Examples of TikaInputStream

org.apache.tika.io.TikaInputStream
Input stream with extended capabilities. The purpose of this class is to allow files and other resources and information to be associated with the {@link InputStream} instance passed through the{@link org.apache.tika.parser.Parser} interface and other similar APIs.
TikaInputStream instances can be created using the various static get() factory methods. Most of these methods take an optional {@link Metadata} argument that is then filled with the available inputmetadata from the given resource. The created TikaInputStream instance keeps track of the original resource used to create it, while behaving otherwise just like a normal, buffered {@link InputStream}. A TikaInputStream instance is also guaranteed to support the {@link #mark(int)} feature.
Code that wants to access the underlying file or other resources associated with a TikaInputStream should first use the {@link #get(InputStream)} factory method to cast or wrap a given{@link InputStream} into a TikaInputStream instance. @since Apache Tika 0.8

Examples of org.apache.tika.io.TikaInputStream

        assertEquals(
          MediaType.application("vnd.ms-powerpoint"),
          d.detect(input, new Metadata())
        );
        
        TikaInputStream tis = TikaInputStream.get(getTestDoc("testPPT.ppt"));
        assertEquals(
          MediaType.application("vnd.ms-powerpoint"),
          d.detect(tis, new Metadata())
        );
        
        assertNotNull(tis.getOpenContainer());
        assertEquals(POIFSFileSystem.class, tis.getOpenContainer().getClass());
        
        // Try some ones that POI doesn't handle, that are still OLE2 based
        tis = TikaInputStream.get(getTestDoc("testWORKS.wps"));
        assertEquals(
           MediaType.application("vnd.ms-works"),

View Full Code Here

Examples of org.apache.tika.io.TikaInputStream

        assertEquals(
          MediaType.application("vnd.oasis.opendocument.formula"),
          d.detect(input, new Metadata())
        );
        
        TikaInputStream tis = TikaInputStream.get(getTestDoc("testOpenOffice2.odf"));
        assertEquals(
          MediaType.application("vnd.oasis.opendocument.formula"),
          d.detect(tis, new Metadata())
        );
        // Doesn't store the zip parser yet
        assertNull(tis.getOpenContainer());
    }

View Full Code Here

Examples of org.apache.tika.io.TikaInputStream

              MediaType.application("vnd.openxmlformats-officedocument.presentationml.presentation"),
              d.detect(input, new Metadata())
        );


        // Try with a tika input stream
        TikaInputStream tis = TikaInputStream.get(getTestDoc("testPPT.pptx"));
        assertEquals(
              MediaType.application("vnd.openxmlformats-officedocument.presentationml.presentation"),
              d.detect(tis, new Metadata())
        );
        
        // There should be an attached OPCPackage as an open container
        assertNotNull(tis.getOpenContainer());
        assertTrue(
              "Open container should be OPCPackage, not " + tis.getOpenContainer().getClass(), 
              tis.getOpenContainer() instanceof OPCPackage
        );
        
        // The underlying TikaInputStream should still be open, and file based
        assertTrue(
              "TikaInputStream should still have a file",
              tis.hasFile()
        );
    }

View Full Code Here

Examples of org.apache.tika.io.TikaInputStream

              tis.hasFile()
        );
    }
    
    public void testDetectIWork() throws Exception {
       TikaInputStream tis;


       tis = TikaInputStream.get(getTestDoc("testKeynote.key"));
       assertEquals(
             MediaType.application("vnd.apple.keynote"),
             d.detect(tis, new Metadata())

View Full Code Here

Examples of org.apache.tika.io.TikaInputStream

             d.detect(tis, new Metadata())
       );
    }
    
    public void testDetectZip() throws Exception {
       TikaInputStream tis;


       tis = TikaInputStream.get(getTestDoc("test-documents.zip"));
       assertEquals(
             MediaType.application("zip"),
             d.detect(tis, new Metadata())

View Full Code Here

Examples of org.apache.tika.io.TikaInputStream

    }


    public MediaType detect(InputStream input, Metadata metadata)
             throws IOException {
        if (TikaInputStream.isTikaInputStream(input)) {
            TikaInputStream stream = TikaInputStream.get(input);
            return detect(stream, metadata);
        } else {
           // We can't do proper detection if we don't
           //  have a TikaInputStream
           return DEFAULT;

View Full Code Here

Examples of org.apache.tika.io.TikaInputStream

      public void write(OutputStream outputStream) throws IOException, WebApplicationException {
        Writer writer = new OutputStreamWriter(outputStream, "UTF-8");


        BodyContentHandler body = new BodyContentHandler(new RichTextContentHandler(writer));


        TikaInputStream tis = TikaInputStream.get(is);


        try {
          tis.getFile();


          parser.parse(tis, body, metadata);
        } catch (SAXException e) {
          throw new WebApplicationException(e);
        } catch (EncryptedDocumentException e) {
          logger.warn(String.format(
                  "%s: Encrypted document",
                  info.getPath()
          ), e);


          throw new WebApplicationException(e, Response.status(422).build());
        } catch (TikaException e) {
          logger.warn(String.format(
            "%s: Text extraction failed",
            info.getPath()
          ), e);


          if (e.getCause()!=null && e.getCause() instanceof WebApplicationException) {
            throw (WebApplicationException) e.getCause();
          }


          if (e.getCause()!=null && e.getCause() instanceof IllegalStateException) {
            throw new WebApplicationException(Response.status(422).build());
          }


          if (e.getCause()!=null && e.getCause() instanceof OldWordFileFormatException) {
            throw new WebApplicationException(Response.status(422).build());
          }


          throw new WebApplicationException(Response.Status.INTERNAL_SERVER_ERROR);
        } finally {
          tis.close();
        }
      }
    };
  }

View Full Code Here

Examples of org.apache.tika.io.TikaInputStream

            throws IOException, SAXException, TikaException {
        XHTMLContentHandler xhtml = new XHTMLContentHandler(handler, metadata);
        xhtml.startDocument();


        final DirectoryNode root;
        TikaInputStream tstream = TikaInputStream.cast(stream);
        if (tstream == null) {
            root = new NPOIFSFileSystem(new CloseShieldInputStream(stream)).getRoot();
        } else {
            final Object container = tstream.getOpenContainer();
            if (container instanceof NPOIFSFileSystem) {
                root = ((NPOIFSFileSystem) container).getRoot();
            } else if (container instanceof DirectoryNode) {
                root = (DirectoryNode) container;
            } else if (tstream.hasFile()) {
                root = new NPOIFSFileSystem(tstream.getFileChannel()).getRoot();
            } else {
                root = new NPOIFSFileSystem(new CloseShieldInputStream(tstream)).getRoot();
            }
        }
        parse(root, context, metadata, xhtml);

View Full Code Here

Examples of org.apache.tika.io.TikaInputStream

    private void handleEmbeddedOLE(PackagePart part, ContentHandler handler, String rel)
            throws IOException, SAXException {
        POIFSFileSystem fs = new POIFSFileSystem(part.getInputStream());
        try {
            Metadata metadata = new Metadata();
            TikaInputStream stream = null;
            metadata.set(Metadata.EMBEDDED_RELATIONSHIP_ID, rel);


            DirectoryNode root = fs.getRoot();
            POIFSDocumentType type = POIFSDocumentType.detectType(root);

View Full Code Here

Examples of org.apache.tika.io.TikaInputStream

            InputStream stream, ContentHandler handler,
            Metadata metadata, ParseContext context)
            throws IOException, SAXException, TikaException {
        TrueTypeFont font;
        TTFParser parser = new TTFParser();
        TikaInputStream tis = TikaInputStream.cast(stream);
        if (tis != null && tis.hasFile()) {
            font = parser.parseTTF(tis.getFile());
        } else {
            font = parser.parseTTF(stream);
        }


        metadata.set(Metadata.CONTENT_TYPE, TYPE.toString());

View Full Code Here

0 1 2 3 4 5

TOP

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.