https://www.mediawiki.org/wiki/Special:Code/MediaWiki/113475
Revision: 113475
Author: maxsem
Date: 2012-03-09 15:54:44 +0000 (Fri, 09 Mar 2012)
Log Message:
-----------
Bug 35083 - OpenSearchXml first sentences extraction produces bad results. Made
extracts at least 100 characters long (configurable).
Modified Paths:
--------------
trunk/extensions/OpenSearchXml/ApiOpenSearchXml.php
trunk/extensions/OpenSearchXml/OpenSearchXml.php
Modified: trunk/extensions/OpenSearchXml/ApiOpenSearchXml.php
===================================================================
--- trunk/extensions/OpenSearchXml/ApiOpenSearchXml.php 2012-03-09 15:49:28 UTC
(rev 113474)
+++ trunk/extensions/OpenSearchXml/ApiOpenSearchXml.php 2012-03-09 15:54:44 UTC
(rev 113475)
@@ -275,6 +275,7 @@
* @access private
*/
function extractStart( $text ) {
+ global $wgOpenSearchDescriptionLength;
$endchars = array(
'([^\d])\.\s', '\!\s', '\?\s', // regular ASCII
'。', // full-width ideographic full-stop
@@ -284,7 +285,7 @@
$endgroup = implode( '|', $endchars );
$end = "(?:$endgroup)";
- $sentence = ".*?$end+";
+ $sentence = ".{{$wgOpenSearchDescriptionLength},}?$end+";
$firstone = "/^($sentence)/u";
$matches = array();
if( preg_match( $firstone, $text, $matches ) ) {
Modified: trunk/extensions/OpenSearchXml/OpenSearchXml.php
===================================================================
--- trunk/extensions/OpenSearchXml/OpenSearchXml.php 2012-03-09 15:49:28 UTC
(rev 113474)
+++ trunk/extensions/OpenSearchXml/OpenSearchXml.php 2012-03-09 15:54:44 UTC
(rev 113475)
@@ -40,6 +40,11 @@
$wgOpenSearchAdvertiseXml = true;
/**
+ * Minimum length of extract in <Description>. Actual extracts will last until
the end of sentence.
+ */
+$wgOpenSearchDescriptionLength = 100;
+
+/**
* @param $urls array
* @return bool
*/
_______________________________________________
MediaWiki-CVS mailing list
[email protected]
https://lists.wikimedia.org/mailman/listinfo/mediawiki-cvs
