Ticket #15549: new-ping-urls-regex.diff

wp-includes/comment.php

@@ -1781 +1781 @@
 
         $pung = get_pung($post_ID);
 
-        // Variables
-        $ltrs = '\w';
-        $gunk = '/#~:.?+=&%@!\-';
-        $punc = '.:?\-';
-        $any = $ltrs . $gunk . $punc;
-
         // Step 1
         // Parsing the post, external links (if any) are stored in the $post_links array
-        // This regexp comes straight from phpfreaks.com
-        // http://www.phpfreaks.com/quickcode/Extract_All_URLs_on_a_Page/15.php
-        preg_match_all("{\b http : [$any] +? (?= [$punc] * [^$any] | $)}x", $content, $post_links_temp);
-
+        $post_links_temp = _extract_pingable_urls( $content );
+        
         // Step 2.
         // Walking thru the links array
         // first we get rid of links pointing to sites, not to specific files
@@ -1802 +1794 @@
         // http://dummy-weblog.org/post.php
         // We don't wanna ping first and second types, even if they have a valid <link/>
 
-        foreach ( (array) $post_links_temp[0] as $link_test ) :
+        foreach ( (array) $post_links_temp as $link_test ) :
                 if ( !in_array($link_test, $pung) && (url_to_postid($link_test) != $post_ID) // If we haven't pung it already and it isn't a link to itself
                                 && !is_local_attachment($link_test) ) : // Also, let's never ping local attachments.
                         if ( $test = @parse_url($link_test) ) {

wp-includes/functions.php

@@ -393 +393 @@
 }
 
 /**
+ * Use RegEx to extract URLs from arbitrary content
+ * 
+ * (?xi)
+ * \b
+ * (                                      # Capture 1: entire matched URL
+ *   (?:
+ *     https?://                          # http or https protocol
+ *     |                                          #   or
+ *     www\d{0,3}[.]                          # "www.", "www1.", "www2." … "www999."
+ *     |                                          #   or
+ *     [a-z0-9.\-]+[.][a-z]{2,4}/         # looks like domain name followed by a slash
+ *   )
+ *   (?:                                          # One or more:
+ *     [^\s()<>]+                         # Run of non-space, non-()<>
+ *     |                                  #   or
+ *     \(([^\s()<>]+|(\([^\s()<>]+\)))*\) # balanced parens, up to 2 levels
+ *   )+
+ *   (?:                                          # End with:
+ *     \(([^\s()<>]+|(\([^\s()<>]+\)))*\) # balanced parens, up to 2 levels
+ *     |                                          #   or
+ *     [^\s`!()\[\]{};:'".,<>?«»“”‘’]     # not a space or one of these punct chars
+ *   )
+ * )
+ * 
+ * @since 3.5.0
+ * 
+ * @param string $content
+ * @return array URLs to ping
+ */
+function _extract_pingable_urls( $content ) {
+        preg_match_all( 
+                "#(?i)\b((?:https?://|www\d{0,3}[.]|[a-z0-9.\-]+[.][a-z]{2,4}/)(?:[^\s()<>]+|\(([^\s()<>]+|(\([^\s()<>]+\)))*\))+(?:\(([^\s()<>]+|(\([^\s()<>]+\)))*\)|[^\s`!()\[\]{};:'\".,<>?«»“”‘’]))#", 
+                $content, 
+                $post_links 
+        );
+        
+        foreach ( $post_links[0] as &$post_link )
+                $post_link = html_entity_decode( $post_link );
+        
+        return $post_links[0];
+}
+
+/**
  * Check content for video and audio links to add as enclosures.
  *
  * Will not add enclosures that have already been added and will
@@ -417 +460 @@
 
         $pung = get_enclosed( $post_ID );
 
-        $ltrs = '\w';
-        $gunk = '/#~:.?+=&%@!\-';
-        $punc = '.:?\-';
-        $any = $ltrs . $gunk . $punc;
+        $post_links_temp = _extract_pingable_urls( $content );
 
-        preg_match_all( "{\b http : [$any] +? (?= [$punc] * [^$any] | $)}x", $content, $post_links_temp );
-
         foreach ( $pung as $link_test ) {
-                if ( !in_array( $link_test, $post_links_temp[0] ) ) { // link no longer in post
+                if ( !in_array( $link_test, $post_links_temp ) ) { // link no longer in post
                         $mids = $wpdb->get_col( $wpdb->prepare("SELECT meta_id FROM $wpdb->postmeta WHERE post_id = %d AND meta_key = 'enclosure' AND meta_value LIKE (%s)", $post_ID, like_escape( $link_test ) . '%') );
                         foreach ( $mids as $mid )
                                 delete_metadata_by_mid( 'post', $mid );
                 }
         }
 
-        foreach ( (array) $post_links_temp[0] as $link_test ) {
+        foreach ( (array) $post_links_temp as $link_test ) {
                 if ( !in_array( $link_test, $pung ) ) { // If we haven't pung it already
                         $test = @parse_url( $link_test );
                         if ( false === $test )