assertContentType("text/html", "html/test1.html", "hoge.htm");
        assertContentType("text/html", "html/test1.shtml", "hoge.shtml");

        assertContentType("application/msword", "test/text1.txt", "hoge.doc");
        assertContentType("application/msword", "extractor/msoffice/test.doc", "hoge.doc");
        assertContentType("application/vnd.ms-excel", "test/text1.txt", "hoge.xls");

        LhaExtractor lhaExtractor = container.getComponent("lhaExtractor");
        PasswordBasedExtractor pdfExtractor = container.getComponent("pdfExtractor");
        extractorFactory.addExtractor("application/msword", tikaExtractor);
        extractorFactory.addExtractor("application/vnd.ms-excel", tikaExtractor);
        extractorFactory.addExtractor("application/vnd.ms-powerpoint", tikaExtractor);

#                                                     ------

suggest.popular.word.seed=0
suggest.popular.word.tags=
suggest.popular.word.fields=
suggest.popular.word.excludes=
suggest.popular.word.size=10
suggest.popular.word.window.size=30
suggest.popular.word.query.freq=10
suggest.min.hit.count=1
suggest.field.contents=_default
suggest.field.tags=label
suggest.field.roles=role

        assertEquals("ぶんるい", extractData.getValues("cp:category")[0]);
        assertEquals("1", extractData.getValues("meta:page-count")[0]);
        assertEquals("application/msword", extractData.getValues("Content-Type")[0]);
    }

    public void test_getTika_mswordx() {
        final InputStream in = ResourceUtil.getResourceAsStream("extractor/msoffice/test.docx");

  /**
   * <a
   * href="https://learn.microsoft.com/en-us/archive/blogs/vsofficedeveloper/office-2007-file-format-mime-types-for-http-content-streaming-2">Microsoft
   * Word</a> documents.
   */
  public static final MediaType MICROSOFT_WORD = createConstant(APPLICATION_TYPE, "msword");

  /**
   * Media type for <a
   * href="https://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP">Dynamic Adaptive

    String INDEX_INDICES_TIMEOUT = "index.indices.timeout";

    /** The key of the configuration. e.g. text/html=html<br>
     * application/msword=word<br>
     * application/vnd.openxmlformats-officedocument.wordprocessingml.document=word<br>
     * application/vnd.ms-excel=excel<br>
     * application/vnd.ms-excel.sheet.2=excel<br>
     * application/vnd.ms-excel.sheet.3=excel<br>

  <!-- http://www.iana.org/assignments/media-types/application/msword -->
  <mime-type type="application/msword">
    <!-- Use DefaultDetector / org.apache.tika.parser.microsoft.POIFSContainerDetector for more reliable detection of OLE2 documents -->
    <alias type="application/vnd.ms-word"/>
    <_comment>Microsoft Word Document</_comment>
    <tika:link>http://en.wikipedia.org/wiki/.doc</tika:link>

  }

  abstract String normalizeWord(String word);

  String normalizeFirstWord(String word) {
    return normalizeWord(word);
  }

  private static String firstCharOnlyToUpper(String word) {
    return word.isEmpty()
        ? word
        : Ascii.toUpperCase(word.charAt(0)) + Ascii.toLowerCase(word.substring(1));
  }

  }

  abstract String normalizeWord(String word);

  String normalizeFirstWord(String word) {
    return normalizeWord(word);
  }

  private static String firstCharOnlyToUpper(String word) {
    return word.isEmpty()
        ? word
        : Ascii.toUpperCase(word.charAt(0)) + Ascii.toLowerCase(word.substring(1));
  }

        }

        public MyToken(final String word, final int startOffset, final int endOffset, final int posInc) {
            this(word, startOffset, endOffset, posInc, Integer.MAX_VALUE); // Integer.MAX_VALUE for seq means unused
        }

        public MyToken(final String word, final int startOffset, final int endOffset, final int posInc, final int seq) {
            this.word = word;
            this.startOffset = startOffset;