jenkins-bot has submitted this change and it was merged.
Change subject: Return assessment level
......................................................................
Return assessment level
Tries to guess image assessment level (quality, featured etc) from Commons
categories.
Change-Id: I137607bb38ee4ceba23b738fd86cc8c713290874
---
M CommonsMetadata_body.php
1 file changed, 44 insertions(+), 1 deletion(-)
Approvals:
MarkTraceur: Looks good to me, approved
jenkins-bot: Verified
diff --git a/CommonsMetadata_body.php b/CommonsMetadata_body.php
index 8caa875..6520062 100755
--- a/CommonsMetadata_body.php
+++ b/CommonsMetadata_body.php
@@ -31,6 +31,19 @@
);
/**
+ * Mapping of category names to assesment levels. Array keys are
regexps which will be
+ * matched case-insensitively against category names; the first match
is returned.
+ * @var array
+ */
+ protected static $assessmentCategories = array(
+ 'poty' => '/^pictures of the year \(.*\)/',
+ 'potd' => '/^pictures of the day \(.*\)/',
+ 'featured' => '/^featured (pictures|sounds) on wikimedia
commons/',
+ 'quality' => '/^quality images/',
+ 'valued' => '/^valued images/',
+ );
+
+ /**
* @param $doc String The html to parse
* @param String|boolean $lang Language code or false for all langs.
* @return Array The properties extracted from the page.
@@ -91,8 +104,16 @@
// For now only get the immediate categories
$categories = self::getCategories( $file );
+
+ $assessments = self::getAssessmentsAndRemoveFromCategories(
$categories );
+
$combinedMeta['Categories'] = array(
'value' => implode( '|', $categories ),
+ 'source' => 'commons-categories',
+ );
+
+ $combinedMeta['Assessments'] = array(
+ 'value' => implode('|', $assessments),
'source' => 'commons-categories',
);
@@ -165,7 +186,29 @@
unset( $categories[$i] );
}
}
- return array_merge( $licenses ); // renumber to avoid holes in
array
+ $categories = array_merge( $categories ); // renumber to avoid
holes in array
+ return $licenses;
+ }
+
+ /**
+ * Matches category names to a category => assessment mapping, removes
the matching categories
+ * and returns the corresponding assessments (valued image, picture of
the day etc).
+ * @param array $categories a list of human-readable category names.
+ * @return array
+ * FIXME categories do not work with Commons-hosted images due to bug
56598
+ */
+ protected static function getAssessmentsAndRemoveFromCategories(
&$categories ) {
+ $assessments = array();
+ foreach ( $categories as $i => $category ) {
+ foreach ( self::$assessmentCategories as
$assessmentType => $regexp ) {
+ if ( preg_match( $regexp, $category ) ) {
+ $assessments[] = $assessmentType;
+ unset( $categories[$i] );
+ }
+ }
+ }
+ $categories = array_merge( $categories ); // renumber to avoid
holes in array
+ return array_unique($assessments); // potd/poty can happen
multiple times
}
/**
--
To view, visit https://gerrit.wikimedia.org/r/90919
To unsubscribe, visit https://gerrit.wikimedia.org/r/settings
Gerrit-MessageType: merged
Gerrit-Change-Id: I137607bb38ee4ceba23b738fd86cc8c713290874
Gerrit-PatchSet: 5
Gerrit-Project: mediawiki/extensions/CommonsMetadata
Gerrit-Branch: master
Gerrit-Owner: Gergő Tisza <[email protected]>
Gerrit-Reviewer: Gergő Tisza <[email protected]>
Gerrit-Reviewer: MarkTraceur <[email protected]>
Gerrit-Reviewer: jenkins-bot
_______________________________________________
MediaWiki-commits mailing list
[email protected]
https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits
