1: /*************************************************
2: * Perl-Compatible Regular Expressions *
3: *************************************************/
4:
5: /* PCRE is a library of functions to support regular expressions whose syntax
6: and semantics are as close as possible to those of the Perl 5 language.
7:
8: Written by Philip Hazel
9: Copyright (c) 1997-2012 University of Cambridge
10:
11: -----------------------------------------------------------------------------
12: Redistribution and use in source and binary forms, with or without
13: modification, are permitted provided that the following conditions are met:
14:
15: * Redistributions of source code must retain the above copyright notice,
16: this list of conditions and the following disclaimer.
17:
18: * Redistributions in binary form must reproduce the above copyright
19: notice, this list of conditions and the following disclaimer in the
20: documentation and/or other materials provided with the distribution.
21:
22: * Neither the name of the University of Cambridge nor the names of its
23: contributors may be used to endorse or promote products derived from
24: this software without specific prior written permission.
25:
26: THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
27: AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
28: IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
29: ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
30: LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
31: CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
32: SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
33: INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
34: CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
35: ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
36: POSSIBILITY OF SUCH DAMAGE.
37: -----------------------------------------------------------------------------
38: */
39:
40:
41: /* This module is a wrapper that provides a POSIX API to the underlying PCRE
42: functions. */
43:
44:
45: #ifdef HAVE_CONFIG_H
46: #include "config.h"
47: #endif
48:
49:
50: /* Ensure that the PCREPOSIX_EXP_xxx macros are set appropriately for
51: compiling these functions. This must come before including pcreposix.h, where
52: they are set for an application (using these functions) if they have not
53: previously been set. */
54:
55: #if defined(_WIN32) && !defined(PCRE_STATIC)
56: # define PCREPOSIX_EXP_DECL extern __declspec(dllexport)
57: # define PCREPOSIX_EXP_DEFN __declspec(dllexport)
58: #endif
59:
60: /* We include pcre.h before pcre_internal.h so that the PCRE library functions
61: are declared as "import" for Windows by defining PCRE_EXP_DECL as "import".
62: This is needed even though pcre_internal.h itself includes pcre.h, because it
63: does so after it has set PCRE_EXP_DECL to "export" if it is not already set. */
64:
65: #include "pcre.h"
66: #include "pcre_internal.h"
67: #include "pcreposix.h"
68:
69:
70: /* Table to translate PCRE compile time error codes into POSIX error codes. */
71:
72: static const int eint[] = {
73: 0, /* no error */
74: REG_EESCAPE, /* \ at end of pattern */
75: REG_EESCAPE, /* \c at end of pattern */
76: REG_EESCAPE, /* unrecognized character follows \ */
77: REG_BADBR, /* numbers out of order in {} quantifier */
78: /* 5 */
79: REG_BADBR, /* number too big in {} quantifier */
80: REG_EBRACK, /* missing terminating ] for character class */
81: REG_ECTYPE, /* invalid escape sequence in character class */
82: REG_ERANGE, /* range out of order in character class */
83: REG_BADRPT, /* nothing to repeat */
84: /* 10 */
85: REG_BADRPT, /* operand of unlimited repeat could match the empty string */
86: REG_ASSERT, /* internal error: unexpected repeat */
87: REG_BADPAT, /* unrecognized character after (? */
88: REG_BADPAT, /* POSIX named classes are supported only within a class */
89: REG_EPAREN, /* missing ) */
90: /* 15 */
91: REG_ESUBREG, /* reference to non-existent subpattern */
92: REG_INVARG, /* erroffset passed as NULL */
93: REG_INVARG, /* unknown option bit(s) set */
94: REG_EPAREN, /* missing ) after comment */
95: REG_ESIZE, /* parentheses nested too deeply */
96: /* 20 */
97: REG_ESIZE, /* regular expression too large */
98: REG_ESPACE, /* failed to get memory */
99: REG_EPAREN, /* unmatched parentheses */
100: REG_ASSERT, /* internal error: code overflow */
101: REG_BADPAT, /* unrecognized character after (?< */
102: /* 25 */
103: REG_BADPAT, /* lookbehind assertion is not fixed length */
104: REG_BADPAT, /* malformed number or name after (?( */
105: REG_BADPAT, /* conditional group contains more than two branches */
106: REG_BADPAT, /* assertion expected after (?( */
107: REG_BADPAT, /* (?R or (?[+-]digits must be followed by ) */
108: /* 30 */
109: REG_ECTYPE, /* unknown POSIX class name */
110: REG_BADPAT, /* POSIX collating elements are not supported */
111: REG_INVARG, /* this version of PCRE is not compiled with PCRE_UTF8 support */
112: REG_BADPAT, /* spare error */
113: REG_BADPAT, /* character value in \x{} or \o{} is too large */
114: /* 35 */
115: REG_BADPAT, /* invalid condition (?(0) */
116: REG_BADPAT, /* \C not allowed in lookbehind assertion */
117: REG_EESCAPE, /* PCRE does not support \L, \l, \N, \U, or \u */
118: REG_BADPAT, /* number after (?C is > 255 */
119: REG_BADPAT, /* closing ) for (?C expected */
120: /* 40 */
121: REG_BADPAT, /* recursive call could loop indefinitely */
122: REG_BADPAT, /* unrecognized character after (?P */
123: REG_BADPAT, /* syntax error in subpattern name (missing terminator) */
124: REG_BADPAT, /* two named subpatterns have the same name */
125: REG_BADPAT, /* invalid UTF-8 string */
126: /* 45 */
127: REG_BADPAT, /* support for \P, \p, and \X has not been compiled */
128: REG_BADPAT, /* malformed \P or \p sequence */
129: REG_BADPAT, /* unknown property name after \P or \p */
130: REG_BADPAT, /* subpattern name is too long (maximum 32 characters) */
131: REG_BADPAT, /* too many named subpatterns (maximum 10,000) */
132: /* 50 */
133: REG_BADPAT, /* repeated subpattern is too long */
134: REG_BADPAT, /* octal value is greater than \377 (not in UTF-8 mode) */
135: REG_BADPAT, /* internal error: overran compiling workspace */
136: REG_BADPAT, /* internal error: previously-checked referenced subpattern not found */
137: REG_BADPAT, /* DEFINE group contains more than one branch */
138: /* 55 */
139: REG_BADPAT, /* repeating a DEFINE group is not allowed */
140: REG_INVARG, /* inconsistent NEWLINE options */
141: REG_BADPAT, /* \g is not followed followed by an (optionally braced) non-zero number */
142: REG_BADPAT, /* a numbered reference must not be zero */
143: REG_BADPAT, /* an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT) */
144: /* 60 */
145: REG_BADPAT, /* (*VERB) not recognized */
146: REG_BADPAT, /* number is too big */
147: REG_BADPAT, /* subpattern name expected */
148: REG_BADPAT, /* digit expected after (?+ */
149: REG_BADPAT, /* ] is an invalid data character in JavaScript compatibility mode */
150: /* 65 */
151: REG_BADPAT, /* different names for subpatterns of the same number are not allowed */
152: REG_BADPAT, /* (*MARK) must have an argument */
153: REG_INVARG, /* this version of PCRE is not compiled with PCRE_UCP support */
154: REG_BADPAT, /* \c must be followed by an ASCII character */
155: REG_BADPAT, /* \k is not followed by a braced, angle-bracketed, or quoted name */
156: /* 70 */
157: REG_BADPAT, /* internal error: unknown opcode in find_fixedlength() */
158: REG_BADPAT, /* \N is not supported in a class */
159: REG_BADPAT, /* too many forward references */
160: REG_BADPAT, /* disallowed UTF-8/16/32 code point (>= 0xd800 && <= 0xdfff) */
161: REG_BADPAT, /* invalid UTF-16 string (should not occur) */
162: /* 75 */
163: REG_BADPAT, /* overlong MARK name */
164: REG_BADPAT, /* character value in \u.... sequence is too large */
165: REG_BADPAT, /* invalid UTF-32 string (should not occur) */
166: REG_BADPAT, /* setting UTF is disabled by the application */
167: REG_BADPAT, /* non-hex character in \\x{} (closing brace missing?) */
168: /* 80 */
169: REG_BADPAT, /* non-octal character in \o{} (closing brace missing?) */
170: REG_BADPAT, /* missing opening brace after \o */
171: REG_BADPAT, /* parentheses too deeply nested */
172: REG_BADPAT, /* invalid range in character class */
173: REG_BADPAT /* group name must start with a non-digit */
174: };
175:
176: /* Table of texts corresponding to POSIX error codes */
177:
178: static const char *const pstring[] = {
179: "", /* Dummy for value 0 */
180: "internal error", /* REG_ASSERT */
181: "invalid repeat counts in {}", /* BADBR */
182: "pattern error", /* BADPAT */
183: "? * + invalid", /* BADRPT */
184: "unbalanced {}", /* EBRACE */
185: "unbalanced []", /* EBRACK */
186: "collation error - not relevant", /* ECOLLATE */
187: "bad class", /* ECTYPE */
188: "bad escape sequence", /* EESCAPE */
189: "empty expression", /* EMPTY */
190: "unbalanced ()", /* EPAREN */
191: "bad range inside []", /* ERANGE */
192: "expression too big", /* ESIZE */
193: "failed to get memory", /* ESPACE */
194: "bad back reference", /* ESUBREG */
195: "bad argument", /* INVARG */
196: "match failed" /* NOMATCH */
197: };
198:
199:
200:
201:
202: /*************************************************
203: * Translate error code to string *
204: *************************************************/
205:
206: PCREPOSIX_EXP_DEFN size_t PCRE_CALL_CONVENTION
207: regerror(int errcode, const regex_t *preg, char *errbuf, size_t errbuf_size)
208: {
209: const char *message, *addmessage;
210: size_t length, addlength;
211:
212: message = (errcode >= (int)(sizeof(pstring)/sizeof(char *)))?
213: "unknown error code" : pstring[errcode];
214: length = strlen(message) + 1;
215:
216: addmessage = " at offset ";
217: addlength = (preg != NULL && (int)preg->re_erroffset != -1)?
218: strlen(addmessage) + 6 : 0;
219:
220: if (errbuf_size > 0)
221: {
222: if (addlength > 0 && errbuf_size >= length + addlength)
223: sprintf(errbuf, "%s%s%-6d", message, addmessage, (int)preg->re_erroffset);
224: else
225: {
226: strncpy(errbuf, message, errbuf_size - 1);
227: errbuf[errbuf_size-1] = 0;
228: }
229: }
230:
231: return length + addlength;
232: }
233:
234:
235:
236:
237: /*************************************************
238: * Free store held by a regex *
239: *************************************************/
240:
241: PCREPOSIX_EXP_DEFN void PCRE_CALL_CONVENTION
242: regfree(regex_t *preg)
243: {
244: (PUBL(free))(preg->re_pcre);
245: }
246:
247:
248:
249:
250: /*************************************************
251: * Compile a regular expression *
252: *************************************************/
253:
254: /*
255: Arguments:
256: preg points to a structure for recording the compiled expression
257: pattern the pattern to compile
258: cflags compilation flags
259:
260: Returns: 0 on success
261: various non-zero codes on failure
262: */
263:
264: PCREPOSIX_EXP_DEFN int PCRE_CALL_CONVENTION
265: regcomp(regex_t *preg, const char *pattern, int cflags)
266: {
267: const char *errorptr;
268: int erroffset;
269: int errorcode;
270: int options = 0;
271: int re_nsub = 0;
272:
273: if ((cflags & REG_ICASE) != 0) options |= PCRE_CASELESS;
274: if ((cflags & REG_NEWLINE) != 0) options |= PCRE_MULTILINE;
275: if ((cflags & REG_DOTALL) != 0) options |= PCRE_DOTALL;
276: if ((cflags & REG_NOSUB) != 0) options |= PCRE_NO_AUTO_CAPTURE;
277: if ((cflags & REG_UTF8) != 0) options |= PCRE_UTF8;
278: if ((cflags & REG_UCP) != 0) options |= PCRE_UCP;
279: if ((cflags & REG_UNGREEDY) != 0) options |= PCRE_UNGREEDY;
280:
281: preg->re_pcre = pcre_compile2(pattern, options, &errorcode, &errorptr,
282: &erroffset, NULL);
283: preg->re_erroffset = erroffset;
284:
285: /* Safety: if the error code is too big for the translation vector (which
286: should not happen, but we all make mistakes), return REG_BADPAT. */
287:
288: if (preg->re_pcre == NULL)
289: {
290: return (errorcode < (int)(sizeof(eint)/sizeof(const int)))?
291: eint[errorcode] : REG_BADPAT;
292: }
293:
294: (void)pcre_fullinfo((const pcre *)preg->re_pcre, NULL, PCRE_INFO_CAPTURECOUNT,
295: &re_nsub);
296: preg->re_nsub = (size_t)re_nsub;
297: return 0;
298: }
299:
300:
301:
302:
303: /*************************************************
304: * Match a regular expression *
305: *************************************************/
306:
307: /* Unfortunately, PCRE requires 3 ints of working space for each captured
308: substring, so we have to get and release working store instead of just using
309: the POSIX structures as was done in earlier releases when PCRE needed only 2
310: ints. However, if the number of possible capturing brackets is small, use a
311: block of store on the stack, to reduce the use of malloc/free. The threshold is
312: in a macro that can be changed at configure time.
313:
314: If REG_NOSUB was specified at compile time, the PCRE_NO_AUTO_CAPTURE flag will
315: be set. When this is the case, the nmatch and pmatch arguments are ignored, and
316: the only result is yes/no/error. */
317:
318: PCREPOSIX_EXP_DEFN int PCRE_CALL_CONVENTION
319: regexec(const regex_t *preg, const char *string, size_t nmatch,
320: regmatch_t pmatch[], int eflags)
321: {
322: int rc, so, eo;
323: int options = 0;
324: int *ovector = NULL;
325: int small_ovector[POSIX_MALLOC_THRESHOLD * 3];
326: BOOL allocated_ovector = FALSE;
327: BOOL nosub =
328: (REAL_PCRE_OPTIONS((const pcre *)preg->re_pcre) & PCRE_NO_AUTO_CAPTURE) != 0;
329:
330: if ((eflags & REG_NOTBOL) != 0) options |= PCRE_NOTBOL;
331: if ((eflags & REG_NOTEOL) != 0) options |= PCRE_NOTEOL;
332: if ((eflags & REG_NOTEMPTY) != 0) options |= PCRE_NOTEMPTY;
333:
334: ((regex_t *)preg)->re_erroffset = (size_t)(-1); /* Only has meaning after compile */
335:
336: /* When no string data is being returned, or no vector has been passed in which
337: to put it, ensure that nmatch is zero. Otherwise, ensure the vector for holding
338: the return data is large enough. */
339:
340: if (nosub || pmatch == NULL) nmatch = 0;
341:
342: else if (nmatch > 0)
343: {
344: if (nmatch <= POSIX_MALLOC_THRESHOLD)
345: {
346: ovector = &(small_ovector[0]);
347: }
348: else
349: {
350: if (nmatch > INT_MAX/(sizeof(int) * 3)) return REG_ESPACE;
351: ovector = (int *)malloc(sizeof(int) * nmatch * 3);
352: if (ovector == NULL) return REG_ESPACE;
353: allocated_ovector = TRUE;
354: }
355: }
356:
357: /* REG_STARTEND is a BSD extension, to allow for non-NUL-terminated strings.
358: The man page from OS X says "REG_STARTEND affects only the location of the
359: string, not how it is matched". That is why the "so" value is used to bump the
360: start location rather than being passed as a PCRE "starting offset". */
361:
362: if ((eflags & REG_STARTEND) != 0)
363: {
364: so = pmatch[0].rm_so;
365: eo = pmatch[0].rm_eo;
366: }
367: else
368: {
369: so = 0;
370: eo = (int)strlen(string);
371: }
372:
373: rc = pcre_exec((const pcre *)preg->re_pcre, NULL, string + so, (eo - so),
374: 0, options, ovector, (int)(nmatch * 3));
375:
376: if (rc == 0) rc = (int)nmatch; /* All captured slots were filled in */
377:
378: /* Successful match */
379:
380: if (rc >= 0)
381: {
382: size_t i;
383: if (!nosub)
384: {
385: for (i = 0; i < (size_t)rc; i++)
386: {
387: pmatch[i].rm_so = ovector[i*2];
388: pmatch[i].rm_eo = ovector[i*2+1];
389: }
390: if (allocated_ovector) free(ovector);
391: for (; i < nmatch; i++) pmatch[i].rm_so = pmatch[i].rm_eo = -1;
392: }
393: return 0;
394: }
395:
396: /* Unsuccessful match */
397:
398: if (allocated_ovector) free(ovector);
399: switch(rc)
400: {
401: /* ========================================================================== */
402: /* These cases are never obeyed. This is a fudge that causes a compile-time
403: error if the vector eint, which is indexed by compile-time error number, is
404: not the correct length. It seems to be the only way to do such a check at
405: compile time, as the sizeof() operator does not work in the C preprocessor.
406: As all the PCRE_ERROR_xxx values are negative, we can use 0 and 1. */
407:
408: case 0:
409: case (sizeof(eint)/sizeof(int) == ERRCOUNT):
410: return REG_ASSERT;
411: /* ========================================================================== */
412:
413: case PCRE_ERROR_NOMATCH: return REG_NOMATCH;
414: case PCRE_ERROR_NULL: return REG_INVARG;
415: case PCRE_ERROR_BADOPTION: return REG_INVARG;
416: case PCRE_ERROR_BADMAGIC: return REG_INVARG;
417: case PCRE_ERROR_UNKNOWN_NODE: return REG_ASSERT;
418: case PCRE_ERROR_NOMEMORY: return REG_ESPACE;
419: case PCRE_ERROR_MATCHLIMIT: return REG_ESPACE;
420: case PCRE_ERROR_BADUTF8: return REG_INVARG;
421: case PCRE_ERROR_BADUTF8_OFFSET: return REG_INVARG;
422: case PCRE_ERROR_BADMODE: return REG_INVARG;
423: default: return REG_ASSERT;
424: }
425: }
426:
427: /* End of pcreposix.c */
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>
![[BACK]](/icons/cvsweb/back.gif){kind=icon}
Return to pcreposix.c CVS log ![[TXT]](/icons/cvsweb/text.gif){kind=icon}
![[DIR]](/icons/cvsweb/dir.gif){kind=icon}
Up to [ELWIX - Embedded LightWeight unIX -] / embedaddon / pcre