/* Extended regular expression matching and search library.
- Copyright (C) 2002, 2003 Free Software Foundation, Inc.
+ Copyright (C) 2002-2018 Free Software Foundation, Inc.
This file is part of the GNU C Library.
Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
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., 51 Franklin Street, Fifth Floor, Boston,
- MA 02110-1301 USA. */
+ License along with the GNU C Library; if not, see
+ <https://www.gnu.org/licenses/>. */
#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>
+#ifndef _REGEX_STANDALONE
+#include <langinfo.h>
+#endif
+#include <locale.h>
+#include <wchar.h>
+#include <wctype.h>
+#include <stdbool.h>
+#include <stdint.h>
+
+/* Properties of integers. Although Gnulib has intprops.h, glibc does
+ without for now. */
+#if !defined(_LIBC) && !defined(_REGEX_STANDALONE)
+# include "intprops.h"
+#else
+/* True if the real type T is signed. */
+# define TYPE_SIGNED(t) (! ((t) 0 < (t) -1))
+
+/* True if adding the nonnegative Idx values A and B would overflow.
+ If false, set *R to A + B. A, B, and R may be evaluated more than
+ once, or zero times. Although this is not a full implementation of
+ Gnulib INT_ADD_WRAPV, it is good enough for glibc regex code.
+ FIXME: This implementation is a fragile stopgap, and this file would
+ be simpler and more robust if intprops.h were migrated into glibc. */
+# define INT_ADD_WRAPV(a, b, r) \
+ (IDX_MAX - (a) < (b) ? true : (*(r) = (a) + (b), false))
+#endif
+
+#ifdef _LIBC
+# include <libc-lock.h>
+# define lock_define(name) __libc_lock_define (, name)
+# define lock_init(lock) (__libc_lock_init (lock), 0)
+# define lock_fini(lock) ((void) 0)
+# define lock_lock(lock) __libc_lock_lock (lock)
+# define lock_unlock(lock) __libc_lock_unlock (lock)
+#elif defined GNULIB_LOCK && !defined USE_UNLOCKED_IO
+# include "glthread/lock.h"
+ /* Use gl_lock_define if empty macro arguments are known to work.
+ Otherwise, fall back on less-portable substitutes. */
+# if ((defined __GNUC__ && !defined __STRICT_ANSI__) \
+ || (defined __STDC_VERSION__ && 199901L <= __STDC_VERSION__))
+# define lock_define(name) gl_lock_define (, name)
+# elif USE_POSIX_THREADS
+# define lock_define(name) pthread_mutex_t name;
+# elif USE_PTH_THREADS
+# define lock_define(name) pth_mutex_t name;
+# elif USE_SOLARIS_THREADS
+# define lock_define(name) mutex_t name;
+# elif USE_WINDOWS_THREADS
+# define lock_define(name) gl_lock_t name;
+# else
+# define lock_define(name)
+# endif
+# define lock_init(lock) glthread_lock_init (&(lock))
+# define lock_fini(lock) glthread_lock_destroy (&(lock))
+# define lock_lock(lock) glthread_lock_lock (&(lock))
+# define lock_unlock(lock) glthread_lock_unlock (&(lock))
+#elif defined GNULIB_PTHREAD && !defined USE_UNLOCKED_IO
+# include <pthread.h>
+# define lock_define(name) pthread_mutex_t name;
+# define lock_init(lock) pthread_mutex_init (&(lock), 0)
+# define lock_fini(lock) pthread_mutex_destroy (&(lock))
+# define lock_lock(lock) pthread_mutex_lock (&(lock))
+# define lock_unlock(lock) pthread_mutex_unlock (&(lock))
+#else
+# define lock_define(name)
+# define lock_init(lock) 0
+# define lock_fini(lock) ((void) 0)
+ /* The 'dfa' avoids an "unused variable 'dfa'" warning from GCC. */
+# define lock_lock(lock) ((void) dfa)
+# define lock_unlock(lock) ((void) 0)
#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
+#if !defined _LIBC && ! (defined isblank || (HAVE_ISBLANK && HAVE_DECL_ISBLANK))
# define isblank(ch) ((ch) == ' ' || (ch) == '\t')
#endif
# 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
+#if (HAVE_LIBINTL_H && ENABLE_NLS) || defined _LIBC
# include <libintl.h>
# ifdef _LIBC
# undef gettext
# define gettext(msgid) \
- INTUSE(__dcgettext) (INTUSE(_libc_intl_domainname), msgid, LC_MESSAGES)
+ __dcgettext (_libc_intl_domainname, msgid, LC_MESSAGES)
# endif
#else
+# undef gettext
# define gettext(msgid) (msgid)
#endif
# 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
+#if (defined MB_CUR_MAX && HAVE_WCTYPE_H && HAVE_ISWCTYPE) || _LIBC
# define RE_ENABLE_I18N
#endif
-#if __GNUC__ >= 3
-# define BE(expr, val) __builtin_expect (expr, val)
+#ifdef __GNUC__
+#define BE(expr, val) __builtin_expect (expr, val)
#else
-# define BE(expr, val) (expr)
-# define inline
+#define BE(expr, val) (expr)
#endif
-/* Number of bits in a byte. */
-#define BYTE_BITS 8
-/* Number of single byte character. */
-#define SBC_MAX 256
+/* Number of ASCII characters. */
+#define ASCII_CHARS 0x80
+
+/* Number of single byte characters. */
+#define SBC_MAX (UCHAR_MAX + 1)
#define COLL_ELEM_LEN_MAX 8
/* Rename to standard API for using out of glibc. */
#ifndef _LIBC
+# undef __wctype
+# undef __iswctype
# define __wctype wctype
+# define __iswalnum iswalnum
# define __iswctype iswctype
+# define __towlower towlower
+# define __towupper towupper
# define __btowc btowc
-# define __mempcpy mempcpy
+# define __mbrtowc mbrtowc
# define __wcrtomb wcrtomb
+# define __regfree regfree
# 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);
+#if __GNUC__ < 3 + (__GNUC_MINOR__ < 1)
+# define __attribute__(arg)
+#endif
+
+#ifndef SSIZE_MAX
+# define SSIZE_MAX ((ssize_t) (SIZE_MAX / 2))
+#endif
+
+/* The type of indexes into strings. This is signed, not size_t,
+ since the API requires indexes to fit in regoff_t anyway, and using
+ signed integers makes the code a bit smaller and presumably faster.
+ The traditional GNU regex implementation uses int for indexes.
+ The POSIX-compatible implementation uses a possibly-wider type.
+ The name 'Idx' is three letters to minimize the hassle of
+ reindenting a lot of regex code that formerly used 'int'. */
+typedef regoff_t Idx;
+#ifdef _REGEX_LARGE_OFFSETS
+# define IDX_MAX SSIZE_MAX
+#else
+# define IDX_MAX INT_MAX
+#endif
+
+/* A hash value, suitable for computing hash tables. */
+typedef __re_size_t re_hashval_t;
+
+/* An integer used to represent a set of bits. It must be unsigned,
+ and must be at least as wide as unsigned int. */
+typedef unsigned long int bitset_word_t;
+/* All bits set in a bitset_word_t. */
+#define BITSET_WORD_MAX ULONG_MAX
+
+/* Number of bits in a bitset_word_t. For portability to hosts with
+ padding bits, do not use '(sizeof (bitset_word_t) * CHAR_BIT)';
+ instead, deduce it directly from BITSET_WORD_MAX. Avoid
+ greater-than-32-bit integers and unconditional shifts by more than
+ 31 bits, as they're not portable. */
+#if BITSET_WORD_MAX == 0xffffffffUL
+# define BITSET_WORD_BITS 32
+#elif BITSET_WORD_MAX >> 31 >> 4 == 1
+# define BITSET_WORD_BITS 36
+#elif BITSET_WORD_MAX >> 31 >> 16 == 1
+# define BITSET_WORD_BITS 48
+#elif BITSET_WORD_MAX >> 31 >> 28 == 1
+# define BITSET_WORD_BITS 60
+#elif BITSET_WORD_MAX >> 31 >> 31 >> 1 == 1
+# define BITSET_WORD_BITS 64
+#elif BITSET_WORD_MAX >> 31 >> 31 >> 9 == 1
+# define BITSET_WORD_BITS 72
+#elif BITSET_WORD_MAX >> 31 >> 31 >> 31 >> 31 >> 3 == 1
+# define BITSET_WORD_BITS 128
+#elif BITSET_WORD_MAX >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 7 == 1
+# define BITSET_WORD_BITS 256
+#elif BITSET_WORD_MAX >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 7 > 1
+# define BITSET_WORD_BITS 257 /* any value > SBC_MAX will do here */
+# if BITSET_WORD_BITS <= SBC_MAX
+# error "Invalid SBC_MAX"
+# endif
+#else
+# error "Add case for new bitset_word_t size"
+#endif
+
+/* Number of bitset_word_t values in a bitset_t. */
+#define BITSET_WORDS ((SBC_MAX + BITSET_WORD_BITS - 1) / BITSET_WORD_BITS)
+
+typedef bitset_word_t bitset_t[BITSET_WORDS];
+typedef bitset_word_t *re_bitset_ptr_t;
+typedef const bitset_word_t *re_const_bitset_ptr_t;
#define PREV_WORD_CONSTRAINT 0x0001
#define PREV_NOTWORD_CONSTRAINT 0x0002
#define NEXT_NEWLINE_CONSTRAINT 0x0020
#define PREV_BEGBUF_CONSTRAINT 0x0040
#define NEXT_ENDBUF_CONSTRAINT 0x0080
-#define DUMMY_CONSTRAINT 0x0100
+#define WORD_DELIM_CONSTRAINT 0x0100
+#define NOT_WORD_DELIM_CONSTRAINT 0x0200
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,
+ INSIDE_NOTWORD = PREV_NOTWORD_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
+ WORD_DELIM = WORD_DELIM_CONSTRAINT,
+ NOT_WORD_DELIM = NOT_WORD_DELIM_CONSTRAINT
} re_context_type;
typedef struct
{
- int alloc;
- int nelem;
- int *elems;
+ Idx alloc;
+ Idx nelem;
+ Idx *elems;
} re_node_set;
typedef enum
{
NON_TYPE = 0,
+ /* Node type, These are used by token, node, tree. */
+ CHARACTER = 1,
+ END_OF_RE = 2,
+ SIMPLE_BRACKET = 3,
+ OP_BACK_REF = 4,
+ OP_PERIOD = 5,
+#ifdef RE_ENABLE_I18N
+ COMPLEX_BRACKET = 6,
+ OP_UTF8_PERIOD = 7,
+#endif /* RE_ENABLE_I18N */
+
+ /* We define EPSILON_BIT as a macro so that OP_OPEN_SUBEXP is used
+ when the debugger shows values of this enum type. */
+#define EPSILON_BIT 8
+ OP_OPEN_SUBEXP = EPSILON_BIT | 0,
+ OP_CLOSE_SUBEXP = EPSILON_BIT | 1,
+ OP_ALT = EPSILON_BIT | 2,
+ OP_DUP_ASTERISK = EPSILON_BIT | 3,
+ ANCHOR = EPSILON_BIT | 4,
+
+ /* Tree type, these are used only by tree. */
+ CONCAT = 16,
+ SUBEXP = 17,
+
/* Token type, these are used only by token. */
+ OP_DUP_PLUS = 18,
+ OP_DUP_QUESTION,
OP_OPEN_BRACKET,
OP_CLOSE_BRACKET,
OP_CHARSET_RANGE,
OP_CLOSE_CHAR_CLASS,
OP_WORD,
OP_NOTWORD,
- BACK_SLASH,
+ OP_SPACE,
+ OP_NOTSPACE,
+ 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
unsigned int non_match : 1;
/* # of multibyte characters. */
- int nmbchars;
+ Idx nmbchars;
/* # of collating symbols. */
- int ncoll_syms;
+ Idx ncoll_syms;
/* # of equivalence classes. */
- int nequiv_classes;
+ Idx nequiv_classes;
/* # of range expressions. */
- int nranges;
+ Idx nranges;
/* # of character classes. */
- int nchar_classes;
+ Idx nchar_classes;
} re_charset_t;
#endif /* RE_ENABLE_I18N */
#ifdef RE_ENABLE_I18N
re_charset_t *mbcset; /* for COMPLEX_BRACKET */
#endif /* RE_ENABLE_I18N */
- int idx; /* for BACK_REF */
+ Idx idx; /* for BACK_REF */
re_context_type ctx_type; /* for ANCHOR */
} opr;
-#if __GNUC__ >= 2
+#if __GNUC__ >= 2 && !defined __STRICT_ANSI__
re_token_type_t type : 8;
#else
re_token_type_t type;
#endif
unsigned int constraint : 10; /* context constraint */
unsigned int duplicated : 1;
+ unsigned int opt_subexp : 1;
#ifdef RE_ENABLE_I18N
+ unsigned int accept_mb : 1;
+ /* These 2 bits can be moved into the union if needed (e.g. if running out
+ of bits; move opr.c to opr.c.c and move the flags to opr.c.flags). */
unsigned int mb_partial : 1;
#endif
+ unsigned int word_char : 1;
} 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)
+#define IS_EPSILON_NODE(type) ((type) & EPSILON_BIT)
struct re_string_t
{
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;
+ Idx *offsets;
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;
+ Idx 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;
+ Idx valid_len;
+ /* The corresponding number of bytes in raw_mbs array. */
+ Idx valid_raw_len;
+ /* The length of the buffers MBS and WCS. */
+ Idx 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;
+ Idx cur_idx;
+ /* length of RAW_MBS array. */
+ Idx raw_len;
+ /* This is RAW_LEN - RAW_MBS_IDX + VALID_LEN - VALID_RAW_LEN. */
+ Idx 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;
+ Idx raw_stop;
+ /* This is RAW_STOP - RAW_MBS_IDX adjusted through OFFSETS. */
+ Idx 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
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;
+ /* Copy of re_dfa_t's word_char. */
+ re_const_bitset_ptr_t word_char;
+ /* true if REG_ICASE. */
+ unsigned char icase;
+ unsigned char is_utf8;
+ unsigned char map_notascii;
+ unsigned char mbs_allocated;
+ unsigned char offsets_needed;
+ unsigned char newline_anchor;
+ unsigned char word_ops_used;
+ int mb_cur_max;
};
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);
+
+
+struct re_dfa_t;
+typedef struct re_dfa_t re_dfa_t;
+
+#ifndef _LIBC
+# define IS_IN(libc) false
+#endif
+
#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)
+ ((idx) == (pstr)->valid_len || (pstr)->wcs[idx] != WEOF)
#define re_string_is_single_byte_char(pstr, idx) \
- ((pstr)->wcs[idx] != WEOF && ((pstr)->len == (idx) \
+ ((pstr)->wcs[idx] != WEOF && ((pstr)->valid_len == (idx) + 1 \
|| (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_skip_bytes(pstr,idx) ((pstr)->cur_idx += (idx))
#define re_string_set_index(pstr,idx) ((pstr)->cur_idx = (idx))
+#if defined _LIBC || HAVE_ALLOCA
+# include <alloca.h>
+#endif
+
+#ifndef _LIBC
+# if HAVE_ALLOCA
+/* The OS usually guarantees only one guard page at the bottom of the stack,
+ and a page size can be as small as 4096 bytes. So we cannot safely
+ allocate anything larger than 4096 bytes. Also care for the possibility
+ of a few compiler-allocated temporary stack slots. */
+# define __libc_use_alloca(n) ((n) < 4032)
+# else
+/* alloca is implemented with malloc, so just use malloc. */
+# define __libc_use_alloca(n) 0
+# undef alloca
+# define alloca(n) malloc (n)
+# endif
+#endif
+
+#ifdef _LIBC
+# define MALLOC_0_IS_NONNULL 1
+#elif !defined MALLOC_0_IS_NONNULL
+# define MALLOC_0_IS_NONNULL 0
+#endif
+
+#ifndef MAX
+# define MAX(a,b) ((a) < (b) ? (b) : (a))
+#endif
+#ifndef MIN
+# define MIN(a,b) ((a) < (b) ? (a) : (b))
+#endif
+
#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 *parent;
struct bin_tree_t *left;
struct bin_tree_t *right;
+ struct bin_tree_t *first;
+ struct bin_tree_t *next;
- /* `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;
+ re_token_t token;
- int first;
- int next;
- re_node_set eclosure;
+ /* 'node_idx' is the index in dfa->nodes, if 'type' == 0.
+ Otherwise 'type' indicate the type of this node. */
+ Idx node_idx;
};
typedef struct bin_tree_t bin_tree_t;
+#define BIN_TREE_STORAGE_SIZE \
+ ((1024 - sizeof (void *)) / sizeof (bin_tree_t))
+
+struct bin_tree_storage_t
+{
+ struct bin_tree_storage_t *next;
+ bin_tree_t data[BIN_TREE_STORAGE_SIZE];
+};
+typedef struct bin_tree_storage_t bin_tree_storage_t;
#define CONTEXT_WORD 1
#define CONTEXT_NEWLINE (CONTEXT_WORD << 1)
#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_WORD_CHAR(ch) (__iswalnum (ch) || (ch) == L'_')
#define IS_WIDE_NEWLINE(ch) ((ch) == WIDE_NEWLINE_CHAR)
#define NOT_SATISFY_PREV_CONSTRAINT(constraint,context) \
struct re_dfastate_t
{
- unsigned int hash;
+ re_hashval_t hash;
re_node_set nodes;
+ re_node_set non_eps_nodes;
+ re_node_set inveclosure;
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;
+ struct re_dfastate_t **trtable, **word_trtable;
+ unsigned int context : 4;
unsigned int halt : 1;
- /* If this state can accept `multi byte'.
+ /* If this state can accept "multi byte".
Note that we refer to multibyte characters, and multi character
- collating elements as `multi byte'. */
+ collating elements as "multi byte". */
unsigned int accept_mb : 1;
/* If this state has backreference node(s). */
unsigned int has_backref : 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;
+ Idx num;
+ Idx alloc;
re_dfastate_t **array;
};
typedef struct
{
- int next_idx;
- int alloc;
+ Idx next_idx;
+ Idx alloc;
re_dfastate_t **array;
} state_array_t;
typedef struct
{
- int node;
- int str_idx; /* The position NODE match at. */
+ Idx node;
+ Idx str_idx; /* The position NODE match at. */
state_array_t path;
} re_sub_match_last_t;
typedef struct
{
- int str_idx;
- int node;
- int next_last_offset;
+ Idx str_idx;
+ Idx node;
state_array_t *path;
- int alasts; /* Allocation size of LASTS. */
- int nlasts; /* The number of LASTS. */
+ Idx alasts; /* Allocation size of LASTS. */
+ Idx 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;
+ Idx node;
+ Idx str_idx;
+ Idx subexp_from;
+ Idx subexp_to;
+ char more;
+ char unused;
+ unsigned short int eps_reachable_subexps_map;
};
typedef struct
{
+ /* The string object corresponding to the input string. */
+ re_string_t input;
+#if defined _LIBC || (defined __STDC_VERSION__ && __STDC_VERSION__ >= 199901L)
+ const re_dfa_t *const dfa;
+#else
+ const re_dfa_t *dfa;
+#endif
/* 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;
+ Idx match_last;
+ Idx last_node;
/* The state log used by the matcher. */
re_dfastate_t **state_log;
- int state_log_top;
+ Idx state_log_top;
/* Back reference cache. */
- int nbkref_ents;
- int abkref_ents;
+ Idx nbkref_ents;
+ Idx abkref_ents;
struct re_backref_cache_entry *bkref_ents;
int max_mb_elem_len;
- int nsub_tops;
- int asub_tops;
+ Idx nsub_tops;
+ Idx 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;
-
+ Idx last_node;
+ Idx last_str_idx;
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;
+ Idx idx;
+ Idx node;
regmatch_t *regs;
re_node_set eps_via_nodes;
};
struct re_fail_stack_t
{
- int num;
- int alloc;
+ Idx num;
+ Idx 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;
+ size_t nodes_alloc;
+ size_t nodes_len;
+ Idx *nexts;
+ Idx *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. */
+ bin_tree_t *str_tree;
+ bin_tree_storage_t *str_tree_storage;
+ re_bitset_ptr_t sb_char;
+ int str_tree_storage_idx;
+
+ /* number of subexpressions 're_nsub' is in regex_t. */
+ re_hashval_t state_hash_mask;
+ Idx init_node;
+ Idx 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
+ bitset_word_t used_bkref_map;
+ bitset_word_t completed_bkref_map;
+
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;
+ unsigned int is_utf8 : 1;
+ unsigned int map_notascii : 1;
+ unsigned int word_ops_used : 1;
+ int mb_cur_max;
+ bitset_t word_char;
+ reg_syntax_t syntax;
+ Idx *subexp_map;
+#ifdef DEBUG
+ char* re_str;
+#endif
+ lock_define (lock)
};
-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
} bracket_elem_t;
-/* Inline functions for bitset operation. */
+/* Functions for bitset_t operation. */
+
+static inline void
+bitset_set (bitset_t set, Idx i)
+{
+ set[i / BITSET_WORD_BITS] |= (bitset_word_t) 1 << i % BITSET_WORD_BITS;
+}
+
static inline void
-bitset_not (set)
- bitset set;
+bitset_clear (bitset_t set, Idx i)
+{
+ set[i / BITSET_WORD_BITS] &= ~ ((bitset_word_t) 1 << i % BITSET_WORD_BITS);
+}
+
+static inline bool
+bitset_contain (const bitset_t set, Idx i)
+{
+ return (set[i / BITSET_WORD_BITS] >> i % BITSET_WORD_BITS) & 1;
+}
+
+static inline void
+bitset_empty (bitset_t set)
+{
+ memset (set, '\0', sizeof (bitset_t));
+}
+
+static inline void
+bitset_set_all (bitset_t set)
+{
+ memset (set, -1, sizeof (bitset_word_t) * (SBC_MAX / BITSET_WORD_BITS));
+ if (SBC_MAX % BITSET_WORD_BITS != 0)
+ set[BITSET_WORDS - 1] =
+ ((bitset_word_t) 1 << SBC_MAX % BITSET_WORD_BITS) - 1;
+}
+
+static inline void
+bitset_copy (bitset_t dest, const bitset_t src)
+{
+ memcpy (dest, src, sizeof (bitset_t));
+}
+
+static inline void
+bitset_not (bitset_t set)
{
int bitset_i;
- for (bitset_i = 0; bitset_i < BITSET_UINTS; ++bitset_i)
+ for (bitset_i = 0; bitset_i < SBC_MAX / BITSET_WORD_BITS; ++bitset_i)
set[bitset_i] = ~set[bitset_i];
+ if (SBC_MAX % BITSET_WORD_BITS != 0)
+ set[BITSET_WORDS - 1] =
+ ((((bitset_word_t) 1 << SBC_MAX % BITSET_WORD_BITS) - 1)
+ & ~set[BITSET_WORDS - 1]);
}
static inline void
-bitset_merge (dest, src)
- bitset dest;
- const bitset src;
+bitset_merge (bitset_t dest, const bitset_t src)
{
int bitset_i;
- for (bitset_i = 0; bitset_i < BITSET_UINTS; ++bitset_i)
+ for (bitset_i = 0; bitset_i < BITSET_WORDS; ++bitset_i)
dest[bitset_i] |= src[bitset_i];
}
static inline void
-bitset_not_merge (dest, src)
- bitset dest;
- const bitset src;
+bitset_mask (bitset_t dest, const bitset_t src)
{
- int i;
- for (i = 0; i < BITSET_UINTS; ++i)
- dest[i] |= ~src[i];
+ int bitset_i;
+ for (bitset_i = 0; bitset_i < BITSET_WORDS; ++bitset_i)
+ dest[bitset_i] &= src[bitset_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;
+/* Functions for re_string. */
+static int
+__attribute__ ((pure, unused))
+re_string_char_size_at (const re_string_t *pstr, Idx idx)
{
int byte_idx;
- if (MB_CUR_MAX == 1)
+ if (pstr->mb_cur_max == 1)
return 1;
- for (byte_idx = 1; idx + byte_idx < pstr->len; ++byte_idx)
+ for (byte_idx = 1; idx + byte_idx < pstr->valid_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;
+static wint_t
+__attribute__ ((pure, unused))
+re_string_wchar_at (const re_string_t *pstr, Idx idx)
{
- if (MB_CUR_MAX == 1)
+ if (pstr->mb_cur_max == 1)
return (wint_t) pstr->mbs[idx];
return (wint_t) pstr->wcs[idx];
}
+# ifdef _LIBC
+# include <locale/weight.h>
+# endif
+
static int
-re_string_elem_size_at (pstr, idx)
- const re_string_t *pstr;
- int idx;
+__attribute__ ((pure, unused))
+re_string_elem_size_at (const re_string_t *pstr, Idx idx)
{
-#ifdef _LIBC
+# 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)
indirect = (const int32_t *) _NL_CURRENT (LC_COLLATE,
_NL_COLLATE_INDIRECTMB);
p = pstr->mbs + idx;
- tmp = findidx (&p);
+ findidx (table, indirect, extra, &p, pstr->len - idx);
return p - pstr->mbs - idx;
}
else
-#endif /* _LIBC */
+# endif /* _LIBC */
return 1;
}
#endif /* RE_ENABLE_I18N */
+#ifndef __GNUC_PREREQ
+# if defined __GNUC__ && defined __GNUC_MINOR__
+# define __GNUC_PREREQ(maj, min) \
+ ((__GNUC__ << 16) + __GNUC_MINOR__ >= ((maj) << 16) + (min))
+# else
+# define __GNUC_PREREQ(maj, min) 0
+# endif
+#endif
+
+#if __GNUC_PREREQ (3,4)
+# undef __attribute_warn_unused_result__
+# define __attribute_warn_unused_result__ \
+ __attribute__ ((__warn_unused_result__))
+#else
+# define __attribute_warn_unused_result__ /* empty */
+#endif
+
+#ifndef FALLTHROUGH
+# if __GNUC__ < 7
+# define FALLTHROUGH ((void) 0)
+# else
+# define FALLTHROUGH __attribute__ ((__fallthrough__))
+# endif
+#endif
+
#endif /* _REGEX_INTERNAL_H */
projects / mirror_smartmontools-debian.git / blobdiff