Map<String, Object> doc = documentReader.read();
                while (doc != null) {
                    if (Thread.currentThread().isInterrupted()) {
                        break;
                    }
                    docs.add(doc);
                    doc = documentReader.read();
                    if (doc == null || docs.size() >= docPerReq) {

     * Returns the result data header.
     * @return The result data header.
     */
    protected String getResultDataHeader() {
        // TODO: Support other XML header types
        return "<?xml version=\"1.0\"?>\n<doc>\n";
    }

    /**
     * Returns the result data body for a single value.
     * @param name The name of the field.
     * @param value The value of the field.
     * @return The result data body.
     */

    // public ResponseData call() throws Exception {
    // String[] urls =
    // new String[] {
    // "http://.../",
    // "http://.../test.pdf",
    // "http://.../test.doc",
    // "http://.../test.xls",
    // "http://.../test.ppt",
    // "http://.../test.txt", };
    // for (String url : urls) {
    // ResponseData responseData = httpClient.doGet(url);

        // Text Formats
        extensionMap.put("odt", "txt");
        extensionMap.put("ott", "txt");
        extensionMap.put("sxw", "txt");
        extensionMap.put("rtf", "txt");
        extensionMap.put("doc", "txt");
        extensionMap.put("docx", "txt");
        extensionMap.put("wpd", "txt");
        extensionMap.put("txt", "txt");
        extensionMap.put("html", "txt");
        // Spreadsheet Formats

        logger.info(content);
        assertTrue(content.contains("テスト"));
    }

    public void test_getTika_msword() {
        final InputStream in = ResourceUtil.getResourceAsStream("extractor/msoffice/test.doc");
        final ExtractData extractData = tikaExtractor.getText(in, null);
        final String content = extractData.getContent();
        CloseableUtil.closeQuietly(in);
        logger.info(content);

    <alias type="application/vnd.ms-word"/>
    <_comment>Microsoft Word Document</_comment>
    <tika:link>http://en.wikipedia.org/wiki/.doc</tika:link>
    <tika:uti>com.microsoft.word.doc</tika:uti>
    <magic priority="50">
      <match value="Microsoft\ Word\ 6.0\ Document" type="string" offset="2080"/>
      <match value="Documento\ Microsoft\ Word\ 6" type="string" offset="2080"/>

				"application/vnd.sbm.cid",
				"application/vnd.sbm.mid2",
				"application/vnd.scribus",
				"application/vnd.sealed.3df",
				"application/vnd.sealed.csf",
				"application/vnd.sealed.doc",
				"application/vnd.sealed.eml",
				"application/vnd.sealed.mht",
				"application/vnd.sealed.net",
				"application/vnd.sealed.ppt",
				"application/vnd.sealed.tiff",
				"application/vnd.sealed.xls",