/[pcre]/code/trunk/pcreposix.c
ViewVC logotype

Diff of /code/trunk/pcreposix.c

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 85 by nigel, Sat Feb 24 21:41:13 2007 UTC revision 96 by nigel, Fri Mar 2 13:10:43 2007 UTC
# Line 6  Line 6 
6  and semantics are as close as possible to those of the Perl 5 language.  and semantics are as close as possible to those of the Perl 5 language.
7    
8                         Written by Philip Hazel                         Written by Philip Hazel
9             Copyright (c) 1997-2005 University of Cambridge             Copyright (c) 1997-2006 University of Cambridge
10    
11  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
12  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 77  static const int eint[] = { Line 77  static const int eint[] = {
77    REG_ASSERT,  /* internal error: code overflow */    REG_ASSERT,  /* internal error: code overflow */
78    REG_BADPAT,  /* unrecognized character after (?< */    REG_BADPAT,  /* unrecognized character after (?< */
79    REG_BADPAT,  /* lookbehind assertion is not fixed length */    REG_BADPAT,  /* lookbehind assertion is not fixed length */
80    REG_BADPAT,  /* malformed number after (?( */    REG_BADPAT,  /* malformed number or name after (?( */
81    REG_BADPAT,  /* conditional group containe more than two branches */    REG_BADPAT,  /* conditional group contains more than two branches */
82    REG_BADPAT,  /* assertion expected after (?( */    REG_BADPAT,  /* assertion expected after (?( */
83    REG_BADPAT,  /* (?R or (?digits must be followed by ) */    REG_BADPAT,  /* (?R or (?digits must be followed by ) */
84    REG_ECTYPE,  /* unknown POSIX class name */    REG_ECTYPE,  /* unknown POSIX class name */
# Line 93  static const int eint[] = { Line 93  static const int eint[] = {
93    REG_BADPAT,  /* closing ) for (?C expected */    REG_BADPAT,  /* closing ) for (?C expected */
94    REG_BADPAT,  /* recursive call could loop indefinitely */    REG_BADPAT,  /* recursive call could loop indefinitely */
95    REG_BADPAT,  /* unrecognized character after (?P */    REG_BADPAT,  /* unrecognized character after (?P */
96    REG_BADPAT,  /* syntax error after (?P */    REG_BADPAT,  /* syntax error in subpattern name (missing terminator) */
97    REG_BADPAT,  /* two named groups have the same name */    REG_BADPAT,  /* two named subpatterns have the same name */
98    REG_BADPAT,  /* invalid UTF-8 string */    REG_BADPAT,  /* invalid UTF-8 string */
99    REG_BADPAT,  /* support for \P, \p, and \X has not been compiled */    REG_BADPAT,  /* support for \P, \p, and \X has not been compiled */
100    REG_BADPAT,  /* malformed \P or \p sequence */    REG_BADPAT,  /* malformed \P or \p sequence */
101    REG_BADPAT   /* unknown property name after \P or \p */    REG_BADPAT,  /* unknown property name after \P or \p */
102      REG_BADPAT,  /* subpattern name is too long (maximum 32 characters) */
103      REG_BADPAT,  /* too many named subpatterns (maximum 10,000) */
104      REG_BADPAT,  /* repeated subpattern is too long */
105      REG_BADPAT,  /* octal value is greater than \377 (not in UTF-8 mode) */
106      REG_BADPAT,  /* internal error: overran compiling workspace */
107      REG_BADPAT,  /* internal error: previously-checked referenced subpattern not found */
108      REG_BADPAT,  /* DEFINE group contains more than one branch */
109      REG_BADPAT,  /* repeating a DEFINE group is not allowed */
110      REG_INVARG,  /* inconsistent NEWLINE options */
111      REG_BADPAT   /* \g is not followed followed by an (optionally braced) non-zero number */
112  };  };
113    
114  /* Table of texts corresponding to POSIX error codes */  /* Table of texts corresponding to POSIX error codes */
# Line 131  static const char *const pstring[] = { Line 141  static const char *const pstring[] = {
141  *          Translate error code to string        *  *          Translate error code to string        *
142  *************************************************/  *************************************************/
143    
144  PCRE_EXPORT size_t  PCRE_DATA_SCOPE size_t
145  regerror(int errcode, const regex_t *preg, char *errbuf, size_t errbuf_size)  regerror(int errcode, const regex_t *preg, char *errbuf, size_t errbuf_size)
146  {  {
147  const char *message, *addmessage;  const char *message, *addmessage;
# Line 166  return length + addlength; Line 176  return length + addlength;
176  *           Free store held by a regex           *  *           Free store held by a regex           *
177  *************************************************/  *************************************************/
178    
179  PCRE_EXPORT void  PCRE_DATA_SCOPE void
180  regfree(regex_t *preg)  regfree(regex_t *preg)
181  {  {
182  (pcre_free)(preg->re_pcre);  (pcre_free)(preg->re_pcre);
# Line 189  Returns:      0 on success Line 199  Returns:      0 on success
199                various non-zero codes on failure                various non-zero codes on failure
200  */  */
201    
202  PCRE_EXPORT int  PCRE_DATA_SCOPE int
203  regcomp(regex_t *preg, const char *pattern, int cflags)  regcomp(regex_t *preg, const char *pattern, int cflags)
204  {  {
205  const char *errorptr;  const char *errorptr;
# Line 197  int erroffset; Line 207  int erroffset;
207  int errorcode;  int errorcode;
208  int options = 0;  int options = 0;
209    
210  if ((cflags & REG_ICASE) != 0) options |= PCRE_CASELESS;  if ((cflags & REG_ICASE) != 0)   options |= PCRE_CASELESS;
211  if ((cflags & REG_NEWLINE) != 0) options |= PCRE_MULTILINE;  if ((cflags & REG_NEWLINE) != 0) options |= PCRE_MULTILINE;
212  if ((cflags & REG_DOTALL) != 0) options |= PCRE_DOTALL;  if ((cflags & REG_DOTALL) != 0)  options |= PCRE_DOTALL;
213    if ((cflags & REG_NOSUB) != 0)   options |= PCRE_NO_AUTO_CAPTURE;
214    if ((cflags & REG_UTF8) != 0)    options |= PCRE_UTF8;
215    
216  preg->re_pcre = pcre_compile2(pattern, options, &errorcode, &errorptr,  preg->re_pcre = pcre_compile2(pattern, options, &errorcode, &errorptr,
217    &erroffset, NULL);    &erroffset, NULL);
# Line 223  substring, so we have to get and release Line 235  substring, so we have to get and release
235  the POSIX structures as was done in earlier releases when PCRE needed only 2  the POSIX structures as was done in earlier releases when PCRE needed only 2
236  ints. However, if the number of possible capturing brackets is small, use a  ints. However, if the number of possible capturing brackets is small, use a
237  block of store on the stack, to reduce the use of malloc/free. The threshold is  block of store on the stack, to reduce the use of malloc/free. The threshold is
238  in a macro that can be changed at configure time. */  in a macro that can be changed at configure time.
239    
240  PCRE_EXPORT int  If REG_NOSUB was specified at compile time, the PCRE_NO_AUTO_CAPTURE flag will
241    be set. When this is the case, the nmatch and pmatch arguments are ignored, and
242    the only result is yes/no/error. */
243    
244    PCRE_DATA_SCOPE int
245  regexec(const regex_t *preg, const char *string, size_t nmatch,  regexec(const regex_t *preg, const char *string, size_t nmatch,
246    regmatch_t pmatch[], int eflags)    regmatch_t pmatch[], int eflags)
247  {  {
# Line 234  int options = 0; Line 250  int options = 0;
250  int *ovector = NULL;  int *ovector = NULL;
251  int small_ovector[POSIX_MALLOC_THRESHOLD * 3];  int small_ovector[POSIX_MALLOC_THRESHOLD * 3];
252  BOOL allocated_ovector = FALSE;  BOOL allocated_ovector = FALSE;
253    BOOL nosub =
254      (((const pcre *)preg->re_pcre)->options & PCRE_NO_AUTO_CAPTURE) != 0;
255    
256  if ((eflags & REG_NOTBOL) != 0) options |= PCRE_NOTBOL;  if ((eflags & REG_NOTBOL) != 0) options |= PCRE_NOTBOL;
257  if ((eflags & REG_NOTEOL) != 0) options |= PCRE_NOTEOL;  if ((eflags & REG_NOTEOL) != 0) options |= PCRE_NOTEOL;
258    
259  ((regex_t *)preg)->re_erroffset = (size_t)(-1);  /* Only has meaning after compile */  ((regex_t *)preg)->re_erroffset = (size_t)(-1);  /* Only has meaning after compile */
260    
261  if (nmatch > 0)  /* When no string data is being returned, ensure that nmatch is zero.
262    Otherwise, ensure the vector for holding the return data is large enough. */
263    
264    if (nosub) nmatch = 0;
265    
266    else if (nmatch > 0)
267    {    {
268    if (nmatch <= POSIX_MALLOC_THRESHOLD)    if (nmatch <= POSIX_MALLOC_THRESHOLD)
269      {      {
# Line 248  if (nmatch > 0) Line 271  if (nmatch > 0)
271      }      }
272    else    else
273      {      {
274        if (nmatch > INT_MAX/(sizeof(int) * 3)) return REG_ESPACE;
275      ovector = (int *)malloc(sizeof(int) * nmatch * 3);      ovector = (int *)malloc(sizeof(int) * nmatch * 3);
276      if (ovector == NULL) return REG_ESPACE;      if (ovector == NULL) return REG_ESPACE;
277      allocated_ovector = TRUE;      allocated_ovector = TRUE;
# Line 262  if (rc == 0) rc = nmatch;    /* All capt Line 286  if (rc == 0) rc = nmatch;    /* All capt
286  if (rc >= 0)  if (rc >= 0)
287    {    {
288    size_t i;    size_t i;
289    for (i = 0; i < (size_t)rc; i++)    if (!nosub)
290      {      {
291      pmatch[i].rm_so = ovector[i*2];      for (i = 0; i < (size_t)rc; i++)
292      pmatch[i].rm_eo = ovector[i*2+1];        {
293          pmatch[i].rm_so = ovector[i*2];
294          pmatch[i].rm_eo = ovector[i*2+1];
295          }
296        if (allocated_ovector) free(ovector);
297        for (; i < nmatch; i++) pmatch[i].rm_so = pmatch[i].rm_eo = -1;
298      }      }
   if (allocated_ovector) free(ovector);  
   for (; i < nmatch; i++) pmatch[i].rm_so = pmatch[i].rm_eo = -1;  
299    return 0;    return 0;
300    }    }
301    

Legend:
Removed from v.85  
changed lines
  Added in v.96

  ViewVC Help
Powered by ViewVC 1.1.5