose hash value is HASH in appropriate
+ position. Return value indicate the error code if failed. */
+
+static reg_errcode_t
+register_state (dfa, newstate, hash)
+ re_dfa_t *dfa;
+ re_dfastate_t *newstate;
+ unsigned int hash;
+{
+ struct re_state_table_entry *spot;
+ spot = dfa->state_table + (hash & dfa->state_hash_mask);
+
+ if (spot->alloc <= spot->num)
+ {
+ re_dfastate_t **new_array;
+ spot->alloc = 2 * spot->num + 2;
+ new_array = re_realloc (spot->array, re_dfastate_t *, spot->alloc);
+ if (BE (new_array == NULL, 0))
+ return REG_ESPACE;
+ spot->array = new_array;
+ }
+ spot->array[spot->num++] = newstate;
+ return REG_NOERROR;
+}
+
+/* Create the new state which is independ of contexts.
+ Return the new state if succeeded, otherwise return NULL. */
+
+static re_dfastate_t *
+create_ci_newstate (dfa, nodes, hash)
+ re_dfa_t *dfa;
+ const re_node_set *nodes;
+ unsigned int hash;
+{
+ int i;
+ reg_errcode_t err;
+ re_dfastate_t *newstate;
+ newstate = create_newstate_common (dfa, nodes, hash);
+ if (BE (newstate == NULL, 0))
+ return NULL;
+ newstate->entrance_nodes = &newstate->nodes;
+
+ for (i = 0 ; i < nodes->nelem ; i++)
+ {
+ re_token_t *node = dfa->nodes + nodes->elems[i];
+ re_token_type_t type = node->type;
+ if (type == CHARACTER && !node->constraint)
+ continue;
+
+ /* If the state has the halt node, the state is a halt state. */
+ else if (type == END_OF_RE)
+ newstate->halt = 1;
+#ifdef RE_ENABLE_I18N
+ else if (type == COMPLEX_BRACKET
+ || (type == OP_PERIOD && MB_CUR_MAX > 1))
+ newstate->accept_mb = 1;
+#endif /* RE_ENABLE_I18N */
+ else if (type == OP_BACK_REF)
+ newstate->has_backref = 1;
+ else if (type == ANCHOR || node->constraint)
+ newstate->has_constraint = 1;
+ }
+ err = register_state (dfa, newstate, hash);
+ if (BE (err != REG_NOERROR, 0))
+ {
+ free_state (newstate);
+ newstate = NULL;
+ }
+ return newstate;
+}
+
+/* Create the new state which is depend on the context CONTEXT.
+ Return the new state if succeeded, otherwise return NULL. */
+
+static re_dfastate_t *
+create_cd_newstate (dfa, nodes, context, hash)
+ re_dfa_t *dfa;
+ const re_node_set *nodes;
+ unsigned int context, hash;
+{
+ int i, nctx_nodes = 0;
+ reg_errcode_t err;
+ re_dfastate_t *newstate;
+
+ newstate = create_newstate_common (dfa, nodes, hash);
+ if (BE (newstate == NULL, 0))
+ return NULL;
+ newstate->context = context;
+ newstate->entrance_nodes = &newstate->nodes;
+
+ for (i = 0 ; i < nodes->nelem ; i++)
+ {
+ unsigned int constraint = 0;
+ re_token_t *node = dfa->nodes + nodes->elems[i];
+ re_token_type_t type = node->type;
+ if (node->constraint)
+ constraint = node->constraint;
+
+ if (type == CHARACTER && !constraint)
+ continue;
+ /* If the state has the halt node, the state is a halt state. */
+ else if (type == END_OF_RE)
+ newstate->halt = 1;
+#ifdef RE_ENABLE_I18N
+ else if (type == COMPLEX_BRACKET
+ || (type == OP_PERIOD && MB_CUR_MAX > 1))
+ newstate->accept_mb = 1;
+#endif /* RE_ENABLE_I18N */
+ else if (type == OP_BACK_REF)
+ newstate->has_backref = 1;
+ else if (type == ANCHOR)
+ constraint = node->opr.ctx_type;
+
+ if (constraint)
+ {
+ if (newstate->entrance_nodes == &newstate->nodes)
+ {
+ newstate->entrance_nodes = re_malloc (re_node_set, 1);
+ if (BE (newstate->entrance_nodes == NULL, 0))
+ {
+ free_state (newstate);
+ return NULL;
+ }
+ re_node_set_init_copy (newstate->entrance_nodes, nodes);
+ nctx_nodes = 0;
+ newstate->has_constraint = 1;
+ }
+
+ if (NOT_SATISFY_PREV_CONSTRAINT (constraint,context))
+ {
+ re_node_set_remove_at (&newstate->nodes, i - nctx_nodes);
+ ++nctx_nodes;
+ }
+ }
+ }
+ err = register_state (dfa, newstate, hash);
+ if (BE (err != REG_NOERROR, 0))
+ {
+ free_state (newstate);
+ newstate = NULL;
+ }
+ return newstate;
+}
+
+static void
+free_state (state)
+ re_dfastate_t *state;
+{
+ if (state->entrance_nodes != &state->nodes)
+ {
+ re_node_set_free (state->entrance_nodes);
+ re_free (state->entrance_nodes);
+ }
+ re_node_set_free (&state->nodes);
+ re_free (state->trtable);
+ re_free (state->trtable_search);
+ re_free (state);
+}
--- /dev/null
+/* Extended regular expression matching and search library.
+ Copyright (C) 2002, 2003 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+ Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, write to the Free
+ Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ 02111-1307 USA. */
+
+#ifndef _REGEX_INTERNAL_H
+#define _REGEX_INTERNAL_H 1
+
+#ifdef HAVE_CONFIG_H
+#include "config.h"
+#endif
+
+#include <assert.h>
+#include <ctype.h>
+#include <limits.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+#if defined HAVE_LOCALE_H || defined _LIBC
+# include <locale.h>
+#endif
+#if defined HAVE_WCHAR_H || defined _LIBC
+# include <wchar.h>
+#endif /* HAVE_WCHAR_H || _LIBC */
+#if defined HAVE_WCTYPE_H || defined _LIBC
+# include <wctype.h>
+#endif /* HAVE_WCTYPE_H || _LIBC */
+
+/* In case that the system doesn't have isblank(). */
+#if !defined _LIBC && !defined HAVE_ISBLANK && !defined isblank
+# define isblank(ch) ((ch) == ' ' || (ch) == '\t')
+#endif
+
+#ifdef _LIBC
+# ifndef _RE_DEFINE_LOCALE_FUNCTIONS
+# define _RE_DEFINE_LOCALE_FUNCTIONS 1
+# include <locale/localeinfo.h>
+# include <locale/elem-hash.h>
+# include <locale/coll-lookup.h>
+# endif
+#endif
+
+/* This is for other GNU distributions with internationalized messages. */
+#if HAVE_LIBINTL_H || defined _LIBC
+# include <libintl.h>
+# ifdef _LIBC
+# undef gettext
+# define gettext(msgid) \
+ INTUSE(__dcgettext) (INTUSE(_libc_intl_domainname), msgid, LC_MESSAGES)
+# endif
+#else
+# define gettext(msgid) (msgid)
+#endif
+
+#ifndef gettext_noop
+/* This define is so xgettext can find the internationalizable
+ strings. */
+# define gettext_noop(String) String
+#endif
+
+#if (defined MB_CUR_MAX && HAVE_LOCALE_H && HAVE_WCTYPE_H && HAVE_WCHAR_H && HAVE_WCRTOMB && HAVE_MBRTOWC && HAVE_WCSCOLL) || _LIBC
+# define RE_ENABLE_I18N
+#endif
+
+#if __GNUC__ >= 3
+# define BE(expr, val) __builtin_expect (expr, val)
+#else
+# define BE(expr, val) (expr)
+# define inline
+#endif
+
+/* Number of bits in a byte. */
+#define BYTE_BITS 8
+/* Number of single byte character. */
+#define SBC_MAX 256
+
+#define COLL_ELEM_LEN_MAX 8
+
+/* The character which represents newline. */
+#define NEWLINE_CHAR '\n'
+#define WIDE_NEWLINE_CHAR L'\n'
+
+/* Rename to standard API for using out of glibc. */
+#ifndef _LIBC
+# define __wctype wctype
+# define __iswctype iswctype
+# define __btowc btowc
+# define __mempcpy mempcpy
+# define __wcrtomb wcrtomb
+# define attribute_hidden
+#endif /* not _LIBC */
+
+extern const char __re_error_msgid[] attribute_hidden;
+extern const size_t __re_error_msgid_idx[] attribute_hidden;
+
+/* Number of bits in an unsinged int. */
+#define UINT_BITS (sizeof (unsigned int) * BYTE_BITS)
+/* Number of unsigned int in an bit_set. */
+#define BITSET_UINTS ((SBC_MAX + UINT_BITS - 1) / UINT_BITS)
+typedef unsigned int bitset[BITSET_UINTS];
+typedef unsigned int *re_bitset_ptr_t;
+
+#define bitset_set(set,i) (set[i / UINT_BITS] |= 1 << i % UINT_BITS)
+#define bitset_clear(set,i) (set[i / UINT_BITS] &= ~(1 << i % UINT_BITS))
+#define bitset_contain(set,i) (set[i / UINT_BITS] & (1 << i % UINT_BITS))
+#define bitset_empty(set) memset (set, 0, sizeof (unsigned int) * BITSET_UINTS)
+#define bitset_set_all(set) \
+ memset (set, 255, sizeof (unsigned int) * BITSET_UINTS)
+#define bitset_copy(dest,src) \
+ memcpy (dest, src, sizeof (unsigned int) * BITSET_UINTS)
+static inline void bitset_not (bitset set);
+static inline void bitset_merge (bitset dest, const bitset src);
+static inline void bitset_not_merge (bitset dest, const bitset src);
+
+#define PREV_WORD_CONSTRAINT 0x0001
+#define PREV_NOTWORD_CONSTRAINT 0x0002
+#define NEXT_WORD_CONSTRAINT 0x0004
+#define NEXT_NOTWORD_CONSTRAINT 0x0008
+#define PREV_NEWLINE_CONSTRAINT 0x0010
+#define NEXT_NEWLINE_CONSTRAINT 0x0020
+#define PREV_BEGBUF_CONSTRAINT 0x0040
+#define NEXT_ENDBUF_CONSTRAINT 0x0080
+#define DUMMY_CONSTRAINT 0x0100
+
+typedef enum
+{
+ INSIDE_WORD = PREV_WORD_CONSTRAINT | NEXT_WORD_CONSTRAINT,
+ WORD_FIRST = PREV_NOTWORD_CONSTRAINT | NEXT_WORD_CONSTRAINT,
+ WORD_LAST = PREV_WORD_CONSTRAINT | NEXT_NOTWORD_CONSTRAINT,
+ LINE_FIRST = PREV_NEWLINE_CONSTRAINT,
+ LINE_LAST = NEXT_NEWLINE_CONSTRAINT,
+ BUF_FIRST = PREV_BEGBUF_CONSTRAINT,
+ BUF_LAST = NEXT_ENDBUF_CONSTRAINT,
+ WORD_DELIM = DUMMY_CONSTRAINT
+} re_context_type;
+
+typedef struct
+{
+ int alloc;
+ int nelem;
+ int *elems;
+} re_node_set;
+
+typedef enum
+{
+ NON_TYPE = 0,
+
+ /* Token type, these are used only by token. */
+ OP_OPEN_BRACKET,
+ OP_CLOSE_BRACKET,
+ OP_CHARSET_RANGE,
+ OP_OPEN_DUP_NUM,
+ OP_CLOSE_DUP_NUM,
+ OP_NON_MATCH_LIST,
+ OP_OPEN_COLL_ELEM,
+ OP_CLOSE_COLL_ELEM,
+ OP_OPEN_EQUIV_CLASS,
+ OP_CLOSE_EQUIV_CLASS,
+ OP_OPEN_CHAR_CLASS,
+ OP_CLOSE_CHAR_CLASS,
+ OP_WORD,
+ OP_NOTWORD,
+ BACK_SLASH,
+
+ /* Tree type, these are used only by tree. */
+ CONCAT,
+ ALT,
+ SUBEXP,
+ SIMPLE_BRACKET,
+#ifdef RE_ENABLE_I18N
+ COMPLEX_BRACKET,
+#endif /* RE_ENABLE_I18N */
+
+ /* Node type, These are used by token, node, tree. */
+ OP_OPEN_SUBEXP,
+ OP_CLOSE_SUBEXP,
+ OP_PERIOD,
+ CHARACTER,
+ END_OF_RE,
+ OP_ALT,
+ OP_DUP_ASTERISK,
+ OP_DUP_PLUS,
+ OP_DUP_QUESTION,
+ OP_BACK_REF,
+ ANCHOR,
+
+ /* Dummy marker. */
+ END_OF_RE_TOKEN_T
+} re_token_type_t;
+
+#ifdef RE_ENABLE_I18N
+typedef struct
+{
+ /* Multibyte characters. */
+ wchar_t *mbchars;
+
+ /* Collating symbols. */
+# ifdef _LIBC
+ int32_t *coll_syms;
+# endif
+
+ /* Equivalence classes. */
+# ifdef _LIBC
+ int32_t *equiv_classes;
+# endif
+
+ /* Range expressions. */
+# ifdef _LIBC
+ uint32_t *range_starts;
+ uint32_t *range_ends;
+# else /* not _LIBC */
+ wchar_t *range_starts;
+ wchar_t *range_ends;
+# endif /* not _LIBC */
+
+ /* Character classes. */
+ wctype_t *char_classes;
+
+ /* If this character set is the non-matching list. */
+ unsigned int non_match : 1;
+
+ /* # of multibyte characters. */
+ int nmbchars;
+
+ /* # of collating symbols. */
+ int ncoll_syms;
+
+ /* # of equivalence classes. */
+ int nequiv_classes;
+
+ /* # of range expressions. */
+ int nranges;
+
+ /* # of character classes. */
+ int nchar_classes;
+} re_charset_t;
+#endif /* RE_ENABLE_I18N */
+
+typedef struct
+{
+ union
+ {
+ unsigned char c; /* for CHARACTER */
+ re_bitset_ptr_t sbcset; /* for SIMPLE_BRACKET */
+#ifdef RE_ENABLE_I18N
+ re_charset_t *mbcset; /* for COMPLEX_BRACKET */
+#endif /* RE_ENABLE_I18N */
+ int idx; /* for BACK_REF */
+ re_context_type ctx_type; /* for ANCHOR */
+ } opr;
+#if __GNUC__ >= 2
+ re_token_type_t type : 8;
+#else
+ re_token_type_t type;
+#endif
+ unsigned int constraint : 10; /* context constraint */
+ unsigned int duplicated : 1;
+#ifdef RE_ENABLE_I18N
+ unsigned int mb_partial : 1;
+#endif
+} re_token_t;
+
+#define IS_EPSILON_NODE(type) \
+ ((type) == OP_ALT || (type) == OP_DUP_ASTERISK || (type) == OP_DUP_PLUS \
+ || (type) == OP_DUP_QUESTION || (type) == ANCHOR \
+ || (type) == OP_OPEN_SUBEXP || (type) == OP_CLOSE_SUBEXP)
+
+#define ACCEPT_MB_NODE(type) \
+ ((type) == COMPLEX_BRACKET || (type) == OP_PERIOD)
+
+struct re_string_t
+{
+ /* Indicate the raw buffer which is the original string passed as an
+ argument of regexec(), re_search(), etc.. */
+ const unsigned char *raw_mbs;
+ /* Store the multibyte string. In case of "case insensitive mode" like
+ REG_ICASE, upper cases of the string are stored, otherwise MBS points
+ the same address that RAW_MBS points. */
+ unsigned char *mbs;
+ /* Store the case sensitive multibyte string. In case of
+ "case insensitive mode", the original string are stored,
+ otherwise MBS_CASE points the same address that MBS points. */
+ unsigned char *mbs_case;
+#ifdef RE_ENABLE_I18N
+ /* Store the wide character string which is corresponding to MBS. */
+ wint_t *wcs;
+ mbstate_t cur_state;
+#endif
+ /* Index in RAW_MBS. Each character mbs[i] corresponds to
+ raw_mbs[raw_mbs_idx + i]. */
+ int raw_mbs_idx;
+ /* The length of the valid characters in the buffers. */
+ int valid_len;
+ /* The length of the buffers MBS, MBS_CASE, and WCS. */
+ int bufs_len;
+ /* The index in MBS, which is updated by re_string_fetch_byte. */
+ int cur_idx;
+ /* This is length_of_RAW_MBS - RAW_MBS_IDX. */
+ int len;
+ /* End of the buffer may be shorter than its length in the cases such
+ as re_match_2, re_search_2. Then, we use STOP for end of the buffer
+ instead of LEN. */
+ int stop;
+
+ /* The context of mbs[0]. We store the context independently, since
+ the context of mbs[0] may be different from raw_mbs[0], which is
+ the beginning of the input string. */
+ unsigned int tip_context;
+ /* The translation passed as a part of an argument of re_compile_pattern. */
+ RE_TRANSLATE_TYPE trans;
+ /* 1 if REG_ICASE. */
+ unsigned int icase : 1;
+};
+typedef struct re_string_t re_string_t;
+/* In case of REG_ICASE, we allocate the buffer dynamically for mbs. */
+#define MBS_ALLOCATED(pstr) (pstr->icase)
+/* In case that we need translation, we allocate the buffer dynamically
+ for mbs_case. Note that mbs == mbs_case if not REG_ICASE. */
+#define MBS_CASE_ALLOCATED(pstr) (pstr->trans != NULL)
+
+
+static reg_errcode_t re_string_allocate (re_string_t *pstr, const char *str,
+ int len, int init_len,
+ RE_TRANSLATE_TYPE trans, int icase);
+static reg_errcode_t re_string_construct (re_string_t *pstr, const char *str,
+ int len, RE_TRANSLATE_TYPE trans,
+ int icase);
+static reg_errcode_t re_string_reconstruct (re_string_t *pstr, int idx,
+ int eflags, int newline);
+static reg_errcode_t re_string_realloc_buffers (re_string_t *pstr,
+ int new_buf_len);
+#ifdef RE_ENABLE_I18N
+static void build_wcs_buffer (re_string_t *pstr);
+static void build_wcs_upper_buffer (re_string_t *pstr);
+#endif /* RE_ENABLE_I18N */
+static void build_upper_buffer (re_string_t *pstr);
+static void re_string_translate_buffer (re_string_t *pstr);
+static void re_string_destruct (re_string_t *pstr);
+#ifdef RE_ENABLE_I18N
+static int re_string_elem_size_at (const re_string_t *pstr, int idx);
+static inline int re_string_char_size_at (const re_string_t *pstr, int idx);
+static inline wint_t re_string_wchar_at (const re_string_t *pstr, int idx);
+#endif /* RE_ENABLE_I18N */
+static unsigned int re_string_context_at (const re_string_t *input, int idx,
+ int eflags, int newline_anchor);
+#define re_string_peek_byte(pstr, offset) \
+ ((pstr)->mbs[(pstr)->cur_idx + offset])
+#define re_string_peek_byte_case(pstr, offset) \
+ ((pstr)->mbs_case[(pstr)->cur_idx + offset])
+#define re_string_fetch_byte(pstr) \
+ ((pstr)->mbs[(pstr)->cur_idx++])
+#define re_string_fetch_byte_case(pstr) \
+ ((pstr)->mbs_case[(pstr)->cur_idx++])
+#define re_string_first_byte(pstr, idx) \
+ ((idx) == (pstr)->len || (pstr)->wcs[idx] != WEOF)
+#define re_string_is_single_byte_char(pstr, idx) \
+ ((pstr)->wcs[idx] != WEOF && ((pstr)->len == (idx) \
+ || (pstr)->wcs[(idx) + 1] != WEOF))
+#define re_string_eoi(pstr) ((pstr)->stop <= (pstr)->cur_idx)
+#define re_string_cur_idx(pstr) ((pstr)->cur_idx)
+#define re_string_get_buffer(pstr) ((pstr)->mbs)
+#define re_string_length(pstr) ((pstr)->len)
+#define re_string_byte_at(pstr,idx) ((pstr)->mbs[idx])
+#define re_string_skip_bytes(pstr,idx) ((pstr)->cur_idx += (idx))
+#define re_string_set_index(pstr,idx) ((pstr)->cur_idx = (idx))
+
+#define re_malloc(t,n) ((t *) malloc ((n) * sizeof (t)))
+#define re_realloc(p,t,n) ((t *) realloc (p, (n) * sizeof (t)))
+#define re_free(p) free (p)
+
+struct bin_tree_t
+{
+ struct bin_tree_t *parent;
+ struct bin_tree_t *left;
+ struct bin_tree_t *right;
+
+ /* `node_idx' is the index in dfa->nodes, if `type' == 0.
+ Otherwise `type' indicate the type of this node. */
+ re_token_type_t type;
+ int node_idx;
+
+ int first;
+ int next;
+ re_node_set eclosure;
+};
+typedef struct bin_tree_t bin_tree_t;
+
+
+#define CONTEXT_WORD 1
+#define CONTEXT_NEWLINE (CONTEXT_WORD << 1)
+#define CONTEXT_BEGBUF (CONTEXT_NEWLINE << 1)
+#define CONTEXT_ENDBUF (CONTEXT_BEGBUF << 1)
+
+#define IS_WORD_CONTEXT(c) ((c) & CONTEXT_WORD)
+#define IS_NEWLINE_CONTEXT(c) ((c) & CONTEXT_NEWLINE)
+#define IS_BEGBUF_CONTEXT(c) ((c) & CONTEXT_BEGBUF)
+#define IS_ENDBUF_CONTEXT(c) ((c) & CONTEXT_ENDBUF)
+#define IS_ORDINARY_CONTEXT(c) ((c) == 0)
+
+#define IS_WORD_CHAR(ch) (isalnum (ch) || (ch) == '_')
+#define IS_NEWLINE(ch) ((ch) == NEWLINE_CHAR)
+#define IS_WIDE_WORD_CHAR(ch) (iswalnum (ch) || (ch) == L'_')
+#define IS_WIDE_NEWLINE(ch) ((ch) == WIDE_NEWLINE_CHAR)
+
+#define NOT_SATISFY_PREV_CONSTRAINT(constraint,context) \
+ ((((constraint) & PREV_WORD_CONSTRAINT) && !IS_WORD_CONTEXT (context)) \
+ || ((constraint & PREV_NOTWORD_CONSTRAINT) && IS_WORD_CONTEXT (context)) \
+ || ((constraint & PREV_NEWLINE_CONSTRAINT) && !IS_NEWLINE_CONTEXT (context))\
+ || ((constraint & PREV_BEGBUF_CONSTRAINT) && !IS_BEGBUF_CONTEXT (context)))
+
+#define NOT_SATISFY_NEXT_CONSTRAINT(constraint,context) \
+ ((((constraint) & NEXT_WORD_CONSTRAINT) && !IS_WORD_CONTEXT (context)) \
+ || (((constraint) & NEXT_NOTWORD_CONSTRAINT) && IS_WORD_CONTEXT (context)) \
+ || (((constraint) & NEXT_NEWLINE_CONSTRAINT) && !IS_NEWLINE_CONTEXT (context)) \
+ || (((constraint) & NEXT_ENDBUF_CONSTRAINT) && !IS_ENDBUF_CONTEXT (context)))
+
+struct re_dfastate_t
+{
+ unsigned int hash;
+ re_node_set nodes;
+ re_node_set *entrance_nodes;
+ struct re_dfastate_t **trtable;
+ struct re_dfastate_t **trtable_search;
+ /* If this state is a special state.
+ A state is a special state if the state is the halt state, or
+ a anchor. */
+ unsigned int context : 2;
+ unsigned int halt : 1;
+ /* If this state can accept `multi byte'.
+ Note that we refer to multibyte characters, and multi character
+ collating elements as `multi byte'. */
+ unsigned int accept_mb : 1;
+ /* If this state has backreference node(s). */
+ unsigned int has_backref : 1;
+ unsigned int has_constraint : 1;
+};
+typedef struct re_dfastate_t re_dfastate_t;
+
+typedef struct
+{
+ /* start <= node < end */
+ int start;
+ int end;
+} re_subexp_t;
+
+struct re_state_table_entry
+{
+ int num;
+ int alloc;
+ re_dfastate_t **array;
+};
+
+/* Array type used in re_sub_match_last_t and re_sub_match_top_t. */
+
+typedef struct
+{
+ int next_idx;
+ int alloc;
+ re_dfastate_t **array;
+} state_array_t;
+
+/* Store information about the node NODE whose type is OP_CLOSE_SUBEXP. */
+
+typedef struct
+{
+ int node;
+ int str_idx; /* The position NODE match at. */
+ state_array_t path;
+} re_sub_match_last_t;
+
+/* Store information about the node NODE whose type is OP_OPEN_SUBEXP.
+ And information about the node, whose type is OP_CLOSE_SUBEXP,
+ corresponding to NODE is stored in LASTS. */
+
+typedef struct
+{
+ int str_idx;
+ int node;
+ int next_last_offset;
+ state_array_t *path;
+ int alasts; /* Allocation size of LASTS. */
+ int nlasts; /* The number of LASTS. */
+ re_sub_match_last_t **lasts;
+} re_sub_match_top_t;
+
+struct re_backref_cache_entry
+{
+ int node;
+ int str_idx;
+ int subexp_from;
+ int subexp_to;
+ int flag;
+};
+
+typedef struct
+{
+ /* EFLAGS of the argument of regexec. */
+ int eflags;
+ /* Where the matching ends. */
+ int match_last;
+ int last_node;
+ /* The string object corresponding to the input string. */
+ re_string_t *input;
+ /* The state log used by the matcher. */
+ re_dfastate_t **state_log;
+ int state_log_top;
+ /* Back reference cache. */
+ int nbkref_ents;
+ int abkref_ents;
+ struct re_backref_cache_entry *bkref_ents;
+ int max_mb_elem_len;
+ int nsub_tops;
+ int asub_tops;
+ re_sub_match_top_t **sub_tops;
+} re_match_context_t;
+
+typedef struct
+{
+ int cur_bkref;
+ int cls_subexp_idx;
+
+ re_dfastate_t **sifted_states;
+ re_dfastate_t **limited_states;
+
+ re_node_set limits;
+
+ int last_node;
+ int last_str_idx;
+ int check_subexp;
+} re_sift_context_t;
+
+struct re_fail_stack_ent_t
+{
+ int idx;
+ int node;
+ regmatch_t *regs;
+ re_node_set eps_via_nodes;
+};
+
+struct re_fail_stack_t
+{
+ int num;
+ int alloc;
+ struct re_fail_stack_ent_t *stack;
+};
+
+struct re_dfa_t
+{
+ re_bitset_ptr_t word_char;
+
+ /* number of subexpressions `re_nsub' is in regex_t. */
+ int subexps_alloc;
+ re_subexp_t *subexps;
+
+ re_token_t *nodes;
+ int nodes_alloc;
+ int nodes_len;
+ bin_tree_t *str_tree;
+ int *nexts;
+ int *org_indices;
+ re_node_set *edests;
+ re_node_set *eclosures;
+ re_node_set *inveclosures;
+ struct re_state_table_entry *state_table;
+ unsigned int state_hash_mask;
+ re_dfastate_t *init_state;
+ re_dfastate_t *init_state_word;
+ re_dfastate_t *init_state_nl;
+ re_dfastate_t *init_state_begbuf;
+ int states_alloc;
+ int init_node;
+ int nbackref; /* The number of backreference in this dfa. */
+ /* Bitmap expressing which backreference is used. */
+ unsigned int used_bkref_map;
+#ifdef DEBUG
+ char* re_str;
+#endif
+ unsigned int has_plural_match : 1;
+ /* If this dfa has "multibyte node", which is a backreference or
+ a node which can accept multibyte character or multi character
+ collating element. */
+ unsigned int has_mb_node : 1;
+};
+typedef struct re_dfa_t re_dfa_t;
+
+static reg_errcode_t re_node_set_alloc (re_node_set *set, int size);
+static reg_errcode_t re_node_set_init_1 (re_node_set *set, int elem);
+static reg_errcode_t re_node_set_init_2 (re_node_set *set, int elem1,
+ int elem2);
+#define re_node_set_init_empty(set) memset (set, '\0', sizeof (re_node_set))
+static reg_errcode_t re_node_set_init_copy (re_node_set *dest,
+ const re_node_set *src);
+static reg_errcode_t re_node_set_add_intersect (re_node_set *dest,
+ const re_node_set *src1,
+ const re_node_set *src2);
+static reg_errcode_t re_node_set_init_union (re_node_set *dest,
+ const re_node_set *src1,
+ const re_node_set *src2);
+static reg_errcode_t re_node_set_merge (re_node_set *dest,
+ const re_node_set *src);
+static int re_node_set_insert (re_node_set *set, int elem);
+static int re_node_set_compare (const re_node_set *set1,
+ const re_node_set *set2);
+static int re_node_set_contains (const re_node_set *set, int elem);
+static void re_node_set_remove_at (re_node_set *set, int idx);
+#define re_node_set_remove(set,id) \
+ (re_node_set_remove_at (set, re_node_set_contains (set, id) - 1))
+#define re_node_set_empty(p) ((p)->nelem = 0)
+#define re_node_set_free(set) re_free ((set)->elems)
+static int re_dfa_add_node (re_dfa_t *dfa, re_token_t token, int mode);
+static re_dfastate_t *re_acquire_state (reg_errcode_t *err, re_dfa_t *dfa,
+ const re_node_set *nodes);
+static re_dfastate_t *re_acquire_state_context (reg_errcode_t *err,
+ re_dfa_t *dfa,
+ const re_node_set *nodes,
+ unsigned int context);
+static void free_state (re_dfastate_t *state);
+\f
+
+typedef enum
+{
+ SB_CHAR,
+ MB_CHAR,
+ EQUIV_CLASS,
+ COLL_SYM,
+ CHAR_CLASS
+} bracket_elem_type;
+
+typedef struct
+{
+ bracket_elem_type type;
+ union
+ {
+ unsigned char ch;
+ unsigned char *name;
+ wchar_t wch;
+ } opr;
+} bracket_elem_t;
+
+
+/* Inline functions for bitset operation. */
+static inline void
+bitset_not (set)
+ bitset set;
+{
+ int bitset_i;
+ for (bitset_i = 0; bitset_i < BITSET_UINTS; ++bitset_i)
+ set[bitset_i] = ~set[bitset_i];
+}
+
+static inline void
+bitset_merge (dest, src)
+ bitset dest;
+ const bitset src;
+{
+ int bitset_i;
+ for (bitset_i = 0; bitset_i < BITSET_UINTS; ++bitset_i)
+ dest[bitset_i] |= src[bitset_i];
+}
+
+static inline void
+bitset_not_merge (dest, src)
+ bitset dest;
+ const bitset src;
+{
+ int i;
+ for (i = 0; i < BITSET_UINTS; ++i)
+ dest[i] |= ~src[i];
+}
+
+#ifdef RE_ENABLE_I18N
+/* Inline functions for re_string. */
+static inline int
+re_string_char_size_at (pstr, idx)
+ const re_string_t *pstr;
+ int idx;
+{
+ int byte_idx;
+ if (MB_CUR_MAX == 1)
+ return 1;
+ for (byte_idx = 1; idx + byte_idx < pstr->len; ++byte_idx)
+ if (pstr->wcs[idx + byte_idx] != WEOF)
+ break;
+ return byte_idx;
+}
+
+static inline wint_t
+re_string_wchar_at (pstr, idx)
+ const re_string_t *pstr;
+ int idx;
+{
+ if (MB_CUR_MAX == 1)
+ return (wint_t) pstr->mbs[idx];
+ return (wint_t) pstr->wcs[idx];
+}
+
+static int
+re_string_elem_size_at (pstr, idx)
+ const re_string_t *pstr;
+ int idx;
+{
+#ifdef _LIBC
+ const unsigned char *p, *extra;
+ const int32_t *table, *indirect;
+ int32_t tmp;
+# include <locale/weight.h>
+ uint_fast32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES);
+
+ if (nrules != 0)
+ {
+ table = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB);
+ extra = (const unsigned char *)
+ _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB);
+ indirect = (const int32_t *) _NL_CURRENT (LC_COLLATE,
+ _NL_COLLATE_INDIRECTMB);
+ p = pstr->mbs + idx;
+ tmp = findidx (&p);
+ return p - pstr->mbs - idx;
+ }
+ else
+#endif /* _LIBC */
+ return 1;
+}
+#endif /* RE_ENABLE_I18N */
+
+#endif /* _REGEX_INTERNAL_H */
--- /dev/null
+/* Extended regular expression matching and search library.
+ Copyright (C) 2002, 2003 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+ Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, write to the Free
+ Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ 02111-1307 USA. */
+
+static reg_errcode_t match_ctx_init (re_match_context_t *cache, int eflags,
+ re_string_t *input, int n);
+static void match_ctx_clean (re_match_context_t *mctx);
+static void match_ctx_free (re_match_context_t *cache);
+static void match_ctx_free_subtops (re_match_context_t *mctx);
+static reg_errcode_t match_ctx_add_entry (re_match_context_t *cache, int node,
+ int str_idx, int from, int to);
+static int search_cur_bkref_entry (re_match_context_t *mctx, int str_idx);
+static void match_ctx_clear_flag (re_match_context_t *mctx);
+static reg_errcode_t match_ctx_add_subtop (re_match_context_t *mctx, int node,
+ int str_idx);
+static re_sub_match_last_t * match_ctx_add_sublast (re_sub_match_top_t *subtop,
+ int node, int str_idx);
+static void sift_ctx_init (re_sift_context_t *sctx, re_dfastate_t **sifted_sts,
+ re_dfastate_t **limited_sts, int last_node,
+ int last_str_idx, int check_subexp);
+static reg_errcode_t re_search_internal (const regex_t *preg,
+ const char *string, int length,
+ int start, int range, int stop,
+ size_t nmatch, regmatch_t pmatch[],
+ int eflags);
+static int re_search_2_stub (struct re_pattern_buffer *bufp,
+ const char *string1, int length1,
+ const char *string2, int length2,
+ int start, int range, struct re_registers *regs,
+ int stop, int ret_len);
+static int re_search_stub (struct re_pattern_buffer *bufp,
+ const char *string, int length, int start,
+ int range, int stop, struct re_registers *regs,
+ int ret_len);
+static unsigned re_copy_regs (struct re_registers *regs, regmatch_t *pmatch,
+ int nregs, int regs_allocated);
+static inline re_dfastate_t *acquire_init_state_context (reg_errcode_t *err,
+ const regex_t *preg,
+ const re_match_context_t *mctx,
+ int idx);
+static reg_errcode_t prune_impossible_nodes (const regex_t *preg,
+ re_match_context_t *mctx);
+static int check_matching (const regex_t *preg, re_match_context_t *mctx,
+ int fl_search, int fl_longest_match);
+static int check_halt_node_context (const re_dfa_t *dfa, int node,
+ unsigned int context);
+static int check_halt_state_context (const regex_t *preg,
+ const re_dfastate_t *state,
+ const re_match_context_t *mctx, int idx);
+static void update_regs (re_dfa_t *dfa, regmatch_t *pmatch, int cur_node,
+ int cur_idx, int nmatch);
+static int proceed_next_node (const regex_t *preg, int nregs, regmatch_t *regs,
+ const re_match_context_t *mctx,
+ int *pidx, int node, re_node_set *eps_via_nodes,
+ struct re_fail_stack_t *fs);
+static reg_errcode_t push_fail_stack (struct re_fail_stack_t *fs,
+ int str_idx, int *dests, int nregs,
+ regmatch_t *regs,
+ re_node_set *eps_via_nodes);
+static int pop_fail_stack (struct re_fail_stack_t *fs, int *pidx, int nregs,
+ regmatch_t *regs, re_node_set *eps_via_nodes);
+static reg_errcode_t set_regs (const regex_t *preg,
+ const re_match_context_t *mctx,
+ size_t nmatch, regmatch_t *pmatch,
+ int fl_backtrack);
+static reg_errcode_t free_fail_stack_return (struct re_fail_stack_t *fs);
+
+#ifdef RE_ENABLE_I18N
+static int sift_states_iter_mb (const regex_t *preg,
+ const re_match_context_t *mctx,
+ re_sift_context_t *sctx,
+ int node_idx, int str_idx, int max_str_idx);
+#endif /* RE_ENABLE_I18N */
+static reg_errcode_t sift_states_backward (const regex_t *preg,
+ re_match_context_t *mctx,
+ re_sift_context_t *sctx);
+static reg_errcode_t update_cur_sifted_state (const regex_t *preg,
+ re_match_context_t *mctx,
+ re_sift_context_t *sctx,
+ int str_idx,
+ re_node_set *dest_nodes);
+static reg_errcode_t add_epsilon_src_nodes (re_dfa_t *dfa,
+ re_node_set *dest_nodes,
+ const re_node_set *candidates);
+static reg_errcode_t sub_epsilon_src_nodes (re_dfa_t *dfa, int node,
+ re_node_set *dest_nodes,
+ const re_node_set *and_nodes);
+static int check_dst_limits (re_dfa_t *dfa, re_node_set *limits,
+ re_match_context_t *mctx, int dst_node,
+ int dst_idx, int src_node, int src_idx);
+static int check_dst_limits_calc_pos (re_dfa_t *dfa, re_match_context_t *mctx,
+ int limit, re_node_set *eclosures,
+ int subexp_idx, int node, int str_idx);
+static reg_errcode_t check_subexp_limits (re_dfa_t *dfa,
+ re_node_set *dest_nodes,
+ const re_node_set *candidates,
+ re_node_set *limits,
+ struct re_backref_cache_entry *bkref_ents,
+ int str_idx);
+static reg_errcode_t sift_states_bkref (const regex_t *preg,
+ re_match_context_t *mctx,
+ re_sift_context_t *sctx,
+ int str_idx, re_node_set *dest_nodes);
+static reg_errcode_t clean_state_log_if_need (re_match_context_t *mctx,
+ int next_state_log_idx);
+static reg_errcode_t merge_state_array (re_dfa_t *dfa, re_dfastate_t **dst,
+ re_dfastate_t **src, int num);
+static re_dfastate_t *transit_state (reg_errcode_t *err, const regex_t *preg,
+ re_match_context_t *mctx,
+ re_dfastate_t *state, int fl_search);
+static reg_errcode_t check_subexp_matching_top (re_dfa_t *dfa,
+ re_match_context_t *mctx,
+ re_node_set *cur_nodes,
+ int str_idx);
+static re_dfastate_t *transit_state_sb (reg_errcode_t *err, const regex_t *preg,
+ re_dfastate_t *pstate,
+ int fl_search,
+ re_match_context_t *mctx);
+#ifdef RE_ENABLE_I18N
+static reg_errcode_t transit_state_mb (const regex_t *preg,
+ re_dfastate_t *pstate,
+ re_match_context_t *mctx);
+#endif /* RE_ENABLE_I18N */
+static reg_errcode_t transit_state_bkref (const regex_t *preg,
+ re_node_set *nodes,
+ re_match_context_t *mctx);
+static reg_errcode_t get_subexp (const regex_t *preg, re_match_context_t *mctx,
+ int bkref_node, int bkref_str_idx);
+static reg_errcode_t get_subexp_sub (const regex_t *preg,
+ re_match_context_t *mctx,
+ re_sub_match_top_t *sub_top,
+ re_sub_match_last_t *sub_last,
+ int bkref_node, int bkref_str);
+static int find_subexp_node (re_dfa_t *dfa, re_node_set *nodes,
+ int subexp_idx, int fl_open);
+static reg_errcode_t check_arrival (const regex_t *preg,
+ re_match_context_t *mctx,
+ state_array_t *path, int top_node,
+ int top_str, int last_node, int last_str,
+ int fl_open);
+static reg_errcode_t check_arrival_add_next_nodes (const regex_t *preg,
+ re_dfa_t *dfa,
+ re_match_context_t *mctx,
+ int str_idx,
+ re_node_set *cur_nodes,
+ re_node_set *next_nodes);
+static reg_errcode_t check_arrival_expand_ecl (re_dfa_t *dfa,
+ re_node_set *cur_nodes,
+ int ex_subexp, int fl_open);
+static reg_errcode_t check_arrival_expand_ecl_sub (re_dfa_t *dfa,
+ re_node_set *dst_nodes,
+ int target, int ex_subexp,
+ int fl_open);
+static reg_errcode_t expand_bkref_cache (const regex_t *preg,
+ re_match_context_t *mctx,
+ re_node_set *cur_nodes, int cur_str,
+ int last_str, int subexp_num,
+ int fl_open);
+static re_dfastate_t **build_trtable (const regex_t *dfa,
+ const re_dfastate_t *state,
+ int fl_search);
+#ifdef RE_ENABLE_I18N
+static int check_node_accept_bytes (const regex_t *preg, int node_idx,
+ const re_string_t *input, int idx);
+# ifdef _LIBC
+static unsigned int find_collation_sequence_value (const unsigned char *mbs,
+ size_t name_len);
+# endif /* _LIBC */
+#endif /* RE_ENABLE_I18N */
+static int group_nodes_into_DFAstates (const regex_t *dfa,
+ const re_dfastate_t *state,
+ re_node_set *states_node,
+ bitset *states_ch);
+static int check_node_accept (const regex_t *preg, const re_token_t *node,
+ const re_match_context_t *mctx, int idx);
+static reg_errcode_t extend_buffers (re_match_context_t *mctx);
+\f
+/* Entry point for POSIX code. */
+
+/* regexec searches for a given pattern, specified by PREG, in the
+ string STRING.
+
+ If NMATCH is zero or REG_NOSUB was set in the cflags argument to
+ `regcomp', we ignore PMATCH. Otherwise, we assume PMATCH has at
+ least NMATCH elements, and we set them to the offsets of the
+ corresponding matched substrings.
+
+ EFLAGS specifies `execution flags' which affect matching: if
+ REG_NOTBOL is set, then ^ does not match at the beginning of the
+ string; if REG_NOTEOL is set, then $ does not match at the end.
+
+ We return 0 if we find a match and REG_NOMATCH if not. */
+
+int
+regexec (preg, string, nmatch, pmatch, eflags)
+ const regex_t *__restrict preg;
+ const char *__restrict string;
+ size_t nmatch;
+ regmatch_t pmatch[];
+ int eflags;
+{
+ reg_errcode_t err;
+ int length = strlen (string);
+ if (preg->no_sub)
+ err = re_search_internal (preg, string, length, 0, length, length, 0,
+ NULL, eflags);
+ else
+ err = re_search_internal (preg, string, length, 0, length, length, nmatch,
+ pmatch, eflags);
+ return err != REG_NOERROR;
+}
+#ifdef _LIBC
+weak_alias (__regexec, regexec)
+#endif
+
+/* Entry points for GNU code. */
+
+/* re_match, re_search, re_match_2, re_search_2
+
+ The former two functions operate on STRING with length LENGTH,
+ while the later two operate on concatenation of STRING1 and STRING2
+ with lengths LENGTH1 and LENGTH2, respectively.
+
+ re_match() matches the compiled pattern in BUFP against the string,
+ starting at index START.
+
+ re_search() first tries matching at index START, then it tries to match
+ starting from index START + 1, and so on. The last start position tried
+ is START + RANGE. (Thus RANGE = 0 forces re_search to operate the same
+ way as re_match().)
+
+ The parameter STOP of re_{match,search}_2 specifies that no match exceeding
+ the first STOP characters of the concatenation of the strings should be
+ concerned.
+
+ If REGS is not NULL, and BUFP->no_sub is not set, the offsets of the match
+ and all groups is stroed in REGS. (For the "_2" variants, the offsets are
+ computed relative to the concatenation, not relative to the individual
+ strings.)
+
+ On success, re_match* functions return the length of the match, re_search*
+ return the position of the start of the match. Return value -1 means no
+ match was found and -2 indicates an internal error. */
+
+int
+re_match (bufp, string, length, start, regs)
+ struct re_pattern_buffer *bufp;
+ const char *string;
+ int length, start;
+ struct re_registers *regs;
+{
+ return re_search_stub (bufp, string, length, start, 0, length, regs, 1);
+}
+#ifdef _LIBC
+weak_alias (__re_match, re_match)
+#endif
+
+int
+re_search (bufp, string, length, start, range, regs)
+ struct re_pattern_buffer *bufp;
+ const char *string;
+ int length, start, range;
+ struct re_registers *regs;
+{
+ return re_search_stub (bufp, string, length, start, range, length, regs, 0);
+}
+#ifdef _LIBC
+weak_alias (__re_search, re_search)
+#endif
+
+int
+re_match_2 (bufp, string1, length1, string2, length2, start, regs, stop)
+ struct re_pattern_buffer *bufp;
+ const char *string1, *string2;
+ int length1, length2, start, stop;
+ struct re_registers *regs;
+{
+ return re_search_2_stub (bufp, string1, length1, string2, length2,
+ start, 0, regs, stop, 1);
+}
+#ifdef _LIBC
+weak_alias (__re_match_2, re_match_2)
+#endif
+
+int
+re_search_2 (bufp, string1, length1, string2, length2, start, range, regs, stop)
+ struct re_pattern_buffer *bufp;
+ const char *string1, *string2;
+ int length1, length2, start, range, stop;
+ struct re_registers *regs;
+{
+ return re_search_2_stub (bufp, string1, length1, string2, length2,
+ start, range, regs, stop, 0);
+}
+#ifdef _LIBC
+weak_alias (__re_search_2, re_search_2)
+#endif
+
+static int
+re_search_2_stub (bufp, string1, length1, string2, length2, start, range, regs,
+ stop, ret_len)
+ struct re_pattern_buffer *bufp;
+ const char *string1, *string2;
+ int length1, length2, start, range, stop, ret_len;
+ struct re_registers *regs;
+{
+ const char *str;
+ int rval;
+ int len = length1 + length2;
+ int free_str = 0;
+
+ if (BE (length1 < 0 || length2 < 0 || stop < 0, 0))
+ return -2;
+
+ /* Concatenate the strings. */
+ if (length2 > 0)
+ if (length1 > 0)
+ {
+ char *s = re_malloc (char, len);
+
+ if (BE (s == NULL, 0))
+ return -2;
+ memcpy (s, string1, length1);
+ memcpy (s + length1, string2, length2);
+ str = s;
+ free_str = 1;
+ }
+ else
+ str = string2;
+ else
+ str = string1;
+
+ rval = re_search_stub (bufp, str, len, start, range, stop, regs,
+ ret_len);
+ if (free_str)
+ re_free ((char *) str);
+ return rval;
+}
+
+/* The parameters have the same meaning as those of re_search.
+ Additional parameters:
+ If RET_LEN is nonzero the length of the match is returned (re_match style);
+ otherwise the position of the match is returned. */
+
+static int
+re_search_stub (bufp, string, length, start, range, stop, regs, ret_len)
+ struct re_pattern_buffer *bufp;
+ const char *string;
+ int length, start, range, stop, ret_len;
+ struct re_registers *regs;
+{
+ reg_errcode_t result;
+ regmatch_t *pmatch;
+ int nregs, rval;
+ int eflags = 0;
+
+ /* Check for out-of-range. */
+ if (BE (start < 0 || start > length, 0))
+ return -1;
+ if (BE (start + range > length, 0))
+ range = length - start;
+ else if (BE (start + range < 0, 0))
+ range = -start;
+
+ eflags |= (bufp->not_bol) ? REG_NOTBOL : 0;
+ eflags |= (bufp->not_eol) ? REG_NOTEOL : 0;
+
+ /* Compile fastmap if we haven't yet. */
+ if (range > 0 && bufp->fastmap != NULL && !bufp->fastmap_accurate)
+ re_compile_fastmap (bufp);
+
+ if (BE (bufp->no_sub, 0))
+ regs = NULL;
+
+ /* We need at least 1 register. */
+ if (regs == NULL)
+ nregs = 1;
+ else if (BE (bufp->regs_allocated == REGS_FIXED &&
+ regs->num_regs < bufp->re_nsub + 1, 0))
+ {
+ nregs = regs->num_regs;
+ if (BE (nregs < 1, 0))
+ {
+ /* Nothing can be copied to regs. */
+ regs = NULL;
+ nregs = 1;
+ }
+ }
+ else
+ nregs = bufp->re_nsub + 1;
+ pmatch = re_malloc (regmatch_t, nregs);
+ if (BE (pmatch == NULL, 0))
+ return -2;
+
+ result = re_search_internal (bufp, string, length, start, range, stop,
+ nregs, pmatch, eflags);
+
+ rval = 0;
+
+ /* I hope we needn't fill ther regs with -1's when no match was found. */
+ if (result != REG_NOERROR)
+ rval = -1;
+ else if (regs != NULL)
+ {
+ /* If caller wants register contents data back, copy them. */
+ bufp->regs_allocated = re_copy_regs (regs, pmatch, nregs,
+ bufp->regs_allocated);
+ if (BE (bufp->regs_allocated == REGS_UNALLOCATED, 0))
+ rval = -2;
+ }
+
+ if (BE (rval == 0, 1))
+ {
+ if (ret_len)
+ {
+ assert (pmatch[0].rm_so == start);
+ rval = pmatch[0].rm_eo - start;
+ }
+ else
+ rval = pmatch[0].rm_so;
+ }
+ re_free (pmatch);
+ return rval;
+}
+
+static unsigned
+re_copy_regs (regs, pmatch, nregs, regs_allocated)
+ struct re_registers *regs;
+ regmatch_t *pmatch;
+ int nregs, regs_allocated;
+{
+ int rval = REGS_REALLOCATE;
+ int i;
+ int need_regs = nregs + 1;
+ /* We need one extra element beyond `num_regs' for the `-1' marker GNU code
+ uses. */
+
+ /* Have the register data arrays been allocated? */
+ if (regs_allocated == REGS_UNALLOCATED)
+ { /* No. So allocate them with malloc. */
+ regs->start = re_malloc (regoff_t, need_regs);
+ if (BE (regs->start == NULL, 0))
+ return REGS_UNALLOCATED;
+ regs->end = re_malloc (regoff_t, need_regs);
+ if (BE (regs->end == NULL, 0))
+ {
+ re_free (regs->start);
+ return REGS_UNALLOCATED;
+ }
+ regs->num_regs = need_regs;
+ }
+ else if (regs_allocated == REGS_REALLOCATE)
+ { /* Yes. If we need more elements than were already
+ allocated, reallocate them. If we need fewer, just
+ leave it alone. */
+ if (need_regs > regs->num_regs)
+ {
+ regs->start = re_realloc (regs->start, regoff_t, need_regs);
+ if (BE (regs->start == NULL, 0))
+ {
+ if (regs->end != NULL)
+ re_free (regs->end);
+ return REGS_UNALLOCATED;
+ }
+ regs->end = re_realloc (regs->end, regoff_t, need_regs);
+ if (BE (regs->end == NULL, 0))
+ {
+ re_free (regs->start);
+ return REGS_UNALLOCATED;
+ }
+ regs->num_regs = need_regs;
+ }
+ }
+ else
+ {
+ assert (regs_allocated == REGS_FIXED);
+ /* This function may not be called with REGS_FIXED and nregs too big. */
+ assert (regs->num_regs >= nregs);
+ rval = REGS_FIXED;
+ }
+
+ /* Copy the regs. */
+ for (i = 0; i < nregs; ++i)
+ {
+ regs->start[i] = pmatch[i].rm_so;
+ regs->end[i] = pmatch[i].rm_eo;
+ }
+ for ( ; i < regs->num_regs; ++i)
+ regs->start[i] = regs->end[i] = -1;
+
+ return rval;
+}
+
+/* Set REGS to hold NUM_REGS registers, storing them in STARTS and
+ ENDS. Subsequent matches using PATTERN_BUFFER and REGS will use
+ this memory for recording register information. STARTS and ENDS
+ must be allocated using the malloc library routine, and must each
+ be at least NUM_REGS * sizeof (regoff_t) bytes long.
+
+ If NUM_REGS == 0, then subsequent matches should allocate their own
+ register data.
+
+ Unless this function is called, the first search or match using
+ PATTERN_BUFFER will allocate its own register data, without
+ freeing the old data. */
+
+void
+re_set_registers (bufp, regs, num_regs, starts, ends)
+ struct re_pattern_buffer *bufp;
+ struct re_registers *regs;
+ unsigned num_regs;
+ regoff_t *starts, *ends;
+{
+ if (num_regs)
+ {
+ bufp->regs_allocated = REGS_REALLOCATE;
+ regs->num_regs = num_regs;
+ regs->start = starts;
+ regs->end = ends;
+ }
+ else
+ {
+ bufp->regs_allocated = REGS_UNALLOCATED;
+ regs->num_regs = 0;
+ regs->start = regs->end = (regoff_t *) 0;
+ }
+}
+#ifdef _LIBC
+weak_alias (__re_set_registers, re_set_registers)
+#endif
+\f
+/* Entry points compatible with 4.2 BSD regex library. We don't define
+ them unless specifically requested. */
+
+#if defined _REGEX_RE_COMP || defined _LIBC
+int
+# ifdef _LIBC
+weak_function
+# endif
+re_exec (s)
+ const char *s;
+{
+ return 0 == regexec (&re_comp_buf, s, 0, NULL, 0);
+}
+#endif /* _REGEX_RE_COMP */
+\f
+static re_node_set empty_set;
+
+/* Internal entry point. */
+
+/* Searches for a compiled pattern PREG in the string STRING, whose
+ length is LENGTH. NMATCH, PMATCH, and EFLAGS have the same
+ mingings with regexec. START, and RANGE have the same meanings
+ with re_search.
+ Return REG_NOERROR if we find a match, and REG_NOMATCH if not,
+ otherwise return the error code.
+ Note: We assume front end functions already check ranges.
+ (START + RANGE >= 0 && START + RANGE <= LENGTH) */
+
+static reg_errcode_t
+re_search_internal (preg, string, length, start, range, stop, nmatch, pmatch,
+ eflags)
+ const regex_t *preg;
+ const char *string;
+ int length, start, range, stop, eflags;
+ size_t nmatch;
+ regmatch_t pmatch[];
+{
+ reg_errcode_t err;
+ re_dfa_t *dfa = (re_dfa_t *)preg->buffer;
+ re_string_t input;
+ int left_lim, right_lim, incr;
+ int fl_longest_match, match_first, match_last = -1;
+ int fast_translate, sb;
+ re_match_context_t mctx;
+ char *fastmap = ((preg->fastmap != NULL && preg->fastmap_accurate
+ && range && !preg->can_be_null) ? preg->fastmap : NULL);
+
+ /* Check if the DFA haven't been compiled. */
+ if (BE (preg->used == 0 || dfa->init_state == NULL
+ || dfa->init_state_word == NULL || dfa->init_state_nl == NULL
+ || dfa->init_state_begbuf == NULL, 0))
+ return REG_NOMATCH;
+
+ re_node_set_init_empty (&empty_set);
+ memset (&mctx, '\0', sizeof (re_match_context_t));
+
+ /* We must check the longest matching, if nmatch > 0. */
+ fl_longest_match = (nmatch != 0 || dfa->nbackref);
+
+ err = re_string_allocate (&input, string, length, dfa->nodes_len + 1,
+ preg->translate, preg->syntax & RE_ICASE);
+ if (BE (err != REG_NOERROR, 0))
+ goto free_return;
+ input.stop = stop;
+
+ err = match_ctx_init (&mctx, eflags, &input, dfa->nbackref * 2);
+ if (BE (err != REG_NOERROR, 0))
+ goto free_return;
+
+ /* We will log all the DFA states through which the dfa pass,
+ if nmatch > 1, or this dfa has "multibyte node", which is a
+ back-reference or a node which can accept multibyte character or
+ multi character collating element. */
+ if (nmatch > 1 || dfa->has_mb_node)
+ {
+ mctx.state_log = re_malloc (re_dfastate_t *, dfa->nodes_len + 1);
+ if (BE (mctx.state_log == NULL, 0))
+ {
+ err = REG_ESPACE;
+ goto free_return;
+ }
+ }
+ else
+ mctx.state_log = NULL;
+
+#ifdef DEBUG
+ /* We assume front-end functions already check them. */
+ assert (start + range >= 0 && start + range <= length);
+#endif
+
+ match_first = start;
+ input.tip_context = ((eflags & REG_NOTBOL) ? CONTEXT_BEGBUF
+ : CONTEXT_NEWLINE | CONTEXT_BEGBUF);
+
+ /* Check incrementally whether of not the input string match. */
+ incr = (range < 0) ? -1 : 1;
+ left_lim = (range < 0) ? start + range : start;
+ right_lim = (range < 0) ? start : start + range;
+ sb = MB_CUR_MAX == 1;
+ fast_translate = sb || !(preg->syntax & RE_ICASE || preg->translate);
+
+ for (;;)
+ {
+ /* At first get the current byte from input string. */
+ if (fastmap)
+ {
+ if (BE (fast_translate, 1))
+ {
+ unsigned RE_TRANSLATE_TYPE t
+ = (unsigned RE_TRANSLATE_TYPE) preg->translate;
+ if (BE (range >= 0, 1))
+ {
+ if (BE (t != NULL, 0))
+ {
+ while (BE (match_first < right_lim, 1)
+ && !fastmap[t[(unsigned char) string[match_first]]])
+ ++match_first;
+ }
+ else
+ {
+ while (BE (match_first < right_lim, 1)
+ && !fastmap[(unsigned char) string[match_first]])
+ ++match_first;
+ }
+ if (BE (match_first == right_lim, 0))
+ {
+ int ch = match_first >= length
+ ? 0 : (unsigned char) string[match_first];
+ if (!fastmap[t ? t[ch] : ch])
+ break;
+ }
+ }
+ else
+ {
+ while (match_first >= left_lim)
+ {
+ int ch = match_first >= length
+ ? 0 : (unsigned char) string[match_first];
+ if (fastmap[t ? t[ch] : ch])
+ break;
+ --match_first;
+ }
+ if (match_first < left_lim)
+ break;
+ }
+ }
+ else
+ {
+ int ch;
+
+ do
+ {
+ /* In this case, we can't determine easily the current byte,
+ since it might be a component byte of a multibyte
+ character. Then we use the constructed buffer
+ instead. */
+ /* If MATCH_FIRST is out of the valid range, reconstruct the
+ buffers. */
+ if (input.raw_mbs_idx + input.valid_len <= match_first
+ || match_first < input.raw_mbs_idx)
+ {
+ err = re_string_reconstruct (&input, match_first, eflags,
+ preg->newline_anchor);
+ if (BE (err != REG_NOERROR, 0))
+ goto free_return;
+ }
+ /* If MATCH_FIRST is out of the buffer, leave it as '\0'.
+ Note that MATCH_FIRST must not be smaller than 0. */
+ ch = ((match_first >= length) ? 0
+ : re_string_byte_at (&input,
+ match_first - input.raw_mbs_idx));
+ if (fastmap[ch])
+ break;
+ match_first += incr;
+ }
+ while (match_first >= left_lim && match_first <= right_lim);
+ if (! fastmap[ch])
+ break;
+ }
+ }
+
+ /* Reconstruct the buffers so that the matcher can assume that
+ the matching starts from the begining of the buffer. */
+ err = re_string_reconstruct (&input, match_first, eflags,
+ preg->newline_anchor);
+ if (BE (err != REG_NOERROR, 0))
+ goto free_return;
+#ifdef RE_ENABLE_I18N
+ /* Eliminate it when it is a component of a multibyte character
+ and isn't the head of a multibyte character. */
+ if (sb || re_string_first_byte (&input, 0))
+#endif
+ {
+ /* It seems to be appropriate one, then use the matcher. */
+ /* We assume that the matching starts from 0. */
+ mctx.state_log_top = mctx.nbkref_ents = mctx.max_mb_elem_len = 0;
+ match_last = check_matching (preg, &mctx, 0, fl_longest_match);
+ if (match_last != -1)
+ {
+ if (BE (match_last == -2, 0))
+ {
+ err = REG_ESPACE;
+ goto free_return;
+ }
+ else
+ {
+ mctx.match_last = match_last;
+ if ((!preg->no_sub && nmatch > 1) || dfa->nbackref)
+ {
+ re_dfastate_t *pstate = mctx.state_log[match_last];
+ mctx.last_node = check_halt_state_context (preg, pstate,
+ &mctx, match_last);
+ }
+ if ((!preg->no_sub && nmatch > 1 && dfa->has_plural_match)
+ || dfa->nbackref)
+ {
+ err = prune_impossible_nodes (preg, &mctx);
+ if (err == REG_NOERROR)
+ break;
+ if (BE (err != REG_NOMATCH, 0))
+ goto free_return;
+ }
+ else
+ break; /* We found a matching. */
+ }
+ }
+ match_ctx_clean (&mctx);
+ }
+ /* Update counter. */
+ match_first += incr;
+ if (match_first < left_lim || right_lim < match_first)
+ break;
+ }
+
+ /* Set pmatch[] if we need. */
+ if (match_last != -1 && nmatch > 0)
+ {
+ int reg_idx;
+
+ /* Initialize registers. */
+ for (reg_idx = 0; reg_idx < nmatch; ++reg_idx)
+ pmatch[reg_idx].rm_so = pmatch[reg_idx].rm_eo = -1;
+
+ /* Set the points where matching start/end. */
+ pmatch[0].rm_so = 0;
+ pmatch[0].rm_eo = mctx.match_last;
+
+ if (!preg->no_sub && nmatch > 1)
+ {
+ err = set_regs (preg, &mctx, nmatch, pmatch,
+ dfa->has_plural_match && dfa->nbackref > 0);
+ if (BE (err != REG_NOERROR, 0))
+ goto free_return;
+ }
+
+ /* At last, add the offset to the each registers, since we slided
+ the buffers so that We can assume that the matching starts from 0. */
+ for (reg_idx = 0; reg_idx < nmatch; ++reg_idx)
+ if (pmatch[reg_idx].rm_so != -1)
+ {
+ pmatch[reg_idx].rm_so += match_first;
+ pmatch[reg_idx].rm_eo += match_first;
+ }
+ }
+ err = (match_last == -1) ? REG_NOMATCH : REG_NOERROR;
+ free_return:
+ re_free (mctx.state_log);
+ if (dfa->nbackref)
+ match_ctx_free (&mctx);
+ re_string_destruct (&input);
+ return err;
+}
+
+static reg_errcode_t
+prune_impossible_nodes (preg, mctx)
+ const regex_t *preg;
+ re_match_context_t *mctx;
+{
+ int halt_node, match_last;
+ reg_errcode_t ret;
+ re_dfa_t *dfa = (re_dfa_t *)preg->buffer;
+ re_dfastate_t **sifted_states;
+ re_dfastate_t **lim_states = NULL;
+ re_sift_context_t sctx;
+#ifdef DEBUG
+ assert (mctx->state_log != NULL);
+#endif
+ match_last = mctx->match_last;
+ halt_node = mctx->last_node;
+ sifted_states = re_malloc (re_dfastate_t *, match_last + 1);
+ if (BE (sifted_states == NULL, 0))
+ {
+ ret = REG_ESPACE;
+ goto free_return;
+ }
+ if (dfa->nbackref)
+ {
+ lim_states = re_malloc (re_dfastate_t *, match_last + 1);
+ if (BE (lim_states == NULL, 0))
+ {
+ ret = REG_ESPACE;
+ goto free_return;
+ }
+ while (1)
+ {
+ memset (lim_states, '\0',
+ sizeof (re_dfastate_t *) * (match_last + 1));
+ match_ctx_clear_flag (mctx);
+ sift_ctx_init (&sctx, sifted_states, lim_states, halt_node,
+ match_last, 0);
+ ret = sift_states_backward (preg, mctx, &sctx);
+ re_node_set_free (&sctx.limits);
+ if (BE (ret != REG_NOERROR, 0))
+ goto free_return;
+ if (sifted_states[0] != NULL || lim_states[0] != NULL)
+ break;
+ do
+ {
+ --match_last;
+ if (match_last < 0)
+ {
+ ret = REG_NOMATCH;
+ goto free_return;
+ }
+ } while (!mctx->state_log[match_last]->halt);
+ halt_node = check_halt_state_context (preg,
+ mctx->state_log[match_last],
+ mctx, match_last);
+ }
+ ret = merge_state_array (dfa, sifted_states, lim_states,
+ match_last + 1);
+ re_free (lim_states);
+ lim_states = NULL;
+ if (BE (ret != REG_NOERROR, 0))
+ goto free_return;
+ }
+ else
+ {
+ sift_ctx_init (&sctx, sifted_states, lim_states, halt_node,
+ match_last, 0);
+ ret = sift_states_backward (preg, mctx, &sctx);
+ re_node_set_free (&sctx.limits);
+ if (BE (ret != REG_NOERROR, 0))
+ goto free_return;
+ }
+ re_free (mctx->state_log);
+ mctx->state_log = sifted_states;
+ sifted_states = NULL;
+ mctx->last_node = halt_node;
+ mctx->match_last = match_last;
+ ret = REG_NOERROR;
+ free_return:
+ re_free (sifted_states);
+ re_free (lim_states);
+ return ret;
+}
+
+/* Acquire an initial state and return it.
+ We must select appropriate initial state depending on the context,
+ since initial states may have constraints like "\<", "^", etc.. */
+
+static inline re_dfastate_t *
+acquire_init_state_context (err, preg, mctx, idx)
+ reg_errcode_t *err;
+ const regex_t *preg;
+ const re_match_context_t *mctx;
+ int idx;
+{
+ re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+
+ *err = REG_NOERROR;
+ if (dfa->init_state->has_constraint)
+ {
+ unsigned int context;
+ context = re_string_context_at (mctx->input, idx - 1, mctx->eflags,
+ preg->newline_anchor);
+ if (IS_WORD_CONTEXT (context))
+ return dfa->init_state_word;
+ else if (IS_ORDINARY_CONTEXT (context))
+ return dfa->init_state;
+ else if (IS_BEGBUF_CONTEXT (context) && IS_NEWLINE_CONTEXT (context))
+ return dfa->init_state_begbuf;
+ else if (IS_NEWLINE_CONTEXT (context))
+ return dfa->init_state_nl;
+ else if (IS_BEGBUF_CONTEXT (context))
+ {
+ /* It is relatively rare case, then calculate on demand. */
+ return re_acquire_state_context (err, dfa,
+ dfa->init_state->entrance_nodes,
+ context);
+ }
+ else
+ /* Must not happen? */
+ return dfa->init_state;
+ }
+ else
+ return dfa->init_state;
+}
+
+/* Check whether the regular expression match input string INPUT or not,
+ and return the index where the matching end, return -1 if not match,
+ or return -2 in case of an error.
+ FL_SEARCH means we must search where the matching starts,
+ FL_LONGEST_MATCH means we want the POSIX longest matching.
+ Note that the matcher assume that the maching starts from the current
+ index of the buffer. */
+
+static int
+check_matching (preg, mctx, fl_search, fl_longest_match)
+ const regex_t *preg;
+ re_match_context_t *mctx;
+ int fl_search, fl_longest_match;
+{
+ re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ reg_errcode_t err;
+ int match = 0;
+ int match_last = -1;
+ int cur_str_idx = re_string_cur_idx (mctx->input);
+ re_dfastate_t *cur_state;
+
+ cur_state = acquire_init_state_context (&err, preg, mctx, cur_str_idx);
+ /* An initial state must not be NULL(invalid state). */
+ if (BE (cur_state == NULL, 0))
+ return -2;
+ if (mctx->state_log != NULL)
+ mctx->state_log[cur_str_idx] = cur_state;
+
+ /* Check OP_OPEN_SUBEXP in the initial state in case that we use them
+ later. E.g. Processing back references. */
+ if (dfa->nbackref)
+ {
+ err = check_subexp_matching_top (dfa, mctx, &cur_state->nodes, 0);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ }
+
+ if (cur_state->has_backref)
+ {
+ err = transit_state_bkref (preg, &cur_state->nodes, mctx);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ }
+
+ /* If the RE accepts NULL string. */
+ if (cur_state->halt)
+ {
+ if (!cur_state->has_constraint
+ || check_halt_state_context (preg, cur_state, mctx, cur_str_idx))
+ {
+ if (!fl_longest_match)
+ return cur_str_idx;
+ else
+ {
+ match_last = cur_str_idx;
+ match = 1;
+ }
+ }
+ }
+
+ while (!re_string_eoi (mctx->input))
+ {
+ cur_state = transit_state (&err, preg, mctx, cur_state,
+ fl_search && !match);
+ if (cur_state == NULL) /* Reached at the invalid state or an error. */
+ {
+ cur_str_idx = re_string_cur_idx (mctx->input);
+ if (BE (err != REG_NOERROR, 0))
+ return -2;
+ if (fl_search && !match)
+ {
+ /* Restart from initial state, since we are searching
+ the point from where matching start. */
+#ifdef RE_ENABLE_I18N
+ if (MB_CUR_MAX == 1
+ || re_string_first_byte (mctx->input, cur_str_idx))
+#endif /* RE_ENABLE_I18N */
+ cur_state = acquire_init_state_context (&err, preg, mctx,
+ cur_str_idx);
+ if (BE (cur_state == NULL && err != REG_NOERROR, 0))
+ return -2;
+ if (mctx->state_log != NULL)
+ mctx->state_log[cur_str_idx] = cur_state;
+ }
+ else if (!fl_longest_match && match)
+ break;
+ else /* (fl_longest_match && match) || (!fl_search && !match) */
+ {
+ if (mctx->state_log == NULL)
+ break;
+ else
+ {
+ int max = mctx->state_log_top;
+ for (; cur_str_idx <= max; ++cur_str_idx)
+ if (mctx->state_log[cur_str_idx] != NULL)
+ break;
+ if (cur_str_idx > max)
+ break;
+ }
+ }
+ }
+
+ if (cur_state != NULL && cur_state->halt)
+ {
+ /* Reached at a halt state.
+ Check the halt state can satisfy the current context. */
+ if (!cur_state->has_constraint
+ || check_halt_state_context (preg, cur_state, mctx,
+ re_string_cur_idx (mctx->input)))
+ {
+ /* We found an appropriate halt state. */
+ match_last = re_string_cur_idx (mctx->input);
+ match = 1;
+ if (!fl_longest_match)
+ break;
+ }
+ }
+ }
+ return match_last;
+}
+
+/* Check NODE match the current context. */
+
+static int check_halt_node_context (dfa, node, context)
+ const re_dfa_t *dfa;
+ int node;
+ unsigned int context;
+{
+ re_token_type_t type = dfa->nodes[node].type;
+ unsigned int constraint = dfa->nodes[node].constraint;
+ if (type != END_OF_RE)
+ return 0;
+ if (!constraint)
+ return 1;
+ if (NOT_SATISFY_NEXT_CONSTRAINT (constraint, context))
+ return 0;
+ return 1;
+}
+
+/* Check the halt state STATE match the current context.
+ Return 0 if not match, if the node, STATE has, is a halt node and
+ match the context, return the node. */
+
+static int
+check_halt_state_context (preg, state, mctx, idx)
+ const regex_t *preg;
+ const re_dfastate_t *state;
+ const re_match_context_t *mctx;
+ int idx;
+{
+ re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ int i;
+ unsigned int context;
+#ifdef DEBUG
+ assert (state->halt);
+#endif
+ context = re_string_context_at (mctx->input, idx, mctx->eflags,
+ preg->newline_anchor);
+ for (i = 0; i < state->nodes.nelem; ++i)
+ if (check_halt_node_context (dfa, state->nodes.elems[i], context))
+ return state->nodes.elems[i];
+ return 0;
+}
+
+/* Compute the next node to which "NFA" transit from NODE("NFA" is a NFA
+ corresponding to the DFA).
+ Return the destination node, and update EPS_VIA_NODES, return -1 in case
+ of errors. */
+
+static int
+proceed_next_node (preg, nregs, regs, mctx, pidx, node, eps_via_nodes, fs)
+ const regex_t *preg;
+ regmatch_t *regs;
+ const re_match_context_t *mctx;
+ int nregs, *pidx, node;
+ re_node_set *eps_via_nodes;
+ struct re_fail_stack_t *fs;
+{
+ re_dfa_t *dfa = (re_dfa_t *)preg->buffer;
+ int i, err, dest_node;
+ dest_node = -1;
+ if (IS_EPSILON_NODE (dfa->nodes[node].type))
+ {
+ re_node_set *cur_nodes = &mctx->state_log[*pidx]->nodes;
+ int ndest, dest_nodes[2];
+ err = re_node_set_insert (eps_via_nodes, node);
+ if (BE (err < 0, 0))
+ return -1;
+ /* Pick up valid destinations. */
+ for (ndest = 0, i = 0; i < dfa->edests[node].nelem; ++i)
+ {
+ int candidate = dfa->edests[node].elems[i];
+ if (!re_node_set_contains (cur_nodes, candidate))
+ continue;
+ dest_nodes[0] = (ndest == 0) ? candidate : dest_nodes[0];
+ dest_nodes[1] = (ndest == 1) ? candidate : dest_nodes[1];
+ ++ndest;
+ }
+ if (ndest <= 1)
+ return ndest == 0 ? -1 : (ndest == 1 ? dest_nodes[0] : 0);
+ /* In order to avoid infinite loop like "(a*)*". */
+ if (re_node_set_contains (eps_via_nodes, dest_nodes[0]))
+ return dest_nodes[1];
+ if (fs != NULL)
+ push_fail_stack (fs, *pidx, dest_nodes, nregs, regs, eps_via_nodes);
+ return dest_nodes[0];
+ }
+ else
+ {
+ int naccepted = 0;
+ re_token_type_t type = dfa->nodes[node].type;
+
+#ifdef RE_ENABLE_I18N
+ if (ACCEPT_MB_NODE (type))
+ naccepted = check_node_accept_bytes (preg, node, mctx->input, *pidx);
+ else
+#endif /* RE_ENABLE_I18N */
+ if (type == OP_BACK_REF)
+ {
+ int subexp_idx = dfa->nodes[node].opr.idx;
+ naccepted = regs[subexp_idx].rm_eo - regs[subexp_idx].rm_so;
+ if (fs != NULL)
+ {
+ if (regs[subexp_idx].rm_so == -1 || regs[subexp_idx].rm_eo == -1)
+ return -1;
+ else if (naccepted)
+ {
+ char *buf = (char *) re_string_get_buffer (mctx->input);
+ if (memcmp (buf + regs[subexp_idx].rm_so, buf + *pidx,
+ naccepted) != 0)
+ return -1;
+ }
+ }
+
+ if (naccepted == 0)
+ {
+ err = re_node_set_insert (eps_via_nodes, node);
+ if (BE (err < 0, 0))
+ return -2;
+ dest_node = dfa->edests[node].elems[0];
+ if (re_node_set_contains (&mctx->state_log[*pidx]->nodes,
+ dest_node))
+ return dest_node;
+ }
+ }
+
+ if (naccepted != 0
+ || check_node_accept (preg, dfa->nodes + node, mctx, *pidx))
+ {
+ dest_node = dfa->nexts[node];
+ *pidx = (naccepted == 0) ? *pidx + 1 : *pidx + naccepted;
+ if (fs && (*pidx > mctx->match_last || mctx->state_log[*pidx] == NULL
+ || !re_node_set_contains (&mctx->state_log[*pidx]->nodes,
+ dest_node)))
+ return -1;
+ re_node_set_empty (eps_via_nodes);
+ return dest_node;
+ }
+ }
+ return -1;
+}
+
+static reg_errcode_t
+push_fail_stack (fs, str_idx, dests, nregs, regs, eps_via_nodes)
+ struct re_fail_stack_t *fs;
+ int str_idx, *dests, nregs;
+ regmatch_t *regs;
+ re_node_set *eps_via_nodes;
+{
+ reg_errcode_t err;
+ int num = fs->num++;
+ if (fs->num == fs->alloc)
+ {
+ struct re_fail_stack_ent_t *new_array;
+ fs->alloc *= 2;
+ new_array = realloc (fs->stack, (sizeof (struct re_fail_stack_ent_t)
+ * fs->alloc));
+ if (new_array == NULL)
+ return REG_ESPACE;
+ fs->stack = new_array;
+ }
+ fs->stack[num].idx = str_idx;
+ fs->stack[num].node = dests[1];
+ fs->stack[num].regs = re_malloc (regmatch_t, nregs);
+ memcpy (fs->stack[num].regs, regs, sizeof (regmatch_t) * nregs);
+ err = re_node_set_init_copy (&fs->stack[num].eps_via_nodes, eps_via_nodes);
+ return err;
+}
+
+static int
+pop_fail_stack (fs, pidx, nregs, regs, eps_via_nodes)
+ struct re_fail_stack_t *fs;
+ int *pidx, nregs;
+ regmatch_t *regs;
+ re_node_set *eps_via_nodes;
+{
+ int num = --fs->num;
+ assert (num >= 0);
+ *pidx = fs->stack[num].idx;
+ memcpy (regs, fs->stack[num].regs, sizeof (regmatch_t) * nregs);
+ re_node_set_free (eps_via_nodes);
+ re_free (fs->stack[num].regs);
+ *eps_via_nodes = fs->stack[num].eps_via_nodes;
+ return fs->stack[num].node;
+}
+
+/* Set the positions where the subexpressions are starts/ends to registers
+ PMATCH.
+ Note: We assume that pmatch[0] is already set, and
+ pmatch[i].rm_so == pmatch[i].rm_eo == -1 (i > 1). */
+
+static reg_errcode_t
+set_regs (preg, mctx, nmatch, pmatch, fl_backtrack)
+ const regex_t *preg;
+ const re_match_context_t *mctx;
+ size_t nmatch;
+ regmatch_t *pmatch;
+ int fl_backtrack;
+{
+ re_dfa_t *dfa = (re_dfa_t *)preg->buffer;
+ int idx, cur_node, real_nmatch;
+ re_node_set eps_via_nodes;
+ struct re_fail_stack_t *fs;
+ struct re_fail_stack_t fs_body = {0, 2, NULL};
+#ifdef DEBUG
+ assert (nmatch > 1);
+ assert (mctx->state_log != NULL);
+#endif
+ if (fl_backtrack)
+ {
+ fs = &fs_body;
+ fs->stack = re_malloc (struct re_fail_stack_ent_t, fs->alloc);
+ }
+ else
+ fs = NULL;
+ cur_node = dfa->init_node;
+ real_nmatch = (nmatch <= preg->re_nsub) ? nmatch : preg->re_nsub + 1;
+ re_node_set_init_empty (&eps_via_nodes);
+ for (idx = pmatch[0].rm_so; idx <= pmatch[0].rm_eo ;)
+ {
+ update_regs (dfa, pmatch, cur_node, idx, real_nmatch);
+ if (idx == pmatch[0].rm_eo && cur_node == mctx->last_node)
+ {
+ int reg_idx;
+ if (fs)
+ {
+ for (reg_idx = 0; reg_idx < nmatch; ++reg_idx)
+ if (pmatch[reg_idx].rm_so > -1 && pmatch[reg_idx].rm_eo == -1)
+ break;
+ if (reg_idx == nmatch)
+ {
+ re_node_set_free (&eps_via_nodes);
+ return free_fail_stack_return (fs);
+ }
+ cur_node = pop_fail_stack (fs, &idx, nmatch, pmatch,
+ &eps_via_nodes);
+ }
+ else
+ {
+ re_node_set_free (&eps_via_nodes);
+ return REG_NOERROR;
+ }
+ }
+
+ /* Proceed to next node. */
+ cur_node = proceed_next_node (preg, nmatch, pmatch, mctx, &idx, cur_node,
+ &eps_via_nodes, fs);
+
+ if (BE (cur_node < 0, 0))
+ {
+ if (cur_node == -2)
+ return REG_ESPACE;
+ if (fs)
+ cur_node = pop_fail_stack (fs, &idx, nmatch, pmatch,
+ &eps_via_nodes);
+ else
+ {
+ re_node_set_free (&eps_via_nodes);
+ return REG_NOMATCH;
+ }
+ }
+ }
+ re_node_set_free (&eps_via_nodes);
+ return free_fail_stack_return (fs);
+}
+
+static reg_errcode_t
+free_fail_stack_return (fs)
+ struct re_fail_stack_t *fs;
+{
+ if (fs)
+ {
+ int fs_idx;
+ for (fs_idx = 0; fs_idx < fs->num; ++fs_idx)
+ {
+ re_node_set_free (&fs->stack[fs_idx].eps_via_nodes);
+ re_free (fs->stack[fs_idx].regs);
+ }
+ re_free (fs->stack);
+ }
+ return REG_NOERROR;
+}
+
+static void
+update_regs (dfa, pmatch, cur_node, cur_idx, nmatch)
+ re_dfa_t *dfa;
+ regmatch_t *pmatch;
+ int cur_node, cur_idx, nmatch;
+{
+ int type = dfa->nodes[cur_node].type;
+ int reg_num;
+ if (type != OP_OPEN_SUBEXP && type != OP_CLOSE_SUBEXP)
+ return;
+ reg_num = dfa->nodes[cur_node].opr.idx + 1;
+ if (reg_num >= nmatch)
+ return;
+ if (type == OP_OPEN_SUBEXP)
+ {
+ /* We are at the first node of this sub expression. */
+ pmatch[reg_num].rm_so = cur_idx;
+ pmatch[reg_num].rm_eo = -1;
+ }
+ else if (type == OP_CLOSE_SUBEXP)
+ /* We are at the first node of this sub expression. */
+ pmatch[reg_num].rm_eo = cur_idx;
+}
+
+#define NUMBER_OF_STATE 1
+
+/* This function checks the STATE_LOG from the SCTX->last_str_idx to 0
+ and sift the nodes in each states according to the following rules.
+ Updated state_log will be wrote to STATE_LOG.
+
+ Rules: We throw away the Node `a' in the STATE_LOG[STR_IDX] if...
+ 1. When STR_IDX == MATCH_LAST(the last index in the state_log):
+ If `a' isn't the LAST_NODE and `a' can't epsilon transit to
+ the LAST_NODE, we throw away the node `a'.
+ 2. When 0 <= STR_IDX < MATCH_LAST and `a' accepts
+ string `s' and transit to `b':
+ i. If 'b' isn't in the STATE_LOG[STR_IDX+strlen('s')], we throw
+ away the node `a'.
+ ii. If 'b' is in the STATE_LOG[STR_IDX+strlen('s')] but 'b' is
+ throwed away, we throw away the node `a'.
+ 3. When 0 <= STR_IDX < n and 'a' epsilon transit to 'b':
+ i. If 'b' isn't in the STATE_LOG[STR_IDX], we throw away the
+ node `a'.
+ ii. If 'b' is in the STATE_LOG[STR_IDX] but 'b' is throwed away,
+ we throw away the node `a'. */
+
+#define STATE_NODE_CONTAINS(state,node) \
+ ((state) != NULL && re_node_set_contains (&(state)->nodes, node))
+
+static reg_errcode_t
+sift_states_backward (preg, mctx, sctx)
+ const regex_t *preg;
+ re_match_context_t *mctx;
+ re_sift_context_t *sctx;
+{
+ reg_errcode_t err;
+ re_dfa_t *dfa = (re_dfa_t *)preg->buffer;
+ int null_cnt = 0;
+ int str_idx = sctx->last_str_idx;
+ re_node_set cur_dest;
+ re_node_set *cur_src; /* Points the state_log[str_idx]->nodes */
+
+#ifdef DEBUG
+ assert (mctx->state_log != NULL && mctx->state_log[str_idx] != NULL);
+#endif
+ cur_src = &mctx->state_log[str_idx]->nodes;
+
+ /* Build sifted state_log[str_idx]. It has the nodes which can epsilon
+ transit to the last_node and the last_node itself. */
+ err = re_node_set_init_1 (&cur_dest, sctx->last_node);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ err = update_cur_sifted_state (preg, mctx, sctx, str_idx, &cur_dest);
+ if (BE (err != REG_NOERROR, 0))
+ goto free_return;
+
+ /* Then check each states in the state_log. */
+ while (str_idx > 0)
+ {
+ int i, ret;
+ /* Update counters. */
+ null_cnt = (sctx->sifted_states[str_idx] == NULL) ? null_cnt + 1 : 0;
+ if (null_cnt > mctx->max_mb_elem_len)
+ {
+ memset (sctx->sifted_states, '\0',
+ sizeof (re_dfastate_t *) * str_idx);
+ re_node_set_free (&cur_dest);
+ return REG_NOERROR;
+ }
+ re_node_set_empty (&cur_dest);
+ --str_idx;
+ cur_src = ((mctx->state_log[str_idx] == NULL) ? &empty_set
+ : &mctx->state_log[str_idx]->nodes);
+
+ /* Then build the next sifted state.
+ We build the next sifted state on `cur_dest', and update
+ `sifted_states[str_idx]' with `cur_dest'.
+ Note:
+ `cur_dest' is the sifted state from `state_log[str_idx + 1]'.
+ `cur_src' points the node_set of the old `state_log[str_idx]'. */
+ for (i = 0; i < cur_src->nelem; i++)
+ {
+ int prev_node = cur_src->elems[i];
+ int naccepted = 0;
+ re_token_type_t type = dfa->nodes[prev_node].type;
+
+ if (IS_EPSILON_NODE(type))
+ continue;
+#ifdef RE_ENABLE_I18N
+ /* If the node may accept `multi byte'. */
+ if (ACCEPT_MB_NODE (type))
+ naccepted = sift_states_iter_mb (preg, mctx, sctx, prev_node,
+ str_idx, sctx->last_str_idx);
+
+#endif /* RE_ENABLE_I18N */
+ /* We don't check backreferences here.
+ See update_cur_sifted_state(). */
+
+ if (!naccepted
+ && check_node_accept (preg, dfa->nodes + prev_node, mctx,
+ str_idx)
+ && STATE_NODE_CONTAINS (sctx->sifted_states[str_idx + 1],
+ dfa->nexts[prev_node]))
+ naccepted = 1;
+
+ if (naccepted == 0)
+ continue;
+
+ if (sctx->limits.nelem)
+ {
+ int to_idx = str_idx + naccepted;
+ if (check_dst_limits (dfa, &sctx->limits, mctx,
+ dfa->nexts[prev_node], to_idx,
+ prev_node, str_idx))
+ continue;
+ }
+ ret = re_node_set_insert (&cur_dest, prev_node);
+ if (BE (ret == -1, 0))
+ {
+ err = REG_ESPACE;
+ goto free_return;
+ }
+ }
+
+ /* Add all the nodes which satisfy the following conditions:
+ - It can epsilon transit to a node in CUR_DEST.
+ - It is in CUR_SRC.
+ And update state_log. */
+ err = update_cur_sifted_state (preg, mctx, sctx, str_idx, &cur_dest);
+ if (BE (err != REG_NOERROR, 0))
+ goto free_return;
+ }
+ err = REG_NOERROR;
+ free_return:
+ re_node_set_free (&cur_dest);
+ return err;
+}
+
+/* Helper functions. */
+
+static inline reg_errcode_t
+clean_state_log_if_need (mctx, next_state_log_idx)
+ re_match_context_t *mctx;
+ int next_state_log_idx;
+{
+ int top = mctx->state_log_top;
+
+ if (next_state_log_idx >= mctx->input->bufs_len
+ || (next_state_log_idx >= mctx->input->valid_len
+ && mctx->input->valid_len < mctx->input->len))
+ {
+ reg_errcode_t err;
+ err = extend_buffers (mctx);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ }
+
+ if (top < next_state_log_idx)
+ {
+ memset (mctx->state_log + top + 1, '\0',
+ sizeof (re_dfastate_t *) * (next_state_log_idx - top));
+ mctx->state_log_top = next_state_log_idx;
+ }
+ return REG_NOERROR;
+}
+
+static reg_errcode_t
+merge_state_array (dfa, dst, src, num)
+ re_dfa_t *dfa;
+ re_dfastate_t **dst;
+ re_dfastate_t **src;
+ int num;
+{
+ int st_idx;
+ reg_errcode_t err;
+ for (st_idx = 0; st_idx < num; ++st_idx)
+ {
+ if (dst[st_idx] == NULL)
+ dst[st_idx] = src[st_idx];
+ else if (src[st_idx] != NULL)
+ {
+ re_node_set merged_set;
+ err = re_node_set_init_union (&merged_set, &dst[st_idx]->nodes,
+ &src[st_idx]->nodes);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ dst[st_idx] = re_acquire_state (&err, dfa, &merged_set);
+ re_node_set_free (&merged_set);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ }
+ }
+ return REG_NOERROR;
+}
+
+static reg_errcode_t
+update_cur_sifted_state (preg, mctx, sctx, str_idx, dest_nodes)
+ const regex_t *preg;
+ re_match_context_t *mctx;
+ re_sift_context_t *sctx;
+ int str_idx;
+ re_node_set *dest_nodes;
+{
+ reg_errcode_t err;
+ re_dfa_t *dfa = (re_dfa_t *)preg->buffer;
+ const re_node_set *candidates;
+ candidates = ((mctx->state_log[str_idx] == NULL) ? &empty_set
+ : &mctx->state_log[str_idx]->nodes);
+
+ /* At first, add the nodes which can epsilon transit to a node in
+ DEST_NODE. */
+ if (dest_nodes->nelem)
+ {
+ err = add_epsilon_src_nodes (dfa, dest_nodes, candidates);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ }
+
+ /* Then, check the limitations in the current sift_context. */
+ if (dest_nodes->nelem && sctx->limits.nelem)
+ {
+ err = check_subexp_limits (dfa, dest_nodes, candidates, &sctx->limits,
+ mctx->bkref_ents, str_idx);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ }
+
+ /* Update state_log. */
+ sctx->sifted_states[str_idx] = re_acquire_state (&err, dfa, dest_nodes);
+ if (BE (sctx->sifted_states[str_idx] == NULL && err != REG_NOERROR, 0))
+ return err;
+
+ if ((mctx->state_log[str_idx] != NULL
+ && mctx->state_log[str_idx]->has_backref))
+ {
+ err = sift_states_bkref (preg, mctx, sctx, str_idx, dest_nodes);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ }
+ return REG_NOERROR;
+}
+
+static reg_errcode_t
+add_epsilon_src_nodes (dfa, dest_nodes, candidates)
+ re_dfa_t *dfa;
+ re_node_set *dest_nodes;
+ const re_node_set *candidates;
+{
+ reg_errcode_t err;
+ int src_idx;
+ re_node_set src_copy;
+
+ err = re_node_set_init_copy (&src_copy, dest_nodes);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ for (src_idx = 0; src_idx < src_copy.nelem; ++src_idx)
+ {
+ err = re_node_set_add_intersect (dest_nodes, candidates,
+ dfa->inveclosures
+ + src_copy.elems[src_idx]);
+ if (BE (err != REG_NOERROR, 0))
+ {
+ re_node_set_free (&src_copy);
+ return err;
+ }
+ }
+ re_node_set_free (&src_copy);
+ return REG_NOERROR;
+}
+
+static reg_errcode_t
+sub_epsilon_src_nodes (dfa, node, dest_nodes, candidates)
+ re_dfa_t *dfa;
+ int node;
+ re_node_set *dest_nodes;
+ const re_node_set *candidates;
+{
+ int ecl_idx;
+ reg_errcode_t err;
+ re_node_set *inv_eclosure = dfa->inveclosures + node;
+ re_node_set except_nodes;
+ re_node_set_init_empty (&except_nodes);
+ for (ecl_idx = 0; ecl_idx < inv_eclosure->nelem; ++ecl_idx)
+ {
+ int cur_node = inv_eclosure->elems[ecl_idx];
+ if (cur_node == node)
+ continue;
+ if (IS_EPSILON_NODE (dfa->nodes[cur_node].type))
+ {
+ int edst1 = dfa->edests[cur_node].elems[0];
+ int edst2 = ((dfa->edests[cur_node].nelem > 1)
+ ? dfa->edests[cur_node].elems[1] : -1);
+ if ((!re_node_set_contains (inv_eclosure, edst1)
+ && re_node_set_contains (dest_nodes, edst1))
+ || (edst2 > 0
+ && !re_node_set_contains (inv_eclosure, edst2)
+ && re_node_set_contains (dest_nodes, edst2)))
+ {
+ err = re_node_set_add_intersect (&except_nodes, candidates,
+ dfa->inveclosures + cur_node);
+ if (BE (err != REG_NOERROR, 0))
+ {
+ re_node_set_free (&except_nodes);
+ return err;
+ }
+ }
+ }
+ }
+ for (ecl_idx = 0; ecl_idx < inv_eclosure->nelem; ++ecl_idx)
+ {
+ int cur_node = inv_eclosure->elems[ecl_idx];
+ if (!re_node_set_contains (&except_nodes, cur_node))
+ {
+ int idx = re_node_set_contains (dest_nodes, cur_node) - 1;
+ re_node_set_remove_at (dest_nodes, idx);
+ }
+ }
+ re_node_set_free (&except_nodes);
+ return REG_NOERROR;
+}
+
+static int
+check_dst_limits (dfa, limits, mctx, dst_node, dst_idx, src_node, src_idx)
+ re_dfa_t *dfa;
+ re_node_set *limits;
+ re_match_context_t *mctx;
+ int dst_node, dst_idx, src_node, src_idx;
+{
+ int lim_idx, src_pos, dst_pos;
+
+ for (lim_idx = 0; lim_idx < limits->nelem; ++lim_idx)
+ {
+ int subexp_idx;
+ struct re_backref_cache_entry *ent;
+ ent = mctx->bkref_ents + limits->elems[lim_idx];
+ subexp_idx = dfa->nodes[ent->node].opr.idx - 1;
+
+ dst_pos = check_dst_limits_calc_pos (dfa, mctx, limits->elems[lim_idx],
+ dfa->eclosures + dst_node,
+ subexp_idx, dst_node, dst_idx);
+ src_pos = check_dst_limits_calc_pos (dfa, mctx, limits->elems[lim_idx],
+ dfa->eclosures + src_node,
+ subexp_idx, src_node, src_idx);
+
+ /* In case of:
+ <src> <dst> ( <subexp> )
+ ( <subexp> ) <src> <dst>
+ ( <subexp1> <src> <subexp2> <dst> <subexp3> ) */
+ if (src_pos == dst_pos)
+ continue; /* This is unrelated limitation. */
+ else
+ return 1;
+ }
+ return 0;
+}
+
+static int
+check_dst_limits_calc_pos (dfa, mctx, limit, eclosures, subexp_idx, node,
+ str_idx)
+ re_dfa_t *dfa;
+ re_match_context_t *mctx;
+ re_node_set *eclosures;
+ int limit, subexp_idx, node, str_idx;
+{
+ struct re_backref_cache_entry *lim = mctx->bkref_ents + limit;
+ int pos = (str_idx < lim->subexp_from ? -1
+ : (lim->subexp_to < str_idx ? 1 : 0));
+ if (pos == 0
+ && (str_idx == lim->subexp_from || str_idx == lim->subexp_to))
+ {
+ int node_idx;
+ for (node_idx = 0; node_idx < eclosures->nelem; ++node_idx)
+ {
+ int node = eclosures->elems[node_idx];
+ re_token_type_t type= dfa->nodes[node].type;
+ if (type == OP_BACK_REF)
+ {
+ int bi = search_cur_bkref_entry (mctx, str_idx);
+ for (; bi < mctx->nbkref_ents; ++bi)
+ {
+ struct re_backref_cache_entry *ent = mctx->bkref_ents + bi;
+ if (ent->str_idx > str_idx)
+ break;
+ if (ent->node == node && ent->subexp_from == ent->subexp_to)
+ {
+ int cpos, dst;
+ dst = dfa->edests[node].elems[0];
+ cpos = check_dst_limits_calc_pos (dfa, mctx, limit,
+ dfa->eclosures + dst,
+ subexp_idx, dst,
+ str_idx);
+ if ((str_idx == lim->subexp_from && cpos == -1)
+ || (str_idx == lim->subexp_to && cpos == 0))
+ return cpos;
+ }
+ }
+ }
+ if (type == OP_OPEN_SUBEXP && subexp_idx == dfa->nodes[node].opr.idx
+ && str_idx == lim->subexp_from)
+ {
+ pos = -1;
+ break;
+ }
+ if (type == OP_CLOSE_SUBEXP && subexp_idx == dfa->nodes[node].opr.idx
+ && str_idx == lim->subexp_to)
+ break;
+ }
+ if (node_idx == eclosures->nelem && str_idx == lim->subexp_to)
+ pos = 1;
+ }
+ return pos;
+}
+
+/* Check the limitations of sub expressions LIMITS, and remove the nodes
+ which are against limitations from DEST_NODES. */
+
+static reg_errcode_t
+check_subexp_limits (dfa, dest_nodes, candidates, limits, bkref_ents, str_idx)
+ re_dfa_t *dfa;
+ re_node_set *dest_nodes;
+ const re_node_set *candidates;
+ re_node_set *limits;
+ struct re_backref_cache_entry *bkref_ents;
+ int str_idx;
+{
+ reg_errcode_t err;
+ int node_idx, lim_idx;
+
+ for (lim_idx = 0; lim_idx < limits->nelem; ++lim_idx)
+ {
+ int subexp_idx;
+ struct re_backref_cache_entry *ent;
+ ent = bkref_ents + limits->elems[lim_idx];
+
+ if (str_idx <= ent->subexp_from || ent->str_idx < str_idx)
+ continue; /* This is unrelated limitation. */
+
+ subexp_idx = dfa->nodes[ent->node].opr.idx - 1;
+ if (ent->subexp_to == str_idx)
+ {
+ int ops_node = -1;
+ int cls_node = -1;
+ for (node_idx = 0; node_idx < dest_nodes->nelem; ++node_idx)
+ {
+ int node = dest_nodes->elems[node_idx];
+ re_token_type_t type= dfa->nodes[node].type;
+ if (type == OP_OPEN_SUBEXP
+ && subexp_idx == dfa->nodes[node].opr.idx)
+ ops_node = node;
+ else if (type == OP_CLOSE_SUBEXP
+ && subexp_idx == dfa->nodes[node].opr.idx)
+ cls_node = node;
+ }
+
+ /* Check the limitation of the open subexpression. */
+ /* Note that (ent->subexp_to = str_idx != ent->subexp_from). */
+ if (ops_node >= 0)
+ {
+ err = sub_epsilon_src_nodes(dfa, ops_node, dest_nodes,
+ candidates);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ }
+ /* Check the limitation of the close subexpression. */
+ for (node_idx = 0; node_idx < dest_nodes->nelem; ++node_idx)
+ {
+ int node = dest_nodes->elems[node_idx];
+ if (!re_node_set_contains (dfa->inveclosures + node, cls_node)
+ && !re_node_set_contains (dfa->eclosures + node, cls_node))
+ {
+ /* It is against this limitation.
+ Remove it form the current sifted state. */
+ err = sub_epsilon_src_nodes(dfa, node, dest_nodes,
+ candidates);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ --node_idx;
+ }
+ }
+ }
+ else /* (ent->subexp_to != str_idx) */
+ {
+ for (node_idx = 0; node_idx < dest_nodes->nelem; ++node_idx)
+ {
+ int node = dest_nodes->elems[node_idx];
+ re_token_type_t type= dfa->nodes[node].type;
+ if (type == OP_CLOSE_SUBEXP || type == OP_OPEN_SUBEXP)
+ {
+ if (subexp_idx != dfa->nodes[node].opr.idx)
+ continue;
+ if ((type == OP_CLOSE_SUBEXP && ent->subexp_to != str_idx)
+ || (type == OP_OPEN_SUBEXP))
+ {
+ /* It is against this limitation.
+ Remove it form the current sifted state. */
+ err = sub_epsilon_src_nodes(dfa, node, dest_nodes,
+ candidates);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ }
+ }
+ }
+ }
+ }
+ return REG_NOERROR;
+}
+
+static reg_errcode_t
+sift_states_bkref (preg, mctx, sctx, str_idx, dest_nodes)
+ const regex_t *preg;
+ re_match_context_t *mctx;
+ re_sift_context_t *sctx;
+ int str_idx;
+ re_node_set *dest_nodes;
+{
+ reg_errcode_t err;
+ re_dfa_t *dfa = (re_dfa_t *)preg->buffer;
+ int node_idx, node;
+ re_sift_context_t local_sctx;
+ const re_node_set *candidates;
+ candidates = ((mctx->state_log[str_idx] == NULL) ? &empty_set
+ : &mctx->state_log[str_idx]->nodes);
+ local_sctx.sifted_states = NULL; /* Mark that it hasn't been initialized. */
+
+ for (node_idx = 0; node_idx < candidates->nelem; ++node_idx)
+ {
+ int cur_bkref_idx = re_string_cur_idx (mctx->input);
+ re_token_type_t type;
+ node = candidates->elems[node_idx];
+ type = dfa->nodes[node].type;
+ if (node == sctx->cur_bkref && str_idx == cur_bkref_idx)
+ continue;
+ /* Avoid infinite loop for the REs like "()\1+". */
+ if (node == sctx->last_node && str_idx == sctx->last_str_idx)
+ continue;
+ if (type == OP_BACK_REF)
+ {
+ int enabled_idx = search_cur_bkref_entry (mctx, str_idx);
+ for (; enabled_idx < mctx->nbkref_ents; ++enabled_idx)
+ {
+ int disabled_idx, subexp_len, to_idx, dst_node;
+ struct re_backref_cache_entry *entry;
+ entry = mctx->bkref_ents + enabled_idx;
+ if (entry->str_idx > str_idx)
+ break;
+ if (entry->node != node)
+ continue;
+ subexp_len = entry->subexp_to - entry->subexp_from;
+ to_idx = str_idx + subexp_len;
+ dst_node = (subexp_len ? dfa->nexts[node]
+ : dfa->edests[node].elems[0]);
+
+ if (to_idx > sctx->last_str_idx
+ || sctx->sifted_states[to_idx] == NULL
+ || !STATE_NODE_CONTAINS (sctx->sifted_states[to_idx],
+ dst_node)
+ || check_dst_limits (dfa, &sctx->limits, mctx, node,
+ str_idx, dst_node, to_idx))
+ continue;
+ {
+ re_dfastate_t *cur_state;
+ entry->flag = 0;
+ for (disabled_idx = enabled_idx + 1;
+ disabled_idx < mctx->nbkref_ents; ++disabled_idx)
+ {
+ struct re_backref_cache_entry *entry2;
+ entry2 = mctx->bkref_ents + disabled_idx;
+ if (entry2->str_idx > str_idx)
+ break;
+ entry2->flag = (entry2->node == node) ? 1 : entry2->flag;
+ }
+
+ if (local_sctx.sifted_states == NULL)
+ {
+ local_sctx = *sctx;
+ err = re_node_set_init_copy (&local_sctx.limits,
+ &sctx->limits);
+ if (BE (err != REG_NOERROR, 0))
+ goto free_return;
+ }
+ local_sctx.last_node = node;
+ local_sctx.last_str_idx = str_idx;
+ err = re_node_set_insert (&local_sctx.limits, enabled_idx);
+ if (BE (err < 0, 0))
+ {
+ err = REG_ESPACE;
+ goto free_return;
+ }
+ cur_state = local_sctx.sifted_states[str_idx];
+ err = sift_states_backward (preg, mctx, &local_sctx);
+ if (BE (err != REG_NOERROR, 0))
+ goto free_return;
+ if (sctx->limited_states != NULL)
+ {
+ err = merge_state_array (dfa, sctx->limited_states,
+ local_sctx.sifted_states,
+ str_idx + 1);
+ if (BE (err != REG_NOERROR, 0))
+ goto free_return;
+ }
+ local_sctx.sifted_states[str_idx] = cur_state;
+ re_node_set_remove (&local_sctx.limits, enabled_idx);
+ /* We must not use the variable entry here, since
+ mctx->bkref_ents might be realloced. */
+ mctx->bkref_ents[enabled_idx].flag = 1;
+ }
+ }
+ enabled_idx = search_cur_bkref_entry (mctx, str_idx);
+ for (; enabled_idx < mctx->nbkref_ents; ++enabled_idx)
+ {
+ struct re_backref_cache_entry *entry;
+ entry = mctx->bkref_ents + enabled_idx;
+ if (entry->str_idx > str_idx)
+ break;
+ if (entry->node == node)
+ entry->flag = 0;
+ }
+ }
+ }
+ err = REG_NOERROR;
+ free_return:
+ if (local_sctx.sifted_states != NULL)
+ {
+ re_node_set_free (&local_sctx.limits);
+ }
+
+ return err;
+}
+
+
+#ifdef RE_ENABLE_I18N
+static int
+sift_states_iter_mb (preg, mctx, sctx, node_idx, str_idx, max_str_idx)
+ const regex_t *preg;
+ const re_match_context_t *mctx;
+ re_sift_context_t *sctx;
+ int node_idx, str_idx, max_str_idx;
+{
+ re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ int naccepted;
+ /* Check the node can accept `multi byte'. */
+ naccepted = check_node_accept_bytes (preg, node_idx, mctx->input, str_idx);
+ if (naccepted > 0 && str_idx + naccepted <= max_str_idx &&
+ !STATE_NODE_CONTAINS (sctx->sifted_states[str_idx + naccepted],
+ dfa->nexts[node_idx]))
+ /* The node can't accept the `multi byte', or the
+ destination was already throwed away, then the node
+ could't accept the current input `multi byte'. */
+ naccepted = 0;
+ /* Otherwise, it is sure that the node could accept
+ `naccepted' bytes input. */
+ return naccepted;
+}
+#endif /* RE_ENABLE_I18N */
+
+\f
+/* Functions for state transition. */
+
+/* Return the next state to which the current state STATE will transit by
+ accepting the current input byte, and update STATE_LOG if necessary.
+ If STATE can accept a multibyte char/collating element/back reference
+ update the destination of STATE_LOG. */
+
+static re_dfastate_t *
+transit_state (err, preg, mctx, state, fl_search)
+ reg_errcode_t *err;
+ const regex_t *preg;
+ re_match_context_t *mctx;
+ re_dfastate_t *state;
+ int fl_search;
+{
+ re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ re_dfastate_t **trtable, *next_state;
+ unsigned char ch;
+ int cur_idx;
+
+ if (re_string_cur_idx (mctx->input) + 1 >= mctx->input->bufs_len
+ || (re_string_cur_idx (mctx->input) + 1 >= mctx->input->valid_len
+ && mctx->input->valid_len < mctx->input->len))
+ {
+ *err = extend_buffers (mctx);
+ if (BE (*err != REG_NOERROR, 0))
+ return NULL;
+ }
+
+ *err = REG_NOERROR;
+ if (state == NULL)
+ {
+ next_state = state;
+ re_string_skip_bytes (mctx->input, 1);
+ }
+ else
+ {
+#ifdef RE_ENABLE_I18N
+ /* If the current state can accept multibyte. */
+ if (state->accept_mb)
+ {
+ *err = transit_state_mb (preg, state, mctx);
+ if (BE (*err != REG_NOERROR, 0))
+ return NULL;
+ }
+#endif /* RE_ENABLE_I18N */
+
+ /* Then decide the next state with the single byte. */
+ if (1)
+ {
+ /* Use transition table */
+ ch = re_string_fetch_byte (mctx->input);
+ trtable = fl_search ? state->trtable_search : state->trtable;
+ if (trtable == NULL)
+ {
+ trtable = build_trtable (preg, state, fl_search);
+ if (fl_search)
+ state->trtable_search = trtable;
+ else
+ state->trtable = trtable;
+ }
+ next_state = trtable[ch];
+ }
+ else
+ {
+ /* don't use transition table */
+ next_state = transit_state_sb (err, preg, state, fl_search, mctx);
+ if (BE (next_state == NULL && err != REG_NOERROR, 0))
+ return NULL;
+ }
+ }
+
+ cur_idx = re_string_cur_idx (mctx->input);
+ /* Update the state_log if we need. */
+ if (mctx->state_log != NULL)
+ {
+ if (cur_idx > mctx->state_log_top)
+ {
+ mctx->state_log[cur_idx] = next_state;
+ mctx->state_log_top = cur_idx;
+ }
+ else if (mctx->state_log[cur_idx] == 0)
+ {
+ mctx->state_log[cur_idx] = next_state;
+ }
+ else
+ {
+ re_dfastate_t *pstate;
+ unsigned int context;
+ re_node_set next_nodes, *log_nodes, *table_nodes = NULL;
+ /* If (state_log[cur_idx] != 0), it implies that cur_idx is
+ the destination of a multibyte char/collating element/
+ back reference. Then the next state is the union set of
+ these destinations and the results of the transition table. */
+ pstate = mctx->state_log[cur_idx];
+ log_nodes = pstate->entrance_nodes;
+ if (next_state != NULL)
+ {
+ table_nodes = next_state->entrance_nodes;
+ *err = re_node_set_init_union (&next_nodes, table_nodes,
+ log_nodes);
+ if (BE (*err != REG_NOERROR, 0))
+ return NULL;
+ }
+ else
+ next_nodes = *log_nodes;
+ /* Note: We already add the nodes of the initial state,
+ then we don't need to add them here. */
+
+ context = re_string_context_at (mctx->input,
+ re_string_cur_idx (mctx->input) - 1,
+ mctx->eflags, preg->newline_anchor);
+ next_state = mctx->state_log[cur_idx]
+ = re_acquire_state_context (err, dfa, &next_nodes, context);
+ /* We don't need to check errors here, since the return value of
+ this function is next_state and ERR is already set. */
+
+ if (table_nodes != NULL)
+ re_node_set_free (&next_nodes);
+ }
+ }
+
+ /* Check OP_OPEN_SUBEXP in the current state in case that we use them
+ later. We must check them here, since the back references in the
+ next state might use them. */
+ if (dfa->nbackref && next_state/* && fl_process_bkref */)
+ {
+ *err = check_subexp_matching_top (dfa, mctx, &next_state->nodes,
+ cur_idx);
+ if (BE (*err != REG_NOERROR, 0))
+ return NULL;
+ }
+
+ /* If the next state has back references. */
+ if (next_state != NULL && next_state->has_backref)
+ {
+ *err = transit_state_bkref (preg, &next_state->nodes, mctx);
+ if (BE (*err != REG_NOERROR, 0))
+ return NULL;
+ next_state = mctx->state_log[cur_idx];
+ }
+ return next_state;
+}
+
+/* Helper functions for transit_state. */
+
+/* From the node set CUR_NODES, pick up the nodes whose types are
+ OP_OPEN_SUBEXP and which have corresponding back references in the regular
+ expression. And register them to use them later for evaluating the
+ correspoding back references. */
+
+static reg_errcode_t
+check_subexp_matching_top (dfa, mctx, cur_nodes, str_idx)
+ re_dfa_t *dfa;
+ re_match_context_t *mctx;
+ re_node_set *cur_nodes;
+ int str_idx;
+{
+ int node_idx;
+ reg_errcode_t err;
+
+ /* TODO: This isn't efficient.
+ Because there might be more than one nodes whose types are
+ OP_OPEN_SUBEXP and whose index is SUBEXP_IDX, we must check all
+ nodes.
+ E.g. RE: (a){2} */
+ for (node_idx = 0; node_idx < cur_nodes->nelem; ++node_idx)
+ {
+ int node = cur_nodes->elems[node_idx];
+ if (dfa->nodes[node].type == OP_OPEN_SUBEXP
+ && dfa->used_bkref_map & (1 << dfa->nodes[node].opr.idx))
+ {
+ err = match_ctx_add_subtop (mctx, node, str_idx);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ }
+ }
+ return REG_NOERROR;
+}
+
+/* Return the next state to which the current state STATE will transit by
+ accepting the current input byte. */
+
+static re_dfastate_t *
+transit_state_sb (err, preg, state, fl_search, mctx)
+ reg_errcode_t *err;
+ const regex_t *preg;
+ re_dfastate_t *state;
+ int fl_search;
+ re_match_context_t *mctx;
+{
+ re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ re_node_set next_nodes;
+ re_dfastate_t *next_state;
+ int node_cnt, cur_str_idx = re_string_cur_idx (mctx->input);
+ unsigned int context;
+
+ *err = re_node_set_alloc (&next_nodes, state->nodes.nelem + 1);
+ if (BE (*err != REG_NOERROR, 0))
+ return NULL;
+ for (node_cnt = 0; node_cnt < state->nodes.nelem; ++node_cnt)
+ {
+ int cur_node = state->nodes.elems[node_cnt];
+ if (check_node_accept (preg, dfa->nodes + cur_node, mctx, cur_str_idx))
+ {
+ *err = re_node_set_merge (&next_nodes,
+ dfa->eclosures + dfa->nexts[cur_node]);
+ if (BE (*err != REG_NOERROR, 0))
+ {
+ re_node_set_free (&next_nodes);
+ return NULL;
+ }
+ }
+ }
+ if (fl_search)
+ {
+#ifdef RE_ENABLE_I18N
+ int not_initial = 0;
+ if (MB_CUR_MAX > 1)
+ for (node_cnt = 0; node_cnt < next_nodes.nelem; ++node_cnt)
+ if (dfa->nodes[next_nodes.elems[node_cnt]].type == CHARACTER)
+ {
+ not_initial = dfa->nodes[next_nodes.elems[node_cnt]].mb_partial;
+ break;
+ }
+ if (!not_initial)
+#endif
+ {
+ *err = re_node_set_merge (&next_nodes,
+ dfa->init_state->entrance_nodes);
+ if (BE (*err != REG_NOERROR, 0))
+ {
+ re_node_set_free (&next_nodes);
+ return NULL;
+ }
+ }
+ }
+ context = re_string_context_at (mctx->input, cur_str_idx, mctx->eflags,
+ preg->newline_anchor);
+ next_state = re_acquire_state_context (err, dfa, &next_nodes, context);
+ /* We don't need to check errors here, since the return value of
+ this function is next_state and ERR is already set. */
+
+ re_node_set_free (&next_nodes);
+ re_string_skip_bytes (mctx->input, 1);
+ return next_state;
+}
+
+#ifdef RE_ENABLE_I18N
+static reg_errcode_t
+transit_state_mb (preg, pstate, mctx)
+ const regex_t *preg;
+ re_dfastate_t *pstate;
+ re_match_context_t *mctx;
+{
+ reg_errcode_t err;
+ re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ int i;
+
+ for (i = 0; i < pstate->nodes.nelem; ++i)
+ {
+ re_node_set dest_nodes, *new_nodes;
+ int cur_node_idx = pstate->nodes.elems[i];
+ int naccepted = 0, dest_idx;
+ unsigned int context;
+ re_dfastate_t *dest_state;
+
+ if (dfa->nodes[cur_node_idx].constraint)
+ {
+ context = re_string_context_at (mctx->input,
+ re_string_cur_idx (mctx->input),
+ mctx->eflags, preg->newline_anchor);
+ if (NOT_SATISFY_NEXT_CONSTRAINT (dfa->nodes[cur_node_idx].constraint,
+ context))
+ continue;
+ }
+
+ /* How many bytes the node can accepts? */
+ if (ACCEPT_MB_NODE (dfa->nodes[cur_node_idx].type))
+ naccepted = check_node_accept_bytes (preg, cur_node_idx, mctx->input,
+ re_string_cur_idx (mctx->input));
+ if (naccepted == 0)
+ continue;
+
+ /* The node can accepts `naccepted' bytes. */
+ dest_idx = re_string_cur_idx (mctx->input) + naccepted;
+ mctx->max_mb_elem_len = ((mctx->max_mb_elem_len < naccepted) ? naccepted
+ : mctx->max_mb_elem_len);
+ err = clean_state_log_if_need (mctx, dest_idx);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+#ifdef DEBUG
+ assert (dfa->nexts[cur_node_idx] != -1);
+#endif
+ /* `cur_node_idx' may point the entity of the OP_CONTEXT_NODE,
+ then we use pstate->nodes.elems[i] instead. */
+ new_nodes = dfa->eclosures + dfa->nexts[pstate->nodes.elems[i]];
+
+ dest_state = mctx->state_log[dest_idx];
+ if (dest_state == NULL)
+ dest_nodes = *new_nodes;
+ else
+ {
+ err = re_node_set_init_union (&dest_nodes,
+ dest_state->entrance_nodes, new_nodes);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ }
+ context = re_string_context_at (mctx->input, dest_idx - 1, mctx->eflags,
+ preg->newline_anchor);
+ mctx->state_log[dest_idx]
+ = re_acquire_state_context (&err, dfa, &dest_nodes, context);
+ if (dest_state != NULL)
+ re_node_set_free (&dest_nodes);
+ if (BE (mctx->state_log[dest_idx] == NULL && err != REG_NOERROR, 0))
+ return err;
+ }
+ return REG_NOERROR;
+}
+#endif /* RE_ENABLE_I18N */
+
+static reg_errcode_t
+transit_state_bkref (preg, nodes, mctx)
+ const regex_t *preg;
+ re_node_set *nodes;
+ re_match_context_t *mctx;
+{
+ reg_errcode_t err;
+ re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ int i;
+ int cur_str_idx = re_string_cur_idx (mctx->input);
+
+ for (i = 0; i < nodes->nelem; ++i)
+ {
+ int dest_str_idx, prev_nelem, bkc_idx;
+ int node_idx = nodes->elems[i];
+ unsigned int context;
+ re_token_t *node = dfa->nodes + node_idx;
+ re_node_set *new_dest_nodes;
+
+ /* Check whether `node' is a backreference or not. */
+ if (node->type != OP_BACK_REF)
+ continue;
+
+ if (node->constraint)
+ {
+ context = re_string_context_at (mctx->input, cur_str_idx,
+ mctx->eflags, preg->newline_anchor);
+ if (NOT_SATISFY_NEXT_CONSTRAINT (node->constraint, context))
+ continue;
+ }
+
+ /* `node' is a backreference.
+ Check the substring which the substring matched. */
+ bkc_idx = mctx->nbkref_ents;
+ err = get_subexp (preg, mctx, node_idx, cur_str_idx);
+ if (BE (err != REG_NOERROR, 0))
+ goto free_return;
+
+ /* And add the epsilon closures (which is `new_dest_nodes') of
+ the backreference to appropriate state_log. */
+#ifdef DEBUG
+ assert (dfa->nexts[node_idx] != -1);
+#endif
+ for (; bkc_idx < mctx->nbkref_ents; ++bkc_idx)
+ {
+ int subexp_len;
+ re_dfastate_t *dest_state;
+ struct re_backref_cache_entry *bkref_ent;
+ bkref_ent = mctx->bkref_ents + bkc_idx;
+ if (bkref_ent->node != node_idx || bkref_ent->str_idx != cur_str_idx)
+ continue;
+ subexp_len = bkref_ent->subexp_to - bkref_ent->subexp_from;
+ new_dest_nodes = (subexp_len == 0
+ ? dfa->eclosures + dfa->edests[node_idx].elems[0]
+ : dfa->eclosures + dfa->nexts[node_idx]);
+ dest_str_idx = (cur_str_idx + bkref_ent->subexp_to
+ - bkref_ent->subexp_from);
+ context = re_string_context_at (mctx->input, dest_str_idx - 1,
+ mctx->eflags, preg->newline_anchor);
+ dest_state = mctx->state_log[dest_str_idx];
+ prev_nelem = ((mctx->state_log[cur_str_idx] == NULL) ? 0
+ : mctx->state_log[cur_str_idx]->nodes.nelem);
+ /* Add `new_dest_node' to state_log. */
+ if (dest_state == NULL)
+ {
+ mctx->state_log[dest_str_idx]
+ = re_acquire_state_context (&err, dfa, new_dest_nodes,
+ context);
+ if (BE (mctx->state_log[dest_str_idx] == NULL
+ && err != REG_NOERROR, 0))
+ goto free_return;
+ }
+ else
+ {
+ re_node_set dest_nodes;
+ err = re_node_set_init_union (&dest_nodes,
+ dest_state->entrance_nodes,
+ new_dest_nodes);
+ if (BE (err != REG_NOERROR, 0))
+ {
+ re_node_set_free (&dest_nodes);
+ goto free_return;
+ }
+ mctx->state_log[dest_str_idx]
+ = re_acquire_state_context (&err, dfa, &dest_nodes, context);
+ re_node_set_free (&dest_nodes);
+ if (BE (mctx->state_log[dest_str_idx] == NULL
+ && err != REG_NOERROR, 0))
+ goto free_return;
+ }
+ /* We need to check recursively if the backreference can epsilon
+ transit. */
+ if (subexp_len == 0
+ && mctx->state_log[cur_str_idx]->nodes.nelem > prev_nelem)
+ {
+ err = check_subexp_matching_top (dfa, mctx, new_dest_nodes,
+ cur_str_idx);
+ if (BE (err != REG_NOERROR, 0))
+ goto free_return;
+ err = transit_state_bkref (preg, new_dest_nodes, mctx);
+ if (BE (err != REG_NOERROR, 0))
+ goto free_return;
+ }
+ }
+ }
+ err = REG_NOERROR;
+ free_return:
+ return err;
+}
+
+/* Enumerate all the candidates which the backreference BKREF_NODE can match
+ at BKREF_STR_IDX, and register them by match_ctx_add_entry().
+ Note that we might collect inappropriate candidates here.
+ However, the cost of checking them strictly here is too high, then we
+ delay these checking for prune_impossible_nodes(). */
+
+static reg_errcode_t
+get_subexp (preg, mctx, bkref_node, bkref_str_idx)
+ const regex_t *preg;
+ re_match_context_t *mctx;
+ int bkref_node, bkref_str_idx;
+{
+ int subexp_num, sub_top_idx;
+ re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ char *buf = (char *) re_string_get_buffer (mctx->input);
+ /* Return if we have already checked BKREF_NODE at BKREF_STR_IDX. */
+ int cache_idx = search_cur_bkref_entry (mctx, bkref_str_idx);
+ for (; cache_idx < mctx->nbkref_ents; ++cache_idx)
+ {
+ struct re_backref_cache_entry *entry = mctx->bkref_ents + cache_idx;
+ if (entry->str_idx > bkref_str_idx)
+ break;
+ if (entry->node == bkref_node)
+ return REG_NOERROR; /* We already checked it. */
+ }
+ subexp_num = dfa->nodes[bkref_node].opr.idx - 1;
+
+ /* For each sub expression */
+ for (sub_top_idx = 0; sub_top_idx < mctx->nsub_tops; ++sub_top_idx)
+ {
+ reg_errcode_t err;
+ re_sub_match_top_t *sub_top = mctx->sub_tops[sub_top_idx];
+ re_sub_match_last_t *sub_last;
+ int sub_last_idx, sl_str;
+ char *bkref_str;
+
+ if (dfa->nodes[sub_top->node].opr.idx != subexp_num)
+ continue; /* It isn't related. */
+
+ sl_str = sub_top->str_idx;
+ bkref_str = buf + bkref_str_idx;
+ /* At first, check the last node of sub expressions we already
+ evaluated. */
+ for (sub_last_idx = 0; sub_last_idx < sub_top->nlasts; ++sub_last_idx)
+ {
+ int sl_str_diff;
+ sub_last = sub_top->lasts[sub_last_idx];
+ sl_str_diff = sub_last->str_idx - sl_str;
+ /* The matched string by the sub expression match with the substring
+ at the back reference? */
+ if (sl_str_diff > 0
+ && memcmp (bkref_str, buf + sl_str, sl_str_diff) != 0)
+ break; /* We don't need to search this sub expression any more. */
+ bkref_str += sl_str_diff;
+ sl_str += sl_str_diff;
+ err = get_subexp_sub (preg, mctx, sub_top, sub_last, bkref_node,
+ bkref_str_idx);
+ if (err == REG_NOMATCH)
+ continue;
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ }
+ if (sub_last_idx < sub_top->nlasts)
+ continue;
+ if (sub_last_idx > 0)
+ ++sl_str;
+ /* Then, search for the other last nodes of the sub expression. */
+ for (; sl_str <= bkref_str_idx; ++sl_str)
+ {
+ int cls_node, sl_str_off;
+ re_node_set *nodes;
+ sl_str_off = sl_str - sub_top->str_idx;
+ /* The matched string by the sub expression match with the substring
+ at the back reference? */
+ if (sl_str_off > 0
+ && memcmp (bkref_str++, buf + sl_str - 1, 1) != 0)
+ break; /* We don't need to search this sub expression any more. */
+ if (mctx->state_log[sl_str] == NULL)
+ continue;
+ /* Does this state have a ')' of the sub expression? */
+ nodes = &mctx->state_log[sl_str]->nodes;
+ cls_node = find_subexp_node (dfa, nodes, subexp_num, 0);
+ if (cls_node == -1)
+ continue; /* No. */
+ if (sub_top->path == NULL)
+ {
+ sub_top->path = calloc (sizeof (state_array_t),
+ sl_str - sub_top->str_idx + 1);
+ if (sub_top->path == NULL)
+ return REG_ESPACE;
+ }
+ /* Can the OP_OPEN_SUBEXP node arrive the OP_CLOSE_SUBEXP node
+ in the current context? */
+ err = check_arrival (preg, mctx, sub_top->path, sub_top->node,
+ sub_top->str_idx, cls_node, sl_str, 0);
+ if (err == REG_NOMATCH)
+ continue;
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ sub_last = match_ctx_add_sublast (sub_top, cls_node, sl_str);
+ if (BE (sub_last == NULL, 0))
+ return REG_ESPACE;
+ err = get_subexp_sub (preg, mctx, sub_top, sub_last, bkref_node,
+ bkref_str_idx);
+ if (err == REG_NOMATCH)
+ continue;
+ }
+ }
+ return REG_NOERROR;
+}
+
+/* Helper functions for get_subexp(). */
+
+/* Check SUB_LAST can arrive to the back reference BKREF_NODE at BKREF_STR.
+ If it can arrive, register the sub expression expressed with SUB_TOP
+ and SUB_LAST. */
+
+static reg_errcode_t
+get_subexp_sub (preg, mctx, sub_top, sub_last, bkref_node, bkref_str)
+ const regex_t *preg;
+ re_match_context_t *mctx;
+ re_sub_match_top_t *sub_top;
+ re_sub_match_last_t *sub_last;
+ int bkref_node, bkref_str;
+{
+ reg_errcode_t err;
+ int to_idx;
+ /* Can the subexpression arrive the back reference? */
+ err = check_arrival (preg, mctx, &sub_last->path, sub_last->node,
+ sub_last->str_idx, bkref_node, bkref_str, 1);
+ if (err != REG_NOERROR)
+ return err;
+ err = match_ctx_add_entry (mctx, bkref_node, bkref_str, sub_top->str_idx,
+ sub_last->str_idx);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ to_idx = bkref_str + sub_last->str_idx - sub_top->str_idx;
+ clean_state_log_if_need (mctx, to_idx);
+ return REG_NOERROR;
+}
+
+/* Find the first node which is '(' or ')' and whose index is SUBEXP_IDX.
+ Search '(' if FL_OPEN, or search ')' otherwise.
+ TODO: This function isn't efficient...
+ Because there might be more than one nodes whose types are
+ OP_OPEN_SUBEXP and whose index is SUBEXP_IDX, we must check all
+ nodes.
+ E.g. RE: (a){2} */
+
+static int
+find_subexp_node (dfa, nodes, subexp_idx, fl_open)
+ re_dfa_t *dfa;
+ re_node_set *nodes;
+ int subexp_idx, fl_open;
+{
+ int cls_idx;
+ for (cls_idx = 0; cls_idx < nodes->nelem; ++cls_idx)
+ {
+ int cls_node = nodes->elems[cls_idx];
+ re_token_t *node = dfa->nodes + cls_node;
+ if (((fl_open && node->type == OP_OPEN_SUBEXP)
+ || (!fl_open && node->type == OP_CLOSE_SUBEXP))
+ && node->opr.idx == subexp_idx)
+ return cls_node;
+ }
+ return -1;
+}
+
+/* Check whether the node TOP_NODE at TOP_STR can arrive to the node
+ LAST_NODE at LAST_STR. We record the path onto PATH since it will be
+ heavily reused.
+ Return REG_NOERROR if it can arrive, or REG_NOMATCH otherwise. */
+
+static reg_errcode_t
+check_arrival (preg, mctx, path, top_node, top_str, last_node, last_str,
+ fl_open)
+ const regex_t *preg;
+ re_match_context_t *mctx;
+ state_array_t *path;
+ int top_node, top_str, last_node, last_str, fl_open;
+{
+ re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ reg_errcode_t err;
+ int subexp_num, backup_cur_idx, str_idx, null_cnt;
+ re_dfastate_t *cur_state = NULL;
+ re_node_set *cur_nodes, next_nodes;
+ re_dfastate_t **backup_state_log;
+ unsigned int context;
+
+ subexp_num = dfa->nodes[top_node].opr.idx;
+ /* Extend the buffer if we need. */
+ if (path->alloc < last_str + mctx->max_mb_elem_len + 1)
+ {
+ re_dfastate_t **new_array;
+ int old_alloc = path->alloc;
+ path->alloc += last_str + mctx->max_mb_elem_len + 1;
+ new_array = re_realloc (path->array, re_dfastate_t *, path->alloc);
+ if (new_array == NULL)
+ return REG_ESPACE;
+ path->array = new_array;
+ memset (new_array + old_alloc, '\0',
+ sizeof (re_dfastate_t *) * (path->alloc - old_alloc));
+ }
+
+ str_idx = path->next_idx == 0 ? top_str : path->next_idx;
+
+ /* Temporary modify MCTX. */
+ backup_state_log = mctx->state_log;
+ backup_cur_idx = mctx->input->cur_idx;
+ mctx->state_log = path->array;
+ mctx->input->cur_idx = str_idx;
+
+ /* Setup initial node set. */
+ context = re_string_context_at (mctx->input, str_idx - 1, mctx->eflags,
+ preg->newline_anchor);
+ if (str_idx == top_str)
+ {
+ err = re_node_set_init_1 (&next_nodes, top_node);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ err = check_arrival_expand_ecl (dfa, &next_nodes, subexp_num, fl_open);
+ if (BE (err != REG_NOERROR, 0))
+ {
+ re_node_set_free (&next_nodes);
+ return err;
+ }
+ }
+ else
+ {
+ cur_state = mctx->state_log[str_idx];
+ if (cur_state && cur_state->has_backref)
+ {
+ err = re_node_set_init_copy (&next_nodes, &cur_state->nodes);
+ if (BE ( err != REG_NOERROR, 0))
+ return err;
+ }
+ else
+ re_node_set_init_empty (&next_nodes);
+ }
+ if (str_idx == top_str || (cur_state && cur_state->has_backref))
+ {
+ if (next_nodes.nelem)
+ {
+ err = expand_bkref_cache (preg, mctx, &next_nodes, str_idx, last_str,
+ subexp_num, fl_open);
+ if (BE ( err != REG_NOERROR, 0))
+ {
+ re_node_set_free (&next_nodes);
+ return err;
+ }
+ }
+ cur_state = re_acquire_state_context (&err, dfa, &next_nodes, context);
+ if (BE (cur_state == NULL && err != REG_NOERROR, 0))
+ {
+ re_node_set_free (&next_nodes);
+ return err;
+ }
+ mctx->state_log[str_idx] = cur_state;
+ }
+
+ for (null_cnt = 0; str_idx < last_str && null_cnt <= mctx->max_mb_elem_len;)
+ {
+ re_node_set_empty (&next_nodes);
+ if (mctx->state_log[str_idx + 1])
+ {
+ err = re_node_set_merge (&next_nodes,
+ &mctx->state_log[str_idx + 1]->nodes);
+ if (BE (err != REG_NOERROR, 0))
+ {
+ re_node_set_free (&next_nodes);
+ return err;
+ }
+ }
+ if (cur_state)
+ {
+ err = check_arrival_add_next_nodes(preg, dfa, mctx, str_idx,
+ &cur_state->nodes, &next_nodes);
+ if (BE (err != REG_NOERROR, 0))
+ {
+ re_node_set_free (&next_nodes);
+ return err;
+ }
+ }
+ ++str_idx;
+ if (next_nodes.nelem)
+ {
+ err = check_arrival_expand_ecl (dfa, &next_nodes, subexp_num,
+ fl_open);
+ if (BE (err != REG_NOERROR, 0))
+ {
+ re_node_set_free (&next_nodes);
+ return err;
+ }
+ err = expand_bkref_cache (preg, mctx, &next_nodes, str_idx, last_str,
+ subexp_num, fl_open);
+ if (BE ( err != REG_NOERROR, 0))
+ {
+ re_node_set_free (&next_nodes);
+ return err;
+ }
+ }
+ context = re_string_context_at (mctx->input, str_idx - 1, mctx->eflags,
+ preg->newline_anchor);
+ cur_state = re_acquire_state_context (&err, dfa, &next_nodes, context);
+ if (BE (cur_state == NULL && err != REG_NOERROR, 0))
+ {
+ re_node_set_free (&next_nodes);
+ return err;
+ }
+ mctx->state_log[str_idx] = cur_state;
+ null_cnt = cur_state == NULL ? null_cnt + 1 : 0;
+ }
+ re_node_set_free (&next_nodes);
+ cur_nodes = (mctx->state_log[last_str] == NULL ? NULL
+ : &mctx->state_log[last_str]->nodes);
+ path->next_idx = str_idx;
+
+ /* Fix MCTX. */
+ mctx->state_log = backup_state_log;
+ mctx->input->cur_idx = backup_cur_idx;
+
+ if (cur_nodes == NULL)
+ return REG_NOMATCH;
+ /* Then check the current node set has the node LAST_NODE. */
+ return (re_node_set_contains (cur_nodes, last_node)
+ || re_node_set_contains (cur_nodes, last_node) ? REG_NOERROR
+ : REG_NOMATCH);
+}
+
+/* Helper functions for check_arrival. */
+
+/* Calculate the destination nodes of CUR_NODES at STR_IDX, and append them
+ to NEXT_NODES.
+ TODO: This function is similar to the functions transit_state*(),
+ however this function has many additional works.
+ Can't we unify them? */
+
+static reg_errcode_t
+check_arrival_add_next_nodes (preg, dfa, mctx, str_idx, cur_nodes, next_nodes)
+ const regex_t *preg;
+ re_dfa_t *dfa;
+ re_match_context_t *mctx;
+ int str_idx;
+ re_node_set *cur_nodes, *next_nodes;
+{
+ int cur_idx;
+ reg_errcode_t err;
+ re_node_set union_set;
+ re_node_set_init_empty (&union_set);
+ for (cur_idx = 0; cur_idx < cur_nodes->nelem; ++cur_idx)
+ {
+ int naccepted = 0;
+ int cur_node = cur_nodes->elems[cur_idx];
+ re_token_type_t type = dfa->nodes[cur_node].type;
+ if (IS_EPSILON_NODE(type))
+ continue;
+#ifdef RE_ENABLE_I18N
+ /* If the node may accept `multi byte'. */
+ if (ACCEPT_MB_NODE (type))
+ {
+ naccepted = check_node_accept_bytes (preg, cur_node, mctx->input,
+ str_idx);
+ if (naccepted > 1)
+ {
+ re_dfastate_t *dest_state;
+ int next_node = dfa->nexts[cur_node];
+ int next_idx = str_idx + naccepted;
+ dest_state = mctx->state_log[next_idx];
+ re_node_set_empty (&union_set);
+ if (dest_state)
+ {
+ err = re_node_set_merge (&union_set, &dest_state->nodes);
+ if (BE (err != REG_NOERROR, 0))
+ {
+ re_node_set_free (&union_set);
+ return err;
+ }
+ err = re_node_set_insert (&union_set, next_node);
+ if (BE (err < 0, 0))
+ {
+ re_node_set_free (&union_set);
+ return REG_ESPACE;
+ }
+ }
+ else
+ {
+ err = re_node_set_insert (&union_set, next_node);
+ if (BE (err < 0, 0))
+ {
+ re_node_set_free (&union_set);
+ return REG_ESPACE;
+ }
+ }
+ mctx->state_log[next_idx] = re_acquire_state (&err, dfa,
+ &union_set);
+ if (BE (mctx->state_log[next_idx] == NULL
+ && err != REG_NOERROR, 0))
+ {
+ re_node_set_free (&union_set);
+ return err;
+ }
+ }
+ }
+#endif /* RE_ENABLE_I18N */
+ if (naccepted
+ || check_node_accept (preg, dfa->nodes + cur_node, mctx,
+ str_idx))
+ {
+ err = re_node_set_insert (next_nodes, dfa->nexts[cur_node]);
+ if (BE (err < 0, 0))
+ {
+ re_node_set_free (&union_set);
+ return REG_ESPACE;
+ }
+ }
+ }
+ re_node_set_free (&union_set);
+ return REG_NOERROR;
+}
+
+/* For all the nodes in CUR_NODES, add the epsilon closures of them to
+ CUR_NODES, however exclude the nodes which are:
+ - inside the sub expression whose number is EX_SUBEXP, if FL_OPEN.
+ - out of the sub expression whose number is EX_SUBEXP, if !FL_OPEN.
+*/
+
+static reg_errcode_t
+check_arrival_expand_ecl (dfa, cur_nodes, ex_subexp, fl_open)
+ re_dfa_t *dfa;
+ re_node_set *cur_nodes;
+ int ex_subexp, fl_open;
+{
+ reg_errcode_t err;
+ int idx, outside_node;
+ re_node_set new_nodes;
+#ifdef DEBUG
+ assert (cur_nodes->nelem);
+#endif
+ err = re_node_set_alloc (&new_nodes, cur_nodes->nelem);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ /* Create a new node set NEW_NODES with the nodes which are epsilon
+ closures of the node in CUR_NODES. */
+
+ for (idx = 0; idx < cur_nodes->nelem; ++idx)
+ {
+ int cur_node = cur_nodes->elems[idx];
+ re_node_set *eclosure = dfa->eclosures + cur_node;
+ outside_node = find_subexp_node (dfa, eclosure, ex_subexp, fl_open);
+ if (outside_node == -1)
+ {
+ /* There are no problematic nodes, just merge them. */
+ err = re_node_set_merge (&new_nodes, eclosure);
+ if (BE (err != REG_NOERROR, 0))
+ {
+ re_node_set_free (&new_nodes);
+ return err;
+ }
+ }
+ else
+ {
+ /* There are problematic nodes, re-calculate incrementally. */
+ err = check_arrival_expand_ecl_sub (dfa, &new_nodes, cur_node,
+ ex_subexp, fl_open);
+ if (BE (err != REG_NOERROR, 0))
+ {
+ re_node_set_free (&new_nodes);
+ return err;
+ }
+ }
+ }
+ re_node_set_free (cur_nodes);
+ *cur_nodes = new_nodes;
+ return REG_NOERROR;
+}
+
+/* Helper function for check_arrival_expand_ecl.
+ Check incrementally the epsilon closure of TARGET, and if it isn't
+ problematic append it to DST_NODES. */
+
+static reg_errcode_t
+check_arrival_expand_ecl_sub (dfa, dst_nodes, target, ex_subexp, fl_open)
+ re_dfa_t *dfa;
+ int target, ex_subexp, fl_open;
+ re_node_set *dst_nodes;
+{
+ int cur_node, type;
+ for (cur_node = target; !re_node_set_contains (dst_nodes, cur_node);)
+ {
+ int err;
+ type = dfa->nodes[cur_node].type;
+
+ if (((type == OP_OPEN_SUBEXP && fl_open)
+ || (type == OP_CLOSE_SUBEXP && !fl_open))
+ && dfa->nodes[cur_node].opr.idx == ex_subexp)
+ {
+ if (!fl_open)
+ {
+ err = re_node_set_insert (dst_nodes, cur_node);
+ if (BE (err == -1, 0))
+ return REG_ESPACE;
+ }
+ break;
+ }
+ err = re_node_set_insert (dst_nodes, cur_node);
+ if (BE (err == -1, 0))
+ return REG_ESPACE;
+ if (dfa->edests[cur_node].nelem == 0)
+ break;
+ if (dfa->edests[cur_node].nelem == 2)
+ {
+ err = check_arrival_expand_ecl_sub (dfa, dst_nodes,
+ dfa->edests[cur_node].elems[1],
+ ex_subexp, fl_open);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ }
+ cur_node = dfa->edests[cur_node].elems[0];
+ }
+ return REG_NOERROR;
+}
+
+
+/* For all the back references in the current state, calculate the
+ destination of the back references by the appropriate entry
+ in MCTX->BKREF_ENTS. */
+
+static reg_errcode_t
+expand_bkref_cache (preg, mctx, cur_nodes, cur_str, last_str, subexp_num,
+ fl_open)
+ const regex_t *preg;
+ re_match_context_t *mctx;
+ int cur_str, last_str, subexp_num, fl_open;
+ re_node_set *cur_nodes;
+{
+ reg_errcode_t err;
+ re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ int cache_idx, cache_idx_start;
+ /* The current state. */
+
+ cache_idx_start = search_cur_bkref_entry (mctx, cur_str);
+ for (cache_idx = cache_idx_start; cache_idx < mctx->nbkref_ents; ++cache_idx)
+ {
+ int to_idx, next_node;
+ struct re_backref_cache_entry *ent = mctx->bkref_ents + cache_idx;
+ if (ent->str_idx > cur_str)
+ break;
+ /* Is this entry ENT is appropriate? */
+ if (!re_node_set_contains (cur_nodes, ent->node))
+ continue; /* No. */
+
+ to_idx = cur_str + ent->subexp_to - ent->subexp_from;
+ /* Calculate the destination of the back reference, and append it
+ to MCTX->STATE_LOG. */
+ if (to_idx == cur_str)
+ {
+ /* The backreference did epsilon transit, we must re-check all the
+ node in the current state. */
+ re_node_set new_dests;
+ reg_errcode_t err2, err3;
+ next_node = dfa->edests[ent->node].elems[0];
+ if (re_node_set_contains (cur_nodes, next_node))
+ continue;
+ err = re_node_set_init_1 (&new_dests, next_node);
+ err2 = check_arrival_expand_ecl (dfa, &new_dests, subexp_num,
+ fl_open);
+ err3 = re_node_set_merge (cur_nodes, &new_dests);
+ re_node_set_free (&new_dests);
+ if (BE (err != REG_NOERROR || err2 != REG_NOERROR
+ || err3 != REG_NOERROR, 0))
+ {
+ err = (err != REG_NOERROR ? err
+ : (err2 != REG_NOERROR ? err2 : err3));
+ return err;
+ }
+ /* TODO: It is still inefficient... */
+ cache_idx = cache_idx_start - 1;
+ continue;
+ }
+ else
+ {
+ re_node_set union_set;
+ next_node = dfa->nexts[ent->node];
+ if (mctx->state_log[to_idx])
+ {
+ int ret;
+ if (re_node_set_contains (&mctx->state_log[to_idx]->nodes,
+ next_node))
+ continue;
+ err = re_node_set_init_copy (&union_set,
+ &mctx->state_log[to_idx]->nodes);
+ ret = re_node_set_insert (&union_set, next_node);
+ if (BE (err != REG_NOERROR || ret < 0, 0))
+ {
+ re_node_set_free (&union_set);
+ err = err != REG_NOERROR ? err : REG_ESPACE;
+ return err;
+ }
+ }
+ else
+ {
+ err = re_node_set_init_1 (&union_set, next_node);
+ if (BE (err != REG_NOERROR, 0))
+ return err;
+ }
+ mctx->state_log[to_idx] = re_acquire_state (&err, dfa, &union_set);
+ re_node_set_free (&union_set);
+ if (BE (mctx->state_log[to_idx] == NULL
+ && err != REG_NOERROR, 0))
+ return err;
+ }
+ }
+ return REG_NOERROR;
+}
+
+/* Build transition table for the state.
+ Return the new table if succeeded, otherwise return NULL. */
+
+static re_dfastate_t **
+build_trtable (preg, state, fl_search)
+ const regex_t *preg;
+ const re_dfastate_t *state;
+ int fl_search;
+{
+ reg_errcode_t err;
+ re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ int i, j, k, ch;
+ int dests_node_malloced = 0, dest_states_malloced = 0;
+ int ndests; /* Number of the destination states from `state'. */
+ re_dfastate_t **trtable;
+ re_dfastate_t **dest_states = NULL, **dest_states_word, **dest_states_nl;
+ re_node_set follows, *dests_node;
+ bitset *dests_ch;
+ bitset acceptable;
+
+ /* We build DFA states which corresponds to the destination nodes
+ from `state'. `dests_node[i]' represents the nodes which i-th
+ destination state contains, and `dests_ch[i]' represents the
+ characters which i-th destination state accepts. */
+#ifdef _LIBC
+ if (__libc_use_alloca ((sizeof (re_node_set) + sizeof (bitset)) * SBC_MAX))
+ dests_node = (re_node_set *)
+ alloca ((sizeof (re_node_set) + sizeof (bitset)) * SBC_MAX);
+ else
+#endif
+ {
+ dests_node = (re_node_set *)
+ malloc ((sizeof (re_node_set) + sizeof (bitset)) * SBC_MAX);
+ if (BE (dests_node == NULL, 0))
+ return NULL;
+ dests_node_malloced = 1;
+ }
+ dests_ch = (bitset *) (dests_node + SBC_MAX);
+
+ /* Initialize transiton table. */
+ trtable = (re_dfastate_t **) calloc (sizeof (re_dfastate_t *), SBC_MAX);
+ if (BE (trtable == NULL, 0))
+ {
+ if (dests_node_malloced)
+ free (dests_node);
+ return NULL;
+ }
+
+ /* At first, group all nodes belonging to `state' into several
+ destinations. */
+ ndests = group_nodes_into_DFAstates (preg, state, dests_node, dests_ch);
+ if (BE (ndests <= 0, 0))
+ {
+ if (dests_node_malloced)
+ free (dests_node);
+ /* Return NULL in case of an error, trtable otherwise. */
+ if (ndests == 0)
+ return trtable;
+ free (trtable);
+ return NULL;
+ }
+
+ err = re_node_set_alloc (&follows, ndests + 1);
+ if (BE (err != REG_NOERROR, 0))
+ goto out_free;
+
+#ifdef _LIBC
+ if (__libc_use_alloca ((sizeof (re_node_set) + sizeof (bitset)) * SBC_MAX
+ + ndests * 3 * sizeof (re_dfastate_t *)))
+ dest_states = (re_dfastate_t **)
+ alloca (ndests * 3 * sizeof (re_dfastate_t *));
+ else
+#endif
+ {
+ dest_states = (re_dfastate_t **)
+ malloc (ndests * 3 * sizeof (re_dfastate_t *));
+ if (BE (dest_states == NULL, 0))
+ {
+out_free:
+ if (dest_states_malloced)
+ free (dest_states);
+ re_node_set_free (&follows);
+ for (i = 0; i < ndests; ++i)
+ re_node_set_free (dests_node + i);
+ free (trtable);
+ if (dests_node_malloced)
+ free (dests_node);
+ return NULL;
+ }
+ dest_states_malloced = 1;
+ }
+ dest_states_word = dest_states + ndests;
+ dest_states_nl = dest_states_word + ndests;
+ bitset_empty (acceptable);
+
+ /* Then build the states for all destinations. */
+ for (i = 0; i < ndests; ++i)
+ {
+ int next_node;
+ re_node_set_empty (&follows);
+ /* Merge the follows of this destination states. */
+ for (j = 0; j < dests_node[i].nelem; ++j)
+ {
+ next_node = dfa->nexts[dests_node[i].elems[j]];
+ if (next_node != -1)
+ {
+ err = re_node_set_merge (&follows, dfa->eclosures + next_node);
+ if (BE (err != REG_NOERROR, 0))
+ goto out_free;
+ }
+ }
+ /* If search flag is set, merge the initial state. */
+ if (fl_search)
+ {
+#ifdef RE_ENABLE_I18N
+ int not_initial = 0;
+ for (j = 0; j < follows.nelem; ++j)
+ if (dfa->nodes[follows.elems[j]].type == CHARACTER)
+ {
+ not_initial = dfa->nodes[follows.elems[j]].mb_partial;
+ break;
+ }
+ if (!not_initial)
+#endif
+ {
+ err = re_node_set_merge (&follows,
+ dfa->init_state->entrance_nodes);
+ if (BE (err != REG_NOERROR, 0))
+ goto out_free;
+ }
+ }
+ dest_states[i] = re_acquire_state_context (&err, dfa, &follows, 0);
+ if (BE (dest_states[i] == NULL && err != REG_NOERROR, 0))
+ goto out_free;
+ /* If the new state has context constraint,
+ build appropriate states for these contexts. */
+ if (dest_states[i]->has_constraint)
+ {
+ dest_states_word[i] = re_acquire_state_context (&err, dfa, &follows,
+ CONTEXT_WORD);
+ if (BE (dest_states_word[i] == NULL && err != REG_NOERROR, 0))
+ goto out_free;
+ dest_states_nl[i] = re_acquire_state_context (&err, dfa, &follows,
+ CONTEXT_NEWLINE);
+ if (BE (dest_states_nl[i] == NULL && err != REG_NOERROR, 0))
+ goto out_free;
+ }
+ else
+ {
+ dest_states_word[i] = dest_states[i];
+ dest_states_nl[i] = dest_states[i];
+ }
+ bitset_merge (acceptable, dests_ch[i]);
+ }
+
+ /* Update the transition table. */
+ /* For all characters ch...: */
+ for (i = 0, ch = 0; i < BITSET_UINTS; ++i)
+ for (j = 0; j < UINT_BITS; ++j, ++ch)
+ if ((acceptable[i] >> j) & 1)
+ {
+ /* The current state accepts the character ch. */
+ if (IS_WORD_CHAR (ch))
+ {
+ for (k = 0; k < ndests; ++k)
+ if ((dests_ch[k][i] >> j) & 1)
+ {
+ /* k-th destination accepts the word character ch. */
+ trtable[ch] = dest_states_word[k];
+ /* There must be only one destination which accepts
+ character ch. See group_nodes_into_DFAstates. */
+ break;
+ }
+ }
+ else /* not WORD_CHAR */
+ {
+ for (k = 0; k < ndests; ++k)
+ if ((dests_ch[k][i] >> j) & 1)
+ {
+ /* k-th destination accepts the non-word character ch. */
+ trtable[ch] = dest_states[k];
+ /* There must be only one destination which accepts
+ character ch. See group_nodes_into_DFAstates. */
+ break;
+ }
+ }
+ }
+ /* new line */
+ if (bitset_contain (acceptable, NEWLINE_CHAR))
+ {
+ /* The current state accepts newline character. */
+ for (k = 0; k < ndests; ++k)
+ if (bitset_contain (dests_ch[k], NEWLINE_CHAR))
+ {
+ /* k-th destination accepts newline character. */
+ trtable[NEWLINE_CHAR] = dest_states_nl[k];
+ /* There must be only one destination which accepts
+ newline. See group_nodes_into_DFAstates. */
+ break;
+ }
+ }
+
+ if (dest_states_malloced)
+ free (dest_states);
+
+ re_node_set_free (&follows);
+ for (i = 0; i < ndests; ++i)
+ re_node_set_free (dests_node + i);
+
+ if (dests_node_malloced)
+ free (dests_node);
+
+ return trtable;
+}
+
+/* Group all nodes belonging to STATE into several destinations.
+ Then for all destinations, set the nodes belonging to the destination
+ to DESTS_NODE[i] and set the characters accepted by the destination
+ to DEST_CH[i]. This function return the number of destinations. */
+
+static int
+group_nodes_into_DFAstates (preg, state, dests_node, dests_ch)
+ const regex_t *preg;
+ const re_dfastate_t *state;
+ re_node_set *dests_node;
+ bitset *dests_ch;
+{
+ reg_errcode_t err;
+ const re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ int i, j, k;
+ int ndests; /* Number of the destinations from `state'. */
+ bitset accepts; /* Characters a node can accept. */
+ const re_node_set *cur_nodes = &state->nodes;
+ bitset_empty (accepts);
+ ndests = 0;
+
+ /* For all the nodes belonging to `state', */
+ for (i = 0; i < cur_nodes->nelem; ++i)
+ {
+ re_token_t *node = &dfa->nodes[cur_nodes->elems[i]];
+ re_token_type_t type = node->type;
+ unsigned int constraint = node->constraint;
+
+ /* Enumerate all single byte character this node can accept. */
+ if (type == CHARACTER)
+ bitset_set (accepts, node->opr.c);
+ else if (type == SIMPLE_BRACKET)
+ {
+ bitset_merge (accepts, node->opr.sbcset);
+ }
+ else if (type == OP_PERIOD)
+ {
+ bitset_set_all (accepts);
+ if (!(preg->syntax & RE_DOT_NEWLINE))
+ bitset_clear (accepts, '\n');
+ if (preg->syntax & RE_DOT_NOT_NULL)
+ bitset_clear (accepts, '\0');
+ }
+ else
+ continue;
+
+ /* Check the `accepts' and sift the characters which are not
+ match it the context. */
+ if (constraint)
+ {
+ if (constraint & NEXT_WORD_CONSTRAINT)
+ for (j = 0; j < BITSET_UINTS; ++j)
+ accepts[j] &= dfa->word_char[j];
+ if (constraint & NEXT_NOTWORD_CONSTRAINT)
+ for (j = 0; j < BITSET_UINTS; ++j)
+ accepts[j] &= ~dfa->word_char[j];
+ if (constraint & NEXT_NEWLINE_CONSTRAINT)
+ {
+ int accepts_newline = bitset_contain (accepts, NEWLINE_CHAR);
+ bitset_empty (accepts);
+ if (accepts_newline)
+ bitset_set (accepts, NEWLINE_CHAR);
+ else
+ continue;
+ }
+ }
+
+ /* Then divide `accepts' into DFA states, or create a new
+ state. */
+ for (j = 0; j < ndests; ++j)
+ {
+ bitset intersec; /* Intersection sets, see below. */
+ bitset remains;
+ /* Flags, see below. */
+ int has_intersec, not_subset, not_consumed;
+
+ /* Optimization, skip if this state doesn't accept the character. */
+ if (type == CHARACTER && !bitset_contain (dests_ch[j], node->opr.c))
+ continue;
+
+ /* Enumerate the intersection set of this state and `accepts'. */
+ has_intersec = 0;
+ for (k = 0; k < BITSET_UINTS; ++k)
+ has_intersec |= intersec[k] = accepts[k] & dests_ch[j][k];
+ /* And skip if the intersection set is empty. */
+ if (!has_intersec)
+ continue;
+
+ /* Then check if this state is a subset of `accepts'. */
+ not_subset = not_consumed = 0;
+ for (k = 0; k < BITSET_UINTS; ++k)
+ {
+ not_subset |= remains[k] = ~accepts[k] & dests_ch[j][k];
+ not_consumed |= accepts[k] = accepts[k] & ~dests_ch[j][k];
+ }
+
+ /* If this state isn't a subset of `accepts', create a
+ new group state, which has the `remains'. */
+ if (not_subset)
+ {
+ bitset_copy (dests_ch[ndests], remains);
+ bitset_copy (dests_ch[j], intersec);
+ err = re_node_set_init_copy (dests_node + ndests, &dests_node[j]);
+ if (BE (err != REG_NOERROR, 0))
+ goto error_return;
+ ++ndests;
+ }
+
+ /* Put the position in the current group. */
+ err = re_node_set_insert (&dests_node[j], cur_nodes->elems[i]);
+ if (BE (err < 0, 0))
+ goto error_return;
+
+ /* If all characters are consumed, go to next node. */
+ if (!not_consumed)
+ break;
+ }
+ /* Some characters remain, create a new group. */
+ if (j == ndests)
+ {
+ bitset_copy (dests_ch[ndests], accepts);
+ err = re_node_set_init_1 (dests_node + ndests, cur_nodes->elems[i]);
+ if (BE (err != REG_NOERROR, 0))
+ goto error_return;
+ ++ndests;
+ bitset_empty (accepts);
+ }
+ }
+ return ndests;
+ error_return:
+ for (j = 0; j < ndests; ++j)
+ re_node_set_free (dests_node + j);
+ return -1;
+}
+
+#ifdef RE_ENABLE_I18N
+/* Check how many bytes the node `dfa->nodes[node_idx]' accepts.
+ Return the number of the bytes the node accepts.
+ STR_IDX is the current index of the input string.
+
+ This function handles the nodes which can accept one character, or
+ one collating element like '.', '[a-z]', opposite to the other nodes
+ can only accept one byte. */
+
+static int
+check_node_accept_bytes (preg, node_idx, input, str_idx)
+ const regex_t *preg;
+ int node_idx, str_idx;
+ const re_string_t *input;
+{
+ const re_dfa_t *dfa = (re_dfa_t *) preg->buffer;
+ const re_token_t *node = dfa->nodes + node_idx;
+ int elem_len = re_string_elem_size_at (input, str_idx);
+ int char_len = re_string_char_size_at (input, str_idx);
+ int i;
+# ifdef _LIBC
+ int j;
+ uint32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES);
+# endif /* _LIBC */
+ if (elem_len <= 1 && char_len <= 1)
+ return 0;
+ if (node->type == OP_PERIOD)
+ {
+ /* '.' accepts any one character except the following two cases. */
+ if ((!(preg->syntax & RE_DOT_NEWLINE) &&
+ re_string_byte_at (input, str_idx) == '\n') ||
+ ((preg->syntax & RE_DOT_NOT_NULL) &&
+ re_string_byte_at (input, str_idx) == '\0'))
+ return 0;
+ return char_len;
+ }
+ else if (node->type == COMPLEX_BRACKET)
+ {
+ const re_charset_t *cset = node->opr.mbcset;
+# ifdef _LIBC
+ const unsigned char *pin = ((char *) re_string_get_buffer (input)
+ + str_idx);
+# endif /* _LIBC */
+ int match_len = 0;
+ wchar_t wc = ((cset->nranges || cset->nchar_classes || cset->nmbchars)
+ ? re_string_wchar_at (input, str_idx) : 0);
+
+ /* match with multibyte character? */
+ for (i = 0; i < cset->nmbchars; ++i)
+ if (wc == cset->mbchars[i])
+ {
+ match_len = char_len;
+ goto check_node_accept_bytes_match;
+ }
+ /* match with character_class? */
+ for (i = 0; i < cset->nchar_classes; ++i)
+ {
+ wctype_t wt = cset->char_classes[i];
+ if (__iswctype (wc, wt))
+ {
+ match_len = char_len;
+ goto check_node_accept_bytes_match;
+ }
+ }
+
+# ifdef _LIBC
+ if (nrules != 0)
+ {
+ unsigned int in_collseq = 0;
+ const int32_t *table, *indirect;
+ const unsigned char *weights, *extra;
+ const char *collseqwc;
+ int32_t idx;
+ /* This #include defines a local function! */
+# include <locale/weight.h>
+
+ /* match with collating_symbol? */
+ if (cset->ncoll_syms)
+ extra = (const unsigned char *)
+ _NL_CURRENT (LC_COLLATE, _NL_COLLATE_SYMB_EXTRAMB);
+ for (i = 0; i < cset->ncoll_syms; ++i)
+ {
+ const unsigned char *coll_sym = extra + cset->coll_syms[i];
+ /* Compare the length of input collating element and
+ the length of current collating element. */
+ if (*coll_sym != elem_len)
+ continue;
+ /* Compare each bytes. */
+ for (j = 0; j < *coll_sym; j++)
+ if (pin[j] != coll_sym[1 + j])
+ break;
+ if (j == *coll_sym)
+ {
+ /* Match if every bytes is equal. */
+ match_len = j;
+ goto check_node_accept_bytes_match;
+ }
+ }
+
+ if (cset->nranges)
+ {
+ if (elem_len <= char_len)
+ {
+ collseqwc = _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQWC);
+ in_collseq = collseq_table_lookup (collseqwc, wc);
+ }
+ else
+ in_collseq = find_collation_sequence_value (pin, elem_len);
+ }
+ /* match with range expression? */
+ for (i = 0; i < cset->nranges; ++i)
+ if (cset->range_starts[i] <= in_collseq
+ && in_collseq <= cset->range_ends[i])
+ {
+ match_len = elem_len;
+ goto check_node_accept_bytes_match;
+ }
+
+ /* match with equivalence_class? */
+ if (cset->nequiv_classes)
+ {
+ const unsigned char *cp = pin;
+ table = (const int32_t *)
+ _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB);
+ weights = (const unsigned char *)
+ _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTMB);
+ extra = (const unsigned char *)
+ _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB);
+ indirect = (const int32_t *)
+ _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTMB);
+ idx = findidx (&cp);
+ if (idx > 0)
+ for (i = 0; i < cset->nequiv_classes; ++i)
+ {
+ int32_t equiv_class_idx = cset->equiv_classes[i];
+ size_t weight_len = weights[idx];
+ if (weight_len == weights[equiv_class_idx])
+ {
+ int cnt = 0;
+ while (cnt <= weight_len
+ && (weights[equiv_class_idx + 1 + cnt]
+ == weights[idx + 1 + cnt]))
+ ++cnt;
+ if (cnt > weight_len)
+ {
+ match_len = elem_len;
+ goto check_node_accept_bytes_match;
+ }
+ }
+ }
+ }
+ }
+ else
+# endif /* _LIBC */
+ {
+ /* match with range expression? */
+#if __GNUC__ >= 2
+ wchar_t cmp_buf[] = {L'\0', L'\0', wc, L'\0', L'\0', L'\0'};
+#else
+ wchar_t cmp_buf[] = {L'\0', L'\0', L'\0', L'\0', L'\0', L'\0'};
+ cmp_buf[2] = wc;
+#endif
+ for (i = 0; i < cset->nranges; ++i)
+ {
+ cmp_buf[0] = cset->range_starts[i];
+ cmp_buf[4] = cset->range_ends[i];
+ if (wcscoll (cmp_buf, cmp_buf + 2) <= 0
+ && wcscoll (cmp_buf + 2, cmp_buf + 4) <= 0)
+ {
+ match_len = char_len;
+ goto check_node_accept_bytes_match;
+ }
+ }
+ }
+ check_node_accept_bytes_match:
+ if (!cset->non_match)
+ return match_len;
+ else
+ {
+ if (match_len > 0)
+ return 0;
+ else
+ return (elem_len > char_len) ? elem_len : char_len;
+ }
+ }
+ return 0;
+}
+
+# ifdef _LIBC
+static unsigned int
+find_collation_sequence_value (mbs, mbs_len)
+ const unsigned char *mbs;
+ size_t mbs_len;
+{
+ uint32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES);
+ if (nrules == 0)
+ {
+ if (mbs_len == 1)
+ {
+ /* No valid character. Match it as a single byte character. */
+ const unsigned char *collseq = (const unsigned char *)
+ _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQMB);
+ return collseq[mbs[0]];
+ }
+ return UINT_MAX;
+ }
+ else
+ {
+ int32_t idx;
+ const unsigned char *extra = (const unsigned char *)
+ _NL_CURRENT (LC_COLLATE, _NL_COLLATE_SYMB_EXTRAMB);
+
+ for (idx = 0; ;)
+ {
+ int mbs_cnt, found = 0;
+ int32_t elem_mbs_len;
+ /* Skip the name of collating element name. */
+ idx = idx + extra[idx] + 1;
+ elem_mbs_len = extra[idx++];
+ if (mbs_len == elem_mbs_len)
+ {
+ for (mbs_cnt = 0; mbs_cnt < elem_mbs_len; ++mbs_cnt)
+ if (extra[idx + mbs_cnt] != mbs[mbs_cnt])
+ break;
+ if (mbs_cnt == elem_mbs_len)
+ /* Found the entry. */
+ found = 1;
+ }
+ /* Skip the byte sequence of the collating element. */
+ idx += elem_mbs_len;
+ /* Adjust for the alignment. */
+ idx = (idx + 3) & ~3;
+ /* Skip the collation sequence value. */
+ idx += sizeof (uint32_t);
+ /* Skip the wide char sequence of the collating element. */
+ idx = idx + sizeof (uint32_t) * (extra[idx] + 1);
+ /* If we found the entry, return the sequence value. */
+ if (found)
+ return *(uint32_t *) (extra + idx);
+ /* Skip the collation sequence value. */
+ idx += sizeof (uint32_t);
+ }
+ }
+}
+# endif /* _LIBC */
+#endif /* RE_ENABLE_I18N */
+
+/* Check whether the node accepts the byte which is IDX-th
+ byte of the INPUT. */
+
+static int
+check_node_accept (preg, node, mctx, idx)
+ const regex_t *preg;
+ const re_token_t *node;
+ const re_match_context_t *mctx;
+ int idx;
+{
+ unsigned char ch;
+ if (node->constraint)
+ {
+ /* The node has constraints. Check whether the current context
+ satisfies the constraints. */
+ unsigned int context = re_string_context_at (mctx->input, idx,
+ mctx->eflags,
+ preg->newline_anchor);
+ if (NOT_SATISFY_NEXT_CONSTRAINT (node->constraint, context))
+ return 0;
+ }
+ ch = re_string_byte_at (mctx->input, idx);
+ if (node->type == CHARACTER)
+ return node->opr.c == ch;
+ else if (node->type == SIMPLE_BRACKET)
+ return bitset_contain (node->opr.sbcset, ch);
+ else if (node->type == OP_PERIOD)
+ return !((ch == '\n' && !(preg->syntax & RE_DOT_NEWLINE))
+ || (ch == '\0' && (preg->syntax & RE_DOT_NOT_NULL)));
+ else
+ return 0;
+}
+
+/* Extend the buffers, if the buffers have run out. */
+
+static reg_errcode_t
+extend_buffers (mctx)
+ re_match_context_t *mctx;
+{
+ reg_errcode_t ret;
+ re_string_t *pstr = mctx->input;
+
+ /* Double the lengthes of the buffers. */
+ ret = re_string_realloc_buffers (pstr, pstr->bufs_len * 2);
+ if (BE (ret != REG_NOERROR, 0))
+ return ret;
+
+ if (mctx->state_log != NULL)
+ {
+ /* And double the length of state_log. */
+ re_dfastate_t **new_array;
+ new_array = re_realloc (mctx->state_log, re_dfastate_t *,
+ pstr->bufs_len * 2);
+ if (BE (new_array == NULL, 0))
+ return REG_ESPACE;
+ mctx->state_log = new_array;
+ }
+
+ /* Then reconstruct the buffers. */
+ if (pstr->icase)
+ {
+#ifdef RE_ENABLE_I18N
+ if (MB_CUR_MAX > 1)
+ build_wcs_upper_buffer (pstr);
+ else
+#endif /* RE_ENABLE_I18N */
+ build_upper_buffer (pstr);
+ }
+ else
+ {
+#ifdef RE_ENABLE_I18N
+ if (MB_CUR_MAX > 1)
+ build_wcs_buffer (pstr);
+ else
+#endif /* RE_ENABLE_I18N */
+ {
+ if (pstr->trans != NULL)
+ re_string_translate_buffer (pstr);
+ else
+ pstr->valid_len = pstr->bufs_len;
+ }
+ }
+ return REG_NOERROR;
+}
+
+\f
+/* Functions for matching context. */
+
+/* Initialize MCTX. */
+
+static reg_errcode_t
+match_ctx_init (mctx, eflags, input, n)
+ re_match_context_t *mctx;
+ int eflags, n;
+ re_string_t *input;
+{
+ mctx->eflags = eflags;
+ mctx->input = input;
+ mctx->match_last = -1;
+ if (n > 0)
+ {
+ mctx->bkref_ents = re_malloc (struct re_backref_cache_entry, n);
+ mctx->sub_tops = re_malloc (re_sub_match_top_t *, n);
+ if (BE (mctx->bkref_ents == NULL || mctx->sub_tops == NULL, 0))
+ return REG_ESPACE;
+ }
+ else
+ mctx->bkref_ents = NULL;
+ mctx->nbkref_ents = 0;
+ mctx->abkref_ents = n;
+ mctx->max_mb_elem_len = 1;
+ mctx->nsub_tops = 0;
+ mctx->asub_tops = n;
+ return REG_NOERROR;
+}
+
+/* Clean the entries which depend on the current input in MCTX.
+ This function must be invoked when the matcher changes the start index
+ of the input, or changes the input string. */
+
+static void
+match_ctx_clean (mctx)
+ re_match_context_t *mctx;
+{
+ match_ctx_free_subtops (mctx);
+ mctx->nsub_tops = 0;
+ mctx->nbkref_ents = 0;
+}
+
+/* Free all the memory associated with MCTX. */
+
+static void
+match_ctx_free (mctx)
+ re_match_context_t *mctx;
+{
+ match_ctx_free_subtops (mctx);
+ re_free (mctx->sub_tops);
+ re_free (mctx->bkref_ents);
+}
+
+/* Free all the memory associated with MCTX->SUB_TOPS. */
+
+static void
+match_ctx_free_subtops (mctx)
+ re_match_context_t *mctx;
+{
+ int st_idx;
+ for (st_idx = 0; st_idx < mctx->nsub_tops; ++st_idx)
+ {
+ int sl_idx;
+ re_sub_match_top_t *top = mctx->sub_tops[st_idx];
+ for (sl_idx = 0; sl_idx < top->nlasts; ++sl_idx)
+ {
+ re_sub_match_last_t *last = top->lasts[sl_idx];
+ re_free (last->path.array);
+ re_free (last);
+ }
+ re_free (top->lasts);
+ if (top->path)
+ {
+ re_free (top->path->array);
+ re_free (top->path);
+ }
+ free (top);
+ }
+}
+
+/* Add a new backreference entry to MCTX.
+ Note that we assume that caller never call this function with duplicate
+ entry, and call with STR_IDX which isn't smaller than any existing entry.
+*/
+
+static reg_errcode_t
+match_ctx_add_entry (mctx, node, str_idx, from, to)
+ re_match_context_t *mctx;
+ int node, str_idx, from, to;
+{
+ if (mctx->nbkref_ents >= mctx->abkref_ents)
+ {
+ struct re_backref_cache_entry* new_entry;
+ new_entry = re_realloc (mctx->bkref_ents, struct re_backref_cache_entry,
+ mctx->abkref_ents * 2);
+ if (BE (new_entry == NULL, 0))
+ {
+ re_free (mctx->bkref_ents);
+ return REG_ESPACE;
+ }
+ mctx->bkref_ents = new_entry;
+ memset (mctx->bkref_ents + mctx->nbkref_ents, '\0',
+ sizeof (struct re_backref_cache_entry) * mctx->abkref_ents);
+ mctx->abkref_ents *= 2;
+ }
+ mctx->bkref_ents[mctx->nbkref_ents].node = node;
+ mctx->bkref_ents[mctx->nbkref_ents].str_idx = str_idx;
+ mctx->bkref_ents[mctx->nbkref_ents].subexp_from = from;
+ mctx->bkref_ents[mctx->nbkref_ents].subexp_to = to;
+ mctx->bkref_ents[mctx->nbkref_ents++].flag = 0;
+ if (mctx->max_mb_elem_len < to - from)
+ mctx->max_mb_elem_len = to - from;
+ return REG_NOERROR;
+}
+
+/* Search for the first entry which has the same str_idx.
+ Note that MCTX->BKREF_ENTS is already sorted by MCTX->STR_IDX. */
+
+static int
+search_cur_bkref_entry (mctx, str_idx)
+ re_match_context_t *mctx;
+ int str_idx;
+{
+ int left, right, mid;
+ right = mctx->nbkref_ents;
+ for (left = 0; left < right;)
+ {
+ mid = (left + right) / 2;
+ if (mctx->bkref_ents[mid].str_idx < str_idx)
+ left = mid + 1;
+ else
+ right = mid;
+ }
+ return left;
+}
+
+static void
+match_ctx_clear_flag (mctx)
+ re_match_context_t *mctx;
+{
+ int i;
+ for (i = 0; i < mctx->nbkref_ents; ++i)
+ {
+ mctx->bkref_ents[i].flag = 0;
+ }
+}
+
+/* Register the node NODE, whose type is OP_OPEN_SUBEXP, and which matches
+ at STR_IDX. */
+
+static reg_errcode_t
+match_ctx_add_subtop (mctx, node, str_idx)
+ re_match_context_t *mctx;
+ int node, str_idx;
+{
+#ifdef DEBUG
+ assert (mctx->sub_tops != NULL);
+ assert (mctx->asub_tops > 0);
+#endif
+ if (mctx->nsub_tops == mctx->asub_tops)
+ {
+ re_sub_match_top_t **new_array;
+ mctx->asub_tops *= 2;
+ new_array = re_realloc (mctx->sub_tops, re_sub_match_top_t *,
+ mctx->asub_tops);
+ if (BE (new_array == NULL, 0))
+ return REG_ESPACE;
+ mctx->sub_tops = new_array;
+ }
+ mctx->sub_tops[mctx->nsub_tops] = calloc (1, sizeof (re_sub_match_top_t));
+ if (mctx->sub_tops[mctx->nsub_tops] == NULL)
+ return REG_ESPACE;
+ mctx->sub_tops[mctx->nsub_tops]->node = node;
+ mctx->sub_tops[mctx->nsub_tops++]->str_idx = str_idx;
+ return REG_NOERROR;
+}
+
+/* Register the node NODE, whose type is OP_CLOSE_SUBEXP, and which matches
+ at STR_IDX, whose corresponding OP_OPEN_SUBEXP is SUB_TOP. */
+
+static re_sub_match_last_t *
+match_ctx_add_sublast (subtop, node, str_idx)
+ re_sub_match_top_t *subtop;
+ int node, str_idx;
+{
+ re_sub_match_last_t *new_entry;
+ if (subtop->nlasts == subtop->alasts)
+ {
+ re_sub_match_last_t **new_array;
+ subtop->alasts = 2 * subtop->alasts + 1;
+ new_array = re_realloc (subtop->lasts, re_sub_match_last_t *,
+ subtop->alasts);
+ if (BE (new_array == NULL, 0))
+ return NULL;
+ subtop->lasts = new_array;
+ }
+ new_entry = calloc (1, sizeof (re_sub_match_last_t));
+ if (BE (new_entry == NULL, 0))
+ return NULL;
+ subtop->lasts[subtop->nlasts] = new_entry;
+ new_entry->node = node;
+ new_entry->str_idx = str_idx;
+ ++subtop->nlasts;
+ return new_entry;
+}
+
+static void
+sift_ctx_init (sctx, sifted_sts, limited_sts, last_node, last_str_idx,
+ check_subexp)
+ re_sift_context_t *sctx;
+ re_dfastate_t **sifted_sts, **limited_sts;
+ int last_node, last_str_idx, check_subexp;
+{
+ sctx->sifted_states = sifted_sts;
+ sctx->limited_states = limited_sts;
+ sctx->last_node = last_node;
+ sctx->last_str_idx = last_str_idx;
+ sctx->check_subexp = check_subexp;
+ sctx->cur_bkref = -1;
+ sctx->cls_subexp_idx = -1;
+ re_node_set_init_empty (&sctx->limits);
+}
*
* Home page of code is: http://smartmontools.sourceforge.net
*
- * Copyright (C) 2006-10 Douglas Gilbert <dougg@torque.net>
+ * Copyright (C) 2006-10 Douglas Gilbert <dgilbert@interlog.com>
* Copyright (C) 2009-10 Christian Franke <smartmontools-support@lists.sourceforge.net>
*
* This program is free software; you can redistribute it and/or modify
#include "extern.h"
#include "scsicmds.h"
#include "atacmds.h" // ataReadHDIdentity()
+#include "knowndrives.h" // lookup_usb_device()
#include "utility.h"
#include "dev_interface.h"
#include "dev_ata_cmd_set.h" // ata_device_with_command_set
#include "dev_tunnelled.h" // tunnelled_device<>
-const char * scsiata_cpp_cvsid = "$Id: scsiata.cpp 3077 2010-03-16 20:48:06Z chrfranke $";
+const char * scsiata_cpp_cvsid = "$Id: scsiata.cpp 3095 2010-04-30 12:33:27Z dpgilbert $";
/* for passing global control variables */
extern smartmonctrl *con;
/////////////////////////////////////////////////////////////////////////////
// USB device type detection
-struct usb_id_entry {
- int vendor_id, product_id, version;
- const char * type;
-};
-
-const char d_sat[] = "sat";
-const char d_cypress[] = "usbcypress";
-const char d_jmicron[] = "usbjmicron";
-const char d_jmicron_x[] = "usbjmicron,x";
-const char d_sunplus[] = "usbsunplus";
-const char d_unsup[] = "unsupported";
-
-// Map USB IDs -> '-d type' string
-const usb_id_entry usb_ids[] = {
- // Cypress
- { 0x04b4, 0x6830, 0x0001, d_unsup }, // Cypress CY7C68300A (AT2)
- { 0x04b4, 0x6830, 0x0240, d_cypress }, // Cypress CY7C68300B/C (AT2LP)
-//{ 0x04b4, 0x6831, -1, d_cypress }, // Cypress CY7C68310 (ISD-300LP)
- // Myson Century
- { 0x04cf, 0x8818, 0xb007, d_unsup }, // Myson Century CS8818
- // Samsung
- { 0x04e8, 0x5f06, -1, d_sat }, // Samsung Story Station
- // Sunplus
- { 0x04fc, 0x0c15, 0xf615, d_sunplus }, // SunPlus SPDIF215
- { 0x04fc, 0x0c25, 0x0103, d_sunplus }, // SunPlus SPDIF225 (USB+SATA->SATA)
- // Iomega
- { 0x059b, 0x0272, -1, d_cypress }, // Iomega LPHD080-0
- { 0x059b, 0x0275, 0x0001, d_unsup }, // Iomega MDHD500-U
- // LaCie
- { 0x059f, 0x0651, -1, d_unsup }, // LaCie hard disk (FA Porsche design)
- { 0x059f, 0x1018, -1, d_sat }, // LaCie hard disk (Neil Poulton design)
- { 0x059f, 0x1019, -1, d_jmicron }, // LaCie Desktop Hard Drive
- // In-System Design
- { 0x05ab, 0x0060, 0x1101, d_cypress }, // In-System/Cypress ISD-300A1
- // Genesys Logic
- { 0x05e3, 0x0702, -1, d_unsup }, // Genesys Logic GL881E
- { 0x05e3, 0x0718, 0x0041, d_sat }, // Genesys Logic ? (TODO: requires '-T permissive')
- // Prolific
- { 0x067b, 0x2507, -1, d_unsup }, // Prolific PL2507 (USB->PATA)
- { 0x067b, 0x3507, 0x0001, d_unsup }, // Prolific PL3507 (USB+IEE1394->PATA)
- // Freecom
- { 0x07ab, 0xfc8e, 0x010f, d_sunplus }, // Freecom Hard Drive XS
- // Toshiba
- { 0x0930, 0x0b03, -1, d_sunplus }, // Toshiba PX1270E-1G16
- { 0x0930, 0x0b09, -1, d_sunplus }, // Toshiba PX1396E-3T01 (similar to Dura Micro 501)
- // Seagate
- { 0x0bc2, 0x2000, -1, d_sat }, // Seagate FreeAgent Go
- { 0x0bc2, 0x2100, -1, d_sat }, // Seagate FreeAgent Go
- { 0x0bc2, 0x2101, -1, d_sat }, // Seagate FreeAgent Go
- { 0x0bc2, 0x2200, -1, d_sat }, // Seagate FreeAgent Go FW
- { 0x0bc2, 0x2300, -1, d_sat }, // Seagate Expansion Portable
- { 0x0bc2, 0x3000, -1, d_sat }, // Seagate FreeAgent Desktop
- { 0x0bc2, 0x3001, -1, d_sat }, // Seagate FreeAgent Desk
- // Dura Micro
- { 0x0c0b, 0xb159, 0x0103, d_sunplus }, // Dura Micro 509
- // Maxtor
- { 0x0d49, 0x7310, 0x0125, d_sat }, // Maxtor OneTouch 4
- { 0x0d49, 0x7350, 0x0125, d_sat }, // Maxtor OneTouch 4 Mini
- { 0x0d49, 0x7410, 0x0122, d_sat }, // Maxtor Basics Desktop
- { 0x0d49, 0x7450, 0x0122, d_sat }, // Maxtor Basics Portable
- // Western Digital
- { 0x1058, 0x0701, 0x0240, d_cypress }, // WD My Passport (IDE)
- { 0x1058, 0x0702, 0x0102, d_sat }, // WD My Passport Portable
- { 0x1058, 0x0704, 0x0175, d_sat }, // WD My Passport Essential
- { 0x1058, 0x0705, 0x0175, d_sat }, // WD My Passport Elite
- { 0x1058, 0x070a, 0x1028, d_sat }, // WD My Passport 070A
- { 0x1058, 0x0906, 0x0012, d_sat }, // WD My Book ES
- { 0x1058, 0x1001, 0x0104, d_sat }, // WD Elements Desktop
- { 0x1058, 0x1003, 0x0175, d_sat }, // WD Elements Desktop WDE1UBK...
- { 0x1058, 0x1010, 0x0105, d_sat }, // WD Elements
- { 0x1058, 0x1100, 0x0165, d_sat }, // WD My Book Essential
- { 0x1058, 0x1102, 0x1028, d_sat }, // WD My Book
- { 0x1058, 0x1110, 0x1030, d_sat }, // WD My Book Essential
- // Initio
- { 0x13fd, 0x0540, -1, d_unsup }, // Initio 316000
- { 0x13fd, 0x1240, 0x0104, d_sat }, // Initio ? (USB->SATA)
- { 0x13fd, 0x1340, 0x0208, d_sat }, // Initio ? (USB+SATA->SATA)
- // JMicron
- { 0x152d, 0x2329, 0x0100, d_jmicron }, // JMicron JM20329 (USB->SATA)
- { 0x152d, 0x2336, 0x0100, d_jmicron_x},// JMicron JM20336 (USB+SATA->SATA, USB->2xSATA)
- { 0x152d, 0x2338, 0x0100, d_jmicron }, // JMicron JM20337/8 (USB->SATA+PATA, USB+SATA->PATA)
- { 0x152d, 0x2339, 0x0100, d_jmicron_x},// JMicron JM20339 (USB->SATA)
- // Verbatim
- { 0x18a5, 0x0215, 0x0001, d_sat }, // Verbatim FW/USB160 - Oxford OXUF934SSA-LQAG (USB+IEE1394->SATA)
- // SunplusIT
- { 0x1bcf, 0x0c31, -1, d_sunplus }, // SunplusIT
- // OnSpec
- { 0x55aa, 0x2b00, 0x0100, d_unsup } // OnSpec ? (USB->PATA)
-};
-
-const unsigned num_usb_ids = sizeof(usb_ids)/sizeof(usb_ids[0]);
-
-
// Format USB ID for error messages
static std::string format_usb_id(int vendor_id, int product_id, int version)
{
const char * smart_interface::get_usb_dev_type_by_id(int vendor_id, int product_id,
int version /*= -1*/)
{
- const usb_id_entry * entry = 0;
- bool state = false;
-
- for (unsigned i = 0; i < num_usb_ids; i++) {
- const usb_id_entry & e = usb_ids[i];
- if (!(vendor_id == e.vendor_id && product_id == e.product_id))
- continue;
-
- // If two entries with same vendor:product ID have different
- // types, use version (if provided by OS) to select entry.
- bool s = (version >= 0 && version == e.version);
- if (entry) {
- if (s <= state) {
- if (s == state && e.type != entry->type) {
- set_err(EINVAL, "USB bridge %s type is ambiguous: '%s' or '%s'",
- format_usb_id(vendor_id, product_id, version).c_str(),
- e.type, entry->type);
- return 0;
- }
- continue;
- }
- }
- state = s;
- entry = &e;
- }
+ usb_dev_info info, info2;
+ int n = lookup_usb_device(vendor_id, product_id, version, info, info2);
- if (!entry) {
+ if (n <= 0) {
set_err(EINVAL, "Unknown USB bridge %s",
format_usb_id(vendor_id, product_id, version).c_str());
return 0;
}
- if (entry->type == d_unsup) {
+
+ if (n > 1) {
+ set_err(EINVAL, "USB bridge %s type is ambiguous: '%s' or '%s'",
+ format_usb_id(vendor_id, product_id, version).c_str(),
+ (!info.usb_type.empty() ? info.usb_type.c_str() : "[unsupported]"),
+ (!info2.usb_type.empty() ? info2.usb_type.c_str() : "[unsupported]"));
+ return 0;
+ }
+
+ if (info.usb_type.empty()) {
set_err(ENOSYS, "Unsupported USB bridge %s",
format_usb_id(vendor_id, product_id, version).c_str());
return 0;
}
- return entry->type;
+
+ // TODO: change return type to std::string
+ static std::string type;
+ type = info.usb_type;
+ return type.c_str();
}
* Copyright (C) 1999-2000 Michael Cornwell <cornwell@acm.org>
*
* Additional SCSI work:
- * Copyright (C) 2003-8 Douglas Gilbert <dougg@torque.net>
+ * Copyright (C) 2003-10 Douglas Gilbert <dgilbert@interlog.com>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
#include "dev_interface.h"
#include "utility.h"
-const char *scsicmds_c_cvsid="$Id: scsicmds.cpp,v 1.98 2009/06/24 04:10:10 dpgilbert Exp $"
+const char *scsicmds_c_cvsid="$Id: scsicmds.cpp 3096 2010-04-30 14:32:49Z chrfranke $"
CONFIG_H_CVSID EXTERN_H_CVSID INT64_H_CVSID SCSICMDS_H_CVSID UTILITY_H_CVSID;
/* for passing global control variables */
* Copyright (C) 2000 Michael Cornwell <cornwell@acm.org>
*
* Additional SCSI work:
- * Copyright (C) 2003-8 Douglas Gilbert <dougg@torque.net>
+ * Copyright (C) 2003-10 Douglas Gilbert <dgilbert@interlog.com>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
#ifndef SCSICMDS_H_
#define SCSICMDS_H_
-#define SCSICMDS_H_CVSID "$Id: scsicmds.h 2924 2009-09-26 20:38:40Z chrfranke $\n"
+#define SCSICMDS_H_CVSID "$Id: scsicmds.h 3095 2010-04-30 12:33:27Z dpgilbert $\n"
#include <stdio.h>
#include <stdlib.h>
* Copyright (C) 2000 Michael Cornwell <cornwell@acm.org>
*
* Additional SCSI work:
- * Copyright (C) 2003-9 Douglas Gilbert <dougg@torque.net>
+ * Copyright (C) 2003-10 Douglas Gilbert <dgilbert@interlog.com>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
#define GBUF_SIZE 65535
-const char * scsiprint_c_cvsid = "$Id: scsiprint.cpp 2861 2009-07-24 16:47:03Z chrfranke $"
+const char * scsiprint_c_cvsid = "$Id: scsiprint.cpp 3095 2010-04-30 12:33:27Z dpgilbert $"
SCSIPRINT_H_CVSID;
// control block which points to external global control variables
* Copyright (C) 2000 Michael Cornwell <cornwell@acm.org>
*
* Additional SCSI work:
- * Copyright (C) 2003-9 Douglas Gilbert <dougg@torque.net>
+ * Copyright (C) 2003-10 Douglas Gilbert <dgilbert@interlog.com>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
#ifndef SCSI_PRINT_H_
#define SCSI_PRINT_H_
-#define SCSIPRINT_H_CVSID "$Id: scsiprint.h,v 1.24 2009/06/21 02:39:32 dpgilbert Exp $\n"
+#define SCSIPRINT_H_CVSID "$Id: scsiprint.h 3096 2010-04-30 14:32:49Z chrfranke $\n"
// Options for scsiPrintMain
// TODO: Move remaining options from con->* to here.
.ig
- Copyright (C) 2002-9 Bruce Allen <smartmontools-support@lists.sourceforge.net>
+ Copyright (C) 2002-10 Bruce Allen <smartmontools-support@lists.sourceforge.net>
- $Id: smartctl.8.in 3072 2010-03-04 21:56:41Z chrfranke $
+ $Id: smartctl.8.in 3119 2010-06-11 16:21:25Z chrfranke $
This program is free software; you can redistribute it and/or modify it
under the terms of the GNU General Public License as published by the Free
California, Santa Cruz. http://ssrc.soe.ucsc.edu/
..
-.TH SMARTCTL 8 CURRENT_CVS_DATE CURRENT_CVS_VERSION CURRENT_CVS_DATE
+.TH SMARTCTL 8 CURRENT_SVN_DATE CURRENT_SVN_VERSION CURRENT_SVN_DATE
.SH NAME
\fBsmartctl\fP \- Control and Monitor Utility for SMART Disks
.B /usr/local/sbin/smartctl
.SH PACKAGE VERSION
-CURRENT_CVS_VERSION released CURRENT_CVS_DATE at CURRENT_CVS_TIME
+CURRENT_SVN_VERSION CURRENT_SVN_DATE CURRENT_SVN_REV
.SH DESCRIPTION
\fBsmartctl\fP controls the Self\-Monitoring, Analysis and Reporting
will execute the corresponding commands in the order: INFORMATION,
ENABLE/DISABLE, DISPLAY DATA, RUN/ABORT TESTS.
-Long options are not supported on all systems. Use
-.B \'smartctl \-h\'
-to see the available options.
-
.TP
.B SHOW INFORMATION OPTIONS:
.TP
.nf
\'\-H \-i \-A \-l error \-l selftest \-l background \-l sasphy\'.
.fi
+.TP
+.B \-\-scan
+Scans for devices and prints each device name, device type and protocol
+([ATA] or [SCSI]) info. May be used in conjunction with \'\-d TYPE\'
+to restrict the scan to a specific TYPE. See also info about platform
+specific device scan and the \fBDEVICESCAN\fP directive on
+\fBsmartd\fP(8) man page.
+.TP
+.B \-\-scan\-open
+Same as \-\-scan, but also tries to open each device before printing
+device info. The device open may change the device type due
+to autodetection (see also \'\-d test\').
.TP
.B RUN\-TIME BEHAVIOR OPTIONS:
.I noserial
\- Do not print the serial number of the device.
-
.TP
.B \-d TYPE, \-\-device=TYPE
Specifies the type of the device. The valid arguments to this option
.B HighPoint RocketRAID controllers are currently ONLY supported under Linux and FreeBSD.
.B cciss controllers are currently ONLY supported under Linux and FreeBSD.
-
.TP
.B \-T TYPE, \-\-tolerance=TYPE
[ATA only] Specifies how tolerant \fBsmartctl\fP should be of ATA and SMART
\- equivalent to giving a large number of \'\-T permissive\' options:
ignore failures of \fBany number\fP of \fBmandatory\fP SMART commands.
Please see the note above.
-
.TP
.B \-b TYPE, \-\-badsum=TYPE
[ATA only] Specifies the action \fBsmartctl\fP should take if a checksum
.I ignore
\- continue silently without issuing a warning.
-
.TP
.B \-r TYPE, \-\-report=TYPE
Intended primarily to help \fBsmartmontools\fP developers understand
reconstructed from the debug report read from stdin.
Then \fBsmartctl\fP internally simulates an ATA device with the same
behaviour. This is does not work for SCSI devices yet.
-
.TP
.B \-n POWERMODE, \-\-nocheck=POWERMODE
[ATA only] Specifies if \fBsmartctl\fP should exit before performing any
connection with the third category.
.TP
.B \-S VALUE, \-\-saveauto=VALUE
-Enables or disables SMART autosave of device vendor\-specific
+[ATA] Enables or disables SMART autosave of device vendor\-specific
Attributes. The valid arguments to this option are \fIon\fP
and \fIoff\fP. Note that this feature is preserved across disk power
cycles, so you should only need to issue it once.
-For SCSI devices this toggles the value of the Global Logging Target
-Save Disabled (GLTSD) bit in the Control Mode Page. Some disk
+The ATA standard does not specify a method to check whether SMART
+autosave is enabled. Unlike SCSI (below), smartctl is unable to print
+a warning if autosave is disabled.
+
+[SCSI] For SCSI devices this toggles the value of the Global Logging
+Target Save Disabled (GLTSD) bit in the Control Mode Page. Some disk
manufacturers set this bit by default. This prevents error counters,
power\-up hours and other useful data from being placed in non\-volatile
storage, so these values may be reset to zero the next time the device
log is not supported, the Summary SMART self-test log is printed.
Please note that some recent (e.g. Samsung) drives report errors only
-in the Comprehensive SMART error log. The Summary SMART error log can
-be read but is always empty.
+in the Extended Comprehensive SMART error log. The Summary SMART error
+log can be read but is always empty.
.I selftest
\- [ATA] prints the SMART self\-test log. The disk maintains a self\-test
\- [ATA only] [NEW EXPERIMENTAL SMARTCTL FEATURE] prints values
and descriptions of the SCT Error Recovery Control settings. These
are equivalent to TLER (as used by Western Digital), CCTL (as used
-by Samsung and Hitachi) and ERC (as used by Seagate. READTIME and
+by Samsung and Hitachi) and ERC (as used by Seagate). READTIME and
WRITETIME arguments (deciseconds) set the specified values. Values of 0
disable the feature, other values less than 65 are probably not
supported. For RAID configurations, this is typically set to
.fi
writes a binary representation of the one sector log 0x11
(SATA Phy Event Counters) to file log.bin.
-
.TP
.B \-v ID,FORMAT[:BYTEORDER][,NAME], \-\-vendorattribute=ID,FORMAT[:BYTEORDER][,NAME]
[ATA only] Sets a vendor\-specific raw value print FORMAT, an optional
Note: a table of hard drive models, listing which Attribute
corresponds to temperature, can be found at:
\fBhttp://www.guzu.net/linux/hddtemp.db\fP
-
.TP
.B \-F TYPE, \-\-firmwarebug=TYPE
[ATA only] Modifies the behavior of \fBsmartctl\fP to compensate for some
(3) strange and impossible values for the ATA error log timestamps.
.I samsung2
-\- In more recent Samsung disks (firmware revisions ending in "\-23")
-the number of ATA errors reported is byte swapped. Enabling this
-option tells \fBsmartctl\fP to evaluate this quantity in
+\- In some Samsung disks the number of ATA errors reported is byte swapped.
+Enabling this option tells \fBsmartctl\fP to evaluate this quantity in
byte\-reversed order. An indication that your Samsung disk needs this
option is that the self\-test log is printed correctly, but there are a
very large number of errors in the SMART error log. This is because
.I swapid
\- Fixes byte swapped ATA identify strings (device name, serial number,
firmware version) returned by some buggy device drivers.
-
.TP
.B \-P TYPE, \-\-presets=TYPE
[ATA only] Specifies whether \fBsmartctl\fP should use any preset options
smartctl \-P showall \'MODEL\' \'FIRMWARE\'
.fi
lists all entries for this MODEL and a specific FIRMWARE version.
-
.TP
.B \-B [+]FILE, \-\-drivedb=[+]FILE
[ATA only] [NEW EXPERIMENTAL SMARTCTL FEATURE] Read the drive database from
specified, then the new entries prepend the built in entries.
If this option is not specified, optional entries are read from the file
-\fB/usr/local/etc/smart_drivedb.h\fP (Windows: \fB./smart_drivedb.conf\fP).
+\fB/usr/local/etc/smart_drivedb.h\fP (Windows: \fBEXEDIR/drivedb-add.h\fP).
.\" BEGIN ENABLE_DRIVEDB
-If \fB/usr/local/share/smartmontools/drivedb.h\fP is present, the
+If \fB/usr/local/share/smartmontools/drivedb.h\fP
+(Windows: \fBEXEDIR/drivedb.h\fP) is present, the
contents of this file is used instead of the built in table.
Run the script \fB/usr/local/sbin/update-smart-drivedb\fP to update this
"", // No warning.
"" // No options preset.
},
+ /* USB ID entry: */
+ {
+ "USB: Device; Bridge", // Info about USB device and bridge name.
+ "0x1234:0xabcd", // Regular expression to match vendor:product ID.
+ "0x0101", // Regular expression to match bcdDevice.
+ "", // Not used.
+ "\-d sat" // String with device type option.
+ },
/* ... */
.fi
The valid arguments to this option are:
.I offline
-\- runs SMART Immediate Offline Test. This immediately
+\- [ATA] runs SMART Immediate Offline Test. This immediately
starts the test described above. This command can be given during
normal system operation. The effects of this test are visible only in
that it updates the SMART Attribute values, and if errors are
found they will appear in the SMART error log, visible with the \'\-l error\'
-option. [In the case of SCSI devices runs the default self test in
-foreground. No entry is placed in the self test log.]
+option.
If the \'\-c\' option to \fBsmartctl\fP shows that the device has the
"Suspend Offline collection upon new command" capability then you can
try to track the progress of the test with \'\-c\', as it will abort
the test.
+.I offline
+\- [SCSI] runs the default self test in foreground. No entry is placed
+in the self test log.
+
.I short
-\- runs SMART Short Self Test (usually under ten minutes).
-[Note: in the case of SCSI devices,
-this command option runs the "Background short" self\-test.]
+\- [ATA] runs SMART Short Self Test (usually under ten minutes).
This command can be given during normal system operation (unless run in
captive mode \- see the \'\-C\' option below). This is a
test in a different category than the immediate or automatic offline
self\-test can be monitored by watching this log during the self\-test; with other disks
use the \'\-c\' option to monitor progress.
+.I short
+\- [SCSI] runs the "Background short" self\-test.
+
.I long
-\- runs SMART Extended Self Test (tens of minutes).
-[Note: in the case of SCSI devices,
-this command option runs the "Background long" self\-test.]
-This is a
+\- [ATA] runs SMART Extended Self Test (tens of minutes). This is a
longer and more thorough version of the Short Self Test described
above. Note that this command can be given during normal
system operation (unless run in captive mode \- see the \'\-C\' option below).
+.I long
+\- [SCSI] runs the "Background long" self\-test.
+
.I conveyance
\- [ATA only] runs a SMART Conveyance Self Test (minutes). This
self\-test routine is intended to identify damage incurred during
Selective self\-tests can be run during normal system operation (unless
done in captive mode \- see the \'\-C\' option below).
-[Note: To use this feature on Linux, the kernel must be compiled with
-the configuration option CONFIG_IDE_TASKFILE_IO enabled. Please report
-unusual or incorrect behavior to the smartmontools\-support mailing list.]
-
The following variants of the selective self\-test command use spans based
on the ranges from past tests already stored on the disk:
default (1 minute), or last non-volatile setting by the next hard reset.
This command also clears the temperature history table. See
\'\-l scttemp\' above for more information about SCT temperature logging.
-
.TP
.B \-C, \-\-captive
-Runs self\-tests in captive mode. This has no effect with \'\-t
-offline\' or if the \'\-t\' option is not used. [Note: in the case of
-SCSI devices, this command option runs the self\-test in "Foreground"
-mode.]
+[ATA] Runs self\-tests in captive mode. This has no effect with \'\-t
+offline\' or if the \'\-t\' option is not used.
\fBWARNING: Tests run in captive mode may busy out the drive for the
length of the test. Only run captive tests on drives without any
mounted partitions!\fP
+[SCSI] Runs the self\-test in "Foreground" mode.
.TP
.B \-X, \-\-abort
Aborts non\-captive SMART Self Tests. Note that this
.fi
An introductory article about smartmontools is \fIMonitoring Hard
Disks with SMART\fP, by Bruce Allen, Linux Journal, January 2004,
-pages 74\-77. This is \fBhttp://www.linuxjournal.com/article.php?sid=6983\fP
+pages 74\-77. This is \fBhttp://www.linuxjournal.com/article/6983\fP
online.
If you would like to understand better how SMART works, and what it
does, a good place to start is with Sections 4.8 and 6.54 of the first
volume of the \'AT Attachment with Packet Interface\-7\' (ATA/ATAPI\-7)
-specification. This documents the SMART functionality which the
-\fBsmartmontools\fP utilities provide access to. You can find
-Revision 4b of this document at
-\fBhttp://www.t13.org/docs2004/d1532v1r4b-ATA-ATAPI-7.pdf\fP .
-Earlier and later versions of this Specification are available from
+specification Revision 4b. This documents the SMART functionality which the
+\fBsmartmontools\fP utilities provide access to.
+This and other versions of this Specification are available from
the T13 web site \fBhttp://www.t13.org/\fP .
.fi
The functioning of SMART was originally defined by the SFF\-8035i
revision 2 and the SFF\-8055i revision 1.4 specifications. These are
-publications of the Small Form Factors (SFF) Committee. Links to
-these documents may be found in the References section of the
-\fBsmartmontools\fP home page at
-\fBhttp://smartmontools.sourceforge.net/\fP .
+publications of the Small Form Factors (SFF) Committee.
+
+Links to these and other documents may be found on the Links page of the
+\fBsmartmontools\fP Wiki at
+\fBhttp://sourceforge.net/apps/trac/smartmontools/wiki/Links\fP .
.SH
SVN ID OF THIS PAGE:
-$Id: smartctl.8.in 3072 2010-03-04 21:56:41Z chrfranke $
-.\" Local Variables:
-.\" mode: nroff
-.\" End:
+$Id: smartctl.8.in 3119 2010-06-11 16:21:25Z chrfranke $
#include "smartctl.h"
#include "utility.h"
-const char * smartctl_cpp_cvsid = "$Id: smartctl.cpp 3065 2010-02-10 22:16:50Z chrfranke $"
+const char * smartctl_cpp_cvsid = "$Id: smartctl.cpp 3119 2010-06-11 16:21:25Z chrfranke $"
CONFIG_H_CVSID EXTERN_H_CVSID SMARTCTL_H_CVSID;
// This is a block containing all the "control variables". We declare
" Show all SMART information for device\n\n"
" -x, --xall\n"
" Show all information for device\n\n"
+" --scan\n"
+" Scan for devices\n\n"
+" --scan-open\n"
+" Scan for devices and try to open each device\n\n"
);
printf(
"================================== SMARTCTL RUN-TIME BEHAVIOR OPTIONS =====\n\n"
" Drive-specific presets: use, ignore, show, showall\n\n"
" -B [+]FILE, --drivedb=[+]FILE (ATA)\n"
" Read and replace [add] drive database from FILE\n"
+" [default is +%s",
+ get_drivedb_path_add()
+ );
#ifdef SMARTMONTOOLS_DRIVEDBDIR
-" [default is "SMARTMONTOOLS_DRIVEDBDIR"/drivedb.h]\n"
-#endif
-"\n"
+ printf(
+ "\n"
+" and then %s",
+ get_drivedb_path_default()
);
+#endif
printf(
+ "]\n\n"
"============================================ DEVICE SELF-TEST OPTIONS =====\n\n"
" -t TEST, --test=TEST\n"
" Run test. TEST: offline short long conveyance select,M-N\n"
static checksum_err_mode_t checksum_err_mode = CHECKSUM_ERR_WARN;
+static void scan_devices(const char * type, bool with_open, const char * pattern);
+
/* Takes command options and sets features to be run */
const char * parse_options(int argc, char** argv,
ata_print_options & ataopts,
// Please update getvalidarglist() if you edit shortopts
const char *shortopts = "h?Vq:d:T:b:r:s:o:S:HcAl:iaxv:P:t:CXF:n:B:";
// Please update getvalidarglist() if you edit longopts
+ enum { opt_scan = 1000, opt_scan_open = 1001 };
struct option longopts[] = {
{ "help", no_argument, 0, 'h' },
{ "usage", no_argument, 0, 'h' },
{ "firmwarebug", required_argument, 0, 'F' },
{ "nocheck", required_argument, 0, 'n' },
{ "drivedb", required_argument, 0, 'B' },
+ { "scan", no_argument, 0, opt_scan },
+ { "scan-open", no_argument, 0, opt_scan_open },
{ 0, 0, 0, 0 }
};
const char * type = 0; // set to -d optarg
bool no_defaultdb = false; // set true on '-B FILE'
+ int scan = 0; // set by --scan, --scan-open
bool badarg = false, captive = false;
int testcnt = 0; // number of self-tests requested
int optchar;
char *arg;
- // This miserable construction is needed to get emacs to do proper indenting. Sorry!
- while (-1 != (optchar =
- getopt_long(argc, argv, shortopts, longopts, NULL)
- )){
+ while ((optchar = getopt_long(argc, argv, shortopts, longopts, 0)) != -1) {
switch (optchar){
case 'V':
con->dont_print = false;
ataopts.smart_selective_args.pending_time = i+1;
}
} else if (!strncmp(optarg,"select",strlen("select"))) {
- testcnt++;
+ if (ataopts.smart_selective_args.num_spans == 0)
+ testcnt++;
// parse range of LBAs to test
uint64_t start, stop; int mode;
if (split_selective_arg(optarg, &start, &stop, &mode)) {
Usage();
EXIT(0);
break;
+
+ case opt_scan:
+ case opt_scan_open:
+ scan = optchar;
+ break;
+
case '?':
default:
con->dont_print = false;
EXIT(FAILCMD);
}
}
+
+ // Special handling of --scan, --scanopen
+ if (scan) {
+ scan_devices(type, (scan == opt_scan_open), argv[optind]);
+ EXIT(0);
+ }
+
// At this point we have processed all command-line options. If the
// print output is switchable, then start with the print output
// turned off
}
}
+// Device scan
+// smartctl [-d type] --scan[-open] [PATTERN]
+void scan_devices(const char * type, bool with_open, const char * pattern)
+{
+ bool dont_print = !(con->reportataioctl || con->reportscsiioctl);
+ smart_device_list devlist;
+
+ con->dont_print = dont_print;
+ bool ok = smi()->scan_smart_devices(devlist, type , pattern);
+ con->dont_print = false;
+
+ if (!ok) {
+ pout("scan_smart_devices: %s\n", smi()->get_errmsg());
+ return;
+ }
+
+ for (unsigned i = 0; i < devlist.size(); i++) {
+ smart_device * dev = devlist.at(i);
+
+ std::string openmsg;
+ if (with_open) {
+ con->dont_print = dont_print;
+ dev = dev->autodetect_open();
+ con->dont_print = false;
+
+ if (dev->is_open())
+ openmsg = " (opened)";
+ else
+ openmsg = strprintf(" (open failed: %s)", dev->get_errmsg());
+ }
+
+ pout("%s -d %s [%s]%s\n", dev->get_info_name(), dev->get_dev_type(),
+ get_protocol_info(dev), openmsg.c_str());
+ if (dev->is_open())
+ dev->close();
+ }
+}
+
// Main program without exception handling
int main_worker(int argc, char **argv)
{
+ // Throw if CPU endianess does not match compile time test.
+ check_endianness();
+
// Initialize interface
smart_interface::init();
if (!smi())
.ig
-Copyright (C) 2002-8 Bruce Allen <smartmontools-support@lists.sourceforge.net>
+Copyright (C) 2002-10 Bruce Allen <smartmontools-support@lists.sourceforge.net>
-$Id: smartd.8.in 3076 2010-03-12 22:23:08Z chrfranke $
+$Id: smartd.8.in 3117 2010-06-08 15:41:04Z chrfranke $
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
Systems Research Center), Jack Baskin School of Engineering,
University of California, Santa Cruz. http://ssrc.soe.ucsc.edu/
..
-.TH SMARTD 8 CURRENT_CVS_DATE CURRENT_CVS_VERSION CURRENT_CVS_DATE
+.TH SMARTD 8 CURRENT_SVN_DATE CURRENT_SVN_VERSION CURRENT_SVN_DATE
.SH NAME
\fBsmartd\fP \- SMART Disk Monitoring Daemon
.B /usr/local/sbin/smartd
.SH PACKAGE VERSION
-CURRENT_CVS_VERSION released CURRENT_CVS_DATE at CURRENT_CVS_TIME
+CURRENT_SVN_VERSION CURRENT_SVN_DATE CURRENT_SVN_REV
.SH DESCRIPTION
\fBsmartd\fP is a daemon that monitors the Self-Monitoring, Analysis
(equivalent to \fBsmartctl -s on\fP) and polls these and SCSI devices
every 30 minutes (configurable), logging SMART errors and changes of
SMART Attributes via the SYSLOG interface. The default location for
-these SYSLOG notifications and warnings is \fB/var/log/messages\fP.
+these SYSLOG notifications and warnings is system-dependent
+(typically \fB/var/log/messages\fP or \fB/var/log/syslog\fP).
To change this default location, please see the \fB\'-l\'\fP
command-line option described below.
details.
\fBsmartd\fP can be configured at start-up using the configuration
-file \fB/usr/local/etc/smartd.conf\fP (Windows: \fB./smartd.conf\fP).
+file \fB/usr/local/etc/smartd.conf\fP (Windows: \fBEXEDIR/smartd.conf\fP).
If the configuration file is subsequently modified, \fBsmartd\fP
can be told to re-read the configuration file by sending it a
\fBHUP\fP signal, for example with the command:
.SH
OPTIONS
-Long options are not supported on all systems. Use \fB\'smartd
-\-h\'\fP to see the available options.
.TP
.B \-A PREFIX, \-\-attributelog=PREFIX
If the PREFIX has the form \'/path/name\' (e.g. \'/var/lib/misc/attrlog\-\'),
then files 'nameMODEL\-SERIAL.ata.csv' are created in directory '/path/'.
The path must be absolute, except if debug mode is enabled.
-
.TP
.B \-B [+]FILE, \-\-drivedb=[+]FILE
[NEW EXPERIMENTAL SMARTD FEATURE] Read the drive database from FILE.
The new database replaces the built in database by default. If \'+\' is
specified, then the new entries prepend the built in entries.
Please see the \fBsmartctl\fP(8) man page for further details.
-
.TP
.B \-c FILE, \-\-configfile=FILE
Read \fBsmartd\fP configuration Directives from FILE, instead of from
-the default location \fB/usr/local/etc/smartd.conf\fP (Windows: \fB./smartd.conf\fP).
+the default location \fB/usr/local/etc/smartd.conf\fP (Windows: \fBEXEDIR/smartd.conf\fP).
If FILE does \fBnot\fP exist, then \fBsmartd\fP will print an error
message and exit with nonzero status. Thus, \'\-c /usr/local/etc/smartd.conf\'
can be used to verify the existence of the default configuration file.
.B echo /dev/hdb \-m user@home \-M test | smartd \-c \- \-q onecheck
.fi
to perform quick and simple checks without a configuration file.
-
.\" BEGIN ENABLE_CAPABILITIES
.TP
.B \-C, \-\-capabilities
Use \fBcapabilities(7)\fP (EXPERIMENTAL).
Warning: Mail notification does not work when used.
-
.\" END ENABLE_CAPABILITIES
.TP
.B \-d, \-\-debug
.TP
.B \-h, \-\-help, \-\-usage
Prints usage message to STDOUT and exits.
-
.TP
.B \-i N, \-\-interval=N
Sets the interval between disk checks to \fIN\fP seconds, where
for the same purpose.
.fi
(Windows: See NOTES below.)
-
.TP
.B \-l FACILITY, \-\-logfacility=FACILITY
Uses syslog facility FACILITY to log the messages from \fBsmartd\fP.
\fIdaemon\fP.
If you would like to have \fBsmartd\fP messages logged somewhere other
-than the default \fB/var/log/messages\fP location, this can typically
-be accomplished with (for example) the following steps:
+than the default location, this can typically be accomplished with
+(for example) the following steps:
.RS 7
.IP \fB[1]\fP 4
Modify the script that starts \fBsmartd\fP to include the \fBsmartd\fP
messages from the event viewer. Use \'\fBsyslogevt -r smartd\fP\'
to register, \'\fBsyslogevt -u smartd\fP\' to unregister and
\'\fBsyslogevt\fP\' for more help.
-
.TP
.B \-n, \-\-no\-fork
Do not fork into background; this is useful when executed from modern
see NOTES below.
On Windows, this option is not available, use \'\-\-service\' instead.
-
.TP
.B \-p NAME, \-\-pidfile=NAME
Writes pidfile \fINAME\fP containing the \fBsmartd\fP Process ID
comma then the integer with no spaces. For example, \fIataioctl,2\fP
The default level is 1, so \'\-r ataioctl,1\' and \'\-r ataioctl\' are
equivalent.
-
.TP
.B \-s PREFIX, \-\-savestates=PREFIX
[NEW EXPERIMENTAL SMARTD FEATURE] [ATA ONLY]
the configuration file (SIGHUP), before smartd shutdown, and after a check
forced by SIGUSR1. After a normal check cycle, a file is only rewritten if
an important change (which usually results in a SYSLOG output) occurred.
-
.TP
.B \-\-service
Cygwin and Windows only: Enables \fBsmartd\fP to run as a Windows service.
The option must be specified in the service command line as the first
argument. It should not be used from console.
See NOTES below for details.
-
.TP
.B \-V, \-\-version, \-\-license, \-\-copyright
Prints version, copyright, license, home page and SVN revision
.fi
Runs the daemon in forked mode. This is the normal way to run
\fBsmartd\fP.
-Entries are logged to SYSLOG (by default
-.B /var/log/messages.)
+Entries are logged to SYSLOG.
.B
smartd -d -i 30
and stop it by using the command:
.nf
.B /usr/local/etc/rc.d/init.d/smartd stop
-
-.fi
-If you want \fBsmartd\fP to start running whenever your machine is
-booted, this can be enabled by using the command:
-.nf
-.B /sbin/chkconfig --add smartd
-.fi
-and disabled using the command:
-.nf
-.B /sbin/chkconfig --del smartd
.fi
.\" DO NOT MODIFY THIS OR THE FOLLOWING TWO LINES. THIS MATERIAL
line. An example file is included with the
.B smartmontools
distribution. You will find this sample configuration file in
-\fB/usr/local/share/doc/smartmontools-5.1/\fP. For security, the configuration file
+\fB/usr/local/share/doc/smartmontools/\fP. For security, the configuration file
should not be writable by anyone but root. The syntax of the file is as
follows:
.IP \(bu 4
these disks. Areca firmware version 1.46 or later which supports
smartmontools must be used; Please see the \fBsmartctl\fP(8) man page
for further details.
-
.TP
.B \-d TYPE
Specifies the type of the device. This Directive may be used multiple
behavior) if the device does not appear to be present when
\fBsmartd\fP is started. This Directive may be used in conjunction
with the other \'\-d\' Directives.
-
.TP
.B \-n POWERMODE[,N][,q]
This \'nocheck\' Directive is used to prevent a disk from being
This prevents a laptop disk from spinning up due to this message.
Both \',N\' and \',q\' can be specified together.
-
.TP
.B \-T TYPE
Specifies how tolerant
Check the SMART health status of the disk. If any Prefailure
Attributes are less than or equal to their threshold values, then disk
failure is predicted in less than 24 hours, and a message at loglevel
-.B \'LOG_CRITICAL\'
+.B \'LOG_CRIT\'
will be logged to syslog. [Please see the
.B smartctl \-H
command-line option.]
.TP
.B \-l TYPE
-Reports increases in the number of errors in one of the two SMART logs. The
+Reports increases in the number of errors in one of three SMART logs. The
valid arguments to this Directive are:
.I error
-\- report if the number of ATA errors reported in the ATA Error Log
+\- report if the number of ATA errors reported in the Summary SMART error log
has increased since the last check.
+.I xerror
+\- [NEW EXPERIMENTAL SMARTD FEATURE] report if the number of ATA errors
+reported in the Extended Comprehensive SMART error log has increased since
+the last check.
+
+If both \'\-l error\' and \'\-l xerror\' are specified, smartd checks
+the maximum of both values.
+
+[Please see the \fBsmartctl \-l xerror\fP command-line option.]
+
.I selftest
\- report if the number of failed tests reported in the SMART
Self-Test Log has increased since the last check, or if the timestamp
issue harmless informational warning messages if it detects characters
in \fBREGEXP\fP that appear to indicate that you have made this
mistake.
-
.TP
.B \-m ADD
Send a warning email to the email address \fBADD\fP if the \'\-H\',
With \'\fBsysmsgbox\fP\', a system modal (always on top) message box
is used. If running as a service, a service notification message box
(always shown on current visible desktop) is used.
-
.TP
.B \-M TYPE
These Directives modify the behavior of the
will \fBblock\fP until the executable PATH returns, so if your
executable hangs, then \fBsmartd\fP will also hang. Some sample
scripts are included in
-/usr/local/share/doc/smartmontools-5.1/examplescripts/.
+/usr/local/share/doc/smartmontools/examplescripts/.
The return status of the executable is recorded by \fBsmartd\fP in
SYSLOG. The executable is not expected to write to STDOUT or
one or more disk sectors could not be read.
.nf
.fi
+\fITemperature\fP: Temperature reached critical limit (see \-W directive).
+.nf
+.fi
\fIFailedHealthCheck\fP: the SMART health status command failed.
.nf
.fi
Some EXAMPLES of scripts that can be used with the \'\-M exec\'
Directive are given below. Some sample scripts are also included in
-/usr/local/share/doc/smartmontools-5.1/examplescripts/.
-
+/usr/local/share/doc/smartmontools/examplescripts/.
.TP
.B \-f
Check for \'failure\' of any Usage Attributes. If these Attributes are
If the optional flag \'!\' is appended, a change of the Normalized
value is considered critical. The report will be logged as LOG_CRIT
and a warning email will be sent if \'-m\' is specified.
-
.TP
.B \-R ID[!]
When tracking, report whenever the \fIRaw\fP value of Attribute
value is considered critical. The report will be logged as
LOG_CRIT and a warning email will be sent if \'-m\' is specified.
An example is \'-R 5!\' to warn when new sectors are reallocated.
-
.TP
.B \-C ID[+]
[ATA only] Report if the current number of pending sectors is
\fBID\fP is 0 to 255 inclusive. To turn off this reporting, use
ID\ =\ 0. If the \fB\-C ID\fP option is not given, then it defaults to
\fB\-C 197\fP (since Attribute 197 is generally used to monitor
-pending sectors).
+pending sectors). If the name of this Attribute is changed by a
+\'\-v 197,FORMAT,NAME\' directive, the default is changed to
+\fB\-C 0\fP.
If \'+\' is specified, a report is only printed if the number of sectors
has increased between two check cycles. Some disks do not reset this
force the sector to reallocate by writing to it (translation: make the
device substitute a spare good sector for the bad one) but at the
price of losing the 512 bytes of data stored there.
-
.TP
.B \-U ID[+]
[ATA only] Report if the number of offline uncorrectable sectors is
\fBID\fP is 0 to 255 inclusive. To turn off this reporting, use
ID\ =\ 0. If the \fB\-U ID\fP option is not given, then it defaults to
\fB\-U 198\fP (since Attribute 198 is generally used to monitor
-offline uncorrectable sectors).
+offline uncorrectable sectors). If the name of this Attribute is changed
+by a \'\-v 198,FORMAT,NAME\' (except \'\-v 198,FORMAT,Offline_Scan_UNC_SectCt\'),
+directive, the default is changed to \fB\-U 0\fP.
If \'+\' is specified, a report is only printed if the number of sectors
has increased since the last check cycle. Some disks do not reset this
to know, because if you have data stored in this disk sector, and you
need to read it, the read will fail. Please see the previous \'\-C\'
option for more details.
-
.TP
.B \-W DIFF[,INFO[,CRIT]]
Report if the current temperature had changed by at least \fBDIFF\fP
Report or Warn if the temperature is greater or equal than one of
\fBINFO\fP or \fBCRIT\fP degrees Celsius.
If the limit \fBCRIT\fP is reached, a message with loglevel
-\fB\'LOG_CRITICAL\'\fP will be logged to syslog and a warning email
+\fB\'LOG_CRIT\'\fP will be logged to syslog and a warning email
will be send if '-m' is specified. If only the limit \fBINFO\fP is
reached, a message with loglevel \fB\'LOG_INFO\'\fP will be logged.
For ATA devices, smartd interprets Attribute 194 as Temperature Celsius
by default. This can be changed to Attribute 9 or 220 by the drive
database or by the \'-v\' directive, see below.
-
.TP
.B \-F TYPE
[ATA only] Modifies the behavior of \fBsmartd\fP to compensate for
used. The valid values are:
.I none
-\- Assume that the device firmware obeys the ATA specifications. This is
-the default, unless the device has presets for \'\-F\' in the device
-database.
+\- Assume that the device firmware obeys the ATA specifications. This
+is the default, unless the device has presets for \'\-F\' in the
+device database.
.I samsung
\- In some Samsung disks (example: model SV4012H Firmware Version:
-RM100-08) some of the two- and four-byte quantities in the SMART data
-structures are byte-swapped (relative to the ATA specification).
+RM100\-08) some of the two\- and four\-byte quantities in the SMART data
+structures are byte\-swapped (relative to the ATA specification).
Enabling this option tells \fBsmartd\fP to evaluate these quantities
-in byte-reversed order. Some signs that your disk needs this option
-are (1) no self-test log printed, even though you have run self-tests;
+in byte\-reversed order. Some signs that your disk needs this option
+are (1) no self\-test log printed, even though you have run self\-tests;
(2) very large numbers of ATA errors reported in the ATA error log;
(3) strange and impossible values for the ATA error log timestamps.
.I samsung2
-\- In more recent Samsung disks (firmware revisions ending in "\-23") the
-number of ATA errors reported is byte swapped. Enabling this option
-tells \fBsmartd\fP to evaluate this quantity in byte-reversed order.
+\- In some Samsung disks the number of ATA errors reported is byte swapped.
+Enabling this option tells \fBsmartd\fP to evaluate this quantity in
+byte\-reversed order.
.I samsung3
\- Some Samsung disks (at least SP2514N with Firmware VF100\-37) report
completed. If this directive is specified, \fBsmartd\fP will not skip the
next scheduled self\-test (see Directive \'\-s\' above) in this case.
-Note that an explicit \'\-F\' Directive will over-ride any preset
+Note that an explicit \'\-F\' Directive will over\-ride any preset
values for \'\-F\' (see the \'\-P\' option below).
[Please see the \fBsmartctl \-F\fP command-line option.]
-
.TP
.B \-v ID,FORMAT[:BYTEORDER][,NAME]
[ATA only] Sets a vendor\-specific raw value print FORMAT, an optional
\- Raw Attribute number 198 (Offline Uncorrectable Sector Count) is not
reset if uncorrectable sector are reallocated. This sets \'-U 198+\'
if no other \'-U\' directive is specified.
-
.TP
.B \-P TYPE
Specifies whether
[Please see the
.B smartctl \-P
command-line option.]
-
.TP
.B \-a
Equivalent to turning on all of the following Directives:
Note that \-a is the default for ATA devices. If none of these other
Directives is given, then \-a is assumed.
-
.TP
.B #
Comment: ignore the remainder of the line.
.fi
Some example scripts are distributed with the smartmontools package,
-in /usr/local/share/doc/smartmontools-5.1/examplescripts/.
+in /usr/local/share/doc/smartmontools/examplescripts/.
Please note that these scripts typically run as root, so any files
that they read/write should not be writable by ordinary users or
.PP
.SH AUTHOR
-\fBBruce Allen\fP smartmontools-support@lists.sourceforge.net
+\fBBruce Allen\fP smartmontools\-support@lists.sourceforge.net
.fi
University of Wisconsin \- Milwaukee Physics Department
.SH CREDITS
.fi
This code was derived from the smartsuite package, written by Michael
-Cornwell, and from the previous ucsc smartsuite package. It extends
-these to cover ATA-5 disks. This code was originally developed as a
+Cornwell, and from the previous UCSC smartsuite package. It extends
+these to cover ATA\-5 disks. This code was originally developed as a
Senior Thesis by Michael Cornwell at the Concurrent Systems Laboratory
(now part of the Storage Systems Research Center), Jack Baskin School
of Engineering, University of California, Santa
Please see the following web site for updates, further documentation, bug
reports and patches: \fBhttp://smartmontools.sourceforge.net/\fP
-.SH SEE ALSO:
+.SH
+SEE ALSO:
\fBsmartd.conf\fP(5), \fBsmartctl\fP(8), \fBsyslogd\fP(8),
\fBsyslog.conf\fP(5), \fBbadblocks\fP(8), \fBide\-smart\fP(8), \fBregex\fP(7).
.fi
An introductory article about smartmontools is \fIMonitoring Hard
Disks with SMART\fP, by Bruce Allen, Linux Journal, January 2004,
-pages 74-77. This is \fBhttp://www.linuxjournal.com/article.php?sid=6983\fP
+pages 74\-77. This is \fBhttp://www.linuxjournal.com/article/6983\fP
online.
If you would like to understand better how SMART works, and what it
does, a good place to start is with Sections 4.8 and 6.54 of the first
-volume of the \'AT Attachment with Packet Interface-7\' (ATA/ATAPI-7)
-specification. This documents the SMART functionality which the
-\fBsmartmontools\fP utilities provide access to. You can find
-Revision 4b of this document at
-\fBhttp://www.t13.org/docs2004/d1532v1r4b-ATA-ATAPI-7.pdf\fP .
-Earlier and later versions of this Specification are available from
+volume of the \'AT Attachment with Packet Interface\-7\' (ATA/ATAPI\-7)
+specification Revision 4b. This documents the SMART functionality which the
+\fBsmartmontools\fP utilities provide access to.
+This and other versions of this Specification are available from
the T13 web site \fBhttp://www.t13.org/\fP .
.fi
-The functioning of SMART was originally defined by the SFF-8035i
-revision 2 and the SFF-8055i revision 1.4 specifications. These are
-publications of the Small Form Factors (SFF) Committee. Links to
-these documents may be found in the References section of the
-smartmontools home page at \fBhttp://smartmontools.sourceforge.net/#references\fP .
+The functioning of SMART was originally defined by the SFF\-8035i
+revision 2 and the SFF\-8055i revision 1.4 specifications. These are
+publications of the Small Form Factors (SFF) Committee.
+
+Links to these and other documents may be found on the Links page of the
+\fBsmartmontools\fP Wiki at
+\fBhttp://sourceforge.net/apps/trac/smartmontools/wiki/Links\fP .
.SH
SVN ID OF THIS PAGE:
-$Id: smartd.8.in 3076 2010-03-12 22:23:08Z chrfranke $
+$Id: smartd.8.in 3117 2010-06-08 15:41:04Z chrfranke $
.ig
-Copyright (C) 2002-8 Bruce Allen <smartmontools-support@lists.sourceforge.net>
+Copyright (C) 2002-10 Bruce Allen <smartmontools-support@lists.sourceforge.net>
-$Id: smartd.conf.5.in 3075 2010-03-12 22:01:44Z chrfranke $
+$Id: smartd.conf.5.in 3117 2010-06-08 15:41:04Z chrfranke $
This program is free software; you can redistribute it and/or modify it
under the terms of the GNU General Public License as published by the Free
Research Center), Jack Baskin School of Engineering, University of
California, Santa Cruz. http://ssrc.soe.ucsc.edu/
..
-.TH SMARTD.CONF 5 CURRENT_CVS_DATE CURRENT_CVS_VERSION CURRENT_CVS_DATE
+.TH SMARTD.CONF 5 CURRENT_SVN_DATE CURRENT_SVN_VERSION CURRENT_SVN_DATE
.SH NAME
\fBsmartd.conf\fP \- SMART Disk Monitoring Daemon Configuration File\fP
.B /usr/local/etc/smartd.conf
.SH PACKAGE VERSION
-CURRENT_CVS_VERSION released CURRENT_CVS_DATE at CURRENT_CVS_TIME
+CURRENT_SVN_VERSION CURRENT_SVN_DATE CURRENT_SVN_REV
.SH DESCRIPTION
\fB/usr/local/etc/smartd.conf\fP is the configuration file for the \fBsmartd\fP
line. An example file is included with the
.B smartmontools
distribution. You will find this sample configuration file in
-\fB/usr/local/share/doc/smartmontools-5.1/\fP. For security, the configuration file
+\fB/usr/local/share/doc/smartmontools/\fP. For security, the configuration file
should not be writable by anyone but root. The syntax of the file is as
follows:
.IP \(bu 4
these disks. Areca firmware version 1.46 or later which supports
smartmontools must be used; Please see the \fBsmartctl\fP(8) man page
for further details.
-
.TP
.B \-d TYPE
Specifies the type of the device. This Directive may be used multiple
behavior) if the device does not appear to be present when
\fBsmartd\fP is started. This Directive may be used in conjunction
with the other \'\-d\' Directives.
-
.TP
.B \-n POWERMODE[,N][,q]
This \'nocheck\' Directive is used to prevent a disk from being
This prevents a laptop disk from spinning up due to this message.
Both \',N\' and \',q\' can be specified together.
-
.TP
.B \-T TYPE
Specifies how tolerant
Check the SMART health status of the disk. If any Prefailure
Attributes are less than or equal to their threshold values, then disk
failure is predicted in less than 24 hours, and a message at loglevel
-.B \'LOG_CRITICAL\'
+.B \'LOG_CRIT\'
will be logged to syslog. [Please see the
.B smartctl \-H
command-line option.]
.TP
.B \-l TYPE
-Reports increases in the number of errors in one of the two SMART logs. The
+Reports increases in the number of errors in one of three SMART logs. The
valid arguments to this Directive are:
.I error
-\- report if the number of ATA errors reported in the ATA Error Log
+\- report if the number of ATA errors reported in the Summary SMART error log
has increased since the last check.
+.I xerror
+\- [NEW EXPERIMENTAL SMARTD FEATURE] report if the number of ATA errors
+reported in the Extended Comprehensive SMART error log has increased since
+the last check.
+
+If both \'\-l error\' and \'\-l xerror\' are specified, smartd checks
+the maximum of both values.
+
+[Please see the \fBsmartctl \-l xerror\fP command-line option.]
+
.I selftest
\- report if the number of failed tests reported in the SMART
Self-Test Log has increased since the last check, or if the timestamp
issue harmless informational warning messages if it detects characters
in \fBREGEXP\fP that appear to indicate that you have made this
mistake.
-
.TP
.B \-m ADD
Send a warning email to the email address \fBADD\fP if the \'\-H\',
With \'\fBsysmsgbox\fP\', a system modal (always on top) message box
is used. If running as a service, a service notification message box
(always shown on current visible desktop) is used.
-
.TP
.B \-M TYPE
These Directives modify the behavior of the
will \fBblock\fP until the executable PATH returns, so if your
executable hangs, then \fBsmartd\fP will also hang. Some sample
scripts are included in
-/usr/local/share/doc/smartmontools-5.1/examplescripts/.
+/usr/local/share/doc/smartmontools/examplescripts/.
The return status of the executable is recorded by \fBsmartd\fP in
SYSLOG. The executable is not expected to write to STDOUT or
one or more disk sectors could not be read.
.nf
.fi
+\fITemperature\fP: Temperature reached critical limit (see \-W directive).
+.nf
+.fi
\fIFailedHealthCheck\fP: the SMART health status command failed.
.nf
.fi
Some EXAMPLES of scripts that can be used with the \'\-M exec\'
Directive are given below. Some sample scripts are also included in
-/usr/local/share/doc/smartmontools-5.1/examplescripts/.
-
+/usr/local/share/doc/smartmontools/examplescripts/.
.TP
.B \-f
Check for \'failure\' of any Usage Attributes. If these Attributes are
If the optional flag \'!\' is appended, a change of the Normalized
value is considered critical. The report will be logged as LOG_CRIT
and a warning email will be sent if \'-m\' is specified.
-
.TP
.B \-R ID[!]
When tracking, report whenever the \fIRaw\fP value of Attribute
value is considered critical. The report will be logged as
LOG_CRIT and a warning email will be sent if \'-m\' is specified.
An example is \'-R 5!\' to warn when new sectors are reallocated.
-
.TP
.B \-C ID[+]
[ATA only] Report if the current number of pending sectors is
\fBID\fP is 0 to 255 inclusive. To turn off this reporting, use
ID\ =\ 0. If the \fB\-C ID\fP option is not given, then it defaults to
\fB\-C 197\fP (since Attribute 197 is generally used to monitor
-pending sectors).
+pending sectors). If the name of this Attribute is changed by a
+\'\-v 197,FORMAT,NAME\' directive, the default is changed to
+\fB\-C 0\fP.
If \'+\' is specified, a report is only printed if the number of sectors
has increased between two check cycles. Some disks do not reset this
force the sector to reallocate by writing to it (translation: make the
device substitute a spare good sector for the bad one) but at the
price of losing the 512 bytes of data stored there.
-
.TP
.B \-U ID[+]
[ATA only] Report if the number of offline uncorrectable sectors is
\fBID\fP is 0 to 255 inclusive. To turn off this reporting, use
ID\ =\ 0. If the \fB\-U ID\fP option is not given, then it defaults to
\fB\-U 198\fP (since Attribute 198 is generally used to monitor
-offline uncorrectable sectors).
+offline uncorrectable sectors). If the name of this Attribute is changed
+by a \'\-v 198,FORMAT,NAME\' (except \'\-v 198,FORMAT,Offline_Scan_UNC_SectCt\'),
+directive, the default is changed to \fB\-U 0\fP.
If \'+\' is specified, a report is only printed if the number of sectors
has increased since the last check cycle. Some disks do not reset this
to know, because if you have data stored in this disk sector, and you
need to read it, the read will fail. Please see the previous \'\-C\'
option for more details.
-
.TP
.B \-W DIFF[,INFO[,CRIT]]
Report if the current temperature had changed by at least \fBDIFF\fP
Report or Warn if the temperature is greater or equal than one of
\fBINFO\fP or \fBCRIT\fP degrees Celsius.
If the limit \fBCRIT\fP is reached, a message with loglevel
-\fB\'LOG_CRITICAL\'\fP will be logged to syslog and a warning email
+\fB\'LOG_CRIT\'\fP will be logged to syslog and a warning email
will be send if '-m' is specified. If only the limit \fBINFO\fP is
reached, a message with loglevel \fB\'LOG_INFO\'\fP will be logged.
For ATA devices, smartd interprets Attribute 194 as Temperature Celsius
by default. This can be changed to Attribute 9 or 220 by the drive
database or by the \'-v\' directive, see below.
-
.TP
.B \-F TYPE
[ATA only] Modifies the behavior of \fBsmartd\fP to compensate for
used. The valid values are:
.I none
-\- Assume that the device firmware obeys the ATA specifications. This is
-the default, unless the device has presets for \'\-F\' in the device
-database.
+\- Assume that the device firmware obeys the ATA specifications. This
+is the default, unless the device has presets for \'\-F\' in the
+device database.
.I samsung
\- In some Samsung disks (example: model SV4012H Firmware Version:
-RM100-08) some of the two- and four-byte quantities in the SMART data
-structures are byte-swapped (relative to the ATA specification).
+RM100\-08) some of the two\- and four\-byte quantities in the SMART data
+structures are byte\-swapped (relative to the ATA specification).
Enabling this option tells \fBsmartd\fP to evaluate these quantities
-in byte-reversed order. Some signs that your disk needs this option
-are (1) no self-test log printed, even though you have run self-tests;
+in byte\-reversed order. Some signs that your disk needs this option
+are (1) no self\-test log printed, even though you have run self\-tests;
(2) very large numbers of ATA errors reported in the ATA error log;
(3) strange and impossible values for the ATA error log timestamps.
.I samsung2
-\- In more recent Samsung disks (firmware revisions ending in "\-23") the
-number of ATA errors reported is byte swapped. Enabling this option
-tells \fBsmartd\fP to evaluate this quantity in byte-reversed order.
+\- In some Samsung disks the number of ATA errors reported is byte swapped.
+Enabling this option tells \fBsmartd\fP to evaluate this quantity in
+byte\-reversed order.
.I samsung3
\- Some Samsung disks (at least SP2514N with Firmware VF100\-37) report
completed. If this directive is specified, \fBsmartd\fP will not skip the
next scheduled self\-test (see Directive \'\-s\' above) in this case.
-Note that an explicit \'\-F\' Directive will over-ride any preset
+Note that an explicit \'\-F\' Directive will over\-ride any preset
values for \'\-F\' (see the \'\-P\' option below).
[Please see the \fBsmartctl \-F\fP command-line option.]
-
.TP
.B \-v ID,FORMAT[:BYTEORDER][,NAME]
[ATA only] Sets a vendor\-specific raw value print FORMAT, an optional
\- Raw Attribute number 198 (Offline Uncorrectable Sector Count) is not
reset if uncorrectable sector are reallocated. This sets \'-U 198+\'
if no other \'-U\' directive is specified.
-
.TP
.B \-P TYPE
Specifies whether
[Please see the
.B smartctl \-P
command-line option.]
-
.TP
.B \-a
Equivalent to turning on all of the following Directives:
Note that \-a is the default for ATA devices. If none of these other
Directives is given, then \-a is assumed.
-
.TP
.B #
Comment: ignore the remainder of the line.
.fi
Some example scripts are distributed with the smartmontools package,
-in /usr/local/share/doc/smartmontools-5.1/examplescripts/.
+in /usr/local/share/doc/smartmontools/examplescripts/.
Please note that these scripts typically run as root, so any files
that they read/write should not be writable by ordinary users or
.PP
.SH AUTHOR
-\fBBruce Allen\fP smartmontools-support@lists.sourceforge.net
+\fBBruce Allen\fP smartmontools\-support@lists.sourceforge.net
.fi
University of Wisconsin \- Milwaukee Physics Department
.SH CREDITS
.fi
This code was derived from the smartsuite package, written by Michael
-Cornwell, and from the previous ucsc smartsuite package. It extends
-these to cover ATA-5 disks. This code was originally developed as a
+Cornwell, and from the previous UCSC smartsuite package. It extends
+these to cover ATA\-5 disks. This code was originally developed as a
Senior Thesis by Michael Cornwell at the Concurrent Systems Laboratory
(now part of the Storage Systems Research Center), Jack Baskin School
of Engineering, University of California, Santa
HOME PAGE FOR SMARTMONTOOLS:
.fi
Please see the following web site for updates, further documentation, bug
-reports and patches:
-.nf
-.B
-http://smartmontools.sourceforge.net/
+reports and patches: \fBhttp://smartmontools.sourceforge.net/\fP
.SH
SEE ALSO:
.SH
SVN ID OF THIS PAGE:
-$Id: smartd.conf.5.in 3075 2010-03-12 22:01:44Z chrfranke $
+$Id: smartd.conf.5.in 3117 2010-06-08 15:41:04Z chrfranke $
#define ARGUSED(x) ((void)(x))
-const char * smartd_cpp_cvsid = "$Id: smartd.cpp 3075 2010-03-12 22:01:44Z chrfranke $"
+const char * smartd_cpp_cvsid = "$Id: smartd.cpp 3101 2010-05-04 16:03:18Z chrfranke $"
CONFIG_H_CVSID EXTERN_H_CVSID;
extern const char *reportbug;
;
// configuration file name
-#define CONFIGFILENAME "smartd.conf"
-
-#ifndef _WIN32
-static const char *configfile = SMARTMONTOOLS_SYSCONFDIR "/" CONFIGFILENAME ;
-#else
-static const char *configfile = "./" CONFIGFILENAME ;
-#endif
+static const char * configfile;
// configuration file "name" if read from stdin
static const char * const configfile_stdin = "<stdin>";
// path of alternate configuration file
bool usage; // Track changes in Usage Attributes
bool selftest; // Monitor number of selftest errors
bool errorlog; // Monitor number of ATA errors
+ bool xerrorlog; // Monitor number of ATA errors (Extended Comprehensive error log)
bool permissive; // Ignore failed SMART commands
char autosave; // 1=disable, 2=enable Autosave Attributes
char autoofflinetest; // 1=disable, 2=enable Auto Offline Test
usage(false),
selftest(false),
errorlog(false),
+ xerrorlog(false),
permissive(false),
autosave(0),
autoofflinetest(0),
setmode(fileno(f), O_TEXT); // Allow files with \r\n
#endif
+ persistent_dev_state new_state;
int good = 0, bad = 0;
char line[256];
while (fgets(line, sizeof(line), f)) {
const char * s = line + strspn(line, " \t");
if (!*s || *s == '#')
continue;
- if (!parse_dev_state_line(line, state))
+ if (!parse_dev_state_line(line, new_state))
bad++;
else
good++;
}
pout("%s: %d invalid line(s) ignored\n", path, bad);
}
+
+ // This sets the values missing in the file to 0.
+ state = new_state;
return true;
}
" -n MODE No check if: never, sleep[,N][,q], standby[,N][,q], idle[,N][,q]\n"
" -H Monitor SMART Health Status, report if failed\n"
" -s REG Do Self-Test at time(s) given by regular expression REG\n"
- " -l TYPE Monitor SMART log. Type is one of: error, selftest\n"
+ " -l TYPE Monitor SMART log. Type is one of: error, selftest, xerror\n"
" -f Monitor 'Usage' Attributes, report failures\n"
" -m ADD Send email warning to address ADD\n"
" -M TYPE Modify email warning behavior (see man page)\n"
PrintOut(LOG_INFO,"\n");
PrintOut(LOG_INFO," -B [+]FILE, --drivedb=[+]FILE\n");
PrintOut(LOG_INFO," Read and replace [add] drive database from FILE\n");
+ PrintOut(LOG_INFO," [default is +%s", get_drivedb_path_add());
#ifdef SMARTMONTOOLS_DRIVEDBDIR
- PrintOut(LOG_INFO," [default is "SMARTMONTOOLS_DRIVEDBDIR"/drivedb.h]\n");
-#endif
PrintOut(LOG_INFO,"\n");
+ PrintOut(LOG_INFO," and then %s", get_drivedb_path_default());
+#endif
+ PrintOut(LOG_INFO,"]\n\n");
PrintOut(LOG_INFO," -c NAME|-, --configfile=NAME|-\n");
- PrintOut(LOG_INFO," Read configuration file NAME or stdin [default is %s]\n\n", configfile);
+ PrintOut(LOG_INFO," Read configuration file NAME or stdin\n");
+ PrintOut(LOG_INFO," [default is %s]\n\n", configfile);
#ifdef HAVE_LIBCAP_NG
PrintOut(LOG_INFO," -C, --capabilities\n");
PrintOut(LOG_INFO," Use capabilities (EXPERIMENTAL).\n"
|| ('a' <= c && c <= 'z'));
}
-// returns <0 on failure
-static int ATAErrorCount(ata_device * device, const char * name,
- unsigned char fix_firmwarebug)
+// Read error count from Summary or Extended Comprehensive SMART error log
+// Return -1 on error
+static int read_ata_error_count(ata_device * device, const char * name,
+ unsigned char fix_firmwarebug, bool extended)
{
- struct ata_smart_errorlog log;
-
- if (ataReadErrorLog(device, &log, fix_firmwarebug)){
- PrintOut(LOG_INFO,"Device: %s, Read SMART Error Log Failed\n",name);
- return -1;
+ if (!extended) {
+ ata_smart_errorlog log;
+ if (ataReadErrorLog(device, &log, fix_firmwarebug)){
+ PrintOut(LOG_INFO,"Device: %s, Read Summary SMART Error Log failed\n",name);
+ return -1;
+ }
+ return (log.error_log_pointer ? log.ata_error_count : 0);
+ }
+ else {
+ ata_smart_exterrlog logx;
+ if (!ataReadExtErrorLog(device, &logx, 1 /*first sector only*/)) {
+ PrintOut(LOG_INFO,"Device: %s, Read Extended Comprehensive SMART Error Log failed\n",name);
+ return -1;
+ }
+ // Some disks use the reserved byte as index, see ataprint.cpp.
+ return (logx.error_log_index || logx.reserved1 ? logx.device_error_count : 0);
}
-
- // return current number of ATA errors
- return log.error_log_pointer?log.ata_error_count:0;
}
// returns <0 if problem. Otherwise, bottom 8 bits are the self test
// do we need to get SMART data?
bool smart_val_ok = false;
- if ( cfg.autoofflinetest || cfg.errorlog || cfg.selftest
+ if ( cfg.autoofflinetest || cfg.selftest
+ || cfg.errorlog || cfg.xerrorlog
|| cfg.usagefailed || cfg.prefail || cfg.usage
|| cfg.tempdiff || cfg.tempinfo || cfg.tempcrit
|| cfg.curr_pending_id || cfg.offl_pending_id ) {
}
// capability check: ATA error log
- if (cfg.errorlog) {
- int val;
+ if (cfg.errorlog || cfg.xerrorlog) {
- // start with service disabled, and re-enable it if all works OK
- cfg.errorlog = false;
state.ataerrorcount=0;
-
- if (!smart_val_ok)
- PrintOut(LOG_INFO, "Device: %s, no SMART Error log (SMART READ DATA failed); disabling -l error\n", name);
- else if (!cfg.permissive && !isSmartErrorLogCapable(&state.smartval, &drive))
- PrintOut(LOG_INFO, "Device: %s, appears to lack SMART Error log; disabling -l error (override with -T permissive Directive)\n", name);
- else if ((val = ATAErrorCount(atadev, name, cfg.fix_firmwarebug)) < 0)
- PrintOut(LOG_INFO, "Device: %s, no SMART Error log; remove -l error Directive from smartd.conf\n", name);
+ if (!(cfg.permissive || (smart_val_ok && isSmartErrorLogCapable(&state.smartval, &drive)))) {
+ PrintOut(LOG_INFO, "Device: %s, no SMART Error Log (%s), ignoring -l [x]error (override with -T permissive)\n",
+ name, (!smart_val_ok ? "SMART READ DATA failed" : "capability missing"));
+ cfg.errorlog = cfg.xerrorlog = false;
+ }
else {
- cfg.errorlog = true;
- state.ataerrorcount=val;
+ int errcnt1 = -1, errcnt2 = -1;
+ if (cfg.errorlog && (errcnt1 = read_ata_error_count(atadev, name, cfg.fix_firmwarebug, false)) < 0) {
+ PrintOut(LOG_INFO, "Device: %s, no Summary SMART Error Log, ignoring -l error\n", name);
+ cfg.errorlog = false;
+ }
+ if (cfg.xerrorlog && (errcnt2 = read_ata_error_count(atadev, name, cfg.fix_firmwarebug, true)) < 0) {
+ PrintOut(LOG_INFO, "Device: %s, no Extended Comprehensive SMART Error Log, ignoring -l xerror\n", name);
+ cfg.xerrorlog = false;
+ }
+ if (cfg.errorlog || cfg.xerrorlog) {
+ if (cfg.errorlog && cfg.xerrorlog && errcnt1 != errcnt2) {
+ PrintOut(LOG_INFO, "Device: %s, SMART Error Logs report different error counts: %d != %d\n",
+ name, errcnt1, errcnt2);
+ }
+ // Record max error count
+ state.ataerrorcount = (errcnt1 >= errcnt2 ? errcnt1 : errcnt2);
+ }
}
}
}
// If no tests available or selected, return
- if (!(cfg.errorlog || cfg.selftest || cfg.smartcheck ||
- cfg.usagefailed || cfg.prefail || cfg.usage ||
- cfg.tempdiff || cfg.tempinfo || cfg.tempcrit )) {
+ if (!( cfg.smartcheck || cfg.selftest
+ || cfg.errorlog || cfg.xerrorlog
+ || cfg.usagefailed || cfg.prefail || cfg.usage
+ || cfg.tempdiff || cfg.tempinfo || cfg.tempcrit)) {
CloseDevice(atadev, name);
return 3;
}
MailWarning(cfg, state, 3, "Device: %s, Self-Test Log error count increased from %d to %d",
name, oldc, newc);
state.must_write = true;
- } else if (oldh!=newh) {
+ }
+ else if (newc > 0 && oldh != newh) {
// more recent error
// a 'more recent' error might actually be a smaller hour number,
// if the hour number has wrapped.
name, newh);
state.must_write = true;
}
-
+
+ // Print info if error entries have disappeared
+ if (oldc > newc)
+ PrintOut(LOG_INFO, "Device: %s, Self-Test Log error count decreased from %d to %d\n",
+ name, oldc, newc);
+
// Needed since self-test error count may DECREASE. Hour might
// also have changed.
state.selflogcount= newc;
CheckSelfTestLogs(cfg, state, SelfTestErrorCount(atadev, name, cfg.fix_firmwarebug));
// check if number of ATA errors has increased
- if (cfg.errorlog) {
+ if (cfg.errorlog || cfg.xerrorlog) {
- int newc, oldc= state.ataerrorcount;
+ int errcnt1 = -1, errcnt2 = -1;
+ if (cfg.errorlog)
+ errcnt1 = read_ata_error_count(atadev, name, cfg.fix_firmwarebug, false);
+ if (cfg.xerrorlog)
+ errcnt2 = read_ata_error_count(atadev, name, cfg.fix_firmwarebug, true);
- // new number of errors
- newc = ATAErrorCount(atadev, name, cfg.fix_firmwarebug);
+ // new number of errors is max of both logs
+ int newc = (errcnt1 >= errcnt2 ? errcnt1 : errcnt2);
// did command fail?
if (newc<0)
MailWarning(cfg, state, 7, "Device: %s, Read SMART Error Log Failed", name);
// has error count increased?
+ int oldc = state.ataerrorcount;
if (newc>oldc){
PrintOut(LOG_CRIT, "Device: %s, ATA error count increased from %d to %d\n",
name, oldc, newc);
name, oldc, newc);
state.must_write = true;
}
-
- // this last line is probably not needed, count always increases
+
if (newc>=0)
state.ataerrorcount=newc;
}
} else if (!strcmp(arg, "error")) {
// track changes in ATA error log
cfg.errorlog = true;
+ } else if (!strcmp(arg, "xerror")) {
+ // track changes in Extended Comprehensive SMART error log
+ cfg.xerrorlog = true;
} else {
badarg = 1;
}
}
// If NO monitoring directives are set, then set all of them.
- if (!(cfg.smartcheck || cfg.usagefailed || cfg.prefail ||
- cfg.usage || cfg.selftest || cfg.errorlog ||
- cfg.tempdiff || cfg.tempinfo || cfg.tempcrit )) {
+ if (!( cfg.smartcheck || cfg.selftest
+ || cfg.errorlog || cfg.xerrorlog
+ || cfg.usagefailed || cfg.prefail || cfg.usage
+ || cfg.tempdiff || cfg.tempinfo || cfg.tempcrit)) {
PrintOut(LOG_INFO,"Drive: %s, implied '-a' Directive on line %d of file %s\n",
cfg.name.c_str(), cfg.lineno, configfile);
// Parses input line, prints usage message and
// version/license/copyright messages
-void ParseOpts(int argc, char **argv){
- int optchar;
- char *tailptr;
- long lchecktime;
+void ParseOpts(int argc, char **argv)
+{
+ // Init default configfile path
+#ifndef _WIN32
+ configfile = SMARTMONTOOLS_SYSCONFDIR"/smartd.conf";
+#else
+ static std::string configfile_str = get_exe_dir() + "/smartd.conf";
+ configfile = configfile_str.c_str();
+#endif
+
// Please update GetValidArgList() if you edit shortopts
static const char shortopts[] = "c:l:q:dDni:p:r:s:A:B:Vh?"
#ifdef HAVE_LIBCAP_NG
"C"
#endif
;
- char *arg;
// Please update GetValidArgList() if you edit longopts
struct option longopts[] = {
{ "configfile", required_argument, 0, 'c' },
bool badarg = false;
bool no_defaultdb = false; // set true on '-B FILE'
- // Parse input options. This horrible construction is so that emacs
- // indents properly. Sorry.
- while (-1 != (optchar =
- getopt_long(argc, argv, shortopts, longopts, NULL)
- )) {
-
+ // Parse input options.
+ int optchar;
+ while ((optchar = getopt_long(argc, argv, shortopts, longopts, NULL)) != -1) {
+ char *arg;
+ char *tailptr;
+ long lchecktime;
+
switch(optchar) {
case 'q':
// when to quit
#include "atacmds.h"
#include "dev_interface.h"
-const char * utility_cpp_cvsid = "$Id: utility.cpp 3022 2010-01-01 17:02:00Z chrfranke $"
+const char * utility_cpp_cvsid = "$Id: utility.cpp 3090 2010-04-28 11:03:11Z chrfranke $"
UTILITY_H_CVSID INT64_H_CVSID;
const char * packet_types[] = {
return "Unknown";
}
+// Runtime check of byte ordering, throws if different from isbigendian().
+void check_endianness()
+{
+ union {
+ // Force compile error if int type is not 32bit.
+ unsigned char c[sizeof(unsigned) == 4 ? 4 : -1];
+ unsigned i;
+ } x = {{1,2,3,4}};
+
+ int big = -1;
+ switch (x.i) {
+ case 0x01020304: big = 1; break;
+ case 0x04030201: big = 0; break;
+ }
-// Returns 1 if machine is big endian, else zero. This is a run-time
-// rather than a compile-time function. We could do it at
-// compile-time but in principle there are architectures that can run
-// with either byte-ordering.
-int isbigendian(){
- short i=0x0100;
- char *tmp=(char *)&i;
- return *tmp;
+ if (big != (isbigendian() ? 1 : 0))
+ throw std::logic_error("CPU endianness does not match compile time test");
}
// Utility function prints date and time and timezone into a character
*
* Home page of code is: http://smartmontools.sourceforge.net
*
- * Copyright (C) 2002-9 Bruce Allen <smartmontools-support@lists.sourceforge.net>
- * Copyright (C) 2008-9 Christian Franke <smartmontools-support@lists.sourceforge.net>
+ * Copyright (C) 2002-10 Bruce Allen <smartmontools-support@lists.sourceforge.net>
+ * Copyright (C) 2008-10 Christian Franke <smartmontools-support@lists.sourceforge.net>
* Copyright (C) 2000 Michael Cornwell <cornwell@acm.org>
*
* This program is free software; you can redistribute it and/or modify
#ifndef UTILITY_H_
#define UTILITY_H_
-#define UTILITY_H_CVSID "$Id: utility.h 3020 2009-12-31 01:11:51Z dlukes $"
+#define UTILITY_H_CVSID "$Id: utility.h 3093 2010-04-30 09:57:36Z chrfranke $"
#include <time.h>
#include <sys/types.h> // for regex.h (according to POSIX)
// appropriate.]
void PrintOut(int priority, const char *fmt, ...) __attribute__ ((format(printf, 2, 3)));
-// run time, determine byte ordering
-int isbigendian();
+// Compile time check of byte ordering
+// (inline const function allows compiler to remove dead code)
+inline bool isbigendian()
+{
+#ifdef WORDS_BIGENDIAN
+ return true;
+#else
+ return false;
+#endif
+}
+
+// Runtime check of byte ordering, throws if different from isbigendian().
+void check_endianness();
// This value follows the peripheral device type value as defined in
// SCSI Primary Commands, ANSI INCITS 301:1997. It is also used in
#define PRINT_ON(control) {if (control->printing_switchable) control->dont_print=false;}
#define PRINT_OFF(control) {if (control->printing_switchable) control->dont_print=true;}
+#ifdef _WIN32
+// Get exe directory
+//(implemented in os_win32.cpp)
+std::string get_exe_dir();
+#endif
+
+
#ifdef OLD_INTERFACE
// possible values for controller_type in extern.h
#define CONTROLLER_UNKNOWN 0x00
projects / mirror_smartmontools-debian.git / commitdiff