Manybubbles has uploaded a new change for review.
https://gerrit.wikimedia.org/r/86781
Change subject: Allow aggressive splitting.
......................................................................
Allow aggressive splitting.
Aggressive splitting splits camelCase, snake_case, terms with numb3rs in
them, hyphen-ated words, etc. It might be good in all wikis but for now
it only works in English.
Bug: 54799
Change-Id: I6a84835ea5d6463f10e4fcd5bb228e1ffd252a28
---
M CirrusSearch.php
M includes/CirrusSearchAnalysisConfigBuilder.php
2 files changed, 22 insertions(+), 1 deletion(-)
git pull ssh://gerrit.wikimedia.org:29418/mediawiki/extensions/CirrusSearch
refs/changes/81/86781/1
diff --git a/CirrusSearch.php b/CirrusSearch.php
index 9743802..f7bf99c 100644
--- a/CirrusSearch.php
+++ b/CirrusSearch.php
@@ -92,6 +92,10 @@
'max_word_len' => 0,
);
+// Should CirrusSearch aggressively split up compound words? Good for
splitting camelCase, snake_case, etc.
+// Changing it requires an in place reindex to take effect. Currently only
available in English.
+$wgCirrusSearchUseAggressiveSplitting = true;
+
$dir = __DIR__ . '/';
/**
* Classes
diff --git a/includes/CirrusSearchAnalysisConfigBuilder.php
b/includes/CirrusSearchAnalysisConfigBuilder.php
index 0da92a6..cd4ed84 100644
--- a/includes/CirrusSearchAnalysisConfigBuilder.php
+++ b/includes/CirrusSearchAnalysisConfigBuilder.php
@@ -84,6 +84,11 @@
),
'lowercase' => array(
'type' => 'lowercase',
+ ),
+
+ 'aggressive_splitting' => array(
+ 'type' => 'word_delimiter',
+ 'stem_english_possessive' => 'false',
// No need
)
),
'tokenizer' => array(
@@ -102,6 +107,7 @@
* Customize the default config for the language.
*/
private function customize( $config ) {
+ global $wgCirrusSearchUseAggressiveSplitting;
switch ( $this->language ) {
// Please add languages in alphabetical order.
case 'el':
@@ -116,8 +122,19 @@
$config[ 'analyzer' ][ 'text' ] = array(
'type' => 'custom',
'tokenizer' => 'standard',
- 'filter' => array( 'standard',
'possessive_english', 'lowercase', 'stop', 'porter_stem', 'asciifolding' )
);
+ $filters = array();
+ $filters[] = 'standard';
+ if ( $wgCirrusSearchUseAggressiveSplitting ) {
+ $filters[] = 'aggressive_splitting';
+ }
+ $filters[] = 'possessive_english';
+ $filters[] = 'lowercase';
+ $filters[] = 'stop';
+ $filters[] = 'porter_stem';
+ $filters[] = 'asciifolding';
+ $config[ 'analyzer' ][ 'text' ][ 'filter' ] = $filters;
+
// Add asciifolding to the the text_plain analyzer as
well
$config[ 'analyzer' ][ 'plain' ][ 'filter' ][] =
'asciifolding';
// Add asciifolding to the prefix queries and
incategory filters
--
To view, visit https://gerrit.wikimedia.org/r/86781
To unsubscribe, visit https://gerrit.wikimedia.org/r/settings
Gerrit-MessageType: newchange
Gerrit-Change-Id: I6a84835ea5d6463f10e4fcd5bb228e1ffd252a28
Gerrit-PatchSet: 1
Gerrit-Project: mediawiki/extensions/CirrusSearch
Gerrit-Branch: master
Gerrit-Owner: Manybubbles <[email protected]>
_______________________________________________
MediaWiki-commits mailing list
[email protected]
https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits
