Package org.apache.tika.parser.microsoft.ooxml

Examples of org.apache.tika.parser.microsoft.ooxml.OOXMLParser

245
246
247
248
249
250
251
252
253
254
255
256
257
                 throw new EncryptedDocumentException();
              }

              // Decrypt the OLE2 stream, and delegate the resulting OOXML
              //  file to the regular OOXML parser for normal handling
              OOXMLParser parser = new OOXMLParser();

              parser.parse(d.getDataStream(root), new EmbeddedContentHandler(
                    new BodyContentHandler(xhtml)),
                    metadata, context);
           } catch (GeneralSecurityException ex) {
              throw new EncryptedDocumentException(ex);
           }
View Full Code Here
256
257
258
259
260
261
262
263
264
265
266
      
       // OfficeParser won't handle it
       assertEquals(false, (new OfficeParser()).getSupportedTypes(new ParseContext()).contains(type));
      
       // OOXMLParser won't handle it
       assertEquals(false, (new OOXMLParser()).getSupportedTypes(new ParseContext()).contains(type));
      
       // AutoDetectParser doesn't break on it
       input = ExcelParserTest.class.getResourceAsStream("/test-documents/testEXCEL.xlsb");

       try {
View Full Code Here
301
302
303
304
305
306
307
308
309
310
311
      
       // OfficeParser will claim to handle it
       assertEquals(true, (new OfficeParser()).getSupportedTypes(new ParseContext()).contains(type));
      
       // OOXMLParser won't handle it
       assertEquals(false, (new OOXMLParser()).getSupportedTypes(new ParseContext()).contains(type));
      
       // AutoDetectParser doesn't break on it
       input = ExcelParserTest.class.getResourceAsStream("/test-documents/testEXCEL_95.xls");

       try {
View Full Code Here
219
220
221
222
223
224
225
226
227
228
229
230
231
                    try {
                        if (!d.verifyPassword(Decryptor.DEFAULT_PASSWORD)) {
                            throw new TikaException("Unable to process: document is encrypted");
                        }

                        OOXMLParser parser = new OOXMLParser();

                        parser.parse(d.getDataStream(filesystem), new EmbeddedContentHandler(
                                        new BodyContentHandler(xhtml)),
                                        metadata, context);
                    } catch (GeneralSecurityException ex) {
                        throw new TikaException("Unable to process encrypted document", ex);
                    }
View Full Code Here
228
229
230
231
232
233
234
235
236
237
238
239
240
                 throw new EncryptedDocumentException();
              }

              // Decrypt the OLE2 stream, and delegate the resulting OOXML
              //  file to the regular OOXML parser for normal handling
              OOXMLParser parser = new OOXMLParser();

              parser.parse(d.getDataStream(root), new EmbeddedContentHandler(
                    new BodyContentHandler(xhtml)),
                    metadata, context);
           } catch (GeneralSecurityException ex) {
              throw new EncryptedDocumentException(ex);
           }
View Full Code Here
183
184
185
186
187
188
189
190
191
192
193
        // Rich Text Format
        addConverter( new RTFParser().getSupportedTypes( parseContext ), RTFConverter.class );

        // MS Open XML Format
        addConverter( new OOXMLParser().getSupportedTypes( parseContext ),
                MSOfficeXMLConverter.class );

        // Open document format
        addConverter( new OpenDocumentParser().getSupportedTypes( parseContext ),
                OpenDocumentConverter.class );
View Full Code Here
559
560
561
562
563
564
565
566
567
568
569
570
571
    InputStream input;
    try {
      input = new FileInputStream(new File(f.fileName()));
      ContentHandler textHandler = new BodyContentHandler(-1);
      Metadata metadata = new Metadata();
      OOXMLParser parser = new OOXMLParser(); 
      ParseContext context = new ParseContext();
      parser.parse(input, textHandler, metadata, context);
      String[] result = textHandler.toString().split(regex);
      for (int i=0; i<result.length && keepRunning; i++) {
        if (interrupt) {
          processInterrupt();
        }
View Full Code Here
230
231
232
233
234
235
236
237
238
239
240
241
242
                    try {
                        if (!d.verifyPassword(Decryptor.DEFAULT_PASSWORD)) {
                            throw new EncryptedDocumentException();
                        }

                        OOXMLParser parser = new OOXMLParser();

                        parser.parse(d.getDataStream(filesystem), new EmbeddedContentHandler(
                                        new BodyContentHandler(xhtml)),
                                        metadata, context);
                    } catch (GeneralSecurityException ex) {
                        throw new EncryptedDocumentException(ex);
                    }
View Full Code Here
183
184
185
186
187
188
189
190
191
192
193
        // Rich Text Format
        addConverter( new RTFParser().getSupportedTypes( parseContext ), RTFConverter.class );

        // MS Open XML Format
        addConverter( new OOXMLParser().getSupportedTypes( parseContext ),
                MSOfficeXMLConverter.class );

        // Open document format
        addConverter( new OpenDocumentParser().getSupportedTypes( parseContext ),
                OpenDocumentConverter.class );
View Full Code Here
244
245
246
247
248
249
250
251
252
253
254
      
       // OfficeParser won't handle it
       assertEquals(false, (new OfficeParser()).getSupportedTypes(new ParseContext()).contains(type));
      
       // OOXMLParser won't handle it
       assertEquals(false, (new OOXMLParser()).getSupportedTypes(new ParseContext()).contains(type));
      
       // AutoDetectParser doesn't break on it
       input = ExcelParserTest.class.getResourceAsStream("/test-documents/testEXCEL.xlsb");

       try {
View Full Code Here
TOP

Related Classes of org.apache.tika.parser.microsoft.ooxml.OOXMLParser

  • cx.fbn.nevernote.threads.IndexRunner
  • org.apache.tika.parser.microsoft.ExcelParserTest
  • org.apache.tika.parser.microsoft.OfficeParser
  • org.apache.tika.parser.ParseContext
  • org.apache.tika.xmp.convert.TikaToXMP
  • java.util.Locale
  • org.apache.tika.exception.TikaException
    • 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.