1 /* Licensed to the Apache Software Foundation (ASF) under one or more
2 * contributor license agreements. See the NOTICE file distributed with
3 * this work for additional information regarding copyright ownership.
4 * The ASF licenses this file to You under the Apache License, Version 2.0
5 * (the "License"); you may not use this file except in compliance with
6 * the License. You may obtain a copy of the License at
8 * http://www.apache.org/licenses/LICENSE-2.0
10 * Unless required by applicable law or agreed to in writing, software
11 * distributed under the License is distributed on an "AS IS" BASIS,
12 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13 * See the License for the specific language governing permissions and
14 * limitations under the License.
19 #include "apr_pools.h"
20 #include "apr_strings.h"
21 #include "ap_config.h"
25 static apr_status_t rxplus_cleanup(void *preg)
27 ap_regfree((ap_regex_t *) preg);
31 AP_DECLARE(ap_rxplus_t*) ap_rxplus_compile(apr_pool_t *pool,
34 /* perl style patterns
35 * add support for more as and when wanted
36 * substitute: s/rx/subs/
37 * match: m/rx/ or just /rx/
40 /* allow any nonalnum delimiter as first or second char.
41 * If we ever use this with non-string pattern we'll need an extra check
44 const char *str = pattern;
46 ap_rxplus_t *ret = apr_pcalloc(pool, sizeof(ap_rxplus_t));
48 enum { SUBSTITUTE = 's', MATCH = 'm'} action = MATCH;
50 if (!apr_isalnum(pattern[0])) {
53 else if (pattern[0] == 's' && !apr_isalnum(pattern[1])) {
58 else if (pattern[0] == 'm' && !apr_isalnum(pattern[1])) {
62 /* TODO: support perl's after/before */
63 /* FIXME: fix these simplminded delims */
65 /* we think there's a delimiter. Allow for it not to be if unmatched */
67 endp = ap_strchr_c(str, delim);
69 if (!endp) { /* there's no delim or flags */
70 if (ap_regcomp(&ret->rx, pattern, 0) == 0) {
71 apr_pool_cleanup_register(pool, &ret->rx, rxplus_cleanup,
72 apr_pool_cleanup_null);
80 /* We have a delimiter. Use it to extract the regexp */
81 rxstr = apr_pstrmemdup(pool, str, endp-str);
83 /* If it's a substitution, we need the replacement string
84 * TODO: possible future enhancement - support other parsing
85 * in the replacement string.
87 if (action == SUBSTITUTE) {
89 if (!*str || (endp = ap_strchr_c(str, delim), !endp)) {
90 /* missing replacement string is an error */
93 ret->subs = apr_pstrmemdup(pool, str, endp-str);
96 /* anything after the current delimiter is flags */
99 case 'i': ret->flags |= AP_REG_ICASE; break;
100 case 'm': ret->flags |= AP_REG_NEWLINE; break;
101 case 'n': ret->flags |= AP_REG_NOMEM; break;
102 case 'g': ret->flags |= AP_REG_MULTI; break;
103 case 's': ret->flags |= AP_REG_DOTALL; break;
104 case '^': ret->flags |= AP_REG_NOTBOL; break;
105 case '$': ret->flags |= AP_REG_NOTEOL; break;
106 default: break; /* we should probably be stricter here */
109 if (ap_regcomp(&ret->rx, rxstr, ret->flags) == 0) {
110 apr_pool_cleanup_register(pool, &ret->rx, rxplus_cleanup,
111 apr_pool_cleanup_null);
116 if (!(ret->flags & AP_REG_NOMEM)) {
117 /* count size of memory required, starting at 1 for the whole-match
118 * Simpleminded should be fine 'cos regcomp already checked syntax
123 case '\\': /* next char is escaped - skip it */
128 case '(': /* unescaped bracket implies memory */
135 ret->pmatch = apr_palloc(pool, ret->nmatch*sizeof(ap_regmatch_t));
140 AP_DECLARE(int) ap_rxplus_exec(apr_pool_t *pool, ap_rxplus_t *rx,
141 const char *pattern, char **newpattern)
144 int startl, oldl, newl, diffsz;
145 const char *remainder;
147 /* snrf process_regexp from mod_headers */
148 if (ap_regexec(&rx->rx, pattern, rx->nmatch, rx->pmatch, rx->flags) != 0) {
150 return 0; /* no match, nothing to do */
154 *newpattern = ap_pregsub(pool, rx->subs, pattern,
155 rx->nmatch, rx->pmatch);
157 return 0; /* FIXME - should we do more to handle error? */
159 startl = rx->pmatch[0].rm_so;
160 oldl = rx->pmatch[0].rm_eo - startl;
161 newl = strlen(*newpattern);
162 diffsz = newl - oldl;
163 remainder = pattern + startl + oldl;
164 if (rx->flags & AP_REG_MULTI) {
165 /* recurse to do any further matches */
166 ret += ap_rxplus_exec(pool, rx, remainder, &subs);
168 /* a further substitution happened */
169 diffsz += strlen(subs) - strlen(remainder);
173 subs = apr_palloc(pool, strlen(pattern) + 1 + diffsz);
174 memcpy(subs, pattern, startl);
175 memcpy(subs+startl, *newpattern, newl);
176 strcpy(subs+startl+newl, remainder);
182 AP_DECLARE(int) ap_rxplus_nmatch(ap_rxplus_t *rx)
184 return (rx->match != NULL) ? rx->nmatch : 0;
188 /* If this blows up on you, see the notes in the header/apidoc
189 * rx->match is a pointer and it's your responsibility to ensure
190 * it hasn't gone out-of-scope since the last ap_rxplus_exec
192 AP_DECLARE(void) ap_rxplus_match(ap_rxplus_t *rx, int n, int *len,
195 if (n >= 0 && n < ap_rxplus_nmatch(rx)) {
196 *match = rx->match + rx->pmatch[n].rm_so;
197 *len = rx->pmatch[n].rm_eo - rx->pmatch[n].rm_so;
204 AP_DECLARE(char*) ap_rxplus_pmatch(apr_pool_t *pool, ap_rxplus_t *rx, int n)
208 ap_rxplus_match(rx, n, &len, &match);
209 return apr_pstrndup(pool, match, len);