import org.codelibs.core.lang.ThreadUtil;
import org.codelibs.fess.crawler.entity.ExtractData;
import org.codelibs.fess.crawler.exception.CrawlerSystemException;
import org.codelibs.fess.crawler.exception.ExtractException;
import org.codelibs.fess.crawler.extractor.Extractor;
import org.codelibs.fess.crawler.extractor.ExtractorFactory;
import org.codelibs.fess.crawler.helper.MimeTypeHelper;

/**

 */
package org.codelibs.fess.crawler.extractor.impl;

import java.io.IOException;
import java.io.InputStream;
import java.util.Map;

import org.apache.poi.hpbf.extractor.PublisherTextExtractor;
import org.codelibs.fess.crawler.entity.ExtractData;
import org.codelibs.fess.crawler.exception.CrawlerSystemException;
import org.codelibs.fess.crawler.exception.ExtractException;

/**
 * Gets a text from . file.

 * either express or implied. See the License for the specific language
 * governing permissions and limitations under the License.
 */
package org.codelibs.fess.crawler.service;

import java.util.List;

import org.codelibs.fess.crawler.entity.AccessResult;
import org.codelibs.fess.crawler.util.AccessResultCallback;

/**
 * Interface for data service operations related to access results.
 *
 * @param <RESULT> the type of access result

import org.codelibs.fess.app.service.RequestHeaderService;
import org.codelibs.fess.app.service.WebAuthenticationService;
import org.codelibs.fess.crawler.client.CrawlerClientFactory;
import org.codelibs.fess.crawler.client.http.Authentication;
import org.codelibs.fess.crawler.client.http.HcHttpClient;
import org.codelibs.fess.helper.SystemHelper;
import org.codelibs.fess.mylasta.direction.FessConfig;

 * governing permissions and limitations under the License.
 */
package org.codelibs.fess.crawler.interval;

/**
 * The {@code IntervalController} interface defines methods for controlling
 * the delay intervals in a web crawler. It includes constants representing
 * different types of processing states and a method to introduce a delay
 * based on the type of processing.
 * <p>

 */
package org.codelibs.fess.crawler.service.impl;

import java.util.ArrayList;
import java.util.HashSet;
import java.util.List;
import java.util.Set;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.TimeUnit;
import java.util.regex.Pattern;
import java.util.stream.Collectors;

import org.codelibs.fess.crawler.entity.OpenSearchUrlFilter;

 */
package org.codelibs.fess.crawler.helper.impl;

import java.util.Set;

import org.apache.logging.log4j.LogManager;
import org.apache.logging.log4j.Logger;
import org.codelibs.fess.crawler.entity.RequestData;
import org.codelibs.fess.crawler.entity.ResponseData;
import org.codelibs.fess.crawler.entity.UrlQueue;
import org.codelibs.fess.crawler.exception.CrawlingAccessException;

 * governing permissions and limitations under the License.
 */
package org.codelibs.fess.crawler.exception;

/**
 * Exception thrown during sitemaps processing in the crawler.
 * This exception extends {@link org.codelibs.fess.crawler.exception.CrawlerSystemException}
 * and provides constructors for wrapping other exceptions or creating a new exception with a message.
 */

 * either express or implied. See the License for the specific language
 * governing permissions and limitations under the License.
 */
package org.codelibs.fess.crawler.client.smb1;

import org.codelibs.fess.crawler.entity.ResponseData;
import org.codelibs.fess.crawler.exception.CrawlingAccessException;
import org.dbflute.utflute.core.PlainTestCase;

/**
 * @author shinsuke
 *
 */

 */
package org.codelibs.fess.crawler.rule.impl;

import java.io.File;

import org.codelibs.core.io.CloseableUtil;
import org.codelibs.core.io.ResourceUtil;
import org.codelibs.fess.crawler.container.StandardCrawlerContainer;
import org.codelibs.fess.crawler.entity.ResponseData;
import org.codelibs.fess.crawler.helper.SitemapsHelper;
import org.codelibs.fess.crawler.rule.Rule;