-const char urlmatch_rcs[] = "$Id: urlmatch.c,v 1.25 2008/04/06 15:18:38 fabiankeil Exp $";
+const char urlmatch_rcs[] = "$Id: urlmatch.c,v 1.33 2008/04/12 14:03:13 fabiankeil Exp $";
/*********************************************************************
*
* File : $Source: /cvsroot/ijbswa/current/urlmatch.c,v $
*
* Revisions :
* $Log: urlmatch.c,v $
+ * Revision 1.33 2008/04/12 14:03:13 fabiankeil
+ * Remove an obvious comment and improve another one.
+ *
+ * Revision 1.32 2008/04/12 12:38:06 fabiankeil
+ * Factor out duplicated code to compile host, path and tag patterns.
+ *
+ * Revision 1.31 2008/04/10 14:41:04 fabiankeil
+ * Ditch url_spec's path member now that it's no longer used.
+ *
+ * Revision 1.30 2008/04/10 04:24:24 fabiankeil
+ * Stop duplicating the plain text representation of the path regex
+ * (and keeping the copy around). Once the regex is compiled it's no
+ * longer useful.
+ *
+ * Revision 1.29 2008/04/10 04:17:56 fabiankeil
+ * In url_match(), check the right member for NULL when determining
+ * whether there's a path regex to execute. Looking for a plain-text
+ * representation works as well, but it looks "interesting" and that
+ * member will be removed soonish anyway.
+ *
+ * Revision 1.28 2008/04/08 16:07:39 fabiankeil
+ * Make it harder to mistake url_match()'s
+ * second parameter for an url_spec.
+ *
+ * Revision 1.27 2008/04/08 15:44:33 fabiankeil
+ * Save a bit of memory (and a few cpu cycles) by not bothering to
+ * compile slash-only path regexes that don't affect the result.
+ *
+ * Revision 1.26 2008/04/07 16:57:18 fabiankeil
+ * - Use free_url_spec() more consistently.
+ * - Let it reset url->dcount just in case.
+ *
* Revision 1.25 2008/04/06 15:18:38 fabiankeil
* Oh well, rename the --enable-pcre-host-patterns option to
* --enable-extended-host-patterns as it's not really PCRE syntax.
const char urlmatch_h_rcs[] = URLMATCH_H_VERSION;
+enum regex_anchoring {NO_ANCHORING, LEFT_ANCHORED, RIGHT_ANCHORED};
+static jb_err compile_host_pattern(struct url_spec *url, const char *host_pattern);
/*********************************************************************
*
}
-#ifdef FEATURE_EXTENDED_HOST_PATTERNS
/*********************************************************************
*
- * Function : compile_host_pattern
+ * Function : compile_pattern
*
- * Description : Parses and compiles a PCRE host pattern..
+ * Description : Compiles a host, domain or TAG pattern.
*
* Parameters :
- * 1 : url = Target url_spec to be filled in.
- * 2 : host_pattern = Host pattern to compile.
+ * 1 : pattern = The pattern to compile.
+ * 2 : anchoring = How the regex should be anchored.
+ * Can be either one of NO_ANCHORING,
+ * LEFT_ANCHORED or RIGHT_ANCHORED.
+ * 3 : url = In case of failures, the spec member is
+ * logged and the structure freed.
+ * 4 : regex = Where the compiled regex should be stored.
*
* Returns : JB_ERR_OK - Success
* JB_ERR_MEMORY - Out of memory
* JB_ERR_PARSE - Cannot parse regex
*
*********************************************************************/
-static jb_err compile_host_pattern(struct url_spec *url, const char *host_pattern)
+static jb_err compile_pattern(const char *pattern, enum regex_anchoring anchoring,
+ struct url_spec *url, regex_t **regex)
{
int errcode;
char rebuf[BUFFER_SIZE];
+ const char *fmt;
+
+ assert(pattern);
+ assert(strlen(pattern) < sizeof(rebuf) - 2);
+
+ if (pattern[0] == '\0')
+ {
+ *regex = NULL;
+ return JB_ERR_OK;
+ }
- assert(host_pattern);
- assert(strlen(host_pattern) < sizeof(rebuf) - 2);
+ switch (anchoring)
+ {
+ case NO_ANCHORING:
+ fmt = "%s";
+ break;
+ case RIGHT_ANCHORED:
+ fmt = "%s$";
+ break;
+ case LEFT_ANCHORED:
+ fmt = "^%s";
+ break;
+ default:
+ log_error(LOG_LEVEL_FATAL,
+ "Invalid anchoring in compile_pattern %d", anchoring);
+ }
- url->host_regex = zalloc(sizeof(*url->host_regex));
- if (NULL == url->host_regex)
+ *regex = zalloc(sizeof(**regex));
+ if (NULL == *regex)
{
free_url_spec(url);
return JB_ERR_MEMORY;
}
- snprintf(rebuf, sizeof(rebuf), "%s$", host_pattern);
+ snprintf(rebuf, sizeof(rebuf), fmt, pattern);
- errcode = regcomp(url->host_regex, rebuf,
- (REG_EXTENDED|REG_NOSUB|REG_ICASE));
+ errcode = regcomp(*regex, rebuf, (REG_EXTENDED|REG_NOSUB|REG_ICASE));
if (errcode)
{
- size_t errlen = regerror(errcode, url->host_regex, rebuf, sizeof(rebuf));
+ size_t errlen = regerror(errcode, *regex, rebuf, sizeof(rebuf));
if (errlen > (sizeof(rebuf) - (size_t)1))
{
errlen = sizeof(rebuf) - (size_t)1;
}
rebuf[errlen] = '\0';
- log_error(LOG_LEVEL_ERROR, "error compiling %s: %s", url->spec, rebuf);
+ log_error(LOG_LEVEL_ERROR, "error compiling %s from %s: %s",
+ pattern, url->spec, rebuf);
free_url_spec(url);
return JB_ERR_PARSE;
}
+
+/*********************************************************************
+ *
+ * Function : compile_url_pattern
+ *
+ * Description : Compiles the three parts of an URL pattern.
+ *
+ * Parameters :
+ * 1 : url = Target url_spec to be filled in.
+ * 2 : buf = The url pattern to compile. Will be messed up.
+ *
+ * Returns : JB_ERR_OK - Success
+ * JB_ERR_MEMORY - Out of memory
+ * JB_ERR_PARSE - Cannot parse regex
+ *
+ *********************************************************************/
+static jb_err compile_url_pattern(struct url_spec *url, char *buf)
+{
+ char *p;
+
+ p = strchr(buf, '/');
+ if (NULL != p)
+ {
+ /*
+ * Only compile the regex if it consists of more than
+ * a single slash, otherwise it wouldn't affect the result.
+ */
+ if (p[1] != '\0')
+ {
+ /*
+ * XXX: does it make sense to compile the slash at the beginning?
+ */
+ jb_err err = compile_pattern(p, LEFT_ANCHORED, url, &url->preg);
+
+ if (JB_ERR_OK != err)
+ {
+ return err;
+ }
+ }
+ *p = '\0';
+ }
+
+ p = strchr(buf, ':');
+ if (NULL != p)
+ {
+ *p++ = '\0';
+ url->port_list = strdup(p);
+ if (NULL == url->port_list)
+ {
+ return JB_ERR_MEMORY;
+ }
+ }
+ else
+ {
+ url->port_list = NULL;
+ }
+
+ if (buf[0] != '\0')
+ {
+ return compile_host_pattern(url, buf);
+ }
+
+ return JB_ERR_OK;
+
+}
+
+
+#ifdef FEATURE_EXTENDED_HOST_PATTERNS
+/*********************************************************************
+ *
+ * Function : compile_host_pattern
+ *
+ * Description : Parses and compiles a host pattern..
+ *
+ * Parameters :
+ * 1 : url = Target url_spec to be filled in.
+ * 2 : host_pattern = Host pattern to compile.
+ *
+ * Returns : JB_ERR_OK - Success
+ * JB_ERR_MEMORY - Out of memory
+ * JB_ERR_PARSE - Cannot parse regex
+ *
+ *********************************************************************/
+static jb_err compile_host_pattern(struct url_spec *url, const char *host_pattern)
+{
+ return compile_pattern(host_pattern, RIGHT_ANCHORED, url, &url->host_regex);
+}
+
#else
/*********************************************************************
* function. If this function succeeds, the
* buffer is copied to url->spec. If this
* function fails, the contents of the buffer
- * are lost forever.
+ * are lost forever. XXX: Why is this const?
*
* Returns : JB_ERR_OK - Success
* JB_ERR_MEMORY - Out of memory
*********************************************************************/
jb_err create_url_spec(struct url_spec * url, const char * buf)
{
- char *p;
- int errcode;
- size_t errlen;
- char rebuf[BUFFER_SIZE];
-
assert(url);
assert(buf);
- /*
- * Zero memory
- */
memset(url, '\0', sizeof(*url));
- /*
- * Save a copy of the orignal specification
- */
- if ((url->spec = strdup(buf)) == NULL)
+ /* Remember the original specification for the CGI pages. */
+ url->spec = strdup(buf);
+ if (NULL == url->spec)
{
return JB_ERR_MEMORY;
}
/* Is it tag pattern? */
if (0 == strncmpic("TAG:", url->spec, 4))
{
- if (NULL == (url->tag_regex = zalloc(sizeof(*url->tag_regex))))
- {
- freez(url->spec);
- return JB_ERR_MEMORY;
- }
-
- /* buf + 4 to skip "TAG:" */
- errcode = regcomp(url->tag_regex, buf + 4, (REG_EXTENDED|REG_NOSUB|REG_ICASE));
- if (errcode)
- {
- errlen = regerror(errcode, url->preg, rebuf, sizeof(rebuf));
- if (errlen > (sizeof(rebuf) - 1))
- {
- errlen = sizeof(rebuf) - 1;
- }
- rebuf[errlen] = '\0';
- log_error(LOG_LEVEL_ERROR, "error compiling %s: %s", url->spec, rebuf);
- free_url_spec(url);
-
- return JB_ERR_PARSE;
- }
- return JB_ERR_OK;
- }
-
- /* Only reached for URL patterns */
- p = strchr(buf, '/');
- if (NULL != p)
- {
- url->path = strdup(p);
- if (NULL == url->path)
- {
- free_url_spec(url);
- return JB_ERR_MEMORY;
- }
- *p = '\0';
+ /* The pattern starts with the first character after "TAG:" */
+ const char *tag_pattern = buf + 4;
+ return compile_pattern(tag_pattern, NO_ANCHORING, url, &url->tag_regex);
}
- else
- {
- url->path = NULL;
- }
- if (url->path)
- {
- if (NULL == (url->preg = zalloc(sizeof(*url->preg))))
- {
- free_url_spec(url);
- return JB_ERR_MEMORY;
- }
-
- snprintf(rebuf, sizeof(rebuf), "^(%s)", url->path);
-
- errcode = regcomp(url->preg, rebuf,
- (REG_EXTENDED|REG_NOSUB|REG_ICASE));
- if (errcode)
- {
- errlen = regerror(errcode, url->preg, rebuf, sizeof(rebuf));
-
- if (errlen > (sizeof(rebuf) - (size_t)1))
- {
- errlen = sizeof(rebuf) - (size_t)1;
- }
- rebuf[errlen] = '\0';
- log_error(LOG_LEVEL_ERROR, "error compiling %s: %s",
- url->spec, rebuf);
- free_url_spec(url);
-
- return JB_ERR_PARSE;
- }
- }
-
- p = strchr(buf, ':');
- if (NULL != p)
- {
- *p++ = '\0';
- url->port_list = strdup(p);
- if (NULL == url->port_list)
- {
- return JB_ERR_MEMORY;
- }
- }
- else
- {
- url->port_list = NULL;
- }
-
- if (buf[0] != '\0')
- {
- return compile_host_pattern(url, buf);
- }
-
- return JB_ERR_OK;
+ /* If it isn't a tag pattern it must be a URL pattern. */
+ return compile_url_pattern(url, (char *)buf);
}
freez(url->dvec);
url->dcount = 0;
#endif /* ndef FEATURE_EXTENDED_HOST_PATTERNS */
- freez(url->path);
freez(url->port_list);
if (url->preg)
{
*
*********************************************************************/
int url_match(const struct url_spec *pattern,
- const struct http_request *url)
+ const struct http_request *http)
{
/* XXX: these should probably be functions. */
-#define PORT_MATCHES ((NULL == pattern->port_list) || match_portlist(pattern->port_list, url->port))
+#define PORT_MATCHES ((NULL == pattern->port_list) || match_portlist(pattern->port_list, http->port))
#ifdef FEATURE_EXTENDED_HOST_PATTERNS
-#define DOMAIN_MATCHES ((NULL == pattern->host_regex) || (0 == regexec(pattern->host_regex, url->host, 0, NULL, 0)))
+#define DOMAIN_MATCHES ((NULL == pattern->host_regex) || (0 == regexec(pattern->host_regex, http->host, 0, NULL, 0)))
#else
-#define DOMAIN_MATCHES ((NULL == pattern->dbuffer) || (0 == domain_match(pattern, url)))
+#define DOMAIN_MATCHES ((NULL == pattern->dbuffer) || (0 == domain_match(pattern, http)))
#endif
-#define PATH_MATCHES ((NULL == pattern->path) || (0 == regexec(pattern->preg, url->path, 0, NULL, 0)))
+#define PATH_MATCHES ((NULL == pattern->preg) || (0 == regexec(pattern->preg, http->path, 0, NULL, 0)))
if (pattern->tag_regex != NULL)
{
projects / privoxy.git / blobdiff