Java HtmlMapper类代码示例

import org.apache.tika.parser.html.HtmlMapper; //导入依赖的package包/类
@SuppressWarnings({ "rawtypes", "unchecked" })
@Override
public void prepare(Map conf, TopologyContext context,
        OutputCollector collector) {

    emitOutlinks = ConfUtils.getBoolean(conf, "parser.emitOutlinks", true);

    urlFilters = URLFilters.fromConf(conf);

    parseFilters = ParseFilters.fromConf(conf);

    upperCaseElementNames = ConfUtils.getBoolean(conf,
            "parser.uppercase.element.names", true);

    extractEmbedded = ConfUtils.getBoolean(conf, "parser.extract.embedded",
            false);

    String htmlmapperClassName = ConfUtils.getString(conf,
            "parser.htmlmapper.classname",
            "org.apache.tika.parser.html.IdentityHtmlMapper");

    try {
        HTMLMapperClass = Class.forName(htmlmapperClassName);
        boolean interfaceOK = HtmlMapper.class
                .isAssignableFrom(HTMLMapperClass);
        if (!interfaceOK) {
            throw new RuntimeException("Class " + htmlmapperClassName
                    + " does not implement HtmlMapper");
        }
    } catch (ClassNotFoundException e) {
        LOG.error("Can't load class {}", htmlmapperClassName);
        throw new RuntimeException("Can't load class "
                + htmlmapperClassName);
    }

    // instanciate Tika
    long start = System.currentTimeMillis();
    tika = new Tika();
    long end = System.currentTimeMillis();

    LOG.debug("Tika loaded in {} msec", end - start);

    this.collector = collector;

    this.eventCounter = context.registerMetric(this.getClass()
            .getSimpleName(), new MultiCountMetric(), 10);

    this.metadataTransfer = MetadataTransfer.getInstance(conf);
}
 

import org.apache.tika.parser.html.HtmlMapper; //导入依赖的package包/类
/**
 * Create a pull-parser from the given {@link TikaInputStream}.
 *
 * @param input the stream to extract from
 * @param document file that is being extracted from
 * @return A pull-parsing reader.
 */
protected Reader extract(final Document document, final TikaInputStream input) throws IOException {
	final Metadata metadata = document.getMetadata();
	final ParseContext context = new ParseContext();
	final AutoDetectParser autoDetectParser = new AutoDetectParser(defaultParser);
	final Parser parser;

	if (null != digester) {
		parser = new DigestingParser(autoDetectParser, digester);
	} else {
		parser = autoDetectParser;
	}

	if (!ocrDisabled) {
		context.set(TesseractOCRConfig.class, ocrConfig);
	}

	context.set(PDFParserConfig.class, pdfConfig);

	// Set a fallback parser that outputs an empty document for empty files,
	// otherwise throws an exception.
	autoDetectParser.setFallback(FallbackParser.INSTANCE);

	// Only include "safe" tags in the HTML output from Tika's HTML parser.
	// This excludes script tags and objects.
	context.set(HtmlMapper.class, DefaultHtmlMapper.INSTANCE);

	final Reader reader;
	final Function<Writer, ContentHandler> handler;

	if (OutputFormat.HTML == outputFormat) {
		handler = (writer) -> new ExpandedTitleContentHandler(new HTML5Serializer(writer));
	} else {

		// The default BodyContentHandler is used when constructing the ParsingReader for text output, but
		// because only the body of embeds is pushed to the content handler further down the line, we can't
		// expect a body tag.
		handler = WriteOutContentHandler::new;
	}

	if (EmbedHandling.SPAWN == embedHandling) {
		context.set(Parser.class, parser);
		context.set(EmbeddedDocumentExtractor.class, new EmbedSpawner(document, context, embedOutput, handler));
	} else if (EmbedHandling.CONCATENATE == embedHandling) {
		context.set(Parser.class, parser);
		context.set(EmbeddedDocumentExtractor.class, new EmbedParser(document, context));
	} else {
		context.set(Parser.class, EmptyParser.INSTANCE);
		context.set(EmbeddedDocumentExtractor.class, new EmbedBlocker());
	}

	if (OutputFormat.HTML == outputFormat) {
		reader = new ParsingReader(parser, input, metadata, context, handler);
	} else {
		reader = new ParsingReader(parser, input, metadata, context);
	}

	return reader;
}