jenkins-bot has submitted this change and it was merged.
Change subject: Skip highlighting we don't need
......................................................................
Skip highlighting we don't need
Slightly changes behavior: only displays redirect or section match if
there isn't a title match. This might not be perfect but it is in line
with our goals to reduce IO usage.
DEPLOYMENT: This does nothing until the upgrade to Elasticsearch 1.3
and experimental-highlight 0.0.11 both of which are scheduled soon.
Change-Id: I58f031adedf875e0f2723c5c21c7ca91e1088893
---
M includes/Search/ResultsType.php
M tests/browser/features/highlighting.feature
2 files changed, 22 insertions(+), 15 deletions(-)
Approvals:
Chad: Looks good to me, approved
jenkins-bot: Verified
diff --git a/includes/Search/ResultsType.php b/includes/Search/ResultsType.php
index 50fac2e..29fc9d9 100644
--- a/includes/Search/ResultsType.php
+++ b/includes/Search/ResultsType.php
@@ -220,13 +220,16 @@
'fragmenter' => 'none',
'number_of_fragments' => 1,
);
- $entireValueInListField = array(
+ $redirectAndHeading = array(
'type' => 'experimental',
'fragmenter' => 'none',
'order' => 'score',
'number_of_fragments' => 1,
+ 'options' => array(
+ 'skip_if_last_matched' => true,
+ )
);
- $singleFragment = array(
+ $remainingText = array(
'type' => 'experimental',
'number_of_fragments' => 1,
'fragmenter' => 'scan',
@@ -246,12 +249,13 @@
// Setting this too low (like 50) can
bury good snippets if the search
// contains common words.
'max_fragments_scored' => 5000,
+ 'skip_if_last_matched' => true,
),
);
if ( !( $this->highlightingConfig &
self::HIGHLIGHT_WITH_DEFAULT_SIMILARITY ) ) {
$entireValue[ 'options' ][ 'default_similarity'
] = false;
- $entireValueInListField[ 'options' ][
'default_similarity' ] = false;
- $singleFragment[ 'options' ][
'default_similarity' ] = false;
+ $redirectAndHeading[ 'options' ][
'default_similarity' ] = false;
+ $remainingText[ 'options' ][
'default_similarity' ] = false;
}
} else {
$entireValue = array(
@@ -259,13 +263,13 @@
'type' => 'fvh',
'order' => 'score',
);
- $entireValueInListField = array(
+ $redirectAndHeading = array(
'number_of_fragments' => 1, // Just one of the
values in the list
'fragment_size' => 10000, // We want the
whole value but more than this is crazy
'type' => 'fvh',
'order' => 'score',
);
- $singleFragment = array(
+ $remainingText = array(
'number_of_fragments' => 1, // Just one fragment
'fragment_size' => $wgCirrusSearchFragmentSize,
'type' => 'fvh',
@@ -273,8 +277,11 @@
);
}
// If there isn't a match just return a match sized chunk from
the beginning of the page.
- $text = $singleFragment;
+ $text = $remainingText;
$text[ 'no_match_size' ] = $text[ 'fragment_size' ];
+ if ( isset( $text[ 'options' ][ 'skip_if_last_matched' ] ) ) {
+ unset( $text[ 'options' ][ 'skip_if_last_matched' ] );
+ }
$config = array(
'pre_tags' => array( Searcher::HIGHLIGHT_PRE ),
@@ -285,14 +292,14 @@
$config[ 'fields' ][ 'title' ] = $entireValue;
}
if ( $this->highlightingConfig & self::HIGHLIGHT_ALT_TITLE ) {
- $config[ 'fields' ][ 'redirect.title' ] =
$entireValueInListField;
- $config[ 'fields' ][ 'heading' ] =
$entireValueInListField;
+ $config[ 'fields' ][ 'redirect.title' ] =
$redirectAndHeading;
+ $config[ 'fields' ][ 'heading' ] = $redirectAndHeading;
}
if ( $this->highlightingConfig & self::HIGHLIGHT_SNIPPET ) {
$config[ 'fields' ][ 'text' ] = $text;
- $config[ 'fields' ][ 'auxiliary_text' ] =
$singleFragment;
+ $config[ 'fields' ][ 'auxiliary_text' ] =
$remainingText;
if ( $this->highlightingConfig &
self::HIGHLIGHT_FILE_TEXT ) {
- $config[ 'fields' ][ 'file_text' ] =
$singleFragment;
+ $config[ 'fields' ][ 'file_text' ] =
$remainingText;
}
}
$config[ 'fields' ] = $this->addMatchedFields( $config[
'fields' ] );
diff --git a/tests/browser/features/highlighting.feature
b/tests/browser/features/highlighting.feature
index da621be..490e2f4 100644
--- a/tests/browser/features/highlighting.feature
+++ b/tests/browser/features/highlighting.feature
@@ -60,7 +60,7 @@
@headings
Scenario: Found words are highlighted in headings even if they contain both
a phrase and a non-phrase
- When I search for "i am a" heading
+ When I search for i "am a heading"
Then *I* *am* *a* *heading* is the highlighted alttitle of the first
search result
@headings
@@ -79,12 +79,12 @@
@headings
Scenario: The highest scoring heading is highlighted AND it doesn't contain
html even if the heading on the page does
- When I search for bold heading
+ When I search for "bold heading"
Then I am a *bold* *heading* is the highlighted alttitle of the first
search result
@headings
Scenario: HTML comments in headings are not highlighted
- When I search for Heading with html comment
+ When I search for "Heading with html comment"
And *Heading* *with* *html* *comment* is the highlighted alttitle of the
first search result
Scenario: Redirects are highlighted
@@ -107,7 +107,7 @@
@references
Scenario: References don't appear in highlighted section titles
- When I search for Reference Section Highlight Test
+ When I search for "Reference Section"
And *Reference* *Section* is the highlighted alttitle of the first search
result
@references
--
To view, visit https://gerrit.wikimedia.org/r/154898
To unsubscribe, visit https://gerrit.wikimedia.org/r/settings
Gerrit-MessageType: merged
Gerrit-Change-Id: I58f031adedf875e0f2723c5c21c7ca91e1088893
Gerrit-PatchSet: 3
Gerrit-Project: mediawiki/extensions/CirrusSearch
Gerrit-Branch: master
Gerrit-Owner: Manybubbles <[email protected]>
Gerrit-Reviewer: Chad <[email protected]>
Gerrit-Reviewer: jenkins-bot <>
_______________________________________________
MediaWiki-commits mailing list
[email protected]
https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits
