org.apache.tika.sax.XHTMLContentHandler Java Examples

/**
 * Reads the contents of the given stream and write it to the given XHTML
 * content handler. The stream is closed once fully processed.
 *
 * @param stream
 *          Stream where is the result of ocr
 * @param xhtml
 *          XHTML content handler
 * @throws SAXException
 *           if the XHTML SAX events could not be handled
 * @throws IOException
 *           if an input error occurred
 */
private void extractOutput(InputStream stream, XHTMLContentHandler xhtml) throws SAXException, IOException {

    xhtml.startDocument();
    xhtml.startElement("div");
    try (Reader reader = new InputStreamReader(stream, UTF_8)) {
        char[] buffer = new char[1024];
        for (int n = reader.read(buffer); n != -1; n = reader.read(buffer)) {
            if (n > 0)
                xhtml.characters(buffer, 0, n);
        }
    }
    xhtml.endElement("div");
    xhtml.endDocument();
}
 

@Override
public void parseInline(final InputStream in, final XHTMLContentHandler xhtml, final ParseContext context,
                        final TesseractOCRConfig config)
		throws IOException, SAXException, TikaException {
	if (null != outputPath) {
		cachedParse(in, xhtml, new Metadata(), context, null == config ?
				context.get(TesseractOCRConfig.class, DEFAULT_CONFIG) : config, true);
	} else {
		super.parseInline(in, xhtml, context, config);
	}
}
 

private void parseToCache(final TikaInputStream tis, final ContentHandler handler, final Metadata metadata,
                          final ParseContext context, final TesseractOCRConfig config, final boolean inline,
                          final Writer writer) throws SAXException, IOException, TikaException {
	final ContentHandler tee = new TeeContentHandler(handler, new WriteOutContentHandler(writer));

	if (inline) {
		super.parseInline(tis, new XHTMLContentHandler(tee, metadata), context, config);
	} else {
		super.parse(tis, tee, metadata, context);
	}
}
 

private void readFully(final Reader reader, final XHTMLContentHandler xhtml) throws IOException, SAXException {
	final char[] buffer = new char[1024];

	for (int n = reader.read(buffer); n != -1; n = reader.read(buffer)) {
		if (n > 0) {
			xhtml.characters(buffer, 0, n);
		}
	}
}
 

@Override
public void parse(InputStream stream, ContentHandler handler, Metadata metadata, ParseContext context)
        throws IOException, SAXException, TikaException {
    TesseractOCRConfig config = context.get(TesseractOCRConfig.class, DEFAULT_CONFIG);

    // If Tesseract is not on the path with the current config, do not try to run OCR
    // getSupportedTypes shouldn't have listed us as handling it, so this should only
    //  occur if someone directly calls this parser, not via DefaultParser or similar
    if (! hasTesseract(config))
        return;

    XHTMLContentHandler xhtml = new XHTMLContentHandler(handler, metadata);

    TemporaryResources tmp = new TemporaryResources();
    File output = null;
    try {
        TikaInputStream tikaStream = TikaInputStream.get(stream, tmp);
        File input = tikaStream.getFile();
        long size = tikaStream.getLength();

        if (size >= config.getMinFileSizeToOcr() && size <= config.getMaxFileSizeToOcr()) {

            output = tmp.createTemporaryFile();
            doOCR(input, output, config);

            // Tesseract appends .txt to output file name
            output = new File(output.getAbsolutePath() + ".txt");

            if (output.exists())
                extractOutput(new FileInputStream(output), xhtml);

        }

        // Temporary workaround for TIKA-1445 - until we can specify
        //  composite parsers with strategies (eg Composite, Try In Turn),
        //  always send the image onwards to the regular parser to have
        //  the metadata for them extracted as well
        _TMP_IMAGE_METADATA_PARSER.parse(tikaStream, handler, metadata, context);
    } finally {
        tmp.dispose();
        if (output != null) {
            output.delete();
        }
    }
}