Thiemo Mättig (WMDE) has uploaded a new change for review.
https://gerrit.wikimedia.org/r/207747
Change subject: Rework regex for quoted keyword:"..." extraction
......................................................................
Rework regex for quoted keyword:"..." extraction
I'm touching this code for two reasons:
1. A trim() to remove the quotes was found in three places. We can
remove the quotes right away in a single place. I think this makes
the code easier to read and maintain.
2. A simple trim() will lead to unexpected results in cases like
keyword:"\"foo\"". The escaped quotes should be part of the search
string but trim() will remove them because trim() can not be limited
to trim only 1 character left and right.
Change-Id: Id6f58010c5a69bd20090bc2c400cf2b93c8da543
---
M includes/Searcher.php
1 file changed, 5 insertions(+), 6 deletions(-)
git pull ssh://gerrit.wikimedia.org:29418/mediawiki/extensions/CirrusSearch
refs/changes/47/207747/1
diff --git a/includes/Searcher.php b/includes/Searcher.php
index 990ab3c..46b8db2 100644
--- a/includes/Searcher.php
+++ b/includes/Searcher.php
@@ -499,12 +499,13 @@
$escaper = $this->escaper;
$fuzzyQuery = $this->fuzzyQuery;
$this->extractSpecialSyntaxFromTerm(
-
'/(?<key>[a-z\\-]{7,15}):\s*(?<value>"(?:[^"]|(?<=\\\)")+"|[^ "]+) ?/',
+
'/(?<key>[a-z\\-]{7,15}):\s*(?:"(?:<quoted>(?:[^"]|(?<=\\\)")+)"|(?<unquoted>\S+))
?/',
function ( $matches ) use ( $searcher, $escaper,
&$filters, &$notFilters, &$boostTemplates,
&$searchContainedSyntax, &$fuzzyQuery,
&$highlightSource ) {
$key = $matches['key'];
- $value = $matches['value']; // Note that if
the user supplied quotes they are not removed
- $value = str_replace( '\"', '"', $value );
+ $value = isset( $matches['quoted'] )
+ ? str_replace( '\"', '"',
$matches['quoted'] )
+ : $matches['unquoted'];
$filterDestination = &$filters;
$keepText = true;
if ( $key[ 0 ] === '-' ) {
@@ -514,14 +515,13 @@
}
switch ( $key ) {
case 'boost-templates':
- $boostTemplates =
Searcher::parseBoostTemplates( trim( $value, '"' ) );
+ $boostTemplates =
Searcher::parseBoostTemplates( $value );
if ( $boostTemplates === null )
{
$boostTemplates =
Searcher::getDefaultBoostTemplates();
}
$searchContainedSyntax = true;
return '';
case 'hastemplate':
- $value = trim( $value, '"' );
// We emulate template syntax
here as best as possible,
// so things in NS_MAIN are
prefixed with ":" and things
// in NS_TEMPLATE don't have a
prefix at all. Since we
@@ -743,7 +743,6 @@
* @return \Elastica\Filter\Query for matching $title to $field
*/
public function matchPage( $field, $title, $underscores = false ) {
- $title = trim( $title, '"' ); // Somtimes title
is wrapped in quotes - throw them away.
if ( $underscores ) {
$title = str_replace( ' ', '_', $title );
} else {
--
To view, visit https://gerrit.wikimedia.org/r/207747
To unsubscribe, visit https://gerrit.wikimedia.org/r/settings
Gerrit-MessageType: newchange
Gerrit-Change-Id: Id6f58010c5a69bd20090bc2c400cf2b93c8da543
Gerrit-PatchSet: 1
Gerrit-Project: mediawiki/extensions/CirrusSearch
Gerrit-Branch: master
Gerrit-Owner: Thiemo Mättig (WMDE) <thiemo.maet...@wikimedia.de>
_______________________________________________
MediaWiki-commits mailing list
MediaWiki-commits@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits
