WikiTextStructure.php

Go to the documentation of this file.

<?php

use HtmlFormatter\HtmlFormatter;

class WikiTextStructure {
    private $openingText;
    private $allText;
    private $auxText = [];
    private $parserOutput;

    private $excludedElementSelectors = [
        // "it looks like you don't have javascript enabled..." – do not need to index
        'audio', 'video',
        // CSS stylesheets aren't content
        'style',
        // The [1] for references from Cite
        'sup.reference',
        // The ↑ next to references in the references section from Cite
        '.mw-cite-backlink',
        // Headings are already indexed in their own field.
        'h1', 'h2', 'h3', 'h4', 'h5', 'h6',
        // Collapsed fields are hidden by default so we don't want them showing up.
        '.autocollapse',
        // Content explicitly decided to be not searchable by editors such
        // as custom navigation templates.
        '.navigation-not-searchable',
        // User-facing interface code prompting the user to act from WikibaseMediaInfo
        '.wbmi-entityview-emptyCaption',
    ];

    private $auxiliaryElementSelectors = [
        // Thumbnail captions aren't really part of the text proper
        '.thumbcaption',
        // Neither are tables
        'table',
        // Common style for "See also:".
        '.rellink',
        // Common style for calling out helpful links at the top of the article.
        '.dablink',
        // New class users can use to mark stuff as auxiliary to searches.
        '.searchaux',
    ];

    public function __construct( ParserOutput $parserOutput ) {
        $this->parserOutput = $parserOutput;
    }

    public function headings() {
        $headings = [];
        $ignoredHeadings = $this->getIgnoredHeadings();
        foreach ( $this->parserOutput->getSections() as $heading ) {
            $heading = $heading[ 'line' ];

            // Some wikis wrap the brackets in a span:
            // https://en.wikipedia.org/wiki/MediaWiki:Cite_reference_link
            $heading = preg_replace( '/<\/?span>/', '', $heading );
            // Normalize [] so the following regexp would work.
            $heading = preg_replace( [ '/&#91;/', '/&#93;/' ], [ '[', ']' ], $heading );
            $heading = preg_replace( '/<sup>\s*\[\s*\d+\s*\]\s*<\/sup>/is', '', $heading );

            // Strip tags from the heading or else we'll display them (escaped) in search results
            $heading = trim( Sanitizer::stripAllTags( $heading ) );

            // Note that we don't take the level of the heading into account - all headings are equal.
            // Except the ones we ignore.
            if ( !in_array( $heading, $ignoredHeadings ) ) {
                $headings[] = $heading;
            }
        }
        return $headings;
    }

    public static function parseSettingsInMessage( $message ) {
        $lines = explode( "\n", $message );
        $lines = preg_replace( '/#.*$/', '', $lines ); // Remove comments
        $lines = array_map( 'trim', $lines );          // Remove extra spaces
        $lines = array_filter( $lines );               // Remove empty lines
        return $lines;
    }

    private function getIgnoredHeadings() {
        static $ignoredHeadings = null;
        if ( $ignoredHeadings === null ) {
            $ignoredHeadings = [];
            $source = wfMessage( 'search-ignored-headings' )->inContentLanguage();
            if ( $source->isBlank() ) {
                // Try old version too, just in case
                $source = wfMessage( 'cirrussearch-ignored-headings' )->inContentLanguage();
            }
            if ( !$source->isDisabled() ) {
                $lines = self::parseSettingsInMessage( $source->plain() );
                $ignoredHeadings = $lines;               // Now we just have headings!
            }
        }
        return $ignoredHeadings;
    }

    private function extractWikitextParts() {
        if ( !is_null( $this->allText ) ) {
            return;
        }
        $text = $this->parserOutput->getText( [
            'enableSectionEditTokens' => false,
            'allowTOC' => false,
        ] );
        if ( $text === '' ) {
            $this->allText = "";
            // empty text - nothing to seek here
            return;
        }
        $opening = null;

        $this->openingText = $this->extractHeadingBeforeFirstHeading( $text );

        $formatter = new HtmlFormatter( $text );

        // Strip elements from the page that we never want in the search text.
        $formatter->remove( $this->excludedElementSelectors );
        $formatter->filterContent();

        // Strip elements from the page that are auxiliary text.  These will still be
        // searched but matches will be ranked lower and non-auxiliary matches will be
        // preferred in highlighting.
        $formatter->remove( $this->auxiliaryElementSelectors );
        $auxiliaryElements = $formatter->filterContent();
        $this->allText = trim( Sanitizer::stripAllTags( $formatter->getText() ) );
        foreach ( $auxiliaryElements as $auxiliaryElement ) {
            $this->auxText[] =
                trim( Sanitizer::stripAllTags( $formatter->getText( $auxiliaryElement ) ) );
        }
    }

    private function extractHeadingBeforeFirstHeading( $text ) {
        $matches = [];
        if ( !preg_match( '/<h[123456]>/', $text, $matches, PREG_OFFSET_CAPTURE ) ) {
            // There isn't a first heading so we interpret this as the article
            // being entirely without heading.
            return null;
        }
        $text = substr( $text, 0, $matches[ 0 ][ 1 ] );
        if ( !$text ) {
            // There isn't any text before the first heading so we declare there isn't
            // a first heading.
            return null;
        }

        $formatter = new HtmlFormatter( $text );
        $formatter->remove( $this->excludedElementSelectors );
        $formatter->remove( $this->auxiliaryElementSelectors );
        $formatter->filterContent();
        $text = trim( Sanitizer::stripAllTags( $formatter->getText() ) );

        if ( !$text ) {
            // There isn't any text after filtering before the first heading so we declare
            // that there isn't a first heading.
            return null;
        }

        return $text;
    }

    public function getOpeningText() {
        $this->extractWikitextParts();
        return $this->openingText;
    }

    public function getMainText() {
        $this->extractWikitextParts();
        return $this->allText;
    }

    public function getAuxiliaryText() {
        $this->extractWikitextParts();
        return $this->auxText;
    }

    public function getDefaultSort() {
        return $this->parserOutput->getProperty( 'defaultsort' );
    }
}