"http://dbflute.org/meta/lastadi10.dtd">
<components namespace="fessCrawler">
	<include path="crawler/container.xml" />

	<component name="contentLengthHelper"
		class="org.codelibs.fess.crawler.helper.ContentLengthHelper" instance="singleton">
	</component>

import org.codelibs.fess.crawler.helper.ContentLengthHelper;
import org.codelibs.fess.crawler.helper.EncodingHelper;
import org.codelibs.fess.crawler.helper.MemoryDataHelper;
import org.codelibs.fess.crawler.helper.RobotsTxtHelper;
import org.codelibs.fess.crawler.helper.SitemapsHelper;
import org.codelibs.fess.crawler.helper.UrlConvertHelper;
import org.codelibs.fess.crawler.helper.impl.LogHelperImpl;

	"http://dbflute.org/meta/lastadi10.dtd">
<components namespace="fessCrawler">
	<include path="crawler/container.xml" />

	<component name="encodingHelper" class="org.codelibs.fess.crawler.helper.EncodingHelper">
		<postConstruct name="addEncodingMapping">
			<arg>"unicode"</arg>
			<arg>"UTF-16LE"</arg>
		</postConstruct>
	</component>

import org.codelibs.fess.crawler.exception.CrawlerSystemException;
import org.codelibs.fess.crawler.exception.MimeTypeException;
import org.codelibs.fess.crawler.helper.MimeTypeHelper;

/**
 * MimeTypeHelperImpl is a helper class that detects the MIME type of a given input stream or filename.
 * It uses the Apache Tika library to detect the MIME type.
 *
 * <p>
 * This class provides methods to:
 * </p>

	</component>
	<component name="urlFilterService" class="org.codelibs.fess.crawler.service.impl.UrlFilterServiceImpl" instance="prototype" >
	</component>

	<!-- Helper -->
	<component name="dataHelper" class="org.codelibs.fess.crawler.helper.MemoryDataHelper" instance="singleton" >
	</component>

import org.codelibs.fess.crawler.exception.CrawlingAccessException;
import org.codelibs.fess.crawler.exception.MaxLengthExceededException;
import org.codelibs.fess.crawler.helper.ContentLengthHelper;
import org.codelibs.fess.crawler.helper.MimeTypeHelper;

import io.minio.BucketExistsArgs;
import io.minio.GetObjectArgs;
import io.minio.GetObjectTagsArgs;
import io.minio.ListObjectsArgs;
import io.minio.MinioClient;

import org.codelibs.fess.crawler.exception.CrawlerSystemException;
import org.codelibs.fess.crawler.exception.CrawlingAccessException;
import org.codelibs.fess.crawler.helper.EncodingHelper;
import org.codelibs.fess.crawler.helper.UrlConvertHelper;
import org.codelibs.fess.crawler.util.CharUtil;
import org.codelibs.fess.crawler.util.XPathAPI;
import org.codelibs.nekohtml.parsers.DOMParser;
import org.w3c.dom.Document;

 * 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.

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

import java.util.List;
import java.util.regex.Pattern;

import org.codelibs.fess.crawler.helper.MemoryDataHelper;
import org.codelibs.fess.crawler.service.UrlFilterService;

import jakarta.annotation.Resource;

/**
 * Implementation of the {@link UrlFilterService} interface.

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

import jakarta.annotation.Resource;

/**
 * An abstract base class for implementing Extractor interfaces.
 * Provides common functionality such as access to CrawlerContainer components