Home | History | Annotate | Download | only in Oniguruma
      1 /**********************************************************************
      2   regposix.c - Oniguruma (regular expression library)
      3 **********************************************************************/
      4 /*-
      5  * Copyright (c) 2002-2008  K.Kosako  <sndgk393 AT ybb DOT ne DOT jp>
      6  * All rights reserved.
      7  *
      8  * (C) Copyright 2015 Hewlett Packard Enterprise Development LP<BR>
      9  *
     10  * Redistribution and use in source and binary forms, with or without
     11  * modification, are permitted provided that the following conditions
     12  * are met:
     13  * 1. Redistributions of source code must retain the above copyright
     14  *    notice, this list of conditions and the following disclaimer.
     15  * 2. Redistributions in binary form must reproduce the above copyright
     16  *    notice, this list of conditions and the following disclaimer in the
     17  *    documentation and/or other materials provided with the distribution.
     18  *
     19  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
     20  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
     21  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
     22  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
     23  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
     24  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
     25  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
     26  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
     27  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
     28  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
     29  * SUCH DAMAGE.
     30  */
     31 
     32 #define regex_t   onig_regex_t
     33 #include "regint.h"
     34 #undef regex_t
     35 #include "onigposix.h"
     36 
     37 #define ONIG_C(reg)    ((onig_regex_t* )((reg)->onig))
     38 #define PONIG_C(reg)   ((onig_regex_t** )(&(reg)->onig))
     39 
     40 /* #define ENC_STRING_LEN(enc,s,len)    len = strlen(s) */
     41 #define ENC_STRING_LEN(enc,s,len) do { \
     42   if (ONIGENC_MBC_MINLEN(enc) == 1) { \
     43     UChar* tmps = (UChar* )(s); \
     44     while (*tmps != 0) tmps++; \
     45     len = (int)(tmps - (UChar* )(s)); \
     46   } \
     47   else { \
     48     len = onigenc_str_bytelen_null(enc, (UChar* )s); \
     49   } \
     50 } while(0)
     51 
     52 typedef struct {
     53   int onig_err;
     54   int posix_err;
     55 } O2PERR;
     56 
     57 static int
     58 onig2posix_error_code(int code)
     59 {
     60   static const O2PERR o2p[] = {
     61     { ONIG_MISMATCH,                                      REG_NOMATCH },
     62     { ONIG_NO_SUPPORT_CONFIG,                             REG_EONIG_INTERNAL },
     63     { ONIGERR_MEMORY,                                     REG_ESPACE  },
     64     { ONIGERR_MATCH_STACK_LIMIT_OVER,                     REG_EONIG_INTERNAL },
     65     { ONIGERR_TYPE_BUG,                                   REG_EONIG_INTERNAL },
     66     { ONIGERR_PARSER_BUG,                                 REG_EONIG_INTERNAL },
     67     { ONIGERR_STACK_BUG,                                  REG_EONIG_INTERNAL },
     68     { ONIGERR_UNDEFINED_BYTECODE,                         REG_EONIG_INTERNAL },
     69     { ONIGERR_UNEXPECTED_BYTECODE,                        REG_EONIG_INTERNAL },
     70     { ONIGERR_DEFAULT_ENCODING_IS_NOT_SETTED,             REG_EONIG_BADARG },
     71     { ONIGERR_SPECIFIED_ENCODING_CANT_CONVERT_TO_WIDE_CHAR, REG_EONIG_BADARG },
     72     { ONIGERR_INVALID_ARGUMENT,                           REG_EONIG_BADARG },
     73     { ONIGERR_END_PATTERN_AT_LEFT_BRACE,                  REG_EBRACE  },
     74     { ONIGERR_END_PATTERN_AT_LEFT_BRACKET,                REG_EBRACK  },
     75     { ONIGERR_EMPTY_CHAR_CLASS,                           REG_ECTYPE  },
     76     { ONIGERR_PREMATURE_END_OF_CHAR_CLASS,                REG_ECTYPE  },
     77     { ONIGERR_END_PATTERN_AT_ESCAPE,                      REG_EESCAPE },
     78     { ONIGERR_END_PATTERN_AT_META,                        REG_EESCAPE },
     79     { ONIGERR_END_PATTERN_AT_CONTROL,                     REG_EESCAPE },
     80     { ONIGERR_META_CODE_SYNTAX,                           REG_BADPAT  },
     81     { ONIGERR_CONTROL_CODE_SYNTAX,                        REG_BADPAT  },
     82     { ONIGERR_CHAR_CLASS_VALUE_AT_END_OF_RANGE,           REG_ECTYPE  },
     83     { ONIGERR_CHAR_CLASS_VALUE_AT_START_OF_RANGE,         REG_ECTYPE  },
     84     { ONIGERR_UNMATCHED_RANGE_SPECIFIER_IN_CHAR_CLASS,    REG_ECTYPE  },
     85     { ONIGERR_TARGET_OF_REPEAT_OPERATOR_NOT_SPECIFIED,    REG_BADRPT  },
     86     { ONIGERR_TARGET_OF_REPEAT_OPERATOR_INVALID,          REG_BADRPT  },
     87     { ONIGERR_NESTED_REPEAT_OPERATOR,                     REG_BADRPT  },
     88     { ONIGERR_UNMATCHED_CLOSE_PARENTHESIS,                REG_EPAREN  },
     89     { ONIGERR_END_PATTERN_WITH_UNMATCHED_PARENTHESIS,     REG_EPAREN  },
     90     { ONIGERR_END_PATTERN_IN_GROUP,                       REG_BADPAT  },
     91     { ONIGERR_UNDEFINED_GROUP_OPTION,                     REG_BADPAT  },
     92     { ONIGERR_INVALID_POSIX_BRACKET_TYPE,                 REG_BADPAT  },
     93     { ONIGERR_INVALID_LOOK_BEHIND_PATTERN,                REG_BADPAT  },
     94     { ONIGERR_INVALID_REPEAT_RANGE_PATTERN,               REG_BADPAT  },
     95     { ONIGERR_TOO_BIG_NUMBER,                             REG_BADPAT  },
     96     { ONIGERR_TOO_BIG_NUMBER_FOR_REPEAT_RANGE,            REG_BADBR   },
     97     { ONIGERR_UPPER_SMALLER_THAN_LOWER_IN_REPEAT_RANGE,   REG_BADBR   },
     98     { ONIGERR_EMPTY_RANGE_IN_CHAR_CLASS,                  REG_ECTYPE  },
     99     { ONIGERR_MISMATCH_CODE_LENGTH_IN_CLASS_RANGE,        REG_ECTYPE  },
    100     { ONIGERR_TOO_MANY_MULTI_BYTE_RANGES,                 REG_ECTYPE  },
    101     { ONIGERR_TOO_SHORT_MULTI_BYTE_STRING,                REG_BADPAT  },
    102     { ONIGERR_TOO_BIG_BACKREF_NUMBER,                     REG_ESUBREG },
    103     { ONIGERR_INVALID_BACKREF,                            REG_ESUBREG },
    104     { ONIGERR_NUMBERED_BACKREF_OR_CALL_NOT_ALLOWED,       REG_BADPAT  },
    105     { ONIGERR_TOO_BIG_WIDE_CHAR_VALUE,                    REG_EONIG_BADWC },
    106     { ONIGERR_TOO_LONG_WIDE_CHAR_VALUE,                   REG_EONIG_BADWC },
    107     { ONIGERR_INVALID_CODE_POINT_VALUE,                   REG_EONIG_BADWC },
    108     { ONIGERR_EMPTY_GROUP_NAME,                           REG_BADPAT },
    109     { ONIGERR_INVALID_GROUP_NAME,                         REG_BADPAT },
    110     { ONIGERR_INVALID_CHAR_IN_GROUP_NAME,                 REG_BADPAT },
    111     { ONIGERR_UNDEFINED_NAME_REFERENCE,                   REG_BADPAT },
    112     { ONIGERR_UNDEFINED_GROUP_REFERENCE,                  REG_BADPAT },
    113     { ONIGERR_MULTIPLEX_DEFINED_NAME,                     REG_BADPAT },
    114     { ONIGERR_MULTIPLEX_DEFINITION_NAME_CALL,             REG_BADPAT },
    115     { ONIGERR_NEVER_ENDING_RECURSION,                     REG_BADPAT },
    116     { ONIGERR_GROUP_NUMBER_OVER_FOR_CAPTURE_HISTORY,      REG_BADPAT },
    117     { ONIGERR_INVALID_CHAR_PROPERTY_NAME,                 REG_BADPAT },
    118     { ONIGERR_NOT_SUPPORTED_ENCODING_COMBINATION,         REG_EONIG_BADARG },
    119     { ONIGERR_OVER_THREAD_PASS_LIMIT_COUNT,               REG_EONIG_THREAD }
    120 
    121   };
    122 
    123   int i;
    124 
    125   if (code >= 0) return 0;
    126 
    127   for (i = 0; i < (int )(sizeof(o2p) / sizeof(o2p[0])); i++) {
    128     if (code == o2p[i].onig_err)
    129       return o2p[i].posix_err;
    130   }
    131 
    132   return REG_EONIG_INTERNAL;  /* but, unknown error code */
    133 }
    134 
    135 extern int
    136 regcomp(regex_t* reg, const char* pattern, int posix_options)
    137 {
    138   int r, len;
    139   OnigSyntaxType* syntax = OnigDefaultSyntax;
    140   OnigOptionType options;
    141 
    142   if ((posix_options & REG_EXTENDED) == 0)
    143     syntax = ONIG_SYNTAX_POSIX_BASIC;
    144 
    145   options = syntax->options;
    146   if ((posix_options & REG_ICASE)   != 0)
    147     ONIG_OPTION_ON(options, ONIG_OPTION_IGNORECASE);
    148   if ((posix_options & REG_NEWLINE) != 0) {
    149     ONIG_OPTION_ON( options, ONIG_OPTION_NEGATE_SINGLELINE);
    150     ONIG_OPTION_OFF(options, ONIG_OPTION_SINGLELINE);
    151   }
    152 
    153   reg->comp_options = posix_options;
    154 
    155   ENC_STRING_LEN(OnigEncDefaultCharEncoding, pattern, len);
    156   r = onig_new(PONIG_C(reg), (UChar* )pattern, (UChar* )(pattern + len),
    157 	       options, OnigEncDefaultCharEncoding, syntax,
    158 	       (OnigErrorInfo* )NULL);
    159   if (r != ONIG_NORMAL) {
    160     return onig2posix_error_code(r);
    161   }
    162 
    163   reg->re_nsub = ONIG_C(reg)->num_mem;
    164   return 0;
    165 }
    166 
    167 extern int
    168 regexec(regex_t* reg, const char* str, size_t nmatch,
    169 	regmatch_t pmatch[], int posix_options)
    170 {
    171   int r, i, len;
    172   UChar* end;
    173   regmatch_t* pm;
    174   OnigOptionType options;
    175 
    176   options = ONIG_OPTION_POSIX_REGION;
    177   if ((posix_options & REG_NOTBOL) != 0) options |= ONIG_OPTION_NOTBOL;
    178   if ((posix_options & REG_NOTEOL) != 0) options |= ONIG_OPTION_NOTEOL;
    179 
    180   if (nmatch == 0 || (reg->comp_options & REG_NOSUB) != 0) {
    181     pm = (regmatch_t* )NULL;
    182     nmatch = 0;
    183   }
    184   else if ((int )nmatch < ONIG_C(reg)->num_mem + 1) {
    185     pm = (regmatch_t* )xmalloc(sizeof(regmatch_t)
    186                                * (ONIG_C(reg)->num_mem + 1));
    187     if (pm == NULL)
    188       return REG_ESPACE;
    189   }
    190   else {
    191     pm = pmatch;
    192   }
    193 
    194   ENC_STRING_LEN(ONIG_C(reg)->enc, str, len);
    195   end = (UChar* )(str + len);
    196   r = onig_search(ONIG_C(reg), (UChar* )str, end, (UChar* )str, end,
    197 		  (OnigRegion* )pm, options);
    198 
    199   if (r >= 0) {
    200     r = 0; /* Match */
    201     if (pm != pmatch && pm != NULL) {
    202       xmemcpy(pmatch, pm, sizeof(regmatch_t) * nmatch);
    203     }
    204   }
    205   else if (r == ONIG_MISMATCH) {
    206     r = REG_NOMATCH;
    207     for (i = 0; i < (int )nmatch; i++)
    208       pmatch[i].rm_so = pmatch[i].rm_eo = ONIG_REGION_NOTPOS;
    209   }
    210   else {
    211     r = onig2posix_error_code(r);
    212   }
    213 
    214   if (pm != pmatch && pm != NULL)
    215     xfree(pm);
    216 
    217 #if 0
    218   if (reg->re_nsub > nmatch - 1)
    219     reg->re_nsub = (nmatch <= 1 ? 0 : nmatch - 1);
    220 #endif
    221 
    222   return r;
    223 }
    224 
    225 extern void
    226 regfree(regex_t* reg)
    227 {
    228   onig_free(ONIG_C(reg));
    229 }
    230 
    231 
    232 extern void
    233 reg_set_encoding(int mb_code)
    234 {
    235   OnigEncoding enc;
    236 
    237   switch (mb_code) {
    238   case REG_POSIX_ENCODING_ASCII:
    239     enc = ONIG_ENCODING_ASCII;
    240     break;
    241   case REG_POSIX_ENCODING_EUC_JP:
    242     enc = ONIG_ENCODING_EUC_JP;
    243     break;
    244   case REG_POSIX_ENCODING_SJIS:
    245     enc = ONIG_ENCODING_SJIS;
    246     break;
    247   case REG_POSIX_ENCODING_UTF8:
    248     enc = ONIG_ENCODING_UTF8;
    249     break;
    250   case REG_POSIX_ENCODING_UTF16_BE:
    251     enc = ONIG_ENCODING_UTF16_BE;
    252     break;
    253   case REG_POSIX_ENCODING_UTF16_LE:
    254     enc = ONIG_ENCODING_UTF16_LE;
    255     break;
    256 
    257   default:
    258     return ;
    259     break;
    260   }
    261 
    262   onigenc_set_default_encoding(enc);
    263 }
    264 
    265 extern int
    266 reg_name_to_group_numbers(regex_t* reg,
    267   const unsigned char* name, const unsigned char* name_end, int** nums)
    268 {
    269   return onig_name_to_group_numbers(ONIG_C(reg), name, name_end, nums);
    270 }
    271 
    272 typedef struct {
    273   int (*func)(const unsigned char*, const unsigned char*,int,int*,regex_t*,void*);
    274   regex_t* reg;
    275   void* arg;
    276 } i_wrap;
    277 
    278 static int
    279 i_wrapper(const UChar* name, const UChar* name_end, int ng, int* gs,
    280 	  onig_regex_t* reg ARG_UNUSED, void* arg)
    281 {
    282   i_wrap* warg = (i_wrap* )arg;
    283 
    284   return (*warg->func)(name, name_end, ng, gs, warg->reg, warg->arg);
    285 }
    286 
    287 extern int
    288 reg_foreach_name(regex_t* reg,
    289  int (*func)(const unsigned char*, const unsigned char*,int,int*,regex_t*,void*),
    290  void* arg)
    291 {
    292   i_wrap warg;
    293 
    294   warg.func = func;
    295   warg.reg  = reg;
    296   warg.arg  = arg;
    297 
    298   return onig_foreach_name(ONIG_C(reg), i_wrapper, &warg);
    299 }
    300 
    301 extern int
    302 reg_number_of_names(regex_t* reg)
    303 {
    304   return onig_number_of_names(ONIG_C(reg));
    305 }
    306