From: Vsevolod Stakhov <vsevolod@rambler-co.ru>
Date: Mon, 1 Jun 2009 19:15:31 +0000 (+0400)
Subject: * Fix url grabbing regexp to match urls like <a href=http://blah.com>
X-Git-Tag: 0.2.7~132^2
X-Git-Url: https://source.dussan.org/?a=commitdiff_plain;h=21844599898aa8b44df6425914995e1b7214ff52;p=rspamd.git

* Fix url grabbing regexp to match urls like <a href=http://blah.com>
---

diff --git a/CMakeLists.txt b/CMakeLists.txt
index f744fff03..2d55565e9 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -7,7 +7,7 @@ PROJECT(rspamd C)
 
 SET(RSPAMD_VERSION_MAJOR 0)
 SET(RSPAMD_VERSION_MINOR 1)
-SET(RSPAMD_VERSION_PATCH 2)
+SET(RSPAMD_VERSION_PATCH 3)
 
 SET(RSPAMD_VERSION         "${RSPAMD_VERSION_MAJOR}.${RSPAMD_VERSION_MINOR}.${RSPAMD_VERSION_PATCH}")
 SET(RSPAMD_MASTER_SITE_URL "http://cebka.pp.ru/hg/rspamd")
diff --git a/src/url.c b/src/url.c
index 875358ae0..e6a4a215c 100644
--- a/src/url.c
+++ b/src/url.c
@@ -64,7 +64,7 @@ static const char *text_url = "((https?|ftp)://)?"
 "(?<![\\s>?!),.'\"\\]:])"
 "(?!@)"
 ")";
-static const char *html_url = "(?: src|href)=\"("
+static const char *html_url = "(?: src|href)=\"?("
 "((https?|ftp)://)?"
 "(\\b(?<![.\\@A-Za-z0-9-])"	
 "(?: [A-Za-z0-9][A-Za-z0-9-]*(?:\\.[A-Za-z0-9-]+)*\\."
@@ -80,7 +80,7 @@ static const char *html_url = "(?: src|href)=\"("
 "(?:[/?][;/?:@&=+\\$,[\\]\\-_.!~*'()A-Za-z0-9#%]*)?" /* path (&query) */
 "(?<![\\s>?!),.'\"\\]:])"
 "(?!@)"
-"))\"";
+"))\"?";
 
 static short url_initialized = 0;
 GRegex *text_re, *html_re;