Details
-
Type:
Bug
-
Status:
Closed
-
Priority:
Minor
-
Resolution: Fixed
-
Affects Version/s: 1.9.5
-
Component/s: Global search
-
Labels:None
-
Affected Branches:MOODLE_19_STABLE
-
Fixed Branches:MOODLE_19_STABLE, MOODLE_20_STABLE
Description
In multibyte language environment (ex. Japanese), we can not seach by multibyte strings on global search.
The cause of this problem is only forgetting character encoding;
Patch is as follows;
In search/Zend/Search/Lucene.php, function find()
public function find($query)
{
if (is_string($query)) {
// (Shirai144): modify from here
// $query = Zend_Search_Lucene_Search_QueryParser::parse($query);
$query = Zend_Search_Lucene_Search_QueryParser::parse($query, 'UTF-8');
// (Shirai144): modify to here
}
Attachments
Issue Links
| This issue blocks: | ||||
| MDL-19426 | Enabling search/cron.php and search/indexer.php to run at command line, and so on |
|
|
|
| This issue has been marked as being related by: | ||||
| MDL-19340 | Advanced search does not work on Global search |
|
|
|
| MDL-19341 | Adding filed 'user_id' at document.php, but it does not return by process_results() |
|
|
|
| MDL-19426 | Enabling search/cron.php and search/indexer.php to run at command line, and so on |
|
|
|
Hi Tatsuva,
This is a valuable input, indeed.
The fact is that we didn't spend a lot of time within the Zend Lucene package itself, mostly because it is an external integration, and merely because any change we make in Lucene distribution might break a bit the upgrading chain for an updated engine from Zend.
As we have no direct connection with Zend team for Lucene, it is not usable for us to feed that miss back, or maybe is it already fixed in the actual distribution.
Any way this is an interesting fix we need to commit for all non english languages using such sequences,
I integrate it ASAP in CVS versions.
Thanks.