Shinsuke Sugaya <******@****.***> 1444529815 +0900

Shinsuke Sugaya <******@****.***> 1444529815 +0900

                    factory.addExtractor("application/zip", zipExtractor);

                })//
        ;

        zipExtractor = container.getComponent("zipExtractor");
    }

    public void test_getText() {
        final InputStream in = ResourceUtil.getResourceAsStream("extractor/zip/test.zip");
        final String content = zipExtractor.getText(in, null).getContent();

        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");

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

import org.apache.commons.compress.archivers.ArchiveInputStream;
import org.apache.commons.compress.archivers.ArchiveStreamFactory;
import org.apache.commons.compress.archivers.zip.ZipArchiveEntry;
import org.apache.logging.log4j.LogManager;
import org.apache.logging.log4j.Logger;
import org.codelibs.fess.crawler.entity.ExtractData;
import org.codelibs.fess.crawler.exception.CrawlerSystemException;

        assertEquals("かいしゃ", extractData.getValues("dc:publisher")[0]);
    }

    public void test_getTika_zip() {
        final InputStream in = ResourceUtil.getResourceAsStream("extractor/zip/test.zip");
        final ExtractData extractData = tikaExtractor.getText(in, null);
        final String content = extractData.getContent();
        CloseableUtil.closeQuietly(in);
        logger.info(content);

 *   <li>Reading content as plain text if all other methods fail</li>
 *   <li>Applying post-extraction filters</li>
 *   <li>Handling Tika exceptions, including zip bomb exceptions</li>
 * </ul>
 *
 * <p>
 * The class also supports configuration options such as:
 * </p>
 * <ul>
 *   <li>Output encoding</li>
 *   <li>Maximum compression ratio and uncompression size</li>

  <mime-type type="application/x-wacz">
    <_comment>web archive frictionless zip</_comment>
    <sub-class-of type="application/x-vnd.datapackage+zip"/>
  </mime-type>
  <mime-type type="application/x-vnd.datapackage+zip">
    <_comment>frictionless data package zip package</_comment>
    <sub-class-of type="application/zip"/>
  </mime-type>
  <mime-type type="application/x-vnd.datapackage+json">

				"application/ecmascript",
				"application/edi-consent",
				"application/edi-x12",
				"application/edifact",
				"application/emma+xml",
				"application/epp+xml",
				"application/epub+zip",
				"application/eshop",
				"application/example",
				"application/fastinfoset",
				"application/fastsoap",
				"application/fits",
				"application/font-tdpfr",
				"application/h224",

 */
package org.codelibs.fess.crawler.helper;

import java.io.BufferedInputStream;
import java.io.BufferedReader;
import java.io.InputStream;
import java.io.InputStreamReader;
import java.util.zip.GZIPInputStream;

import javax.xml.XMLConstants;
import javax.xml.parsers.SAXParser;
import javax.xml.parsers.SAXParserFactory;

import org.apache.logging.log4j.LogManager;
import org.apache.logging.log4j.Logger;