Package org.apache.tika.sax

Examples of org.apache.tika.sax.WriteOutContentHandler

54
55
56
57
58
59
60
61
62
63
64
       
        Metadata metadata = new Metadata();
        StringWriter writer = new StringWriter();
        tika.getParser().parse(
                     new FileInputStream(file),
                     new WriteOutContentHandler(writer),
                     metadata,
                     new ParseContext());
        String content = writer.toString();

        assertEquals("application/rtf", metadata.get(Metadata.CONTENT_TYPE));
View Full Code Here
290
291
292
293
294
295
296
297
298
299
300
      
        Metadata metadata = new Metadata();
        StringWriter writer = new StringWriter();
        tika.getParser().parse(
                     new FileInputStream(file),
                     new WriteOutContentHandler(writer),
                     metadata,
                     new ParseContext());
        String content = writer.toString();
        return new Result(content, metadata);
    }
View Full Code Here
132
133
134
135
136
137
138
139
140
141
142
    logRequest(logger, info, metadata);

    return new StreamingOutput() {
      public void write(OutputStream outputStream) throws IOException, WebApplicationException {
        BodyContentHandler body = new BodyContentHandler(new WriteOutContentHandler(outputStream) {
          @Override
          public void startElement(String uri, String localName, String qName, Attributes attributes) throws SAXException {
            super.startElement(uri, localName, qName, attributes);

            if ("img".equals(localName) && attributes.getValue("alt")!=null) {
View Full Code Here
101
102
103
104
105
106
107
108
109
110
111
    try {
      Metadata metadata = metadataProcessor.prepareMetadata();
      ParseContext parseContext = parseContextProvider.getParseContext( name, value );

      StringWriter writer = new StringWriter();
      WriteOutContentHandler contentHandler = new WriteOutContentHandler( writer );

      Parser parser = new AutoDetectParser();
      parser.parse( in, contentHandler, metadata, parseContext );
      luceneOptions.addFieldToDocument( name, writer.toString(), document );
View Full Code Here
141
142
143
144
145
146
147
     * the given writer.
     *
     * @param writer writer
     */
    public BoilerpipeContentHandler(Writer writer) {
        this(new WriteOutContentHandler(writer));
    }
View Full Code Here
40
41
42
43
44
45
46
47
48
49
50
        Metadata metadata = new Metadata();
        StringWriter writer = new StringWriter();
        parser.parse(
                new ByteArrayInputStream(text.getBytes("UTF-8")),
                new WriteOutContentHandler(writer),
                metadata,
                new ParseContext());
        String content = writer.toString();

        assertEquals("text/plain", metadata.get(Metadata.CONTENT_TYPE));
View Full Code Here
181
182
183
184
185
186
187
188
189
190
191
    public void testCP866() throws Exception {
        Metadata metadata = new Metadata();
        StringWriter writer = new StringWriter();
        parser.parse(
                TXTParserTest.class.getResourceAsStream("/test-documents/russian.cp866.txt"),
                new WriteOutContentHandler(writer),
                metadata,
                new ParseContext());

        assertEquals("text/plain", metadata.get(Metadata.CONTENT_TYPE));
        assertEquals("IBM866", metadata.get(Metadata.CONTENT_ENCODING));
View Full Code Here
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
     * @throws IOException if the document can not be read
     * @throws TikaException if the document can not be parsed
     */
    public String parseToString(InputStream stream, Metadata metadata)
            throws IOException, TikaException {
        WriteOutContentHandler handler =
            new WriteOutContentHandler(maxStringLength);
        try {
            ParseContext context = new ParseContext();
            context.set(Parser.class, parser);
            parser.parse(
                    stream, new BodyContentHandler(handler), metadata, context);
        } catch (SAXException e) {
            if (!handler.isWriteLimitReached(e)) {
                // This should never happen with BodyContentHandler...
                throw new TikaException("Unexpected SAX processing failure", e);
            }
        } finally {
            stream.close();
        }
        return handler.toString();
    }
View Full Code Here
77
78
79
80
81
82
83
84
85
86
87
       
        Metadata metadata = new Metadata();
        StringWriter writer = new StringWriter();
        tika.getParser().parse(
                     new FileInputStream(file),
                     new WriteOutContentHandler(writer),
                     metadata,
                     new ParseContext());
        String content = writer.toString();

        assertEquals("application/rtf", metadata.get(Metadata.CONTENT_TYPE));
View Full Code Here
588
589
590
591
592
593
594
595
596
597
598
      
        Metadata metadata = new Metadata();
        StringWriter writer = new StringWriter();
        tika.getParser().parse(
                     new FileInputStream(file),
                     new WriteOutContentHandler(writer),
                     metadata,
                     new ParseContext());
        String content = writer.toString();
        return new Result(content, metadata);
    }
View Full Code Here
TOP

Related Classes of org.apache.tika.sax.WriteOutContentHandler

  • bixo.examples.crawl.SimpleBodyContentHandler
  • org.apache.jackrabbit.core.query.lucene.LazyTextExtractorField$ParsingTask
  • org.apache.jackrabbit.oak.plugins.index.lucene.LuceneIndexEditor
  • org.apache.jackrabbit.oak.plugins.index.lucene.LuceneIndexUpdate
  • org.apache.tika.parser.AutoDetectParserTest
  • org.apache.tika.parser.html.BoilerpipeContentHandler
  • org.apache.tika.parser.html.HtmlParser
  • org.apache.tika.parser.html.HtmlParserTest
  • org.apache.tika.parser.microsoft.ExcelParserTest
  • org.apache.tika.parser.microsoft.PowerPointParserTest
    • Copyright © 2018 www.massapicom. All rights reserved.
    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.