tags/pcre-4.0/pcreposix.c

/*************************************************
*      Perl-Compatible Regular Expressions       *
*************************************************/

/*
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. 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>

           Copyright (c) 1997-2003 University of Cambridge

-----------------------------------------------------------------------------
Permission is granted to anyone to use this software for any purpose on any
computer system, and to redistribute it freely, subject to the following
restrictions:

1. This software is distributed in the hope that it will be useful,
   but WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

2. The origin of this software must not be misrepresented, either by
   explicit claim or by omission.

3. Altered versions must be plainly marked as such, and must not be
   misrepresented as being the original software.

4. If PCRE is embedded in any software that is released under the GNU
   General Purpose Licence (GPL), then the terms of that licence shall
   supersede any condition above with which it is incompatible.
-----------------------------------------------------------------------------
*/

#include "internal.h"
#include "pcreposix.h"
#include "stdlib.h"


/* Corresponding tables of PCRE error messages and POSIX error codes. */

static const char *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 };

static int eint[] = {
  REG_EESCAPE, /* "\\ at end of pattern" */
  REG_EESCAPE, /* "\\c at end of pattern" */
  REG_EESCAPE, /* "unrecognized character follows \\" */
  REG_BADBR,   /* "numbers out of order in {} quantifier" */
  REG_BADBR,   /* "number too big in {} quantifier" */
  REG_EBRACK,  /* "missing terminating ] for character class" */
  REG_ECTYPE,  /* "invalid escape sequence in character class" */
  REG_ERANGE,  /* "range out of order in character class" */
  REG_BADRPT,  /* "nothing to repeat" */
  REG_BADRPT,  /* "operand of unlimited repeat could match the empty string" */
  REG_ASSERT,  /* "internal error: unexpected repeat" */
  REG_BADPAT,  /* "unrecognized character after (?" */
  REG_BADPAT,  /* "POSIX named classes are supported only within a class" */
  REG_EPAREN,  /* "missing )" */
  REG_ESUBREG, /* "reference to non-existent subpattern" */
  REG_INVARG,  /* "erroffset passed as NULL" */
  REG_INVARG,  /* "unknown option bit(s) set" */
  REG_EPAREN,  /* "missing ) after comment" */
  REG_ESIZE,   /* "parentheses nested too deeply" */
  REG_ESIZE,   /* "regular expression too large" */
  REG_ESPACE,  /* "failed to get memory" */
  REG_EPAREN,  /* "unmatched brackets" */
  REG_ASSERT,  /* "internal error: code overflow" */
  REG_BADPAT,  /* "unrecognized character after (?<" */
  REG_BADPAT,  /* "lookbehind assertion is not fixed length" */
  REG_BADPAT,  /* "malformed number after (?(" */
  REG_BADPAT,  /* "conditional group containe more than two branches" */
  REG_BADPAT,  /* "assertion expected after (?(" */
  REG_BADPAT,  /* "(?R or (?digits must be followed by )" */
  REG_ECTYPE,  /* "unknown POSIX class name" */
  REG_BADPAT,  /* "POSIX collating elements are not supported" */
  REG_INVARG,  /* "this version of PCRE is not compiled with PCRE_UTF8 support" */
  REG_BADPAT,  /* "spare error" */
  REG_BADPAT,  /* "character value in \x{...} sequence is too large" */
  REG_BADPAT,  /* "invalid condition (?(0)" */
  REG_BADPAT,  /* "\\C not allowed in lookbehind assertion" */
  REG_EESCAPE, /* "PCRE does not support \\L, \\l, \\N, \\P, \\p, \\U, \\u, or \\X" */
  REG_BADPAT,  /* "number after (?C is > 255" */
  REG_BADPAT,  /* "closing ) for (?C expected" */
  REG_BADPAT,  /* "recursive call could loop indefinitely" */
  REG_BADPAT,  /* "unrecognized character after (?P" */
  REG_BADPAT,  /* "syntax error after (?P" */
  REG_BADPAT   /* "two named groups have the same name" */
};

/* Table of texts corresponding to POSIX error codes */

static const char *pstring[] = {
  "",                                /* Dummy for value 0 */
  "internal error",                  /* REG_ASSERT */
  "invalid repeat counts in {}",     /* BADBR      */
  "pattern error",                   /* BADPAT     */
  "? * + invalid",                   /* BADRPT     */
  "unbalanced {}",                   /* EBRACE     */
  "unbalanced []",                   /* EBRACK     */
  "collation error - not relevant",  /* ECOLLATE   */
  "bad class",                       /* ECTYPE     */
  "bad escape sequence",             /* EESCAPE    */
  "empty expression",                /* EMPTY      */
  "unbalanced ()",                   /* EPAREN     */
  "bad range inside []",             /* ERANGE     */
  "expression too big",              /* ESIZE      */
  "failed to get memory",            /* ESPACE     */
  "bad back reference",              /* ESUBREG    */
  "bad argument",                    /* INVARG     */
  "match failed"                     /* NOMATCH    */
};


/*************************************************
*          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;
}


/*************************************************
*          Translate error code to string        *
*************************************************/

size_t
regerror(int errcode, const regex_t *preg, char *errbuf, size_t errbuf_size)
{
const char *message, *addmessage;
size_t length, addlength;

message = (errcode >= (int)(sizeof(pstring)/sizeof(char *)))?
  "unknown error code" : pstring[errcode];
length = strlen(message) + 1;

addmessage = " at offset ";
addlength = (preg != NULL && (int)preg->re_erroffset != -1)?
  strlen(addmessage) + 6 : 0;

if (errbuf_size > 0)
  {
  if (addlength > 0 && errbuf_size >= length + addlength)
    sprintf(errbuf, "%s%s%-6d", message, addmessage, (int)preg->re_erroffset);
  else
    {
    strncpy(errbuf, message, errbuf_size - 1);
    errbuf[errbuf_size-1] = 0;
    }
  }

return length + addlength;
}


/*************************************************
*           Free store held by a regex           *
*************************************************/

void
regfree(regex_t *preg)
{
(pcre_free)(preg->re_pcre);
}


/*************************************************
*            Compile a regular expression        *
*************************************************/

/*
Arguments:
  preg        points to a structure for recording the compiled expression
  pattern     the pattern to compile
  cflags      compilation flags

Returns:      0 on success
              various non-zero codes on failure
*/

int
regcomp(regex_t *preg, const char *pattern, int cflags)
{
const char *errorptr;
int erroffset;
int options = 0;

if ((cflags & REG_ICASE) != 0) options |= PCRE_CASELESS;
if ((cflags & REG_NEWLINE) != 0) options |= PCRE_MULTILINE;

preg->re_pcre = pcre_compile(pattern, options, &errorptr, &erroffset, NULL);
preg->re_erroffset = erroffset;

if (preg->re_pcre == NULL) return pcre_posix_error_code(errorptr);

preg->re_nsub = pcre_info(preg->re_pcre, NULL, NULL);
return 0;
}


/*************************************************
*              Match a regular expression        *
*************************************************/

/* Unfortunately, PCRE requires 3 ints of working space for each captured
substring, so we have to get and release working store instead of just using
the POSIX structures as was done in earlier releases when PCRE needed only 2
ints. However, if the number of possible capturing brackets is small, use a
block of store on the stack, to reduce the use of malloc/free. The threshold is
in a macro that can be changed at configure time. */

int
regexec(regex_t *preg, const char *string, size_t nmatch,
  regmatch_t pmatch[], int eflags)
{
int rc;
int options = 0;
int *ovector = NULL;
int small_ovector[POSIX_MALLOC_THRESHOLD * 3];
BOOL allocated_ovector = FALSE;

if ((eflags & REG_NOTBOL) != 0) options |= PCRE_NOTBOL;
if ((eflags & REG_NOTEOL) != 0) options |= PCRE_NOTEOL;

preg->re_erroffset = (size_t)(-1);   /* Only has meaning after compile */

if (nmatch > 0)
  {
  if (nmatch <= POSIX_MALLOC_THRESHOLD)
    {
    ovector = &(small_ovector[0]);
    }
  else
    {
    ovector = (int *)malloc(sizeof(int) * nmatch * 3);
    if (ovector == NULL) return REG_ESPACE;
    allocated_ovector = TRUE;
    }
  }

rc = pcre_exec(preg->re_pcre, NULL, string, (int)strlen(string), 0, options,
  ovector, nmatch * 3);

if (rc == 0) rc = nmatch;    /* All captured slots were filled in */

if (rc >= 0)
  {
  size_t i;
  for (i = 0; i < (size_t)rc; i++)
    {
    pmatch[i].rm_so = ovector[i*2];
    pmatch[i].rm_eo = ovector[i*2+1];
    }
  if (allocated_ovector) free(ovector);
  for (; i < nmatch; i++) pmatch[i].rm_so = pmatch[i].rm_eo = -1;
  return 0;
  }

else
  {
  if (allocated_ovector) free(ovector);
  switch(rc)
    {
    case PCRE_ERROR_NOMATCH: return REG_NOMATCH;
    case PCRE_ERROR_NULL: return REG_INVARG;
    case PCRE_ERROR_BADOPTION: return REG_INVARG;
    case PCRE_ERROR_BADMAGIC: return REG_INVARG;
    case PCRE_ERROR_UNKNOWN_NODE: return REG_ASSERT;
    case PCRE_ERROR_NOMEMORY: return REG_ESPACE;
    default: return REG_ASSERT;
    }
  }
}

/* End of pcreposix.c */
1	/*************************************************
2	* Perl-Compatible Regular Expressions *
3	*************************************************/
4
5	/*
6	This is a library of functions to support regular expressions whose syntax
7	and semantics are as close as possible to those of the Perl 5 language. See
8	the file Tech.Notes for some information on the internals.
9
10	This module is a wrapper that provides a POSIX API to the underlying PCRE
11	functions.
12
13	Written by: Philip Hazel <ph10@cam.ac.uk>
14
15	Copyright (c) 1997-2003 University of Cambridge
16
17	-----------------------------------------------------------------------------
18	Permission is granted to anyone to use this software for any purpose on any
19	computer system, and to redistribute it freely, subject to the following
20	restrictions:
21
22	1. This software is distributed in the hope that it will be useful,
23	but WITHOUT ANY WARRANTY; without even the implied warranty of
24	MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
25
26	2. The origin of this software must not be misrepresented, either by
27	explicit claim or by omission.
28
29	3. Altered versions must be plainly marked as such, and must not be
30	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
38	#include "internal.h"
39	#include "pcreposix.h"
40	#include "stdlib.h"
41
42
43
44	/* Corresponding tables of PCRE error messages and POSIX error codes. */
45
46	static const char *estring[] = {
47	ERR1, ERR2, ERR3, ERR4, ERR5, ERR6, ERR7, ERR8, ERR9, ERR10,
48	ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19, ERR20,
49	ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR29, ERR29, ERR30,
50	ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39, ERR40,
51	ERR41, ERR42, ERR43 };
52
53	static int eint[] = {
54	REG_EESCAPE, /* "\\ at end of pattern" */
55	REG_EESCAPE, /* "\\c at end of pattern" */
56	REG_EESCAPE, /* "unrecognized character follows \\" */
57	REG_BADBR, /* "numbers out of order in {} quantifier" */
58	REG_BADBR, /* "number too big in {} quantifier" */
59	REG_EBRACK, /* "missing terminating ] for character class" */
60	REG_ECTYPE, /* "invalid escape sequence in character class" */
61	REG_ERANGE, /* "range out of order in character class" */
62	REG_BADRPT, /* "nothing to repeat" */
63	REG_BADRPT, /* "operand of unlimited repeat could match the empty string" */
64	REG_ASSERT, /* "internal error: unexpected repeat" */
65	REG_BADPAT, /* "unrecognized character after (?" */
66	REG_BADPAT, /* "POSIX named classes are supported only within a class" */
67	REG_EPAREN, /* "missing )" */
68	REG_ESUBREG, /* "reference to non-existent subpattern" */
69	REG_INVARG, /* "erroffset passed as NULL" */
70	REG_INVARG, /* "unknown option bit(s) set" */
71	REG_EPAREN, /* "missing ) after comment" */
72	REG_ESIZE, /* "parentheses nested too deeply" */
73	REG_ESIZE, /* "regular expression too large" */
74	REG_ESPACE, /* "failed to get memory" */
75	REG_EPAREN, /* "unmatched brackets" */
76	REG_ASSERT, /* "internal error: code overflow" */
77	REG_BADPAT, /* "unrecognized character after (?<" */
78	REG_BADPAT, /* "lookbehind assertion is not fixed length" */
79	REG_BADPAT, /* "malformed number after (?(" */
80	REG_BADPAT, /* "conditional group containe more than two branches" */
81	REG_BADPAT, /* "assertion expected after (?(" */
82	REG_BADPAT, /* "(?R or (?digits must be followed by )" */
83	REG_ECTYPE, /* "unknown POSIX class name" */
84	REG_BADPAT, /* "POSIX collating elements are not supported" */
85	REG_INVARG, /* "this version of PCRE is not compiled with PCRE_UTF8 support" */
86	REG_BADPAT, /* "spare error" */
87	REG_BADPAT, /* "character value in \x{...} sequence is too large" */
88	REG_BADPAT, /* "invalid condition (?(0)" */
89	REG_BADPAT, /* "\\C not allowed in lookbehind assertion" */
90	REG_EESCAPE, /* "PCRE does not support \\L, \\l, \\N, \\P, \\p, \\U, \\u, or \\X" */
91	REG_BADPAT, /* "number after (?C is > 255" */
92	REG_BADPAT, /* "closing ) for (?C expected" */
93	REG_BADPAT, /* "recursive call could loop indefinitely" */
94	REG_BADPAT, /* "unrecognized character after (?P" */
95	REG_BADPAT, /* "syntax error after (?P" */
96	REG_BADPAT /* "two named groups have the same name" */
97	};
98
99	/* Table of texts corresponding to POSIX error codes */
100
101	static const char *pstring[] = {
102	"", /* Dummy for value 0 */
103	"internal error", /* REG_ASSERT */
104	"invalid repeat counts in {}", /* BADBR */
105	"pattern error", /* BADPAT */
106	"? * + invalid", /* BADRPT */
107	"unbalanced {}", /* EBRACE */
108	"unbalanced []", /* EBRACK */
109	"collation error - not relevant", /* ECOLLATE */
110	"bad class", /* ECTYPE */
111	"bad escape sequence", /* EESCAPE */
112	"empty expression", /* EMPTY */
113	"unbalanced ()", /* EPAREN */
114	"bad range inside []", /* ERANGE */
115	"expression too big", /* ESIZE */
116	"failed to get memory", /* ESPACE */
117	"bad back reference", /* ESUBREG */
118	"bad argument", /* INVARG */
119	"match failed" /* NOMATCH */
120	};
121
122
123
124
125	/*************************************************
126	* Translate PCRE text code to int *
127	*************************************************/
128
129	/* PCRE compile-time errors are given as strings defined as macros. We can just
130	look them up in a table to turn them into POSIX-style error codes. */
131
132	static int
133	pcre_posix_error_code(const char *s)
134	{
135	size_t i;
136	for (i = 0; i < sizeof(estring)/sizeof(char *); i++)
137	if (strcmp(s, estring[i]) == 0) return eint[i];
138	return REG_ASSERT;
139	}
140
141
142
143	/*************************************************
144	* Translate error code to string *
145	*************************************************/
146
147	size_t
148	regerror(int errcode, const regex_t preg, char errbuf, size_t errbuf_size)
149	{
150	const char message, addmessage;
151	size_t length, addlength;
152
153	message = (errcode >= (int)(sizeof(pstring)/sizeof(char *)))?
154	"unknown error code" : pstring[errcode];
155	length = strlen(message) + 1;
156
157	addmessage = " at offset ";
158	addlength = (preg != NULL && (int)preg->re_erroffset != -1)?
159	strlen(addmessage) + 6 : 0;
160
161	if (errbuf_size > 0)
162	{
163	if (addlength > 0 && errbuf_size >= length + addlength)
164	sprintf(errbuf, "%s%s%-6d", message, addmessage, (int)preg->re_erroffset);
165	else
166	{
167	strncpy(errbuf, message, errbuf_size - 1);
168	errbuf[errbuf_size-1] = 0;
169	}
170	}
171
172	return length + addlength;
173	}
174
175
176
177
178	/*************************************************
179	* Free store held by a regex *
180	*************************************************/
181
182	void
183	regfree(regex_t *preg)
184	{
185	(pcre_free)(preg->re_pcre);
186	}
187
188
189
190
191	/*************************************************
192	* Compile a regular expression *
193	*************************************************/
194
195	/*
196	Arguments:
197	preg points to a structure for recording the compiled expression
198	pattern the pattern to compile
199	cflags compilation flags
200
201	Returns: 0 on success
202	various non-zero codes on failure
203	*/
204
205	int
206	regcomp(regex_t preg, const char pattern, int cflags)
207	{
208	const char *errorptr;
209	int erroffset;
210	int options = 0;
211
212	if ((cflags & REG_ICASE) != 0) options \|= PCRE_CASELESS;
213	if ((cflags & REG_NEWLINE) != 0) options \|= PCRE_MULTILINE;
214
215	preg->re_pcre = pcre_compile(pattern, options, &errorptr, &erroffset, NULL);
216	preg->re_erroffset = erroffset;
217
218	if (preg->re_pcre == NULL) return pcre_posix_error_code(errorptr);
219
220	preg->re_nsub = pcre_info(preg->re_pcre, NULL, NULL);
221	return 0;
222	}
223
224
225
226
227	/*************************************************
228	* Match a regular expression *
229	*************************************************/
230
231	/* Unfortunately, PCRE requires 3 ints of working space for each captured
232	substring, so we have to get and release working store instead of just using
233	the POSIX structures as was done in earlier releases when PCRE needed only 2
234	ints. However, if the number of possible capturing brackets is small, use a
235	block of store on the stack, to reduce the use of malloc/free. The threshold is
236	in a macro that can be changed at configure time. */
237
238	int
239	regexec(regex_t preg, const char string, size_t nmatch,
240	regmatch_t pmatch[], int eflags)
241	{
242	int rc;
243	int options = 0;
244	int *ovector = NULL;
245	int small_ovector[POSIX_MALLOC_THRESHOLD * 3];
246	BOOL allocated_ovector = FALSE;
247
248	if ((eflags & REG_NOTBOL) != 0) options \|= PCRE_NOTBOL;
249	if ((eflags & REG_NOTEOL) != 0) options \|= PCRE_NOTEOL;
250
251	preg->re_erroffset = (size_t)(-1); /* Only has meaning after compile */
252
253	if (nmatch > 0)
254	{
255	if (nmatch <= POSIX_MALLOC_THRESHOLD)
256	{
257	ovector = &(small_ovector[0]);
258	}
259	else
260	{
261	ovector = (int )malloc(sizeof(int) nmatch * 3);
262	if (ovector == NULL) return REG_ESPACE;
263	allocated_ovector = TRUE;
264	}
265	}
266
267	rc = pcre_exec(preg->re_pcre, NULL, string, (int)strlen(string), 0, options,
268	ovector, nmatch * 3);
269
270	if (rc == 0) rc = nmatch; /* All captured slots were filled in */
271
272	if (rc >= 0)
273	{
274	size_t i;
275	for (i = 0; i < (size_t)rc; i++)
276	{
277	pmatch[i].rm_so = ovector[i*2];
278	pmatch[i].rm_eo = ovector[i*2+1];
279	}
280	if (allocated_ovector) free(ovector);
281	for (; i < nmatch; i++) pmatch[i].rm_so = pmatch[i].rm_eo = -1;
282	return 0;
283	}
284
285	else
286	{
287	if (allocated_ovector) free(ovector);
288	switch(rc)
289	{
290	case PCRE_ERROR_NOMATCH: return REG_NOMATCH;
291	case PCRE_ERROR_NULL: return REG_INVARG;
292	case PCRE_ERROR_BADOPTION: return REG_INVARG;
293	case PCRE_ERROR_BADMAGIC: return REG_INVARG;
294	case PCRE_ERROR_UNKNOWN_NODE: return REG_ASSERT;
295	case PCRE_ERROR_NOMEMORY: return REG_ESPACE;
296	default: return REG_ASSERT;
297	}
298	}
299	}
300
301	/* End of pcreposix.c */