--- /dev/null
+/**********************************************************************\r
+ regposix.c - Oniguruma (regular expression library)\r
+**********************************************************************/\r
+/*-\r
+ * Copyright (c) 2002-2008 K.Kosako <sndgk393 AT ybb DOT ne DOT jp>\r
+ * All rights reserved.\r
+ *\r
+ * Copyright (c) 2015, Hewlett Packard Enterprise Development, L.P.<BR>\r
+ *\r
+ * Redistribution and use in source and binary forms, with or without\r
+ * modification, are permitted provided that the following conditions\r
+ * are met:\r
+ * 1. Redistributions of source code must retain the above copyright\r
+ * notice, this list of conditions and the following disclaimer.\r
+ * 2. Redistributions in binary form must reproduce the above copyright\r
+ * notice, this list of conditions and the following disclaimer in the\r
+ * documentation and/or other materials provided with the distribution.\r
+ *\r
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND\r
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE\r
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE\r
+ * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE\r
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL\r
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS\r
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)\r
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT\r
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY\r
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF\r
+ * SUCH DAMAGE.\r
+ */\r
+\r
+#define regex_t onig_regex_t\r
+#include "regint.h"\r
+#undef regex_t\r
+#include "onigposix.h"\r
+\r
+#define ONIG_C(reg) ((onig_regex_t* )((reg)->onig))\r
+#define PONIG_C(reg) ((onig_regex_t** )(&(reg)->onig))\r
+\r
+/* #define ENC_STRING_LEN(enc,s,len) len = strlen(s) */\r
+#define ENC_STRING_LEN(enc,s,len) do { \\r
+ if (ONIGENC_MBC_MINLEN(enc) == 1) { \\r
+ UChar* tmps = (UChar* )(s); \\r
+ while (*tmps != 0) tmps++; \\r
+ len = (int)(tmps - (UChar* )(s)); \\r
+ } \\r
+ else { \\r
+ len = onigenc_str_bytelen_null(enc, (UChar* )s); \\r
+ } \\r
+} while(0)\r
+\r
+typedef struct {\r
+ int onig_err;\r
+ int posix_err;\r
+} O2PERR;\r
+\r
+static int\r
+onig2posix_error_code(int code)\r
+{\r
+ static const O2PERR o2p[] = {\r
+ { ONIG_MISMATCH, REG_NOMATCH },\r
+ { ONIG_NO_SUPPORT_CONFIG, REG_EONIG_INTERNAL },\r
+ { ONIGERR_MEMORY, REG_ESPACE },\r
+ { ONIGERR_MATCH_STACK_LIMIT_OVER, REG_EONIG_INTERNAL },\r
+ { ONIGERR_TYPE_BUG, REG_EONIG_INTERNAL },\r
+ { ONIGERR_PARSER_BUG, REG_EONIG_INTERNAL },\r
+ { ONIGERR_STACK_BUG, REG_EONIG_INTERNAL },\r
+ { ONIGERR_UNDEFINED_BYTECODE, REG_EONIG_INTERNAL },\r
+ { ONIGERR_UNEXPECTED_BYTECODE, REG_EONIG_INTERNAL },\r
+ { ONIGERR_DEFAULT_ENCODING_IS_NOT_SETTED, REG_EONIG_BADARG },\r
+ { ONIGERR_SPECIFIED_ENCODING_CANT_CONVERT_TO_WIDE_CHAR, REG_EONIG_BADARG },\r
+ { ONIGERR_INVALID_ARGUMENT, REG_EONIG_BADARG },\r
+ { ONIGERR_END_PATTERN_AT_LEFT_BRACE, REG_EBRACE },\r
+ { ONIGERR_END_PATTERN_AT_LEFT_BRACKET, REG_EBRACK },\r
+ { ONIGERR_EMPTY_CHAR_CLASS, REG_ECTYPE },\r
+ { ONIGERR_PREMATURE_END_OF_CHAR_CLASS, REG_ECTYPE },\r
+ { ONIGERR_END_PATTERN_AT_ESCAPE, REG_EESCAPE },\r
+ { ONIGERR_END_PATTERN_AT_META, REG_EESCAPE },\r
+ { ONIGERR_END_PATTERN_AT_CONTROL, REG_EESCAPE },\r
+ { ONIGERR_META_CODE_SYNTAX, REG_BADPAT },\r
+ { ONIGERR_CONTROL_CODE_SYNTAX, REG_BADPAT },\r
+ { ONIGERR_CHAR_CLASS_VALUE_AT_END_OF_RANGE, REG_ECTYPE },\r
+ { ONIGERR_CHAR_CLASS_VALUE_AT_START_OF_RANGE, REG_ECTYPE },\r
+ { ONIGERR_UNMATCHED_RANGE_SPECIFIER_IN_CHAR_CLASS, REG_ECTYPE },\r
+ { ONIGERR_TARGET_OF_REPEAT_OPERATOR_NOT_SPECIFIED, REG_BADRPT },\r
+ { ONIGERR_TARGET_OF_REPEAT_OPERATOR_INVALID, REG_BADRPT },\r
+ { ONIGERR_NESTED_REPEAT_OPERATOR, REG_BADRPT },\r
+ { ONIGERR_UNMATCHED_CLOSE_PARENTHESIS, REG_EPAREN },\r
+ { ONIGERR_END_PATTERN_WITH_UNMATCHED_PARENTHESIS, REG_EPAREN },\r
+ { ONIGERR_END_PATTERN_IN_GROUP, REG_BADPAT },\r
+ { ONIGERR_UNDEFINED_GROUP_OPTION, REG_BADPAT },\r
+ { ONIGERR_INVALID_POSIX_BRACKET_TYPE, REG_BADPAT },\r
+ { ONIGERR_INVALID_LOOK_BEHIND_PATTERN, REG_BADPAT },\r
+ { ONIGERR_INVALID_REPEAT_RANGE_PATTERN, REG_BADPAT },\r
+ { ONIGERR_TOO_BIG_NUMBER, REG_BADPAT },\r
+ { ONIGERR_TOO_BIG_NUMBER_FOR_REPEAT_RANGE, REG_BADBR },\r
+ { ONIGERR_UPPER_SMALLER_THAN_LOWER_IN_REPEAT_RANGE, REG_BADBR },\r
+ { ONIGERR_EMPTY_RANGE_IN_CHAR_CLASS, REG_ECTYPE },\r
+ { ONIGERR_MISMATCH_CODE_LENGTH_IN_CLASS_RANGE, REG_ECTYPE },\r
+ { ONIGERR_TOO_MANY_MULTI_BYTE_RANGES, REG_ECTYPE },\r
+ { ONIGERR_TOO_SHORT_MULTI_BYTE_STRING, REG_BADPAT },\r
+ { ONIGERR_TOO_BIG_BACKREF_NUMBER, REG_ESUBREG },\r
+ { ONIGERR_INVALID_BACKREF, REG_ESUBREG },\r
+ { ONIGERR_NUMBERED_BACKREF_OR_CALL_NOT_ALLOWED, REG_BADPAT },\r
+ { ONIGERR_TOO_BIG_WIDE_CHAR_VALUE, REG_EONIG_BADWC },\r
+ { ONIGERR_TOO_LONG_WIDE_CHAR_VALUE, REG_EONIG_BADWC },\r
+ { ONIGERR_INVALID_CODE_POINT_VALUE, REG_EONIG_BADWC },\r
+ { ONIGERR_EMPTY_GROUP_NAME, REG_BADPAT },\r
+ { ONIGERR_INVALID_GROUP_NAME, REG_BADPAT },\r
+ { ONIGERR_INVALID_CHAR_IN_GROUP_NAME, REG_BADPAT },\r
+ { ONIGERR_UNDEFINED_NAME_REFERENCE, REG_BADPAT },\r
+ { ONIGERR_UNDEFINED_GROUP_REFERENCE, REG_BADPAT },\r
+ { ONIGERR_MULTIPLEX_DEFINED_NAME, REG_BADPAT },\r
+ { ONIGERR_MULTIPLEX_DEFINITION_NAME_CALL, REG_BADPAT },\r
+ { ONIGERR_NEVER_ENDING_RECURSION, REG_BADPAT },\r
+ { ONIGERR_GROUP_NUMBER_OVER_FOR_CAPTURE_HISTORY, REG_BADPAT },\r
+ { ONIGERR_INVALID_CHAR_PROPERTY_NAME, REG_BADPAT },\r
+ { ONIGERR_NOT_SUPPORTED_ENCODING_COMBINATION, REG_EONIG_BADARG },\r
+ { ONIGERR_OVER_THREAD_PASS_LIMIT_COUNT, REG_EONIG_THREAD }\r
+\r
+ };\r
+\r
+ int i;\r
+\r
+ if (code >= 0) return 0;\r
+\r
+ for (i = 0; i < (int )(sizeof(o2p) / sizeof(o2p[0])); i++) {\r
+ if (code == o2p[i].onig_err)\r
+ return o2p[i].posix_err;\r
+ }\r
+\r
+ return REG_EONIG_INTERNAL; /* but, unknown error code */\r
+}\r
+\r
+extern int\r
+regcomp(regex_t* reg, const char* pattern, int posix_options)\r
+{\r
+ int r, len;\r
+ OnigSyntaxType* syntax = OnigDefaultSyntax;\r
+ OnigOptionType options;\r
+\r
+ if ((posix_options & REG_EXTENDED) == 0)\r
+ syntax = ONIG_SYNTAX_POSIX_BASIC;\r
+\r
+ options = syntax->options;\r
+ if ((posix_options & REG_ICASE) != 0)\r
+ ONIG_OPTION_ON(options, ONIG_OPTION_IGNORECASE);\r
+ if ((posix_options & REG_NEWLINE) != 0) {\r
+ ONIG_OPTION_ON( options, ONIG_OPTION_NEGATE_SINGLELINE);\r
+ ONIG_OPTION_OFF(options, ONIG_OPTION_SINGLELINE);\r
+ }\r
+\r
+ reg->comp_options = posix_options;\r
+\r
+ ENC_STRING_LEN(OnigEncDefaultCharEncoding, pattern, len);\r
+ r = onig_new(PONIG_C(reg), (UChar* )pattern, (UChar* )(pattern + len),\r
+ options, OnigEncDefaultCharEncoding, syntax,\r
+ (OnigErrorInfo* )NULL);\r
+ if (r != ONIG_NORMAL) {\r
+ return onig2posix_error_code(r);\r
+ }\r
+\r
+ reg->re_nsub = ONIG_C(reg)->num_mem;\r
+ return 0;\r
+}\r
+\r
+extern int\r
+regexec(regex_t* reg, const char* str, size_t nmatch,\r
+ regmatch_t pmatch[], int posix_options)\r
+{\r
+ int r, i, len;\r
+ UChar* end;\r
+ regmatch_t* pm;\r
+ OnigOptionType options;\r
+\r
+ options = ONIG_OPTION_POSIX_REGION;\r
+ if ((posix_options & REG_NOTBOL) != 0) options |= ONIG_OPTION_NOTBOL;\r
+ if ((posix_options & REG_NOTEOL) != 0) options |= ONIG_OPTION_NOTEOL;\r
+\r
+ if (nmatch == 0 || (reg->comp_options & REG_NOSUB) != 0) {\r
+ pm = (regmatch_t* )NULL;\r
+ nmatch = 0;\r
+ }\r
+ else if ((int )nmatch < ONIG_C(reg)->num_mem + 1) {\r
+ pm = (regmatch_t* )xmalloc(sizeof(regmatch_t)\r
+ * (ONIG_C(reg)->num_mem + 1));\r
+ if (pm == NULL)\r
+ return REG_ESPACE;\r
+ }\r
+ else {\r
+ pm = pmatch;\r
+ }\r
+\r
+ ENC_STRING_LEN(ONIG_C(reg)->enc, str, len);\r
+ end = (UChar* )(str + len);\r
+ r = onig_search(ONIG_C(reg), (UChar* )str, end, (UChar* )str, end,\r
+ (OnigRegion* )pm, options);\r
+\r
+ if (r >= 0) {\r
+ r = 0; /* Match */\r
+ if (pm != pmatch && pm != NULL) {\r
+ xmemcpy(pmatch, pm, sizeof(regmatch_t) * nmatch);\r
+ }\r
+ }\r
+ else if (r == ONIG_MISMATCH) {\r
+ r = REG_NOMATCH;\r
+ for (i = 0; i < (int )nmatch; i++)\r
+ pmatch[i].rm_so = pmatch[i].rm_eo = ONIG_REGION_NOTPOS;\r
+ }\r
+ else {\r
+ r = onig2posix_error_code(r);\r
+ }\r
+\r
+ if (pm != pmatch && pm != NULL)\r
+ xfree(pm);\r
+\r
+#if 0\r
+ if (reg->re_nsub > nmatch - 1)\r
+ reg->re_nsub = (nmatch <= 1 ? 0 : nmatch - 1);\r
+#endif\r
+\r
+ return r;\r
+}\r
+\r
+extern void\r
+regfree(regex_t* reg)\r
+{\r
+ onig_free(ONIG_C(reg));\r
+}\r
+\r
+\r
+extern void\r
+reg_set_encoding(int mb_code)\r
+{\r
+ OnigEncoding enc;\r
+\r
+ switch (mb_code) {\r
+ case REG_POSIX_ENCODING_ASCII:\r
+ enc = ONIG_ENCODING_ASCII;\r
+ break;\r
+ case REG_POSIX_ENCODING_EUC_JP:\r
+ enc = ONIG_ENCODING_EUC_JP;\r
+ break;\r
+ case REG_POSIX_ENCODING_SJIS:\r
+ enc = ONIG_ENCODING_SJIS;\r
+ break;\r
+ case REG_POSIX_ENCODING_UTF8:\r
+ enc = ONIG_ENCODING_UTF8;\r
+ break;\r
+ case REG_POSIX_ENCODING_UTF16_BE:\r
+ enc = ONIG_ENCODING_UTF16_BE;\r
+ break;\r
+ case REG_POSIX_ENCODING_UTF16_LE:\r
+ enc = ONIG_ENCODING_UTF16_LE;\r
+ break;\r
+\r
+ default:\r
+ return ;\r
+ break;\r
+ }\r
+\r
+ onigenc_set_default_encoding(enc);\r
+}\r
+\r
+extern int\r
+reg_name_to_group_numbers(regex_t* reg,\r
+ const unsigned char* name, const unsigned char* name_end, int** nums)\r
+{\r
+ return onig_name_to_group_numbers(ONIG_C(reg), name, name_end, nums);\r
+}\r
+\r
+typedef struct {\r
+ int (*func)(const unsigned char*, const unsigned char*,int,int*,regex_t*,void*);\r
+ regex_t* reg;\r
+ void* arg;\r
+} i_wrap;\r
+\r
+static int\r
+i_wrapper(const UChar* name, const UChar* name_end, int ng, int* gs,\r
+ onig_regex_t* reg ARG_UNUSED, void* arg)\r
+{\r
+ i_wrap* warg = (i_wrap* )arg;\r
+\r
+ return (*warg->func)(name, name_end, ng, gs, warg->reg, warg->arg);\r
+}\r
+\r
+extern int\r
+reg_foreach_name(regex_t* reg,\r
+ int (*func)(const unsigned char*, const unsigned char*,int,int*,regex_t*,void*),\r
+ void* arg)\r
+{\r
+ i_wrap warg;\r
+\r
+ warg.func = func;\r
+ warg.reg = reg;\r
+ warg.arg = arg;\r
+\r
+ return onig_foreach_name(ONIG_C(reg), i_wrapper, &warg);\r
+}\r
+\r
+extern int\r
+reg_number_of_names(regex_t* reg)\r
+{\r
+ return onig_number_of_names(ONIG_C(reg));\r
+}\r