X-Git-Url: http://www.privoxy.org/gitweb/?a=blobdiff_plain;ds=sidebyside;f=urlmatch.c;h=09db9295748f6384fbfa603911b64a12dd30e6ee;hb=b61d80241944fbc1b9dac07b12e721c4d24bf970;hp=a031348a8bd3c187015cae5e44df42e6746d5eae;hpb=20414be38c51b53dbe43fce6ce8883952b05221d;p=privoxy.git diff --git a/urlmatch.c b/urlmatch.c index a031348a..09db9295 100644 --- a/urlmatch.c +++ b/urlmatch.c @@ -1,4 +1,4 @@ -const char urlmatch_rcs[] = "$Id: urlmatch.c,v 1.30 2008/04/10 04:24:24 fabiankeil Exp $"; +const char urlmatch_rcs[] = "$Id: urlmatch.c,v 1.37 2008/04/17 14:53:29 fabiankeil Exp $"; /********************************************************************* * * File : $Source: /cvsroot/ijbswa/current/urlmatch.c,v $ @@ -33,6 +33,30 @@ const char urlmatch_rcs[] = "$Id: urlmatch.c,v 1.30 2008/04/10 04:24:24 fabianke * * Revisions : * $Log: urlmatch.c,v $ + * Revision 1.37 2008/04/17 14:53:29 fabiankeil + * Move simplematch() into urlmatch.c as it's only + * used to match (old-school) domain patterns. + * + * Revision 1.36 2008/04/14 18:19:48 fabiankeil + * Remove now-pointless cast in create_url_spec(). + * + * Revision 1.35 2008/04/14 18:11:21 fabiankeil + * The compiler might not notice it, but the buffer passed to + * create_url_spec() is modified later on and thus shouldn't + * be declared immutable. + * + * Revision 1.34 2008/04/13 13:32:07 fabiankeil + * Factor URL pattern compilation out of create_url_spec(). + * + * 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 @@ -219,6 +243,8 @@ const char urlmatch_rcs[] = "$Id: urlmatch.c,v 1.30 2008/04/10 04:24:24 fabianke const char urlmatch_h_rcs[] = URLMATCH_H_VERSION; +enum regex_anchoring {NO_ANCHORING, LEFT_ANCHORED, RIGHT_ANCHORED}; +static jb_err compile_host_pattern(struct url_spec *url, const char *host_pattern); /********************************************************************* * @@ -666,51 +692,79 @@ jb_err parse_http_request(const char *req, } -#ifdef FEATURE_EXTENDED_HOST_PATTERNS /********************************************************************* * - * Function : compile_host_pattern + * Function : compile_pattern * - * Description : Parses and compiles a PCRE host pattern.. + * Description : Compiles a host, domain or TAG pattern. * * Parameters : - * 1 : url = Target url_spec to be filled in. - * 2 : host_pattern = Host pattern to compile. + * 1 : pattern = The pattern to compile. + * 2 : anchoring = How the regex should be anchored. + * Can be either one of NO_ANCHORING, + * LEFT_ANCHORED or RIGHT_ANCHORED. + * 3 : url = In case of failures, the spec member is + * logged and the structure freed. + * 4 : regex = Where the compiled regex should be stored. * * Returns : JB_ERR_OK - Success * JB_ERR_MEMORY - Out of memory * JB_ERR_PARSE - Cannot parse regex * *********************************************************************/ -static jb_err compile_host_pattern(struct url_spec *url, const char *host_pattern) +static jb_err compile_pattern(const char *pattern, enum regex_anchoring anchoring, + struct url_spec *url, regex_t **regex) { int errcode; char rebuf[BUFFER_SIZE]; + const char *fmt; + + assert(pattern); + assert(strlen(pattern) < sizeof(rebuf) - 2); + + if (pattern[0] == '\0') + { + *regex = NULL; + return JB_ERR_OK; + } - assert(host_pattern); - assert(strlen(host_pattern) < sizeof(rebuf) - 2); + switch (anchoring) + { + case NO_ANCHORING: + fmt = "%s"; + break; + case RIGHT_ANCHORED: + fmt = "%s$"; + 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; @@ -720,6 +774,94 @@ static jb_err compile_host_pattern(struct url_spec *url, const char *host_patter } + +/********************************************************************* + * + * 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 /********************************************************************* @@ -808,6 +950,144 @@ static jb_err compile_host_pattern(struct url_spec *url, const char *host_patter } +/********************************************************************* + * + * Function : simplematch + * + * Description : String matching, with a (greedy) '*' wildcard that + * stands for zero or more arbitrary characters and + * character classes in [], which take both enumerations + * and ranges. + * + * Parameters : + * 1 : pattern = pattern for matching + * 2 : text = text to be matched + * + * Returns : 0 if match, else nonzero + * + *********************************************************************/ +static int simplematch(const char *pattern, const char *text) +{ + const unsigned char *pat = (const unsigned char *)pattern; + const unsigned char *txt = (const unsigned char *)text; + const unsigned char *fallback = pat; + int wildcard = 0; + + unsigned char lastchar = 'a'; + unsigned i; + unsigned char charmap[32]; + + while (*txt) + { + + /* EOF pattern but !EOF text? */ + if (*pat == '\0') + { + if (wildcard) + { + pat = fallback; + } + else + { + return 1; + } + } + + /* '*' in the pattern? */ + if (*pat == '*') + { + + /* The pattern ends afterwards? Speed up the return. */ + if (*++pat == '\0') + { + return 0; + } + + /* Else, set wildcard mode and remember position after '*' */ + wildcard = 1; + fallback = pat; + } + + /* Character range specification? */ + if (*pat == '[') + { + memset(charmap, '\0', sizeof(charmap)); + + while (*++pat != ']') + { + if (!*pat) + { + return 1; + } + else if (*pat == '-') + { + if ((*++pat == ']') || *pat == '\0') + { + return(1); + } + for (i = lastchar; i <= *pat; i++) + { + charmap[i / 8] |= (unsigned char)(1 << (i % 8)); + } + } + else + { + charmap[*pat / 8] |= (unsigned char)(1 << (*pat % 8)); + lastchar = *pat; + } + } + } /* -END- if Character range specification */ + + + /* + * Char match, or char range match? + */ + if ( (*pat == *txt) + || (*pat == '?') + || ((*pat == ']') && (charmap[*txt / 8] & (1 << (*txt % 8)))) ) + { + /* + * Sucess: Go ahead + */ + pat++; + } + else if (!wildcard) + { + /* + * No match && no wildcard: No luck + */ + return 1; + } + else if (pat != fallback) + { + /* + * Increment text pointer if in char range matching + */ + if (*pat == ']') + { + txt++; + } + /* + * Wildcard mode && nonmatch beyond fallback: Rewind pattern + */ + pat = fallback; + /* + * Restart matching from current text pointer + */ + continue; + } + txt++; + } + + /* Cut off extra '*'s */ + if(*pat == '*') pat++; + + /* If this is the pattern's end, fine! */ + return(*pat); + +} + + /********************************************************************* * * Function : simple_domaincmp @@ -951,25 +1231,16 @@ static int domain_match(const struct url_spec *pattern, const struct http_reques * written to system log) * *********************************************************************/ -jb_err create_url_spec(struct url_spec * url, const char * buf) +jb_err create_url_spec(struct url_spec *url, 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; } @@ -977,92 +1248,13 @@ jb_err create_url_spec(struct url_spec * url, const char * buf) /* 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; + /* 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); } - /* Only reached for URL patterns */ - 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: mostly duplicated code, should be factored out. */ - url->preg = zalloc(sizeof(*url->preg)); - if (NULL == url->preg) - { - free_url_spec(url); - return JB_ERR_MEMORY; - } - - snprintf(rebuf, sizeof(rebuf), "^(%s)", p); - 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 = '\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; - + /* If it isn't a tag pattern it must be a URL pattern. */ + return compile_url_pattern(url, buf); }