Merge commit 'br3nda/0.8.x-hashtags' into 0.8.x

classes/Notice.php

@@ -97,13 +97,21 @@ class Notice extends Memcached_DataObject
     function saveTags()
     {
         /* extract all #hastags */
-        $count = preg_match_all('/(?:^|\s)#([A-Za-z0-9_\-\.]{1,64})/', strtolower($this->content), $match);
+        $count = preg_match_all('/(?:^|\s)#([\pL\pN_\-\.]{1,64})/', strtolower($this->content), $match);
         if (!$count) {
             return true;
         }
+        
+        //turn each into their canonical tag
+        //this is needed to remove dupes before saving e.g. #hash.tag = #hashtag
+        $hashtags = array();
+        for($i=0; $i<count($match[1]); $i++) {
+             $hashtags[] = common_canonical_tag($match[1][$i]);
+        }
 
+ 
         /* Add them to the database */
-        foreach(array_unique($match[1]) as $hashtag) {
+        foreach(array_unique($hashtags) as $hashtag) {
             /* elide characters we don't want in the tag */
             $this->saveTag($hashtag);
         }
@@ -112,8 +120,6 @@ class Notice extends Memcached_DataObject
 
     function saveTag($hashtag)
     {
-        $hashtag = common_canonical_tag($hashtag);
-
         $tag = new Notice_tag();
         $tag->notice_id = $this->id;
         $tag->tag = $hashtag;

lib/router.php

@@ -211,7 +211,7 @@ class Router
                     array('tag' => '[a-zA-Z0-9]+'));
         $m->connect('tag/:tag',
                     array('action' => 'tag'),
-                    array('tag' => '[a-zA-Z0-9]+'));
+                    array('tag' => '[\pL\pN_\-\.]{1,64}'));
 
         $m->connect('peopletag/:tag',
                     array('action' => 'peopletag'),

lib/util.php

@@ -404,7 +404,7 @@ function common_render_text($text)
 
     $r = preg_replace('/[\x{0}-\x{8}\x{b}-\x{c}\x{e}-\x{19}]/', '', $r);
     $r = common_replace_urls_callback($r, 'common_linkify');
-    $r = preg_replace('/(^|\(|\[|\s+)#([A-Za-z0-9_\-\.]{1,64})/e', "'\\1#'.common_tag_link('\\2')", $r);
+    $r = preg_replace('/(^|\(|\[|\s+)#([\pL\pN_\-\.]{1,64})/e', "'\\1#'.common_tag_link('\\2')", $r);
     // XXX: machine tags
     return $r;
 }