import java.io.Serializable;

/**
 * The Sitemap interface represents a sitemap entity with location and last modification date.
 * It extends the Serializable interface to allow sitemap objects to be serialized.
 */
public interface Sitemap extends Serializable {

    /**
     * Retrieves the location (URL) of the sitemap.
     *
     * @return the location of the sitemap as a String.
     */
    String getLoc();

            final Set<RequestData> requestDataSet = new LinkedHashSet<>();
            for (final Sitemap sitemap : sitemapSet.getSitemaps()) {
                if (sitemap != null) {
                    requestDataSet.add(RequestDataBuilder.newRequestData().get().url(sitemap.getLoc()).build()); // TODO priority
                }
            }

        sitemapList.add(sitemap);
    }

    /**
     * Removes a sitemap from this set.
     * @param sitemap the sitemap to remove
     */
    public void removeSitemap(final Sitemap sitemap) {
        sitemapList.remove(sitemap);
    }

    /**
     * Gets all sitemaps in this set as an array.
     * @return an array of sitemaps
     */
    public Sitemap[] getSitemaps() {

<?xml version="1.0" encoding="UTF-8"?>
<sitemapindex xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
   <sitemap>
      <loc>http://www.example.com/sitemap1.xml.gz</loc>
      <lastmod>2004-10-01T18:23:17+00:00</lastmod>
   </sitemap>
   <sitemap>
      <loc>http://www.example.com/sitemap2.xml.gz</loc>
      <lastmod>2005-01-01</lastmod>
   </sitemap>

        sitemapsRule.addRule("url", Pattern.compile(".*sitemap.*"));
        assertTrue(sitemapsRule.match(responseData));
        InputStream is = responseData.getResponseBody();
        assertTrue(is instanceof InputStream);
        CloseableUtil.closeQuietly(responseData);
    }

    private void assertMatchFalse(ResponseData responseData) {
        sitemapsRule.addRule("url", Pattern.compile(".*sitemap.*"));

package org.codelibs.fess.crawler.entity;

import org.codelibs.core.lang.StringUtil;

/**
 * Represents a Sitemap file entry, conforming to the Sitemap XML format.
 * This class holds information about a single Sitemap, including its location and last modification timestamp.
 * It implements the {@link Sitemap} interface.
 *
 * <p>
 * A Sitemap file provides search engines with a list of URLs available for crawling.

        accessResult1.setContentLength(Long.valueOf(10));
        accessResult1.setCreateTime(System.currentTimeMillis());
        accessResult1.setExecutionTime(10);
        accessResult1.setHttpStatusCode(200);
        accessResult1.setLastModified(System.currentTimeMillis());
        accessResult1.setMethod("GET");
        accessResult1.setMimeType("text/plain");

 * 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 class SitemapsRule extends RegexRule {
    /**

import java.io.ByteArrayInputStream;
import java.io.InputStream;

import org.codelibs.core.io.ResourceUtil;
import org.codelibs.fess.crawler.container.StandardCrawlerContainer;
import org.codelibs.fess.crawler.entity.Sitemap;
import org.codelibs.fess.crawler.entity.SitemapSet;
import org.codelibs.fess.crawler.entity.SitemapUrl;
import org.codelibs.fess.crawler.exception.CrawlingAccessException;
import org.dbflute.utflute.core.PlainTestCase;

    private static final String SPRING_PROFILES_ACTIVE = "spring.profiles.active";

    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);
            }
        }