* Exim - an Internet mail transport agent *
*************************************************/
-/* Copyright (c) Tom Kistner <tom@duncanthrax.net> 2003-2015
+/*
+ * Copyright (c) The Exim Maintainers 2016 - 2023
+ * Copyright (c) Tom Kistner <tom@duncanthrax.net> 2003-2015
* License: GPL
- * Copyright (c) The Exim Maintainers 2016 - 2018
+ * SPDX-License-Identifier: GPL-2.0-or-later
*/
/* Code for matching regular expressions against headers and body.
/* Structure to hold a list of Regular expressions */
typedef struct pcre_list {
- pcre2_code *re;
- uschar *pcre_text;
- struct pcre_list *next;
+ const pcre2_code * re;
+ uschar * pcre_text;
+ struct pcre_list * next;
} pcre_list;
uschar regex_match_string_buffer[1024];
extern FILE *mime_stream;
extern uschar *mime_current_boundary;
+
static pcre_list *
-compile(const uschar * list)
+compile(const uschar * list, BOOL cacheable, int * cntp)
{
-int sep = 0;
-uschar *regex_string;
-pcre_list *re_list_head = NULL;
-pcre_list *ri;
+int sep = 0, cnt = 0;
+uschar * regex_string;
+pcre_list * re_list_head = NULL, * ri;
/* precompile our regexes */
while ((regex_string = string_nextinlist(&list, &sep, NULL, 0)))
if (strcmpic(regex_string, US"false") != 0 && Ustrcmp(regex_string, "0") != 0)
{
- pcre2_code * re;
- int err;
- PCRE2_SIZE pcre_erroffset;
-
/* compile our regular expression */
- if (!(re = pcre2_compile( (PCRE2_SPTR) regex_string, PCRE2_ZERO_TERMINATED,
- 0, &err, &pcre_erroffset, pcre_cmp_ctx)))
+ uschar * errstr;
+ const pcre2_code * re = regex_compile(regex_string,
+ cacheable ? MCS_CACHEABLE : MCS_NOFLAGS, &errstr, pcre_gen_cmp_ctx);
+
+ if (!re)
{
- uschar errbuf[128];
- pcre2_get_error_message(err, errbuf, sizeof(errbuf));
- log_write(0, LOG_MAIN,
- "regex acl condition warning - error in regex '%s': %s at offset %ld, skipped.",
- regex_string, errbuf, (long)pcre_erroffset);
+ log_write(0, LOG_MAIN, "regex acl condition warning - %s, skipped", errstr);
continue;
}
- ri = store_get(sizeof(pcre_list), FALSE);
+ ri = store_get(sizeof(pcre_list), GET_UNTAINTED);
ri->re = re;
ri->pcre_text = regex_string;
ri->next = re_list_head;
re_list_head = ri;
+ cnt++;
}
+if (cntp) *cntp = cnt;
return re_list_head;
}
+
+/* Check list of REs against buffer, returning OK for (first) match,
+else FAIL. On match return allocated result strings in regex_vars[].
+
+We use the perm-pool for that, so that our caller can release
+other allocations.
+*/
static int
matcher(pcre_list * re_list_head, uschar * linebuffer, int len)
{
int n;
/* try matcher on the line */
- if ((n = pcre2_match(ri->re, (PCRE2_SPTR)linebuffer, len, 0, 0, md, pcre_mtc_ctx)) > 0)
+ if ((n = pcre2_match(ri->re, (PCRE2_SPTR)linebuffer, len, 0, 0, md, pcre_gen_mtc_ctx)) > 0)
{
+ int save_pool = store_pool;
+ store_pool = POOL_PERM;
+
Ustrncpy(regex_match_string_buffer, ri->pcre_text,
sizeof(regex_match_string_buffer)-1);
regex_match_string = regex_match_string_buffer;
for (int nn = 1; nn < n; nn++)
{
- PCRE2_UCHAR * cstr;
- PCRE2_SIZE cslen;
- pcre2_substring_get_bynumber(md, nn, &cstr, &cslen);
- regex_vars[nn-1] = CUS cstr;
+ PCRE2_SIZE * ovec = pcre2_get_ovector_pointer(md);
+ int off = nn * 2;
+ int len = ovec[off + 1] - ovec[off];
+ regex_vars[nn-1] = string_copyn(linebuffer + ovec[off], len);
}
+ store_pool = save_pool;
return OK;
}
}
-pcre2_match_data_free(md);
+/* pcre2_match_data_free(md); gen ctx needs no free */
return FAIL;
}
+
+/* reset expansion variables */
+void
+regex_vars_clear(void)
+{
+regex_match_string = NULL;
+for (int i = 0; i < REGEX_VARS; i++) regex_vars[i] = NULL;
+}
+
+
+
int
-regex(const uschar **listptr)
+regex(const uschar ** listptr, BOOL cacheable)
{
unsigned long mbox_size;
-FILE *mbox_file;
-pcre_list *re_list_head;
-uschar *linebuffer;
+FILE * mbox_file;
+pcre_list * re_list_head;
+uschar * linebuffer;
long f_pos = 0;
-int ret = FAIL;
+int ret = FAIL, cnt, lcount = REGEX_LOOPCOUNT_STORE_RESET;
+rmark reset_point;
-/* reset expansion variable */
-regex_match_string = NULL;
+regex_vars_clear();
if (!mime_stream) /* We are in the DATA ACL */
{
mbox_file = mime_stream;
}
-/* precompile our regexes */
-if (!(re_list_head = compile(*listptr)))
- return FAIL; /* no regexes -> nothing to do */
-
-/* match each line against all regexes */
-linebuffer = store_get(32767, TRUE); /* tainted */
-while (fgets(CS linebuffer, 32767, mbox_file))
+reset_point = store_mark();
{
- if ( mime_stream && mime_current_boundary /* check boundary */
- && Ustrncmp(linebuffer, "--", 2) == 0
- && Ustrncmp((linebuffer+2), mime_current_boundary,
- Ustrlen(mime_current_boundary)) == 0)
- break; /* found boundary */
-
- if ((ret = matcher(re_list_head, linebuffer, (int)Ustrlen(linebuffer))) == OK)
- goto done;
+ /* precompile our regexes */
+ if ((re_list_head = compile(*listptr, cacheable, &cnt)))
+ {
+ /* match each line against all regexes */
+ linebuffer = store_get(32767, GET_TAINTED);
+ while (fgets(CS linebuffer, 32767, mbox_file))
+ {
+ if ( mime_stream && mime_current_boundary /* check boundary */
+ && Ustrncmp(linebuffer, "--", 2) == 0
+ && Ustrncmp((linebuffer+2), mime_current_boundary,
+ Ustrlen(mime_current_boundary)) == 0)
+ break; /* found boundary */
+
+ if ((ret = matcher(re_list_head, linebuffer, (int)Ustrlen(linebuffer))) == OK)
+ break;
+
+ if ((lcount -= cnt) <= 0)
+ {
+ store_reset(reset_point); reset_point = store_mark();
+ lcount = REGEX_LOOPCOUNT_STORE_RESET;
+ }
+ }
+ }
}
-/* no matches ... */
+store_reset(reset_point);
-done:
if (!mime_stream)
(void)fclose(mbox_file);
else
int
-mime_regex(const uschar **listptr)
+mime_regex(const uschar **listptr, BOOL cacheable)
{
-pcre_list *re_list_head = NULL;
-FILE *f;
-uschar *mime_subject = NULL;
+pcre_list * re_list_head = NULL;
+FILE * f;
+uschar * mime_subject = NULL;
int mime_subject_len = 0;
-int ret;
-
-/* reset expansion variable */
-regex_match_string = NULL;
+int ret = FAIL;
+rmark reset_point;
-/* precompile our regexes */
-if (!(re_list_head = compile(*listptr)))
- return FAIL; /* no regexes -> nothing to do */
+regex_vars_clear();
/* check if the file is already decoded */
if (!mime_decoded_filename)
return DEFER;
}
-/* get 32k memory, tainted */
-mime_subject = store_get(32767, TRUE);
+reset_point = store_mark();
+ {
+ /* precompile our regexes */
+ if ((re_list_head = compile(*listptr, cacheable, NULL)))
+ {
+ /* get 32k memory, tainted */
+ mime_subject = store_get(32767, GET_TAINTED);
-mime_subject_len = fread(mime_subject, 1, 32766, f);
+ mime_subject_len = fread(mime_subject, 1, 32766, f);
-ret = matcher(re_list_head, mime_subject, mime_subject_len);
+ ret = matcher(re_list_head, mime_subject, mime_subject_len);
+ }
+ }
+store_reset(reset_point);
(void)fclose(f);
return ret;
}
git://git.exim.org / exim.git / blobdiff