Code Coverage |
||||||||||
Classes and Traits |
Functions and Methods |
Lines |
||||||||
| Total | |
0.00% |
0 / 1 |
|
0.00% |
0 / 4 |
CRAP | |
0.00% |
0 / 31 |
| EntitySearchUtils | |
0.00% |
0 / 1 |
|
0.00% |
0 / 4 |
156 | |
0.00% |
0 / 31 |
| makeConstScoreQuery | |
0.00% |
0 / 1 |
2 | |
0.00% |
0 / 4 |
|||
| normalizeId | |
0.00% |
0 / 1 |
30 | |
0.00% |
0 / 9 |
|||
| parseOrNull | |
0.00% |
0 / 1 |
6 | |
0.00% |
0 / 4 |
|||
| findTermForDisplay | |
0.00% |
0 / 1 |
20 | |
0.00% |
0 / 14 |
|||
| <?php | |
| namespace Wikibase\Search\Elastic; | |
| use Elastica\Query\ConstantScore; | |
| use Elastica\Query\MatchQuery; | |
| use Wikibase\DataModel\Entity\EntityIdParser; | |
| use Wikibase\DataModel\Entity\EntityIdParsingException; | |
| use Wikibase\DataModel\Term\Term; | |
| use Wikibase\Lib\TermLanguageFallbackChain; | |
| /** | |
| * Utilities useful for entity searches. | |
| */ | |
| final class EntitySearchUtils { | |
| /** | |
| * Create constant score query for a field. | |
| * @param string $field | |
| * @param string|double $boost | |
| * @param string $text | |
| * @return ConstantScore | |
| */ | |
| public static function makeConstScoreQuery( $field, $boost, $text ) { | |
| $csquery = new ConstantScore(); | |
| $csquery->setFilter( new MatchQuery( $field, $text ) ); | |
| $csquery->setBoost( $boost ); | |
| return $csquery; | |
| } | |
| /** | |
| * If the text looks like ID, normalize it to ID title | |
| * Cases handled: | |
| * - q42 | |
| * - (q42) | |
| * - leading/trailing spaces | |
| * - http://www.wikidata.org/entity/Q42 | |
| * @param string $text | |
| * @param EntityIdParser $idParser | |
| * @return string Normalized ID or original string | |
| */ | |
| public static function normalizeId( $text, EntityIdParser $idParser ) { | |
| // TODO: this is a bit hacky, better way would be to make the field case-insensitive | |
| // or add new subfiled which is case-insensitive | |
| $text = strtoupper( str_replace( [ '(', ')' ], '', trim( $text ) ) ); | |
| $id = self::parseOrNull( $text, $idParser ); | |
| if ( $id ) { | |
| return $id->getSerialization(); | |
| } | |
| if ( preg_match( '/\b(\w+)$/', $text, $matches ) && $matches[1] ) { | |
| $id = self::parseOrNull( $matches[1], $idParser ); | |
| if ( $id ) { | |
| return $id->getSerialization(); | |
| } | |
| } | |
| return $text; | |
| } | |
| /** | |
| * Parse entity ID or return null | |
| * @param string $text | |
| * @param EntityIdParser $idParser | |
| * @return null|\Wikibase\DataModel\Entity\EntityId | |
| */ | |
| public static function parseOrNull( $text, EntityIdParser $idParser ) { | |
| try { | |
| $id = $idParser->parse( $text ); | |
| } catch ( EntityIdParsingException $ex ) { | |
| return null; | |
| } | |
| return $id; | |
| } | |
| /** | |
| * Locate label for display among the source data, basing on fallback chain. | |
| * @param array $sourceData | |
| * @param string $field | |
| * @param TermLanguageFallbackChain $termFallbackChain | |
| * @return null|Term | |
| */ | |
| public static function findTermForDisplay( $sourceData, $field, TermLanguageFallbackChain $termFallbackChain ) { | |
| if ( empty( $sourceData[$field] ) ) { | |
| return null; | |
| } | |
| $data = $sourceData[$field]; | |
| $first = reset( $data ); | |
| if ( is_array( $first ) ) { | |
| // If we have multiple, like for labels, extract the first one | |
| $labels_data = array_map( | |
| static function ( $data ) { | |
| return $data[0] ?? null; | |
| }, | |
| $data | |
| ); | |
| } else { | |
| $labels_data = $data; | |
| } | |
| // Drop empty ones | |
| $labels_data = array_filter( $labels_data ); | |
| $preferredValue = $termFallbackChain->extractPreferredValueOrAny( $labels_data ); | |
| if ( $preferredValue ) { | |
| return new Term( $preferredValue['language'], $preferredValue['value'] ); | |
| } | |
| return null; | |
| } | |
| } |