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

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

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

/**
 * @author shinsuke
 *
 */
public interface Extractor {

    ExtractData getText(InputStream in, Map<String, String> params);

    default int getWeight() {

		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"
		class="org.codelibs.fess.crawler.extractor.impl.PdfExtractor" />
	<component name="lhaExtractor"

	"http://dbflute.org/meta/lastadi10.dtd">
<components namespace="fessCrawler">
	<include path="crawler/container.xml" />
	<component name="tikaExtractor"
		class="org.codelibs.fess.crawler.extractor.impl.TikaExtractor">
		<property name="maxCompressionRatio">2</property>
		<property name="maxUncompressionSize">10000000</property>
	</component>

        assertContentType("application/pdf", "extractor/test.pdf", "hoge.pdf");

        assertContentType("application/gzip", "extractor/gz/test.tar.gz", "hoge.tar.gz");
        assertContentType("application/zip", "extractor/zip/test.zip", "hoge.zip");
        assertContentType("application/x-lharc", "extractor/lha/test.lzh", "hoge.lzh"); // TODO is it correct?

        assertContentType("application/xml", "extractor/test.mm", "hoge.mm");

    protected Map<String, Extractor> extractorMap = new HashMap<>();

    public void addExtractor(final String key, final Extractor extractor) {
        if (StringUtil.isBlank(key)) {
            throw new CrawlerSystemException("The key is null.");
        }
        if (extractor == null) {
            throw new CrawlerSystemException("The extractor is null.");
        }

            final Extractor extractor = extractorFactory.getExtractor(mimeType);
            if (extractor != null) {
                try (COSInputStream is = embeddedFile.createInputStream()) {
                    final Map<String, String> map = new HashMap<>();
                    map.put(ExtractData.RESOURCE_NAME_KEY, filename);
                    final String content = extractor.getText(is, map).getContent();

        final CommandExtractor extractor = new CommandExtractor();
        extractor.executionTimeout = 1000L;
        extractor.command = getCommand(scriptFile);
        final Map<String, String> params = new HashMap<String, String>();
        try {
            final ExtractData data = extractor.getText(new FileInputStream(contentFile), params);
            fail(data.toString());

import org.apache.logging.log4j.LogManager;
import org.apache.logging.log4j.Logger;
import org.codelibs.fess.crawler.entity.ResponseData;
import org.codelibs.fess.crawler.extractor.Extractor;
import org.codelibs.fess.crawler.extractor.ExtractorFactory;
import org.codelibs.fess.exception.FessSystemException;
import org.codelibs.fess.mylasta.direction.FessConfig;
import org.codelibs.fess.util.ComponentUtil;

 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND,
 * either express or implied. See the License for the specific language
 * governing permissions and limitations under the License.
 */
package org.codelibs.fess.crawler.extractor.impl;

import java.io.IOException;
import java.io.InputStream;

import org.apache.commons.compress.archivers.ArchiveStreamFactory;
import org.apache.commons.compress.compressors.CompressorStreamFactory;

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;
import org.codelibs.fess.crawler.util.IgnoreCloseInputStream;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;