Package org.apache.tika.mime

Examples of org.apache.tika.mime.MediaType

264
265
266
267
268
269
270
271
272
273
274
                return COMP_OBJ;
            } else if (names.contains("CONTENTS") && names.contains("\u0001CompObj")) {
               // CompObj is a general kind of OLE2 embedding, but this may be an old Works file
               // If we have the Directory, check
               if (root != null) {
                  MediaType type = processCompObjFormatType(root);
                  if (type == WPS) {
                     return WPS;
                  } else {
                     // Assume it's a general CompObj embedded resource
                     return COMP_OBJ;
View Full Code Here
328
329
330
331
332
333
334
335
336
337
338
                NodeList mimes = node.getElementsByTagName("mime");
                if (mimes.getLength() > 0) {
                    Set<MediaType> types = new HashSet<MediaType>();
                    for (int j = 0; j < mimes.getLength(); j++) {
                        String mime = getText(mimes.item(j));
                        MediaType type = MediaType.parse(mime);
                        if (type != null) {
                            types.add(type);
                        } else {
                            throw new TikaException(
                                    "Invalid media type name: " + mime);
View Full Code Here
613
614
615
616
617
618
619
620
621
622
623
        for (MediaType type : registry.getTypes()) {
            System.out.println(type);
            for (MediaType alias : registry.getAliases(type)) {
                System.out.println("  alias:     " + alias);
            }
            MediaType supertype = registry.getSupertype(type);
            if (supertype != null) {
                System.out.println("  supertype: " + supertype);
            }
            Parser p = parsers.get(type);
            if (p != null) {
View Full Code Here
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
            if (name == null) {
                name = "file" + count++;
            }

            MediaType contentType = detector.detect(inputStream, metadata);

            if (name.indexOf('.')==-1 && contentType!=null) {
                try {
                    name += config.getMimeRepository().forName(
                            contentType.toString()).getExtension();
                } catch (MimeTypeException e) {
                    e.printStackTrace();
                }
            }
View Full Code Here
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
            TikaInputStream tis = TikaInputStream.get(input, tmp);

            byte[] prefix = new byte[1024]; // enough for all known formats
            int length = tis.peek(prefix);

            MediaType type = detectArchiveFormat(prefix, length);
            if (PackageParser.isZipArchive(type)
                    && TikaInputStream.isTikaInputStream(input)) {
                return detectZipFormat(tis);
            } else if (!type.equals(MediaType.OCTET_STREAM)) {
                return type;
            } else {
                return detectCompressorFormat(prefix, length);
            }
        } finally {
View Full Code Here
130
131
132
133
134
135
136
137
138
139
140
    private static MediaType detectZipFormat(TikaInputStream tis) {
        try {
            ZipFile zip = new ZipFile(tis.getFile()); // TODO: hasFile()?
            try {
                MediaType type = detectOpenDocument(zip);
                if (type == null) {
                    type = detectOfficeOpenXML(zip, tis);
                }
                if (type == null) {
                    type = detectIWork(zip);
View Full Code Here
285
286
287
288
289
290
291
292
293
294
295
       };
      
       // Check we found the parser
       CompositeParser parser = (CompositeParser)tika.getParser();
       for (String type : mimetypes) {
          MediaType mt = MediaType.parse(type);
          assertNotNull("Parser not found for " + type, parser.getParsers().get(mt) );
       }
      
       // Have each file parsed, and check
       for (int i=0; i<testFiles.length; i++) {
View Full Code Here
289
290
291
292
293
294
295
296
    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
244
245
246
247
248
249
250
251
252
253
254
255
256
257
             "/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();
       }
      
       // OfficeParser won't handle it
View Full Code Here
289
290
291
292
293
294
295
296
297
298
299
300
301
302
             "/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();
       }
      
       // OfficeParser will claim to handle it
View Full Code Here
TOP

Related Classes of org.apache.tika.mime.MediaType

  • bixo.utils.HttpUtils
  • com.cloudera.cdk.morphline.solrcell.SolrCellBuilder$SolrCell
  • com.cloudera.cdk.morphline.tika.DetectMimeTypeBuilder$DetectMimeType
  • com.digitalpebble.behemoth.tika.TikaProcessor
  • controllers.CodeApp
  • dk.i2m.drupal.util.EntityUtils
  • edu.isi.karma.rdf.GenericRDFGenerator
  • io.lumify.tikaMimeType.TikaMimeTypeMapper
  • org.apache.cxf.jaxrs.ext.search.tika.TikaContentExtractor
  • org.apache.jackrabbit.oak.http.AcceptHeader
    • 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.