Package org.apache.tika.detect

Examples of org.apache.tika.detect.Detector

144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
    }

    if (mediaType !=null) {
      metadata.add(org.apache.tika.metadata.HttpHeaders.CONTENT_TYPE, mediaType.toString());

      final Detector detector = parser.getDetector();

      parser.setDetector(new Detector() {
        public MediaType detect(InputStream inputStream, Metadata metadata) throws IOException {
          String ct = metadata.get(org.apache.tika.metadata.HttpHeaders.CONTENT_TYPE);

          if (ct!=null) {
            return MediaType.parse(ct);
          } else {
            return detector.detect(inputStream, metadata);
          }
        }
      });
    }
  }
View Full Code Here
190
191
192
193
194
195
196
197
198
199
200
       ForkParser parser = new ForkParser(
             ForkParserIntegrationTest.class.getClassLoader(),
             tika.getParser());
      
       ParseContext context = new ParseContext();
       context.set(Detector.class, new Detector() {
          public MediaType detect(InputStream input, Metadata metadata) {
             return MediaType.OCTET_STREAM;
          }
       });
View Full Code Here
38
39
40
41
42
43
44
45
46
47
48
49
50
51
   private static final String file = "testWINMAIL.dat";
  
   @Test
   public void testBasics() throws Exception {
      TikaInputStream stream = getTestFile(file);
      Detector detector = new DefaultDetector();
      try {
         assertEquals(
                 MediaType.application("vnd.ms-tnef"),
                 detector.detect(stream, new Metadata()));
     } finally {
         stream.close();
     }
   }
View Full Code Here
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
     *  (an OOXML container with binary blobs), but we
     *  shouldn't break on these files either (TIKA-826) 
     */
    @Test
    public void testExcelXLSB() throws Exception {
       Detector detector = new DefaultDetector();
       AutoDetectParser parser = new AutoDetectParser();
      
       InputStream input = ExcelParserTest.class.getResourceAsStream(
             "/test-documents/testEXCEL.xlsb");
       Metadata m = new Metadata();
       m.add(Metadata.RESOURCE_NAME_KEY, "excel.xlsb");
      
       // Should be detected correctly
       MediaType type = null;
       try {
          type = detector.detect(input, m);
          assertEquals("application/vnd.ms-excel.sheet.binary.macroenabled.12", type.toString());
       } finally {
          input.close();
       }
      
View Full Code Here
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
     * We don't currently support the old Excel 95 .xls file format,
     *  but we shouldn't break on these files either (TIKA-976) 
     */
    @Test
    public void testExcel95() throws Exception {
       Detector detector = new DefaultDetector();
       AutoDetectParser parser = new AutoDetectParser();
      
       InputStream input = ExcelParserTest.class.getResourceAsStream(
             "/test-documents/testEXCEL_95.xls");
       Metadata m = new Metadata();
       m.add(Metadata.RESOURCE_NAME_KEY, "excel_95.xls");
      
       // Should be detected correctly
       MediaType type = null;
       try {
          type = detector.detect(input, m);
          assertEquals("application/vnd.ms-excel", type.toString());
       } finally {
          input.close();
       }
      
View Full Code Here
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
    }

    if (mediaType !=null) {
      metadata.add(org.apache.tika.metadata.HttpHeaders.CONTENT_TYPE, mediaType.toString());

      final Detector detector = parser.getDetector();

      parser.setDetector(new Detector() {
        public MediaType detect(InputStream inputStream, Metadata metadata) throws IOException {
          String ct = metadata.get(org.apache.tika.metadata.HttpHeaders.CONTENT_TYPE);

          if (ct!=null) {
            return MediaType.parse(ct);
          } else {
            return detector.detect(inputStream, metadata);
          }
        }
      });
    }
  }
View Full Code Here
109
110
111
112
113
114
115
116
117
118
119
    String[] segments = uriInfo.getRequestUri().toASCIIString()
        .split("/" + opkey + "/" + pathkey + "/");
    final String filename = segments[segments.length - 1];
    logger.info("resource :" + segments[segments.length - 1]);

    final Detector detector = createDetector(httpHeaders);
    final AutoDetectParser parser = new AutoDetectParser(detector);
    final ParseContext context = new ParseContext();
    context.set(Parser.class, parser);
    final org.apache.tika.metadata.Metadata metadata = new org.apache.tika.metadata.Metadata();
    setMetadataFromHeader(parser, metadata, httpHeaders);
View Full Code Here
238
239
240
241
242
243
244
245
246
247
248
  @Produces({ MediaType.APPLICATION_JSON })
  @Path("/{opkey}")
  public StreamingOutput getMetadata(final InputStream is,
      @PathParam("opkey") final String opkey,
      @Context HttpHeaders httpHeaders) throws Exception {
    final Detector detector = createDetector(httpHeaders);
    final AutoDetectParser parser = new AutoDetectParser(detector);
    final ParseContext context = new ParseContext();
    context.set(Parser.class, parser);
    final org.apache.tika.metadata.Metadata metadata = new org.apache.tika.metadata.Metadata();
    setMetadataFromHeader(parser, metadata, httpHeaders);
View Full Code Here
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
        && !mediaType
            .equals(javax.ws.rs.core.MediaType.APPLICATION_OCTET_STREAM_TYPE)) {
      metadata.add(org.apache.tika.metadata.HttpHeaders.CONTENT_TYPE,
          mediaType.toString());

      final Detector detector = parser.getDetector();

      parser.setDetector(new Detector() {
        public org.apache.tika.mime.MediaType detect(
            InputStream inputStream,
            org.apache.tika.metadata.Metadata metadata)
            throws IOException {
          String ct = metadata
              .get(org.apache.tika.metadata.HttpHeaders.CONTENT_TYPE);
          logger.info("Content type " + ct);
          if (ct != null) {
            return org.apache.tika.mime.MediaType.parse(ct);
          } else {
            return detector.detect(inputStream, metadata);
          }
        }
      });
    }
  }
View Full Code Here
411
412
413
414
415
416
417
418
419
420
421
    if (mediaType == null
        || mediaType
            .equals(javax.ws.rs.core.MediaType.APPLICATION_OCTET_STREAM_TYPE))
      return (new TikaConfig()).getMimeRepository();
    else
      return new Detector() {

        public org.apache.tika.mime.MediaType detect(
            InputStream inputStream,
            org.apache.tika.metadata.Metadata metadata)
            throws IOException {
View Full Code Here
TOP

Related Classes of org.apache.tika.detect.Detector

  • com.nidhinova.tika.server.TikaService
  • org.apache.tika.bundle.BundleIT
  • org.apache.tika.config.TikaConfig
  • org.apache.tika.example.DumpTikaConfigExample
  • org.apache.tika.parser.fork.ForkParserIntegrationTest
  • org.apache.tika.parser.microsoft.ExcelParserTest
  • org.apache.tika.parser.microsoft.TNEFParserTest
  • org.apache.tika.parser.rtf.RTFEmbObjHandler
  • org.apache.tika.server.TikaResource
  • org.jasig.portal.io.xml.JaxbPortalDataHandlerService
    • 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.