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

/**
 * ChildUrlsException is thrown when having child urls.
 *
 * @author shinsuke
 *
 */
public class ChildUrlsException extends CrawlerSystemException {

    private static final long serialVersionUID = 1L;

    private final Set<RequestData> childUrlList;

    public ChildUrlsException(final Set<RequestData> childUrlList, final String description) {

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

/**
 * @author shinsuke
 *
 */
public class CrawlingAccessException extends CrawlerSystemException {

    private static final long serialVersionUID = 1L;

    public static final String DEBUG = "DEBUG";

    public static final String INFO = "ERROR";

    public static final String WARN = "WARN";

import org.codelibs.core.lang.StringUtil;
import org.codelibs.fess.crawler.Constants;
import org.codelibs.fess.crawler.entity.AccessResultData;
import org.codelibs.fess.crawler.exception.CrawlerSystemException;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.xml.sax.Attributes;
import org.xml.sax.InputSource;
import org.xml.sax.helpers.DefaultHandler;

/**
 * @author shinsuke
 *
 */

import org.codelibs.fess.crawler.client.http.impl.AuthenticationImpl;
import org.codelibs.fess.crawler.client.http.ntlm.JcifsEngine;
import org.codelibs.fess.crawler.exception.CrawlerSystemException;
import org.codelibs.fess.es.config.bsentity.BsWebAuthentication;
import org.codelibs.fess.es.config.exentity.CrawlingConfig.ConfigName;
import org.codelibs.fess.es.config.exentity.CrawlingConfig.Param.Config;

import java.util.regex.Matcher;
import java.util.regex.Pattern;
import java.util.stream.Collectors;

import org.codelibs.fess.crawler.container.CrawlerContainer;
import org.codelibs.fess.crawler.exception.CrawlerSystemException;
import org.codelibs.fess.crawler.filter.UrlFilter;
import org.codelibs.fess.crawler.service.UrlFilterService;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

import jakarta.annotation.Resource;

 * either express or implied. See the License for the specific language
 * governing permissions and limitations under the License.
 */

/**
 * @author shinsuke
 *
 */
public class SitemapsException extends CrawlerSystemException {
    private static final long serialVersionUID = 1L;

    public SitemapsException(final String message, final Throwable cause) {
        super(message, cause);
    }

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

/**
 * @author shinsuke
 *
 */
public class MimeTypeException extends CrawlerSystemException {

    private static final long serialVersionUID = 1L;

    public MimeTypeException(final String message, final Throwable cause) {
        super(message, cause);
    }

import org.codelibs.core.io.InputStreamUtil;
import org.codelibs.core.lang.StringUtil;
import org.codelibs.fess.crawler.Constants;
import org.codelibs.fess.crawler.entity.ExtractData;
import org.codelibs.fess.crawler.exception.CrawlerSystemException;
import org.codelibs.fess.crawler.exception.ExtractException;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

/**
 * @author shinsuke
 *
 */

import org.apache.pdfbox.text.PDFTextStripper;
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 javax.xml.xpath.XPathNodes;

import org.codelibs.core.lang.StringUtil;
import org.codelibs.core.stream.StreamUtil;
import org.codelibs.fess.crawler.entity.ExtractData;
import org.codelibs.fess.crawler.exception.CrawlerSystemException;
import org.codelibs.fess.crawler.util.XPathAPI;
import org.codelibs.nekohtml.parsers.DOMParser;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.w3c.dom.Document;
import org.w3c.dom.Node;