/[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 75 by nigel, Sat Feb 24 21:40:37 2007 UTC revision 166 by ph10, Wed May 9 14:48:28 2007 UTC
# Line 2  Line 2 
2  *      Perl-Compatible Regular Expressions       *  *      Perl-Compatible Regular Expressions       *
3  *************************************************/  *************************************************/
4    
5  /*  /* PCRE is a library of functions to support regular expressions whose syntax
6  This is a library of functions to support regular expressions whose syntax  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. See  
 the file Tech.Notes for some information on the internals.  
   
 This module is a wrapper that provides a POSIX API to the underlying PCRE  
 functions.  
   
 Written by: Philip Hazel <ph10@cam.ac.uk>  
7    
8             Copyright (c) 1997-2004 University of Cambridge                         Written by Philip Hazel
9               Copyright (c) 1997-2007 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 43  POSSIBILITY OF SUCH DAMAGE. Line 37  POSSIBILITY OF SUCH DAMAGE.
37  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
38  */  */
39    
40  #include "internal.h"  
41    /* This module is a wrapper that provides a POSIX API to the underlying PCRE
42    functions. */
43    
44    
45    #include "pcre_internal.h"
46  #include "pcreposix.h"  #include "pcreposix.h"
47  #include "stdlib.h"  #include "stdlib.h"
48    
49    
50    
51  /* Corresponding tables of PCRE error messages and POSIX error codes. */  /* Table to translate PCRE compile time error codes into POSIX error codes. */
   
 static const char *const estring[] = {  
   ERR1,  ERR2,  ERR3,  ERR4,  ERR5,  ERR6,  ERR7,  ERR8,  ERR9,  ERR10,  
   ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19, ERR20,  
   ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR29, ERR29, ERR30,  
   ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39, ERR40,  
   ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47 };  
52    
53  static const int eint[] = {  static const int eint[] = {
54    REG_EESCAPE, /* "\\ at end of pattern" */    0,           /* no error */
55    REG_EESCAPE, /* "\\c at end of pattern" */    REG_EESCAPE, /* \ at end of pattern */
56    REG_EESCAPE, /* "unrecognized character follows \\" */    REG_EESCAPE, /* \c at end of pattern */
57    REG_BADBR,   /* "numbers out of order in {} quantifier" */    REG_EESCAPE, /* unrecognized character follows \ */
58    REG_BADBR,   /* "number too big in {} quantifier" */    REG_BADBR,   /* numbers out of order in {} quantifier */
59    REG_EBRACK,  /* "missing terminating ] for character class" */    REG_BADBR,   /* number too big in {} quantifier */
60    REG_ECTYPE,  /* "invalid escape sequence in character class" */    REG_EBRACK,  /* missing terminating ] for character class */
61    REG_ERANGE,  /* "range out of order in character class" */    REG_ECTYPE,  /* invalid escape sequence in character class */
62    REG_BADRPT,  /* "nothing to repeat" */    REG_ERANGE,  /* range out of order in character class */
63    REG_BADRPT,  /* "operand of unlimited repeat could match the empty string" */    REG_BADRPT,  /* nothing to repeat */
64    REG_ASSERT,  /* "internal error: unexpected repeat" */    REG_BADRPT,  /* operand of unlimited repeat could match the empty string */
65    REG_BADPAT,  /* "unrecognized character after (?" */    REG_ASSERT,  /* internal error: unexpected repeat */
66    REG_BADPAT,  /* "POSIX named classes are supported only within a class" */    REG_BADPAT,  /* unrecognized character after (? */
67    REG_EPAREN,  /* "missing )" */    REG_BADPAT,  /* POSIX named classes are supported only within a class */
68    REG_ESUBREG, /* "reference to non-existent subpattern" */    REG_EPAREN,  /* missing ) */
69    REG_INVARG,  /* "erroffset passed as NULL" */    REG_ESUBREG, /* reference to non-existent subpattern */
70    REG_INVARG,  /* "unknown option bit(s) set" */    REG_INVARG,  /* erroffset passed as NULL */
71    REG_EPAREN,  /* "missing ) after comment" */    REG_INVARG,  /* unknown option bit(s) set */
72    REG_ESIZE,   /* "parentheses nested too deeply" */    REG_EPAREN,  /* missing ) after comment */
73    REG_ESIZE,   /* "regular expression too large" */    REG_ESIZE,   /* parentheses nested too deeply */
74    REG_ESPACE,  /* "failed to get memory" */    REG_ESIZE,   /* regular expression too large */
75    REG_EPAREN,  /* "unmatched brackets" */    REG_ESPACE,  /* failed to get memory */
76    REG_ASSERT,  /* "internal error: code overflow" */    REG_EPAREN,  /* unmatched brackets */
77    REG_BADPAT,  /* "unrecognized character after (?<" */    REG_ASSERT,  /* internal error: code overflow */
78    REG_BADPAT,  /* "lookbehind assertion is not fixed length" */    REG_BADPAT,  /* unrecognized character after (?< */
79    REG_BADPAT,  /* "malformed number after (?(" */    REG_BADPAT,  /* lookbehind assertion is not fixed length */
80    REG_BADPAT,  /* "conditional group containe more than two branches" */    REG_BADPAT,  /* malformed number or name after (?( */
81    REG_BADPAT,  /* "assertion expected after (?(" */    REG_BADPAT,  /* conditional group contains more than two branches */
82    REG_BADPAT,  /* "(?R or (?digits must be followed by )" */    REG_BADPAT,  /* assertion expected after (?( */
83    REG_ECTYPE,  /* "unknown POSIX class name" */    REG_BADPAT,  /* (?R or (?[+-]digits must be followed by ) */
84    REG_BADPAT,  /* "POSIX collating elements are not supported" */    REG_ECTYPE,  /* unknown POSIX class name */
85    REG_INVARG,  /* "this version of PCRE is not compiled with PCRE_UTF8 support" */    REG_BADPAT,  /* POSIX collating elements are not supported */
86    REG_BADPAT,  /* "spare error" */    REG_INVARG,  /* this version of PCRE is not compiled with PCRE_UTF8 support */
87    REG_BADPAT,  /* "character value in \x{...} sequence is too large" */    REG_BADPAT,  /* spare error */
88    REG_BADPAT,  /* "invalid condition (?(0)" */    REG_BADPAT,  /* character value in \x{...} sequence is too large */
89    REG_BADPAT,  /* "\\C not allowed in lookbehind assertion" */    REG_BADPAT,  /* invalid condition (?(0) */
90    REG_EESCAPE, /* "PCRE does not support \\L, \\l, \\N, \\U, or \\u" */    REG_BADPAT,  /* \C not allowed in lookbehind assertion */
91    REG_BADPAT,  /* "number after (?C is > 255" */    REG_EESCAPE, /* PCRE does not support \L, \l, \N, \U, or \u */
92    REG_BADPAT,  /* "closing ) for (?C expected" */    REG_BADPAT,  /* number after (?C is > 255 */
93    REG_BADPAT,  /* "recursive call could loop indefinitely" */    REG_BADPAT,  /* closing ) for (?C expected */
94    REG_BADPAT,  /* "unrecognized character after (?P" */    REG_BADPAT,  /* recursive call could loop indefinitely */
95    REG_BADPAT,  /* "syntax error after (?P" */    REG_BADPAT,  /* unrecognized character after (?P */
96    REG_BADPAT,  /* "two named groups have the same name" */    REG_BADPAT,  /* syntax error in subpattern name (missing terminator) */
97    REG_BADPAT,  /* "invalid UTF-8 string" */    REG_BADPAT,  /* two named subpatterns have the same name */
98    REG_BADPAT,  /* "support for \\P, \\p, and \\X has not been compiled" */    REG_BADPAT,  /* invalid UTF-8 string */
99    REG_BADPAT,  /* "malformed \\P or \\p sequence" */    REG_BADPAT,  /* support for \P, \p, and \X has not been compiled */
100    REG_BADPAT   /* "unknown property name after \\P or \\p" */    REG_BADPAT,  /* malformed \P or \p sequence */
101      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      REG_BADPAT   /* (?+ or (?- must be followed by a non-zero number */
113  };  };
114    
115  /* Table of texts corresponding to POSIX error codes */  /* Table of texts corresponding to POSIX error codes */
# Line 135  static const char *const pstring[] = { Line 139  static const char *const pstring[] = {
139    
140    
141  /*************************************************  /*************************************************
 *          Translate PCRE text code to int       *  
 *************************************************/  
   
 /* PCRE compile-time errors are given as strings defined as macros. We can just  
 look them up in a table to turn them into POSIX-style error codes. */  
   
 static int  
 pcre_posix_error_code(const char *s)  
 {  
 size_t i;  
 for (i = 0; i < sizeof(estring)/sizeof(char *); i++)  
   if (strcmp(s, estring[i]) == 0) return eint[i];  
 return REG_ASSERT;  
 }  
   
   
   
 /*************************************************  
142  *          Translate error code to string        *  *          Translate error code to string        *
143  *************************************************/  *************************************************/
144    
145  EXPORT size_t  PCREPOSIX_EXP_DEFN size_t
146  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)
147  {  {
148  const char *message, *addmessage;  const char *message, *addmessage;
# Line 191  return length + addlength; Line 177  return length + addlength;
177  *           Free store held by a regex           *  *           Free store held by a regex           *
178  *************************************************/  *************************************************/
179    
180  EXPORT void  PCREPOSIX_EXP_DEFN void
181  regfree(regex_t *preg)  regfree(regex_t *preg)
182  {  {
183  (pcre_free)(preg->re_pcre);  (pcre_free)(preg->re_pcre);
# Line 214  Returns:      0 on success Line 200  Returns:      0 on success
200                various non-zero codes on failure                various non-zero codes on failure
201  */  */
202    
203  EXPORT int  PCREPOSIX_EXP_DEFN int
204  regcomp(regex_t *preg, const char *pattern, int cflags)  regcomp(regex_t *preg, const char *pattern, int cflags)
205  {  {
206  const char *errorptr;  const char *errorptr;
207  int erroffset;  int erroffset;
208    int errorcode;
209  int options = 0;  int options = 0;
210    
211  if ((cflags & REG_ICASE) != 0) options |= PCRE_CASELESS;  if ((cflags & REG_ICASE) != 0)   options |= PCRE_CASELESS;
212  if ((cflags & REG_NEWLINE) != 0) options |= PCRE_MULTILINE;  if ((cflags & REG_NEWLINE) != 0) options |= PCRE_MULTILINE;
213    if ((cflags & REG_DOTALL) != 0)  options |= PCRE_DOTALL;
214    if ((cflags & REG_NOSUB) != 0)   options |= PCRE_NO_AUTO_CAPTURE;
215    if ((cflags & REG_UTF8) != 0)    options |= PCRE_UTF8;
216    
217  preg->re_pcre = pcre_compile(pattern, options, &errorptr, &erroffset, NULL);  preg->re_pcre = pcre_compile2(pattern, options, &errorcode, &errorptr,
218      &erroffset, NULL);
219  preg->re_erroffset = erroffset;  preg->re_erroffset = erroffset;
220    
221  if (preg->re_pcre == NULL) return pcre_posix_error_code(errorptr);  if (preg->re_pcre == NULL) return eint[errorcode];
222    
223  preg->re_nsub = pcre_info((const pcre *)preg->re_pcre, NULL, NULL);  preg->re_nsub = pcre_info((const pcre *)preg->re_pcre, NULL, NULL);
224  return 0;  return 0;
# Line 245  substring, so we have to get and release Line 236  substring, so we have to get and release
236  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
237  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
238  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
239  in a macro that can be changed at configure time. */  in a macro that can be changed at configure time.
240    
241    If REG_NOSUB was specified at compile time, the PCRE_NO_AUTO_CAPTURE flag will
242    be set. When this is the case, the nmatch and pmatch arguments are ignored, and
243    the only result is yes/no/error. */
244    
245  EXPORT int  PCREPOSIX_EXP_DEFN int
246  regexec(const regex_t *preg, const char *string, size_t nmatch,  regexec(const regex_t *preg, const char *string, size_t nmatch,
247    regmatch_t pmatch[], int eflags)    regmatch_t pmatch[], int eflags)
248  {  {
# Line 256  int options = 0; Line 251  int options = 0;
251  int *ovector = NULL;  int *ovector = NULL;
252  int small_ovector[POSIX_MALLOC_THRESHOLD * 3];  int small_ovector[POSIX_MALLOC_THRESHOLD * 3];
253  BOOL allocated_ovector = FALSE;  BOOL allocated_ovector = FALSE;
254    BOOL nosub =
255      (((const pcre *)preg->re_pcre)->options & PCRE_NO_AUTO_CAPTURE) != 0;
256    
257  if ((eflags & REG_NOTBOL) != 0) options |= PCRE_NOTBOL;  if ((eflags & REG_NOTBOL) != 0) options |= PCRE_NOTBOL;
258  if ((eflags & REG_NOTEOL) != 0) options |= PCRE_NOTEOL;  if ((eflags & REG_NOTEOL) != 0) options |= PCRE_NOTEOL;
259    
260  ((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 */
261    
262  if (nmatch > 0)  /* When no string data is being returned, ensure that nmatch is zero.
263    Otherwise, ensure the vector for holding the return data is large enough. */
264    
265    if (nosub) nmatch = 0;
266    
267    else if (nmatch > 0)
268    {    {
269    if (nmatch <= POSIX_MALLOC_THRESHOLD)    if (nmatch <= POSIX_MALLOC_THRESHOLD)
270      {      {
# Line 270  if (nmatch > 0) Line 272  if (nmatch > 0)
272      }      }
273    else    else
274      {      {
275        if (nmatch > INT_MAX/(sizeof(int) * 3)) return REG_ESPACE;
276      ovector = (int *)malloc(sizeof(int) * nmatch * 3);      ovector = (int *)malloc(sizeof(int) * nmatch * 3);
277      if (ovector == NULL) return REG_ESPACE;      if (ovector == NULL) return REG_ESPACE;
278      allocated_ovector = TRUE;      allocated_ovector = TRUE;
# Line 284  if (rc == 0) rc = nmatch;    /* All capt Line 287  if (rc == 0) rc = nmatch;    /* All capt
287  if (rc >= 0)  if (rc >= 0)
288    {    {
289    size_t i;    size_t i;
290    for (i = 0; i < (size_t)rc; i++)    if (!nosub)
291      {      {
292      pmatch[i].rm_so = ovector[i*2];      for (i = 0; i < (size_t)rc; i++)
293      pmatch[i].rm_eo = ovector[i*2+1];        {
294          pmatch[i].rm_so = ovector[i*2];
295          pmatch[i].rm_eo = ovector[i*2+1];
296          }
297        if (allocated_ovector) free(ovector);
298        for (; i < nmatch; i++) pmatch[i].rm_so = pmatch[i].rm_eo = -1;
299      }      }
   if (allocated_ovector) free(ovector);  
   for (; i < nmatch; i++) pmatch[i].rm_so = pmatch[i].rm_eo = -1;  
300    return 0;    return 0;
301    }    }
302    

Legend:
Removed from v.75  
changed lines
  Added in v.166

  ViewVC Help
Powered by ViewVC 1.1.5