Examples of it.unimi.dsi.mg4j.document.DocumentCollection

it.unimi.dsi.mg4j.document.DocumentCollection
A collection of documents.
Classes implementing this interface have additional responsibilities w.r.t. {@link it.unimi.dsi.mg4j.document.DocumentSequence} in that theymust provide random access to the documents, and guarantee the possibility of multiple calls to {@link it.unimi.dsi.mg4j.document.DocumentSequence#iterator()}.
Note, however, that the objects returned by {@link it.unimi.dsi.mg4j.document.DocumentSequence#iterator() iterator()}, {@link #stream(int)} and {@link #document(int)} are, unless explicitly stated otherwise,mutually exclusive. They share a single resource managed by the collection (and disposed by a call to {@link java.io.Closeable#close() close()}), so each time a stream or a document are returned by some method, the ones previously returned are no longer valid, and access to their methods will cause unpredictable behaviour. If you need many documents, you can {@linkplain FlyweightPrototype#copy() obtain a flyweight copy}of the collection.
Warning: implementations of this class are not required to be thread-safe, but they provide {@linkplain FlyweightPrototype flyweight copies}. The {@link #copy()} method is strengthened so to return a instance of this class.

      });


    final JSAPResult jsapResult = jsap.parse( arg );
    if ( jsap.messagePrinted() ) return;


    final DocumentCollection documentCollection = (DocumentCollection)(jsapResult.userSpecified( "collection" ) ? AbstractDocumentSequence.load( jsapResult.getString( "collection" ) ) : 
      jsapResult.userSpecified( "objectCollection" ) ? jsapResult.getObject( "objectCollection" ): null );
    final List<? extends CharSequence> titleList = (List<? extends CharSequence>) ( 
        jsapResult.userSpecified( "titleList" ) ? BinIO.loadObject( jsapResult.getString( "titleList" ) ) :
          jsapResult.userSpecified( "titleFile" ) ? new FileLinesList( jsapResult.getString( "titleFile" ), "UTF-8" ) :
            null );    final String[] basenameWeight = jsapResult.getStringArray( "basenameWeight" );

View Full Code Here

        ObjectArrayList<ResultItem> resultItems = new ObjectArrayList<ResultItem>();


        if ( ! results.isEmpty() ) {
          SelectedInterval[] selectedInterval = null;


          final DocumentCollection collection = documentCollection != null ? documentCollection.copy() : null;


          for( int i = 0; i < results.size(); i++ ) {
            DocumentScoreInfo<Reference2ObjectMap<Index,SelectedInterval[]>> dsi = results.get( i );
            LOGGER.debug( "Intervals for item " + i );
            final ResultItem resultItem = new ResultItem( dsi.document, dsi.score );
            resultItems.add( resultItem );


            if ( collection != null ) {
              final Document document = collection.document( dsi.document );
              // If both collection and title list are present, we override the collection title (cfr. Query)
              resultItem.title = StringEscapeUtils.escapeHtml( titleList != null ? titleList.get( resultItem.doc ).toString() : document.title().toString() );
              if ( useUri ) {
                if ( document.uri() != null ) resultItem.uri = StringEscapeUtils.escapeHtml( document.uri().toString() );
              }
              else {
                if ( document.uri() != null ) {
                  String stringUri = document.uri().toString();
                  // TODO: this is a quick patch to get the file server running with relative files
                  final String documentUri = URLEncoder.encode( derelativise
                  ? new File( stringUri.startsWith( "file:" ) ? stringUri.substring( 5 ) : stringUri ).getAbsoluteFile().toURI().toASCIIString()
                      : document.uri().toString(), "UTF-8" );
                  resultItem.uri = StringEscapeUtils.escapeHtml( "./Item?doc=" + resultItem.doc + "&m=" + urlEncodedMimeType + "&uri=" + documentUri );
                }
                else resultItem.uri = StringEscapeUtils.escapeHtml( "./Item?doc=" + resultItem.doc + "&m=" + urlEncodedMimeType );
              }
              
              MarkingMutableString snippet = new MarkingMutableString( TextMarker.HTML_STRONG, MarkingMutableString.HTML_ESCAPE ); 
              
              for( int j = 0; j < sortedIndex.length; j++ ) {
                if ( ! sortedIndex[ j ].hasPositions || dsi.info == null ) continue;
                selectedInterval = dsi.info.get( sortedIndex[ j ] );
                if ( selectedInterval != null ) {
                  final int field = documentCollection.factory().fieldIndex( sortedIndex[ j ].field );
                  // If the field is not present (e.g., because of parallel indexing) or it is not text we skip
                  if ( field == -1 || documentCollection.factory().fieldType( field ) != DocumentFactory.FieldType.TEXT ) continue;
                  LOGGER.debug( "Found intervals for " + sortedIndex[ j ].field + " (" + field + ")" );
                  final Reader content = (Reader)document.content( field );
                  snippet.startField( selectedInterval ).appendAndMark( document.wordReader( field ).setReader( content ) ).endField();
                }
                if ( LOGGER.isDebugEnabled() ) LOGGER.debug( sortedIndex[ j ].field + ": " + ( selectedInterval == null ? null : Arrays.asList( selectedInterval ) ) ); 
                document.close();
              }
              
              resultItem.text = snippet; 
            }
            else {
              if ( titleList != null ) {
                // TODO: this is a bit radical
                resultItem.title = resultItem.uri = titleList.get( resultItem.doc );
              }
              else {
                resultItem.title = "Document #" +  resultItem.doc;
                resultItem.uri = new MutableString( "./Item?doc=" ).append( resultItem.doc ).append( "&m=" ).append( urlEncodedMimeType );
              }
              
              MutableString text = new MutableString();
              for( Iterator<Index> j = indexMap.values().iterator(); j.hasNext(); ) {
                final Index index = j.next();
                selectedInterval = dsi.info.get( index );
                if ( selectedInterval != null )
                  text.append( "<p>" ).append( index.field ).append( ": " ).append( Arrays.asList( selectedInterval ) );
                LOGGER.debug( index.field + ": " + ( selectedInterval == null ? null : Arrays.asList( selectedInterval ) ) ); 
              }
              resultItem.text = text;
            }
          }
          
          if ( collection != null ) collection.close();
        }


        
        // Note that if we pass an array to the template we lose the possibility of measuring its length.
        context.put( "result", resultItems );

View Full Code Here

    return HttpQueryServer.setLiberalResourceLoading( super.loadConfiguration( config ) );
  }


  public Template handleRequest( final HttpServletRequest request, final HttpServletResponse response, final Context context ) throws Exception {
    if ( request.getParameter( "doc" ) != null ) {
      DocumentCollection collection = (DocumentCollection)getServletContext().getAttribute( "collection" );
      response.setContentType( request.getParameter( "m" ) );
      response.setCharacterEncoding( "UTF-8" );
      final Document document = collection.document( Integer.parseInt( request.getParameter( "doc" ) ) );
      final DocumentFactory factory = collection.factory();
      final ObjectArrayList<String> fields = new ObjectArrayList<String>();
      final int numberOfFields = factory.numberOfFields();
      
      LOGGER.debug( "ParsingFactory declares " + numberOfFields + " fields"  );

View Full Code Here

  private static final int skip = Integer.getInteger( InputStreamItem.class.getName() + ".skip", 0 ).intValue();
  
  protected void doGet( final HttpServletRequest request, final HttpServletResponse response ) throws IOException {
    try {
      if ( request.getParameter( "m" ) != null && request.getParameter( "doc" ) != null ) {
        DocumentCollection collection = (DocumentCollection)getServletContext().getAttribute( "collection" );
        if ( collection == null ) LOGGER.fatal( "The servlet context does not contain a document collection." );
        response.setContentType( request.getParameter( "m" ) );
        response.setCharacterEncoding( "UTF-8" );
        InputStream rawContent = collection.stream( Integer.parseInt( request.getParameter( "doc" ) ) );
        for( int i = skip; i-- != 0; ) rawContent.reset();
        IOUtils.copy( rawContent, response.getOutputStream() );
      }
    } catch( RuntimeException e ) {
      e.printStackTrace();

View Full Code Here

TOP

Related Classes of it.unimi.dsi.mg4j.document.DocumentCollection

it.unimi.dsi.mg4j.query.GenericItem

it.unimi.dsi.mg4j.query.InputStreamItem

it.unimi.dsi.mg4j.query.Query

it.unimi.dsi.mg4j.query.QueryServlet

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.