final Extractor extractor = extractorFactory.getExtractor(mimeType);
                    if (extractor != null) {
                        try {
                            final Map<String, String> map = new HashMap<>();
                            map.put(ExtractData.RESOURCE_NAME_KEY, filename);
                            buf.append(extractor.getText(new IgnoreCloseInputStream(ais), map).getContent());

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

import java.io.File;
import java.io.IOException;
import java.util.List;

import org.codelibs.fess.crawler.container.CrawlerContainer;
import org.codelibs.fess.crawler.exception.CrawlerSystemException;
import org.codelibs.fess.crawler.extractor.Extractor;
import org.codelibs.fess.crawler.extractor.ExtractorFactory;

                    final Extractor extractor = extractorFactory.getExtractor(mimeType);
                    if (extractor != null) {
                        try {
                            final Map<String, String> map = new HashMap<>();
                            map.put(ExtractData.RESOURCE_NAME_KEY, filename);
                            buf.append(extractor.getText(new IgnoreCloseInputStream(ais), map).getContent());

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

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

import org.apache.poi.hdgf.extractor.VisioTextExtractor;
import org.codelibs.fess.crawler.entity.ExtractData;
import org.codelibs.fess.crawler.exception.CrawlerSystemException;

 * 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.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;

 * 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.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.entity.ResponseData;
import org.codelibs.fess.crawler.entity.ResultData;
import org.codelibs.fess.crawler.exception.CrawlerSystemException;
import org.codelibs.fess.crawler.extractor.ExtractorFactory;
import org.codelibs.fess.crawler.extractor.impl.TikaExtractor;
import org.dbflute.utflute.core.PlainTestCase;

/**
 * @author shinsuke
 *
 */
public class TextTransformerTest extends PlainTestCase {

 * 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.logging.log4j.LogManager;
import org.apache.logging.log4j.Logger;
import org.codelibs.core.io.ResourceUtil;

        final BufferedInputStream bis = new BufferedInputStream(in);
        final String encoding = htmlExtractor.getEncoding(bis);
        CloseableUtil.closeQuietly(bis);
        assertEquals("UTF-8", encoding);
    }

    public void test_getEncoding_sjis() {
        final InputStream in = ResourceUtil.getResourceAsStream("extractor/test_sjis.html");

 * 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;