DCausse has uploaded a new change for review. (
https://gerrit.wikimedia.org/r/398230 )
Change subject: [WIP] Fetch redirect namespace during prefix search
......................................................................
[WIP] Fetch redirect namespace during prefix search
Use the experimental highlighter to fetch redirect namesapces.
WIP: because Erik implemented a more generic solution that does
not depend on the exp highlighter.
Bug: T115756
Change-Id: I2f48f06216752386df342b05a96b9fa69d9a3bc4
---
M includes/CirrusSearch.php
M includes/CompletionSuggester.php
M includes/Search/ResultsType.php
3 files changed, 34 insertions(+), 15 deletions(-)
git pull ssh://gerrit.wikimedia.org:29418/mediawiki/extensions/CirrusSearch
refs/changes/30/398230/1
diff --git a/includes/CirrusSearch.php b/includes/CirrusSearch.php
index b5c2725..36f539f 100644
--- a/includes/CirrusSearch.php
+++ b/includes/CirrusSearch.php
@@ -650,7 +650,9 @@
return SearchSuggestionSet::fromTitles(
$status->getValue() );
}
$results = array_filter( array_map(
- [ FancyTitleResultsType::class,
'chooseBestTitleOrRedirect' ],
+ function( $result ) {
+ return
FancyTitleResultsType::chooseBestTitleOrRedirect( $result, $this->namespaces );
+ },
$status->getValue() ) );
return SearchSuggestionSet::fromTitles( $results );
}
diff --git a/includes/CompletionSuggester.php b/includes/CompletionSuggester.php
index e553aca..98245d1 100644
--- a/includes/CompletionSuggester.php
+++ b/includes/CompletionSuggester.php
@@ -285,7 +285,8 @@
foreach ( $prefixResults->getResults() as $res ) {
$pageId = $this->config->makePageId( $res->getId() );
- $title =
FancyTitleResultsType::chooseBestTitleOrRedirect(
$rType->transformOneElasticResult( $res ) );
+ $title =
FancyTitleResultsType::chooseBestTitleOrRedirect(
$rType->transformOneElasticResult( $res ),
+
$this->prefixSearchRequestBuilder->getSearchContext()->getNamespaces() );
if ( $title === false ) {
continue;
}
diff --git a/includes/Search/ResultsType.php b/includes/Search/ResultsType.php
index 372ef36..318ef0a 100644
--- a/includes/Search/ResultsType.php
+++ b/includes/Search/ResultsType.php
@@ -118,6 +118,9 @@
/** @var string */
private $matchedAnalyzer;
+ /** @var boolean */
+ private $useExperimentalHighlighter;
+
/**
* Build result type. The matchedAnalyzer is required to detect if
the match
* was from the title or a redirect (and is kind of a leaky
abstraction.)
@@ -125,7 +128,9 @@
* @param string $matchedAnalyzer the analyzer used to match the title
*/
public function __construct( $matchedAnalyzer ) {
+ global $wgCirrusSearchUseExperimentalHighlighter;
$this->matchedAnalyzer = $matchedAnalyzer;
+ $this->useExperimentalHighlighter =
$wgCirrusSearchUseExperimentalHighlighter;
}
/**
@@ -133,9 +138,8 @@
* @return array|null
*/
public function getHighlightingConfiguration( array $highlightSource ) {
- global $wgCirrusSearchUseExperimentalHighlighter;
- if ( $wgCirrusSearchUseExperimentalHighlighter ) {
+ if ( $this->useExperimentalHighlighter ) {
// This is much less esoteric then the plain
highlighter based
// invocation but does the same thing. The magic is
that the none
// fragmenter still fragments on multi valued fields.
@@ -148,6 +152,9 @@
'type' => 'experimental',
'fragmenter' => 'none',
'order' => 'score',
+ 'options' => [
+ 'fetch_fields' => [
'redirect.namespace' ]
+ ]
];
} else {
// This is similar to the FullTextResults type but
against the near_match and
@@ -189,6 +196,7 @@
public function transformElasticsearchResult( SearchContext $context,
\Elastica\ResultSet $resultSet ) {
$results = [];
foreach ( $resultSet->getResults() as $r ) {
+ wfDebug( print_r( $r, true ) );
$results[] = $this->transformOneElasticResult( $r );
}
return $results;
@@ -197,15 +205,20 @@
/**
* Finds best title or redirect
* @param array $match array returned by self::transformOneElasticResult
+ * @param array $namespaceFilter filter on these namespaces
* @return \Title|false choose best
*/
- public static function chooseBestTitleOrRedirect( array $match ) {
+ public static function chooseBestTitleOrRedirect( array $match,
$namespaceFilter = [] ) {
if ( isset( $match['titleMatch'] ) ) {
return $match['titleMatch'];
} else {
- if ( isset( $match['redirectMatches'][0] ) ) {
- // TODO maybe dig around in the redirect
matches and find the best one?
- return $match['redirectMatches'][0];
+ if ( isset( $match['redirectMatches'] ) ) {
+ foreach( $match['redirectMatches'] as $redir ) {
+ if ( empty( $namespaceFilter ) ||
in_array( $redir->getNamespace(), $namespaceFilter ) ) {
+ // TODO maybe dig around in the
redirect matches and find the best one?
+ return $redir;
+ }
+ }
}
}
@@ -251,20 +264,23 @@
$highlights["redirect.title.{$this->matchedAnalyzer}_asciifolding"] );
}
if ( count( $redirectHighlights ) !== 0 ) {
- foreach ( $redirectHighlights as $redirectTitle ) {
+ for ( $i = 0; $i < count( $redirectHighlights ); $i++ )
{
+ // Without the experimental highlighter we
don't really know the namespace
+ // of the redirect.
+ $namespace = $r->namespace;
+ $redirectTitle = $redirectHighlights[$i];
+ if ( $this->useExperimentalHighlighter ) {
+ // fetched fields are just after the
highlight snippet
+ $namespace = $redirectHighlights[++$i];
+ }
// The match was against a redirect so we
should replace the $title with one that
// represents the redirect.
// The first step is to strip the actual
highlighting from the title.
$redirectTitle = str_replace( [
Searcher::HIGHLIGHT_PRE, Searcher::HIGHLIGHT_POST ],
'', $redirectTitle );
- // Instead of getting the redirect's real
namespace we're going to just use the namespace
- // of the title. This is not great but OK
given that we can't find cross namespace
- // redirects properly any way.
- // TODO: ask the highlighter to return the
namespace for this kind of matches
- // this would perhaps help to partially fix
T115756
$redirectTitle =
- TitleHelper::makeRedirectTitle( $r,
$redirectTitle, $r->namespace );
+ TitleHelper::makeRedirectTitle( $r,
$redirectTitle, $namespace );
$resultForTitle['redirectMatches'][] =
$redirectTitle;
}
}
--
To view, visit https://gerrit.wikimedia.org/r/398230
To unsubscribe, visit https://gerrit.wikimedia.org/r/settings
Gerrit-MessageType: newchange
Gerrit-Change-Id: I2f48f06216752386df342b05a96b9fa69d9a3bc4
Gerrit-PatchSet: 1
Gerrit-Project: mediawiki/extensions/CirrusSearch
Gerrit-Branch: master
Gerrit-Owner: DCausse <[email protected]>
_______________________________________________
MediaWiki-commits mailing list
[email protected]
https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits
