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

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

import org.apache.poi.hslf.usermodel.HSLFShape;
import org.apache.poi.hslf.usermodel.HSLFSlideShow;
import org.apache.poi.hslf.usermodel.HSLFTextParagraph;
import org.apache.poi.sl.extractor.SlideShowExtractor;
import org.codelibs.fess.crawler.entity.ExtractData;

<!DOCTYPE components PUBLIC "-//DBFLUTE//DTD LastaDi 1.0//EN"
	"http://dbflute.org/meta/lastadi10.dtd">
<components namespace="fessCrawler">
	<include path="crawler/container.xml" />

	<component name="logHelper"
		class="org.codelibs.fess.crawler.helper.impl.LogHelperImpl">
	</component>

<!DOCTYPE components PUBLIC "-//DBFLUTE//DTD LastaDi 1.0//EN"
	"http://dbflute.org/meta/lastadi10.dtd">
<components namespace="fessCrawler">
	<include path="crawler/container.xml" />

	<component name="mimeTypeHelper"
		class="org.codelibs.fess.crawler.helper.impl.MimeTypeHelperImpl" instance="singleton">
	</component>

<!DOCTYPE components PUBLIC "-//DBFLUTE//DTD LastaDi 1.0//EN"
	"http://dbflute.org/meta/lastadi10.dtd">
<components namespace="fessCrawler">
	<include path="crawler/container.xml" />

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

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

import java.io.InputStream;

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

/**

<!DOCTYPE components PUBLIC "-//DBFLUTE//DTD LastaDi 1.0//EN"
	"http://dbflute.org/meta/lastadi10.dtd">
<components namespace="fessCrawler">
	<include path="crawler/container.xml" />

	<!-- Converts child links to normalize it. -->
	<component name="urlConvertHelper" class="org.codelibs.fess.crawler.helper.UrlConvertHelper">
	</component>

 */
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;

		class="org.codelibs.fess.crawler.extractor.impl.MsExcelExtractor" />
	<component name="msPowerPointExtractor"
		class="org.codelibs.fess.crawler.extractor.impl.MsPowerPointExtractor" />
	<component name="msPublisherExtractor"
		class="org.codelibs.fess.crawler.extractor.impl.MsPublisherExtractor" />
	<component name="msVisioExtractor"
		class="org.codelibs.fess.crawler.extractor.impl.MsVisioExtractor" />
	<component name="pdfExtractor"

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

import org.codelibs.fess.crawler.container.StandardCrawlerContainer;
import org.codelibs.fess.crawler.helper.MemoryDataHelper;
import org.codelibs.fess.crawler.service.impl.UrlFilterServiceImpl;
import org.dbflute.utflute.core.PlainTestCase;

/**
 * @author shinsuke
 *
 */

import org.apache.tika.mime.MimeTypesFactory;
import org.codelibs.core.lang.StringUtil;
import org.codelibs.fess.crawler.entity.ExtractData;
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.