Package org.apache.tika.TikaTest

Examples of org.apache.tika.TikaTest.TrackingHandler

43
44
45
46
47
48
49
50
51
52
53
             "testEXCEL.xls", "testWORD.doc", "testPPT.ppt",
             "testVISIO.vsd", "test-outlook.msg"
       };
       for(String file : files) {
          // Process it without recursing
          TrackingHandler handler = process(file, extractor, false);
         
          // Won't have fired
          assertEquals(0, handler.filenames.size());
          assertEquals(0, handler.mediaTypes.size());
         
View Full Code Here
63
64
65
66
67
68
69
70
71
72
73
     *  office files in them
     */
    @Test
    public void testEmbeddedImages() throws Exception {
       ContainerExtractor extractor = new ParserContainerExtractor();
       TrackingHandler handler;
      
       // Excel with 1 image
       handler = process("testEXCEL_1img.xls", extractor, false);
       assertEquals(1, handler.filenames.size());
       assertEquals(1, handler.mediaTypes.size());
View Full Code Here
116
117
118
119
120
121
122
123
124
125
126
     *           -> image
     */
    @Test
    public void testEmbeddedOfficeFiles() throws Exception {
       ContainerExtractor extractor = new ParserContainerExtractor();
       TrackingHandler handler;
      
      
       // Excel with a word doc and a powerpoint doc, both of which have images in them
       // Without recursion, should see both documents + the images
       handler = process("testEXCEL_embeded.xls", extractor, false);
View Full Code Here
325
326
327
328
329
330
331
332
333
334
335
    }

    @Test
    public void testEmbeddedOfficeFilesXML() throws Exception {
        ContainerExtractor extractor = new ParserContainerExtractor();
        TrackingHandler handler;

        handler = process("EmbeddedDocument.docx", extractor, false);
        assertTrue(handler.filenames.contains("Microsoft_Office_Excel_97-2003_Worksheet1.bin"));
        assertEquals(2, handler.filenames.size());
    }
View Full Code Here
335
336
337
338
339
340
341
342
343
344
345
    }

    @Test
    public void testPowerpointImages() throws Exception {
        ContainerExtractor extractor = new ParserContainerExtractor();
        TrackingHandler handler;

        handler = process("pictures.ppt", extractor, false);
        assertTrue(handler.mediaTypes.contains(new MediaType("image", "jpeg")));
        assertTrue(handler.mediaTypes.contains(new MediaType("image", "png")));
    }
View Full Code Here
55
56
57
58
59
60
61
62
63
64
65
            TikaInputStream stream = TikaInputStream.get(input);

            assertEquals(true, extractor.isSupported(stream));

            // Process it
            TrackingHandler handler = new TrackingHandler();
            extractor.extract(stream, null, handler);

            assertEquals(2, handler.filenames.size());
        } finally {
            input.close();
View Full Code Here
52
53
54
55
56
57
58
59
60
61
62
        TikaInputStream stream = getTestFile(filename);
        try {
            assertEquals(true, extractor.isSupported(stream));

            // Process it
            TrackingHandler handler = new TrackingHandler();
            if(recurse) {
                extractor.extract(stream, extractor, handler);
            } else {
                extractor.extract(stream, null, handler);
            }
View Full Code Here
73
74
75
76
77
78
79
80
81
82
83
    public void testBodyAndAttachments() throws Exception {
       ContainerExtractor extractor = new ParserContainerExtractor();
      
       // Process it with recursing
       // Will have the message body RTF and the attachments
       TrackingHandler handler = process(file, extractor, true);
       assertEquals(6, handler.filenames.size());
       assertEquals(6, handler.mediaTypes.size());
      
       // We know the filenames for all of them
       assertEquals("message.rtf", handler.filenames.get(0));
View Full Code Here
TOP

Related Classes of org.apache.tika.TikaTest.TrackingHandler

  • org.apache.tika.parser.microsoft.AbstractPOIContainerExtractionTest
  • org.apache.tika.parser.microsoft.ooxml.OOXMLContainerExtractionTest
  • org.apache.tika.parser.microsoft.POIContainerExtractionTest
  • org.apache.tika.parser.microsoft.TNEFParserTest
  • org.apache.tika.parser.xml.FictionBookParserTest
    • 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.