Package org.apache.tika.sax

Examples of org.apache.tika.sax.WriteOutContentHandler

109
110
111
112
113
114
115
116
117
118
119
        parser.parse(new InputSource(Utils.getUTF8Reader(stream, metadata)));
        xhtml.endDocument();
    }

    private ContentHandler getTitleHandler(final Metadata metadata) {
        return new WriteOutContentHandler() {
            @Override
            public void endElement(String u, String l, String n) {
                metadata.set(Metadata.TITLE, toString());
            }
        };
View Full Code Here
118
119
120
121
122
123
124
125
126
127
128
            }
        };
    }

    private ContentHandler getMetaHandler(final Metadata metadata) {
        return new WriteOutContentHandler() {
            @Override
            public void startElement(
                    String uri, String local, String name, Attributes atts)
                    throws SAXException {
                    if (atts.getValue("http-equiv") != null) {
View Full Code Here
59
60
61
62
63
64
65
66
67
68
69
       
        Metadata metadata = new Metadata();
        StringWriter writer = new StringWriter();
        parser.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
293
294
295
296
297
298
299
300
301
302
303
      
        Metadata metadata = new Metadata();
        StringWriter writer = new StringWriter();
        parser.parse(
                     new FileInputStream(file),
                     new WriteOutContentHandler(writer),
                     metadata,
                     new ParseContext());
        String content = writer.toString();
        return new Result(content, metadata);
    }
View Full Code Here
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
        Blob blob = (Blob) value;
        InputStream is = null;

        // TODO make write limit configurable
        WriteOutContentHandler woh = new WriteOutContentHandler(500 * 1000); // 500K limit (Tika default: 100K)
        BodyContentHandler ch = new BodyContentHandler(woh);

        try {
            is = repository.getTable(table).getInputStream(record, fieldType.getName(), indexes);

            Metadata metadata = new Metadata();
            metadata.add(Metadata.CONTENT_TYPE, blob.getMediaType());
            if (blob.getName() != null) {
                metadata.add(Metadata.RESOURCE_NAME_KEY, blob.getName());
            }

            ParseContext parseContext = new ParseContext();

            tikaParser.parse(is, ch, metadata, parseContext);
        } catch (Throwable t) {
            if (woh.isWriteLimitReached(t)) {
                // ok, we'll just add use the partial result
                if (log.isInfoEnabled()) {
                    log.info("Blob extraction: write limit reached. Field '" + fieldType.getName() + "', record '"
                            + record.getId() + "'.");
                }
View Full Code Here
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
     * @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
107
108
109
110
111
112
113
114
115
116
117
        parser.setContentHandler(new XHTMLDowngradeHandler(handler));
        parser.parse(new InputSource(Utils.getUTF8Reader(stream, metadata)));
    }

    private ContentHandler getTitleHandler(final Metadata metadata) {
        return new WriteOutContentHandler() {
            @Override
            public void endElement(String u, String l, String n) {
                metadata.set(Metadata.TITLE, toString());
            }
        };
View Full Code Here
116
117
118
119
120
121
122
123
124
125
126
            }
        };
    }

    private ContentHandler getMetaHandler(final Metadata metadata) {
        return new WriteOutContentHandler() {
            @Override
            public void startElement(
                    String uri, String local, String name, Attributes atts)
                    throws SAXException {
                    if (atts.getValue("http-equiv") != null) {
View Full Code Here
39
40
41
42
43
44
45
46
47
48
49
        Metadata metadata = new Metadata();
        StringWriter writer = new StringWriter();
        parser.parse(
                new ByteArrayInputStream(text.getBytes("UTF-8")),
                new WriteOutContentHandler(writer),
                metadata);
        String content = writer.toString();

        assertEquals("text/plain", metadata.get(Metadata.CONTENT_TYPE));
        assertEquals("en", metadata.get(Metadata.CONTENT_LANGUAGE));
View Full Code Here
42
43
44
45
46
47
48
49
50
51
52
        Metadata metadata = new Metadata();
        StringWriter writer = new StringWriter();
        parser.parse(
                new ByteArrayInputStream(text.getBytes("ISO-8859-1")),
                new WriteOutContentHandler(writer),
                metadata,
                new ParseContext());
        String content = writer.toString();

        assertEquals("text/plain; charset=ISO-8859-1", metadata.get(Metadata.CONTENT_TYPE));
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.