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

Contents of /code/trunk/pcreposix.c

Parent Directory Parent Directory | Revision Log Revision Log


Revision 15 - (hide annotations) (download)
Sat Feb 24 21:38:25 2007 UTC (7 years, 1 month ago) by nigel
File MIME type: text/plain
File size: 7563 byte(s)
Load pcre-1.06 into code/trunk.

1 nigel 3 /*************************************************
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 nigel 15 Copyright (c) 1998 University of Cambridge
16 nigel 3
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     */
33    
34     #include "internal.h"
35     #include "pcreposix.h"
36     #include "stdlib.h"
37    
38    
39    
40     /* Corresponding tables of PCRE error messages and POSIX error codes. */
41    
42 nigel 7 static const char *estring[] = {
43 nigel 3 ERR1, ERR2, ERR3, ERR4, ERR5, ERR6, ERR7, ERR8, ERR9, ERR10,
44     ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19, ERR20,
45     ERR21, ERR22, ERR23 };
46    
47     static int eint[] = {
48     REG_EESCAPE, /* "\\ at end of pattern" */
49     REG_EESCAPE, /* "\\c at end of pattern" */
50     REG_EESCAPE, /* "unrecognized character follows \\" */
51     REG_BADBR, /* "numbers out of order in {} quantifier" */
52     REG_BADBR, /* "number too big in {} quantifier" */
53     REG_EBRACK, /* "missing terminating ] for character class" */
54     REG_ECTYPE, /* "invalid escape sequence in character class" */
55     REG_ERANGE, /* "range out of order in character class" */
56     REG_BADRPT, /* "nothing to repeat" */
57     REG_BADRPT, /* "operand of unlimited repeat could match the empty string" */
58     REG_ASSERT, /* "internal error: unexpected repeat" */
59     REG_BADPAT, /* "unrecognized character after (?" */
60     REG_ESIZE, /* "too many capturing parenthesized sub-patterns" */
61     REG_EPAREN, /* "missing )" */
62     REG_ESUBREG, /* "back reference to non-existent subpattern" */
63     REG_INVARG, /* "erroffset passed as NULL" */
64     REG_INVARG, /* "unknown option bit(s) set" */
65     REG_EPAREN, /* "missing ) after comment" */
66     REG_ESIZE, /* "too many sets of parentheses" */
67     REG_ESIZE, /* "regular expression too large" */
68     REG_ESPACE, /* "failed to get memory" */
69     REG_EPAREN, /* "unmatched brackets" */
70     REG_ASSERT /* "internal error: code overflow" */
71     };
72    
73     /* Table of texts corresponding to POSIX error codes */
74    
75 nigel 7 static const char *pstring[] = {
76 nigel 3 "", /* Dummy for value 0 */
77     "internal error", /* REG_ASSERT */
78     "invalid repeat counts in {}", /* BADBR */
79     "pattern error", /* BADPAT */
80     "? * + invalid", /* BADRPT */
81     "unbalanced {}", /* EBRACE */
82     "unbalanced []", /* EBRACK */
83     "collation error - not relevant", /* ECOLLATE */
84     "bad class", /* ECTYPE */
85     "bad escape sequence", /* EESCAPE */
86     "empty expression", /* EMPTY */
87     "unbalanced ()", /* EPAREN */
88     "bad range inside []", /* ERANGE */
89     "expression too big", /* ESIZE */
90     "failed to get memory", /* ESPACE */
91     "bad back reference", /* ESUBREG */
92     "bad argument", /* INVARG */
93     "match failed" /* NOMATCH */
94     };
95    
96    
97    
98    
99     /*************************************************
100     * Translate PCRE text code to int *
101     *************************************************/
102    
103     /* PCRE compile-time errors are given as strings defined as macros. We can just
104     look them up in a table to turn them into POSIX-style error codes. */
105    
106     static int
107     pcre_posix_error_code(const char *s)
108     {
109 nigel 7 size_t i;
110 nigel 3 for (i = 0; i < sizeof(estring)/sizeof(char *); i++)
111     if (strcmp(s, estring[i]) == 0) return eint[i];
112     return REG_ASSERT;
113     }
114    
115    
116    
117     /*************************************************
118     * Translate error code to string *
119     *************************************************/
120    
121     size_t
122     regerror(int errcode, const regex_t *preg, char *errbuf, size_t errbuf_size)
123     {
124 nigel 7 const char *message, *addmessage;
125     size_t length, addlength;
126 nigel 3
127 nigel 7 message = (errcode >= (int)(sizeof(pstring)/sizeof(char *)))?
128 nigel 3 "unknown error code" : pstring[errcode];
129 nigel 7 length = strlen(message) + 1;
130 nigel 3
131 nigel 7 addmessage = " at offset ";
132     addlength = (preg != NULL && (int)preg->re_erroffset != -1)?
133     strlen(addmessage) + 6 : 0;
134 nigel 3
135     if (errbuf_size > 0)
136     {
137 nigel 7 if (addlength > 0 && errbuf_size >= length + addlength)
138 nigel 3 sprintf(errbuf, "%s%s%-6d", message, addmessage, preg->re_erroffset);
139     else
140     {
141     strncpy(errbuf, message, errbuf_size - 1);
142     errbuf[errbuf_size-1] = 0;
143     }
144     }
145    
146 nigel 7 return length + addlength;
147 nigel 3 }
148    
149    
150    
151    
152     /*************************************************
153     * Free store held by a regex *
154     *************************************************/
155    
156     void
157     regfree(regex_t *preg)
158     {
159     (pcre_free)(preg->re_pcre);
160     }
161    
162    
163    
164    
165     /*************************************************
166     * Compile a regular expression *
167     *************************************************/
168    
169     /*
170     Arguments:
171     preg points to a structure for recording the compiled expression
172     pattern the pattern to compile
173     cflags compilation flags
174    
175     Returns: 0 on success
176     various non-zero codes on failure
177     */
178    
179     int
180     regcomp(regex_t *preg, const char *pattern, int cflags)
181     {
182 nigel 7 const char *errorptr;
183 nigel 3 int erroffset;
184     int options = 0;
185    
186     if ((cflags & REG_ICASE) != 0) options |= PCRE_CASELESS;
187     if ((cflags & REG_NEWLINE) != 0) options |= PCRE_MULTILINE;
188    
189     preg->re_pcre = pcre_compile(pattern, options, &errorptr, &erroffset);
190     preg->re_erroffset = erroffset;
191    
192     if (preg->re_pcre == NULL) return pcre_posix_error_code(errorptr);
193    
194     preg->re_nsub = pcre_info(preg->re_pcre, NULL, NULL);
195     return 0;
196     }
197    
198    
199    
200    
201     /*************************************************
202     * Match a regular expression *
203     *************************************************/
204    
205     int
206     regexec(regex_t *preg, const char *string, size_t nmatch,
207     regmatch_t pmatch[], int eflags)
208     {
209     int rc;
210     int options = 0;
211    
212     if ((eflags & REG_NOTBOL) != 0) options |= PCRE_NOTBOL;
213     if ((eflags & REG_NOTEOL) != 0) options |= PCRE_NOTEOL;
214    
215 nigel 15 preg->re_erroffset = (size_t)(-1); /* Only has meaning after compile */
216 nigel 3
217     rc = pcre_exec(preg->re_pcre, NULL, string, (int)strlen(string), options,
218     (int *)pmatch, nmatch * 2);
219    
220     if (rc == 0) return 0; /* All pmatch were filled in */
221    
222     if (rc > 0)
223     {
224 nigel 7 size_t i;
225 nigel 3 for (i = rc; i < nmatch; i++) pmatch[i].rm_so = pmatch[i].rm_eo = -1;
226     return 0;
227     }
228    
229     else switch(rc)
230     {
231     case PCRE_ERROR_NOMATCH: return REG_NOMATCH;
232     case PCRE_ERROR_BADREF: return REG_ESUBREG;
233     case PCRE_ERROR_NULL: return REG_INVARG;
234     case PCRE_ERROR_BADOPTION: return REG_INVARG;
235     case PCRE_ERROR_BADMAGIC: return REG_INVARG;
236     case PCRE_ERROR_UNKNOWN_NODE: return REG_ASSERT;
237     case PCRE_ERROR_NOMEMORY: return REG_ESPACE;
238     default: return REG_ASSERT;
239     }
240     }
241    
242     /* End of pcreposix.c */

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12