/**
 * SitemapsRule is a rule implementation that extends RegexRule to determine if a given response data
 * represents a valid sitemap. It uses a SitemapsHelper to validate the response body as an InputStream.
 * The rule checks if the URL matches the defined regex pattern and then validates the content as a sitemap.
 * If any exception occurs during the sitemap validation, it logs the error and returns false.
 *
 */

    public static void setup() {
        final List<String> list = new ArrayList<String>();

        final String values = System.getProperty(SPRING_PROFILES_ACTIVE);
        if (values != null) {
            for (final String value : values.split(",")) {
                list.add(value);
            }
        }

        if (ResourceUtil.getResourceNoException("org/h2/Driver.class") != null) {

    /**
     * How frequently the page is likely to change. This value provides general
     * information to search engines and may not correlate exactly to how often
     * they crawl the page. Valid values are:
     * <ul>
     * <li>always</li>
     * <li>hourly</li>
     * <li>daily</li>
     * <li>weekly</li>
     * <li>monthly</li>
     * <li>yearly</li>
     * <li>never</li>
     * </ul>

     * @param tagFieldNames the names of the fields to use as tags
     * @param roleFieldName the name of the field to use for roles
     * @param readingConverter the converter to use for reading values
     * @param normalizer the normalizer to use for normalizing values
     * @return a list of suggest items parsed from the query log
     */
    List<SuggestItem> parseQueryLog(QueryLog queryLog, String[] fields, String[] tagFieldNames, String roleFieldName,

    /**
     * Checks if the given character is a valid URL character.
     *
     * Valid URL characters include:
     * - Lowercase letters (a-z)
     * - Uppercase letters (A-Z)
     * - Digits (0-9)
     * - Special characters: . - * _ : / + % = & ? # [ ] @ ~ ! $ ' ( ) , ;
     *
     * @param c the character to check
     * @return {@code true} if the character is a valid URL character, {@code false} otherwise
     */

        this.content = content;
    }

    /**
     * Puts multiple values for a given key in the metadata.
     *
     * @param key the metadata key
     * @param values the values to associate with the key
     */
    public void putValues(final String key, final String[] values) {
        metadata.put(key, values);
    }

    /**
     * Puts a single value for a given key in the metadata.
     *

     */
    public void setValue(final String value) {
        this.value = value;
    }

    /**
     * Checks if the request header is valid.
     * A header is considered valid if its name is not blank and its value is not null.
     * @return true if the header is valid, false otherwise.
     */
    public boolean isValid() {
        if (StringUtil.isBlank(name) || value == null) {
            return false;
        }

import java.nio.charset.Charset;
import java.nio.charset.StandardCharsets;

/**
 * Constants used in the fess-crawler.
 * This class provides a collection of constant values for HTTP methods, status codes,
 * transformer names, boolean values, character encodings, date/time formats, and XML features.
 * It is designed to avoid the instantiation.
 */
public final class Constants {
    /**
     * The GET method.
     */

    /**
     *  Use an XPath string to select a nodelist.
     *  XPath namespace prefixes are resolved from the contextNode.
     *
     *  @param contextNode The node to start searching from.
     *  @param expression A valid XPath string.
     *  @return A XPathNodes, should never be null.
     *
     * @throws XPathExpressionException if an XPath expression error occurs.
     */

        nBestExamples = settings.get(NBEST_EXAMPLES);
    }

    private static String parse(String rule, Set<String> dup) {
        String[] values = CSVUtil.parse(rule);
        if (dup.add(values[0]) == false) {
            throw new IllegalArgumentException("Found duplicate term [" + values[0] + "] in user dictionary.");
        }
        return rule;
    }