assertEquals("UTF-8", encodingHelper.normalize(enc));

        enc = "Shift_JIS";
        assertEquals("Shift_JIS", encodingHelper.normalize(enc));

        enc = "S-JIS";
        assertEquals("Shift_JIS", encodingHelper.normalize(enc));
    }

    public void test_normalize_null() {
        String enc;

        enc = null;
        assertNull(encodingHelper.normalize(enc));

        enc = "";

    /**
     * PDF config file path.
     */
    public static final String TIKA_PDF_CONFIG = "tika.pdf.config";

    /**
     * A parameter key to normalize a text.
     */
    public static final String NORMALIZE_TEXT = "normalize_text";

    private static final String FILE_PASSWORD = "fess.file.password";

    /**
     * Output encoding.
     */

<!DOCTYPE components PUBLIC "-//DBFLUTE//DTD LastaDi 1.0//EN"
	"http://dbflute.org/meta/lastadi10.dtd">
<components namespace="fessCrawler">
	<include path="crawler/container.xml" />

	<!-- Converts child links to normalize it. -->
	<component name="urlConvertHelper" class="org.codelibs.fess.crawler.helper.UrlConvertHelper">
	</component>

        // Default constructor
    }

    /**
     * Normalizes an encoding string to its preferred form.
     *
     * @param enc the encoding string to normalize
     * @return the normalized encoding or the default encoding if the input is blank
     */
    public String normalize(final String enc) {
        if (StringUtil.isBlank(enc)) {
            return defaultEncoding;
        }

         * This method ensures the URL is in a consistent format for FTP operations.
         *
         * @param s The URL string to normalize
         * @return The normalized URL string
         */
        protected String normalize(final String s) {
            if (s == null) {
                return null;
            }
            String url = s.replaceAll("/+", "/").replace("ftp:/", "ftp://");

 * It uses the Apache Tika library to detect the MIME type.
 *
 * <p>
 * This class provides methods to:
 * </p>
 * <ul>
 *   <li>Detect the MIME type based on the input stream and filename.</li>
 *   <li>Normalize the filename to handle special characters.</li>
 *   <li>Configure whether to use the filename for MIME type detection.</li>
 *   <li>Configure whether to use the filename for MIME type detection when the stream is octet-stream.</li>

import org.apache.logging.log4j.Logger;
import org.codelibs.core.lang.StringUtil;

/**
 * Utility class for text normalization and processing.
 *
 * This class provides methods to normalize text by reading characters from a provided Reader
 * and processing them according to specific rules. The main functionality is encapsulated
 * within the nested {@link TextNormalizeContext} class.
 *

        assertEquals("2012-05-18T22:45:00Z", extractData.getValues("dcterms:modified")[0]);
        assertEquals("3", extractData.getValues("meta:character-count")[0]);
        assertEquals("Normal", extractData.getValues("extended-properties:Template")[0]);
        assertEquals("たいとる", extractData.getValues("dc:title")[0]);
        assertEquals("3000000000", extractData.getValues("extended-properties:TotalTime")[0]);

     *
     * @param encoding the encoding name to normalize
     * @return the normalized encoding name
     */
    protected String normalizeEncoding(final String encoding) {
        try {
            final EncodingHelper encodingHelper = crawlerContainer.getComponent("encodingHelper");
            return encodingHelper.normalize(encoding);
        } catch (final Exception e) {
            // NOP
        }

          documentation, if provided along with the Derivative Works; or,
          within a display generated by the Derivative Works, if and
          wherever such third-party notices normally appear. The contents
          of the NOTICE file are for informational purposes only and
          do not modify the License. You may add Your own attribution