DjVuHandler.php

Go to the documentation of this file.

<?php
use MediaWiki\MediaWikiServices;
use MediaWiki\Shell\Shell;
 
class DjVuHandler extends ImageHandler {
    private const EXPENSIVE_SIZE_LIMIT = 10485760; // 10MiB
 
    // Constants for getHandlerState
    private const STATE_DJVU_IMAGE = 'djvuImage';
    private const STATE_TEXT_TREE = 'djvuTextTree';
    private const STATE_META_TREE = 'djvuMetaTree';
 
    public function isEnabled() {
        global $wgDjvuRenderer, $wgDjvuDump, $wgDjvuToXML;
        if ( !$wgDjvuRenderer || ( !$wgDjvuDump && !$wgDjvuToXML ) ) {
            wfDebug( "DjVu is disabled, please set \$wgDjvuRenderer and \$wgDjvuDump" );
 
            return false;
        } else {
            return true;
        }
    }
 
    public function mustRender( $file ) {
        return true;
    }
 
    public function isExpensiveToThumbnail( $file ) {
        return $file->getSize() > static::EXPENSIVE_SIZE_LIMIT;
    }
 
    public function isMultiPage( $file ) {
        return true;
    }
 
    public function getParamMap() {
        return [
            'img_width' => 'width',
            'img_page' => 'page',
        ];
    }
 
    public function validateParam( $name, $value ) {
        if ( $name === 'page' && trim( $value ) !== (string)intval( $value ) ) {
            // Extra junk on the end of page, probably actually a caption
            // e.g. [[File:Foo.djvu|thumb|Page 3 of the document shows foo]]
            return false;
        }
        if ( in_array( $name, [ 'width', 'height', 'page' ] ) ) {
            if ( $value <= 0 ) {
                return false;
            } else {
                return true;
            }
        } else {
            return false;
        }
    }
 
    public function makeParamString( $params ) {
        $page = $params['page'] ?? 1;
        if ( !isset( $params['width'] ) ) {
            return false;
        }
 
        return "page{$page}-{$params['width']}px";
    }
 
    public function parseParamString( $str ) {
        $m = false;
        if ( preg_match( '/^page(\d+)-(\d+)px$/', $str, $m ) ) {
            return [ 'width' => $m[2], 'page' => $m[1] ];
        } else {
            return false;
        }
    }
 
    protected function getScriptParams( $params ) {
        return [
            'width' => $params['width'],
            'page' => $params['page'],
        ];
    }
 
    public function doTransform( $image, $dstPath, $dstUrl, $params, $flags = 0 ) {
        global $wgDjvuRenderer, $wgDjvuPostProcessor;
 
        if ( !$this->normaliseParams( $image, $params ) ) {
            return new TransformParameterError( $params );
        }
        $width = $params['width'];
        $height = $params['height'];
        $page = $params['page'];
 
        if ( $flags & self::TRANSFORM_LATER ) {
            $params = [
                'width' => $width,
                'height' => $height,
                'page' => $page
            ];
 
            return new ThumbnailImage( $image, $dstUrl, $dstPath, $params );
        }
 
        if ( !wfMkdirParents( dirname( $dstPath ), null, __METHOD__ ) ) {
            return new MediaTransformError(
                'thumbnail_error',
                $width,
                $height,
                wfMessage( 'thumbnail_dest_directory' )
            );
        }
 
        // Get local copy source for shell scripts
        // Thumbnail extraction is very inefficient for large files.
        // Provide a way to pool count limit the number of downloaders.
        if ( $image->getSize() >= 1e7 ) { // 10MB
            $work = new PoolCounterWorkViaCallback( 'GetLocalFileCopy', sha1( $image->getName() ),
                [
                    'doWork' => static function () use ( $image ) {
                        return $image->getLocalRefPath();
                    }
                ]
            );
            $srcPath = $work->execute();
        } else {
            $srcPath = $image->getLocalRefPath();
        }
 
        if ( $srcPath === false ) { // Failed to get local copy
            wfDebugLog( 'thumbnail',
                sprintf( 'Thumbnail failed on %s: could not get local copy of "%s"',
                    wfHostname(), $image->getName() ) );
 
            return new MediaTransformError( 'thumbnail_error',
                $params['width'], $params['height'],
                wfMessage( 'filemissing' )
            );
        }
 
        # Use a subshell (brackets) to aggregate stderr from both pipeline commands
        # before redirecting it to the overall stdout. This works in both Linux and Windows XP.
        $cmd = '(' . Shell::escape(
            $wgDjvuRenderer,
            "-format=ppm",
            "-page={$page}",
            "-size={$params['physicalWidth']}x{$params['physicalHeight']}",
            $srcPath );
        if ( $wgDjvuPostProcessor ) {
            $cmd .= " | {$wgDjvuPostProcessor}";
        }
        $cmd .= ' > ' . Shell::escape( $dstPath ) . ') 2>&1';
        wfDebug( __METHOD__ . ": $cmd" );
        $retval = '';
        $err = wfShellExec( $cmd, $retval );
 
        $removed = $this->removeBadFile( $dstPath, $retval );
        if ( $retval != 0 || $removed ) {
            $this->logErrorForExternalProcess( $retval, $err, $cmd );
            return new MediaTransformError( 'thumbnail_error', $width, $height, $err );
        } else {
            $params = [
                'width' => $width,
                'height' => $height,
                'page' => $page
            ];
 
            return new ThumbnailImage( $image, $dstUrl, $dstPath, $params );
        }
    }
 
    private function getDjVuImage( $state, $path ) {
        $deja = $state->getHandlerState( self::STATE_DJVU_IMAGE );
        if ( !$deja ) {
            $deja = new DjVuImage( $path );
            $state->setHandlerState( self::STATE_DJVU_IMAGE, $deja );
        }
        return $deja;
    }
 
    private function getXMLMetadata( File $file ) {
        $unser = $file->getMetadataArray();
        if ( isset( $unser['error'] ) ) {
            return false;
        } elseif ( isset( $unser['xml'] ) ) {
            return $unser['xml'];
        } elseif ( isset( $unser['_error'] )
            && is_string( $unser['_error'] )
            && substr( $unser['_error'], 0, 3 ) === '<?xml'
        ) {
            // Old style. Not serialized but instead just a raw string of XML.
            return $unser['_error'];
        } else {
            return false;
        }
    }
 
    public function getMetaTree( $image, $gettext = false ) {
        if ( $gettext && $image->getHandlerState( self::STATE_TEXT_TREE ) ) {
            return $image->getHandlerState( self::STATE_TEXT_TREE );
        }
        if ( !$gettext && $image->getHandlerState( self::STATE_META_TREE ) ) {
            return $image->getHandlerState( self::STATE_META_TREE );
        }
 
        $xml = $this->getXMLMetadata( $image );
        if ( !$xml ) {
            wfDebug( "DjVu XML metadata is invalid or missing, should have been fixed in upgradeRow" );
 
            return false;
        }
 
        $trees = $this->extractTreesFromXML( $xml );
        $image->setHandlerState( self::STATE_TEXT_TREE, $trees['TextTree'] );
        $image->setHandlerState( self::STATE_META_TREE, $trees['MetaTree'] );
 
        if ( $gettext ) {
            return $trees['TextTree'];
        } else {
            return $trees['MetaTree'];
        }
    }
 
    protected function extractTreesFromXML( $xml ) {
        Wikimedia\suppressWarnings();
        try {
            // Set to false rather than null to avoid further attempts
            $metaTree = false;
            $textTree = false;
            $tree = new SimpleXMLElement( $xml, LIBXML_PARSEHUGE );
            if ( $tree->getName() == 'mw-djvu' ) {
                foreach ( $tree->children() as $b ) {
                    if ( $b->getName() == 'DjVuTxt' ) {
                        // @todo File::djvuTextTree and File::dejaMetaTree are declared
                        // dynamically. Add a public File::$data to facilitate this?
                        $textTree = $b;
                    } elseif ( $b->getName() == 'DjVuXML' ) {
                        $metaTree = $b;
                    }
                }
            } else {
                $metaTree = $tree;
            }
        } catch ( Exception $e ) {
            wfDebug( "Bogus multipage XML metadata" );
        }
        Wikimedia\restoreWarnings();
 
        return [ 'MetaTree' => $metaTree, 'TextTree' => $textTree ];
    }
 
    public function getThumbType( $ext, $mime, $params = null ) {
        global $wgDjvuOutputExtension;
        static $mime;
        if ( !isset( $mime ) ) {
            $magic = MediaWiki\MediaWikiServices::getInstance()->getMimeAnalyzer();
            $mime = $magic->getMimeTypeFromExtensionOrNull( $wgDjvuOutputExtension );
        }
 
        return [ $wgDjvuOutputExtension, $mime ];
    }
 
    public function getSizeAndMetadata( $state, $path ) {
        wfDebug( "Getting DjVu metadata for $path" );
 
        $djvuImage = $this->getDjVuImage( $state, $path );
        $xml = $djvuImage->retrieveMetaData();
        if ( $xml === false ) {
            // Special value so that we don't repetitively try and decode a broken file.
            $metadata = [ 'error' => 'Error extracting metadata' ];
        } else {
            $metadata = [ 'xml' => $xml ];
        }
        return [ 'metadata' => $metadata ] + $djvuImage->getImageSize();
    }
 
    public function getMetadataType( $image ) {
        return 'djvuxml';
    }
 
    public function isFileMetadataValid( $image ) {
        return $image->getMetadataArray() ? self::METADATA_GOOD : self::METADATA_BAD;
    }
 
    public function pageCount( File $image ) {
        $info = $this->getDimensionInfo( $image );
 
        return $info ? $info['pageCount'] : false;
    }
 
    public function getPageDimensions( File $image, $page ) {
        $index = $page - 1; // MW starts pages at 1
 
        $info = $this->getDimensionInfo( $image );
        if ( $info && isset( $info['dimensionsByPage'][$index] ) ) {
            return $info['dimensionsByPage'][$index];
        }
 
        return false;
    }
 
    protected function getDimensionInfo( File $file ) {
        $cache = MediaWikiServices::getInstance()->getMainWANObjectCache();
        return $cache->getWithSetCallback(
            $cache->makeKey( 'file-djvu', 'dimensions', $file->getSha1() ),
            $cache::TTL_INDEFINITE,
            function () use ( $file ) {
                $tree = $this->getMetaTree( $file );
                return $this->getDimensionInfoFromMetaTree( $tree );
            },
            [ 'pcTTL' => $cache::TTL_INDEFINITE ]
        );
    }
 
    protected function getDimensionInfoFromMetaTree( $metatree ) {
        if ( !$metatree ) {
            return false;
        }
 
        $dimsByPage = [];
        $count = count( $metatree->xpath( '//OBJECT' ) );
        for ( $i = 0; $i < $count; $i++ ) {
            $o = $metatree->BODY[0]->OBJECT[$i];
            if ( $o ) {
                $dimsByPage[$i] = [
                    'width' => (int)$o['width'],
                    'height' => (int)$o['height'],
                ];
            } else {
                $dimsByPage[$i] = false;
            }
        }
 
        return [ 'pageCount' => $count, 'dimensionsByPage' => $dimsByPage ];
    }
 
    public function getPageText( File $image, $page ) {
        $tree = $this->getMetaTree( $image, true );
        if ( !$tree ) {
            return false;
        }
 
        $o = $tree->BODY[0]->PAGE[$page - 1];
        return $o ? (string)$o['value'] : false;
    }
}