12 |
|
|
13 |
Written by: Philip Hazel <ph10@cam.ac.uk> |
Written by: Philip Hazel <ph10@cam.ac.uk> |
14 |
|
|
15 |
Copyright (c) 1997 University of Cambridge |
Copyright (c) 1997-2000 University of Cambridge |
16 |
|
|
17 |
----------------------------------------------------------------------------- |
----------------------------------------------------------------------------- |
18 |
Permission is granted to anyone to use this software for any purpose on any |
Permission is granted to anyone to use this software for any purpose on any |
28 |
|
|
29 |
3. Altered versions must be plainly marked as such, and must not be |
3. Altered versions must be plainly marked as such, and must not be |
30 |
misrepresented as being the original software. |
misrepresented as being the original software. |
31 |
|
|
32 |
|
4. If PCRE is embedded in any software that is released under the GNU |
33 |
|
General Purpose Licence (GPL), then the terms of that licence shall |
34 |
|
supersede any condition above with which it is incompatible. |
35 |
----------------------------------------------------------------------------- |
----------------------------------------------------------------------------- |
36 |
*/ |
*/ |
37 |
|
|
43 |
|
|
44 |
/* Corresponding tables of PCRE error messages and POSIX error codes. */ |
/* Corresponding tables of PCRE error messages and POSIX error codes. */ |
45 |
|
|
46 |
static char *estring[] = { |
static const char *estring[] = { |
47 |
ERR1, ERR2, ERR3, ERR4, ERR5, ERR6, ERR7, ERR8, ERR9, ERR10, |
ERR1, ERR2, ERR3, ERR4, ERR5, ERR6, ERR7, ERR8, ERR9, ERR10, |
48 |
ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19, ERR20, |
ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19, ERR20, |
49 |
ERR21, ERR22, ERR23 }; |
ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR29, ERR29, ERR30, |
50 |
|
ERR31 }; |
51 |
|
|
52 |
static int eint[] = { |
static int eint[] = { |
53 |
REG_EESCAPE, /* "\\ at end of pattern" */ |
REG_EESCAPE, /* "\\ at end of pattern" */ |
72 |
REG_ESIZE, /* "regular expression too large" */ |
REG_ESIZE, /* "regular expression too large" */ |
73 |
REG_ESPACE, /* "failed to get memory" */ |
REG_ESPACE, /* "failed to get memory" */ |
74 |
REG_EPAREN, /* "unmatched brackets" */ |
REG_EPAREN, /* "unmatched brackets" */ |
75 |
REG_ASSERT /* "internal error: code overflow" */ |
REG_ASSERT, /* "internal error: code overflow" */ |
76 |
|
REG_BADPAT, /* "unrecognized character after (?<" */ |
77 |
|
REG_BADPAT, /* "lookbehind assertion is not fixed length" */ |
78 |
|
REG_BADPAT, /* "malformed number after (?(" */ |
79 |
|
REG_BADPAT, /* "conditional group containe more than two branches" */ |
80 |
|
REG_BADPAT, /* "assertion expected after (?(" */ |
81 |
|
REG_BADPAT, /* "(?p must be followed by )" */ |
82 |
|
REG_ECTYPE, /* "unknown POSIX class name" */ |
83 |
|
REG_BADPAT, /* "POSIX collating elements are not supported" */ |
84 |
|
REG_INVARG, /* "this version of PCRE is not compiled with PCRE_UTF8 support" */ |
85 |
|
REG_BADPAT, /* "characters with values > 255 are not yet supported in classes" */ |
86 |
|
REG_BADPAT, /* "character value in \x{...} sequence is too large" */ |
87 |
|
REG_BADPAT /* "invalid condition (?(0)" */ |
88 |
}; |
}; |
89 |
|
|
90 |
/* Table of texts corresponding to POSIX error codes */ |
/* Table of texts corresponding to POSIX error codes */ |
91 |
|
|
92 |
static char *pstring[] = { |
static const char *pstring[] = { |
93 |
"", /* Dummy for value 0 */ |
"", /* Dummy for value 0 */ |
94 |
"internal error", /* REG_ASSERT */ |
"internal error", /* REG_ASSERT */ |
95 |
"invalid repeat counts in {}", /* BADBR */ |
"invalid repeat counts in {}", /* BADBR */ |
123 |
static int |
static int |
124 |
pcre_posix_error_code(const char *s) |
pcre_posix_error_code(const char *s) |
125 |
{ |
{ |
126 |
int i; |
size_t i; |
127 |
for (i = 0; i < sizeof(estring)/sizeof(char *); i++) |
for (i = 0; i < sizeof(estring)/sizeof(char *); i++) |
128 |
if (strcmp(s, estring[i]) == 0) return eint[i]; |
if (strcmp(s, estring[i]) == 0) return eint[i]; |
129 |
return REG_ASSERT; |
return REG_ASSERT; |
138 |
size_t |
size_t |
139 |
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) |
140 |
{ |
{ |
141 |
char *message, *addmessage; |
const char *message, *addmessage; |
142 |
int length, adlength; |
size_t length, addlength; |
143 |
|
|
144 |
message = (errcode >= sizeof(pstring)/sizeof(char *))? |
message = (errcode >= (int)(sizeof(pstring)/sizeof(char *)))? |
145 |
"unknown error code" : pstring[errcode]; |
"unknown error code" : pstring[errcode]; |
146 |
|
length = strlen(message) + 1; |
147 |
|
|
148 |
length = (int)strlen(message) + 1; |
addmessage = " at offset "; |
149 |
|
addlength = (preg != NULL && (int)preg->re_erroffset != -1)? |
150 |
if (preg != NULL && (int)preg->re_erroffset != -1) |
strlen(addmessage) + 6 : 0; |
|
{ |
|
|
addmessage = " at offset "; |
|
|
adlength = (int)strlen(addmessage) + 6; |
|
|
} |
|
|
else adlength = 0; |
|
151 |
|
|
152 |
if (errbuf_size > 0) |
if (errbuf_size > 0) |
153 |
{ |
{ |
154 |
if (adlength > 0 && errbuf_size >= length + adlength) |
if (addlength > 0 && errbuf_size >= length + addlength) |
155 |
sprintf(errbuf, "%s%s%-6d", message, addmessage, preg->re_erroffset); |
sprintf(errbuf, "%s%s%-6d", message, addmessage, (int)preg->re_erroffset); |
156 |
else |
else |
157 |
{ |
{ |
158 |
strncpy(errbuf, message, errbuf_size - 1); |
strncpy(errbuf, message, errbuf_size - 1); |
160 |
} |
} |
161 |
} |
} |
162 |
|
|
163 |
return length + adlength; |
return length + addlength; |
164 |
} |
} |
165 |
|
|
166 |
|
|
196 |
int |
int |
197 |
regcomp(regex_t *preg, const char *pattern, int cflags) |
regcomp(regex_t *preg, const char *pattern, int cflags) |
198 |
{ |
{ |
199 |
char *errorptr; |
const char *errorptr; |
200 |
int erroffset; |
int erroffset; |
201 |
int options = 0; |
int options = 0; |
202 |
|
|
203 |
if ((cflags & REG_ICASE) != 0) options |= PCRE_CASELESS; |
if ((cflags & REG_ICASE) != 0) options |= PCRE_CASELESS; |
204 |
if ((cflags & REG_NEWLINE) != 0) options |= PCRE_MULTILINE; |
if ((cflags & REG_NEWLINE) != 0) options |= PCRE_MULTILINE; |
205 |
|
|
206 |
preg->re_pcre = pcre_compile(pattern, options, &errorptr, &erroffset); |
preg->re_pcre = pcre_compile(pattern, options, &errorptr, &erroffset, NULL); |
207 |
preg->re_erroffset = erroffset; |
preg->re_erroffset = erroffset; |
208 |
|
|
209 |
if (preg->re_pcre == NULL) return pcre_posix_error_code(errorptr); |
if (preg->re_pcre == NULL) return pcre_posix_error_code(errorptr); |
219 |
* Match a regular expression * |
* Match a regular expression * |
220 |
*************************************************/ |
*************************************************/ |
221 |
|
|
222 |
|
/* Unfortunately, PCRE requires 3 ints of working space for each captured |
223 |
|
substring, so we have to get and release working store instead of just using |
224 |
|
the POSIX structures as was done in earlier releases when PCRE needed only 2 |
225 |
|
ints. */ |
226 |
|
|
227 |
int |
int |
228 |
regexec(regex_t *preg, const char *string, size_t nmatch, |
regexec(regex_t *preg, const char *string, size_t nmatch, |
229 |
regmatch_t pmatch[], int eflags) |
regmatch_t pmatch[], int eflags) |
230 |
{ |
{ |
231 |
int rc; |
int rc; |
232 |
int options = 0; |
int options = 0; |
233 |
|
int *ovector = NULL; |
234 |
|
|
235 |
if ((eflags & REG_NOTBOL) != 0) options |= PCRE_NOTBOL; |
if ((eflags & REG_NOTBOL) != 0) options |= PCRE_NOTBOL; |
236 |
if ((eflags & REG_NOTEOL) != 0) options |= PCRE_NOTEOL; |
if ((eflags & REG_NOTEOL) != 0) options |= PCRE_NOTEOL; |
237 |
|
|
238 |
preg->re_erroffset = -1; /* Only has meaning after compile */ |
preg->re_erroffset = (size_t)(-1); /* Only has meaning after compile */ |
239 |
|
|
240 |
|
if (nmatch > 0) |
241 |
|
{ |
242 |
|
ovector = (int *)malloc(sizeof(int) * nmatch * 3); |
243 |
|
if (ovector == NULL) return REG_ESPACE; |
244 |
|
} |
245 |
|
|
246 |
rc = pcre_exec(preg->re_pcre, NULL, string, (int)strlen(string), options, |
rc = pcre_exec(preg->re_pcre, NULL, string, (int)strlen(string), 0, options, |
247 |
(int *)pmatch, nmatch * 2); |
ovector, nmatch * 3); |
248 |
|
|
249 |
if (rc == 0) return 0; /* All pmatch were filled in */ |
if (rc == 0) rc = nmatch; /* All captured slots were filled in */ |
250 |
|
|
251 |
if (rc > 0) |
if (rc >= 0) |
252 |
{ |
{ |
253 |
int i; |
size_t i; |
254 |
for (i = rc; i < nmatch; i++) pmatch[i].rm_so = pmatch[i].rm_eo = -1; |
for (i = 0; i < rc; i++) |
255 |
|
{ |
256 |
|
pmatch[i].rm_so = ovector[i*2]; |
257 |
|
pmatch[i].rm_eo = ovector[i*2+1]; |
258 |
|
} |
259 |
|
if (ovector != NULL) free(ovector); |
260 |
|
for (; i < nmatch; i++) pmatch[i].rm_so = pmatch[i].rm_eo = -1; |
261 |
return 0; |
return 0; |
262 |
} |
} |
263 |
|
|
264 |
else switch(rc) |
else |
265 |
{ |
{ |
266 |
case PCRE_ERROR_NOMATCH: return REG_NOMATCH; |
if (ovector != NULL) free(ovector); |
267 |
case PCRE_ERROR_BADREF: return REG_ESUBREG; |
switch(rc) |
268 |
case PCRE_ERROR_NULL: return REG_INVARG; |
{ |
269 |
case PCRE_ERROR_BADOPTION: return REG_INVARG; |
case PCRE_ERROR_NOMATCH: return REG_NOMATCH; |
270 |
case PCRE_ERROR_BADMAGIC: return REG_INVARG; |
case PCRE_ERROR_NULL: return REG_INVARG; |
271 |
case PCRE_ERROR_UNKNOWN_NODE: return REG_ASSERT; |
case PCRE_ERROR_BADOPTION: return REG_INVARG; |
272 |
case PCRE_ERROR_NOMEMORY: return REG_ESPACE; |
case PCRE_ERROR_BADMAGIC: return REG_INVARG; |
273 |
default: return REG_ASSERT; |
case PCRE_ERROR_UNKNOWN_NODE: return REG_ASSERT; |
274 |
|
case PCRE_ERROR_NOMEMORY: return REG_ESPACE; |
275 |
|
default: return REG_ASSERT; |
276 |
|
} |
277 |
} |
} |
278 |
} |
} |
279 |
|
|