<?xml version="1.0" encoding="UTF-8"?>
<metadata>
  <groupId>org.codelibs.fess</groupId>
  <artifactId>fess-crawler-smbj</artifactId>
  <versioning>
    <latest>14.15.0</latest>
    <release>14.15.0</release>
    <versions>
      <version>14.14.0</version>
      <version>14.15.0</version>
    </versions>
    <lastUpdated>20240526053206</lastUpdated>
  </versioning>

<!DOCTYPE components PUBLIC "-//DBFLUTE//DTD LastaDi 1.0//EN" 
	"http://dbflute.org/meta/lastadi10.dtd">
<components namespace="fessCrawler">
	<component name="urlQueueService"
		class="org.codelibs.fess.crawler.service.FessUrlQueueService">
		<arg>crawlerConfig</arg>
	</component>

import org.codelibs.fess.crawler.Constants;
import org.codelibs.fess.crawler.builder.RequestDataBuilder;
import org.codelibs.fess.crawler.client.AbstractCrawlerClient;
import org.codelibs.fess.crawler.client.AccessTimeoutTarget;
import org.codelibs.fess.crawler.entity.RequestData;
import org.codelibs.fess.crawler.entity.ResponseData;
import org.codelibs.fess.crawler.exception.ChildUrlsException;

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

import org.codelibs.fess.crawler.Constants;
import org.codelibs.fess.crawler.container.StandardCrawlerContainer;
import org.codelibs.fess.crawler.entity.AccessResultDataImpl;
import org.codelibs.fess.crawler.entity.ResponseData;
import org.codelibs.fess.crawler.entity.ResultData;
import org.codelibs.fess.crawler.exception.CrawlerSystemException;

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

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

import org.codelibs.core.misc.Base64Util;
import org.codelibs.fess.crawler.exception.OpenSearchAccessException;
import org.opensearch.core.xcontent.ToXContent;
import org.opensearch.core.xcontent.XContentBuilder;

/**

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.exception.MaxLengthExceededException;
import org.codelibs.fess.crawler.extractor.Extractor;
import org.codelibs.fess.crawler.extractor.ExtractorFactory;
import org.codelibs.fess.crawler.helper.MimeTypeHelper;

#!/bin/bash

touch $(ls -d ./fess-*/logs)/fess-crawler.log
tail -f ./fess-*/logs/*.log &

mvn test -P integrationTests -Dtest.fess.url="http://localhost:8080" -Dtest.search_engine.url="http://localhost:9201"
ret=$?

if [ $ret != 0 ] ; then
  for f in `find ./target -type f | grep surefire-reports | grep -v /TEST-` ; do
    cat $f
  done
fi

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.codelibs.fess.crawler.extractor.Extractor;
import org.codelibs.fess.crawler.util.TextUtil;
import org.xml.sax.ContentHandler;
import org.xml.sax.SAXException;

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

import java.util.HashMap;
import java.util.Map;

import org.codelibs.core.lang.StringUtil;
import org.codelibs.fess.crawler.exception.CrawlerSystemException;

/**
 * Helper class for managing content length limits based on MIME types.

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

import java.io.InputStream;

import org.apache.logging.log4j.LogManager;
import org.apache.logging.log4j.Logger;
import org.codelibs.core.io.CloseableUtil;
import org.codelibs.core.io.ResourceUtil;
import org.codelibs.fess.crawler.container.StandardCrawlerContainer;
import org.codelibs.fess.crawler.exception.CrawlerSystemException;