String mimeType = (String) data.getProperty().getMetadata().get(LumifyProperties.MIME_TYPE.getPropertyName());

        checkNotNull(mimeType, LumifyProperties.MIME_TYPE.getPropertyName() + " is a required metadata field");

        Charset charset = Charset.forName("UTF-8");

        String text = extractText(in, mimeType, metadata);

        ExistingElementMutation<Vertex> m = data.getElement().prepareMutation();

        // TODO set("url", extractUrl(metadata));

            }

        }

        DefaultDetector detector = new DefaultDetector();

        MediaType mediaType = detector.detect(new BufferedInputStream(in), metadata);

        mimeType = mediaType.toString();

        if (mimeType != null) {

            return mimeType;

        }

        if (fetchedDatum.getContentType().startsWith("text/html")) {

          init();

          InputStream is = new ByteArrayInputStream(fetchedDatum.getContentBytes());

         

          try {

            _parser.parse(is, _handler, metadata, new ParseContext());

    @Test

    public void testNotTerminating() throws Exception {

        DelayParser parser = new DelayParser(true);

        InputStream is = Mockito.mock(InputStream.class);

       

        BaseContentExtractor contentExtractor = Mockito.mock(BaseContentExtractor.class);

        BaseLinkExtractor linkExtractor = Mockito.mock(BaseLinkExtractor.class);

       

        Callable<ParsedDatum> c = new TikaCallable(parser, contentExtractor, linkExtractor, is, md);

    @Test

    public void testTerminating() throws Exception {

        Parser parser = new DelayParser(false);

        InputStream is = Mockito.mock(InputStream.class);

       

        BaseContentExtractor contentExtractor = Mockito.mock(BaseContentExtractor.class);

        BaseLinkExtractor linkExtractor = Mockito.mock(BaseLinkExtractor.class);

        Mockito.when(linkExtractor.getLinks()).thenReturn(new Outlink[0]);

       

       

        // Now, if the FetchedDatum mime-type is application/mbox, we want to parse it and

        // output the results

        if (fetchedDatum.getContentType().equals("application/mbox")) {

          ParseContext context = new ParseContext();

          InputStream is = new ByteArrayInputStream(fetchedDatum.getContentBytes());

         

          try {

            _parser.parse(is, _handler, metadata, context);

            // _content now has all of the body text, and metadata has the header info.

            if (emailAddress == null) {

              LOGGER.warn("No email address for message: " + messageId);

              return;

            }

            String address = null;

            String name = null;

            Matcher addressMatcher = FULL_EMAIL_ADDRESS_PATTERN.matcher(emailAddress);

            if (addressMatcher.matches()) {

              name = addressMatcher.group(1);

              address = addressMatcher.group(2);

            } else {

              addressMatcher = SIMPLE_EMAIL_ADDRESS_PATTERN.matcher(emailAddress);

              if (addressMatcher.matches()) {

                address = addressMatcher.group(1);

              } else {

                LOGGER.warn("Email address has invalid format: " + emailAddress);

                return;

              }

            }

            // Now we might need to remain the address, if this user has aliases.

            if (EMAIL_ALIASES.containsKey(address)) {

              address = EMAIL_ALIASES.get(address);

            }

            Tuple tuple = new Tuple(messageId, address, name, 0.0);

            functionCall.getOutputCollector().add(tuple);

            if (replyId != null) {

              double score = analyzeReply(_content.toString());

              if (score > 0.0) {

                tuple = new Tuple(replyId, null, null, score);

          LOGGER.trace(String.format("Parsing %s", fetchedDatum.getUrl()));

        }

       

        // Provide clues to the parser about the format of the content.

        String charset = getCharset(fetchedDatum);

       

        InputStream is = new ByteArrayInputStream(fetchedDatum.getContentBytes(), 0, fetchedDatum.getContentLength());

        try {

          URL baseUrl = getContentLocation(fetchedDatum);

            Callable<ParsedDatum> c = new TikaCallable(_parser, _contentExtractor, _linkExtractor, is, metadata, isExtractLanguage(), _parseContext);

            FutureTask<ParsedDatum> task = new FutureTask<ParsedDatum>(c);

            Thread t = new Thread(task);

            t.start();

    //<start id="tika"/>

    InputStream input = new FileInputStream(

            new File("src/test/resources/pdfBox-sample.pdf"));//<co id="tika.is"/>

    ContentHandler textHandler = new BodyContentHandler();//<co id="tika.handler"/>

    Parser parser = new AutoDetectParser();//<co id="tika.parser"/>

    ParseContext context = new ParseContext();

    parser.parse(input, textHandler, metadata, context);//<co id="tika.parse"/>

    System.out.println("Body: " + textHandler.toString());//<co id="tika.body"/>

    /*

<calloutlist>

    <callout arearefs="tika.is"><para>Create the <classname>InputStream</classname> to read in the content</para></callout>

    <callout arearefs="tika.handler"><para>The <classname>BodyContentHandler</classname> is a Tika-provided <classname>ContentHandler</classname> that extracts just the "body" of the InputStream</para></callout>

    InputStream input = new ByteArrayInputStream(html.getBytes(Charset.forName("UTF-8")));

    ContentHandler text = new BodyContentHandler();//<co id="html.text.co"/>

    LinkContentHandler links = new LinkContentHandler();//<co id="html.link.co"/>

    ContentHandler handler = new TeeContentHandler(links, text);//<co id="html.merge"/>

    Parser parser = new HtmlParser();//<co id="html.parser"/>

    ParseContext context = new ParseContext();

    parser.parse(input, handler, metadata, context);//<co id="html.parse"/>

    System.out.println("Body: " + text.toString());

    System.out.println("Links: " + links.getLinks());

    /*

    <calloutlist>

        <callout arearefs="html.text.co"><para>Construct a ContentHandler that will just extract between the body tags.</para></callout>

        if (!in.markSupported()) {

            in = new java.io.BufferedInputStream(in);

        }

        try {

            String mime = this.mimeTypes.detect(in, metadata).toString();

            assertEquals(urlOrFileName + " is not properly detected: detected.", expected, mime);

            //Add resource name and test again

            mime = this.mimeTypes.detect(in, metadata).toString();

            assertEquals(urlOrFileName + " is not properly detected after adding resource name.", expected, mime);

        } finally {

            in.close();

        }