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

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

    //                                            ----------

    /**
     * Creates a new scheduled job from a crawler configuration.
     *
     * @param type the crawler type (web, file, or data)
     * @param id the crawler configuration ID
     * @param name the name for the new job (base64 encoded)
     * @return HTML response for the job creation form
     */
    @Execute

    }

    /**
     * Stores crawling information and parameters for the specified session.
     * Creates a new crawling info record if none exists or if create flag is true.
     * Also stores any accumulated information parameters and clears the info map.
     *
     * @param sessionId the session ID for the crawling information
     * @param create if true, creates a new crawling info regardless of existing records

import java.util.LinkedList;
import java.util.List;
import java.util.Map;
import java.util.Queue;

import org.codelibs.fess.crawler.Constants;
import org.codelibs.fess.crawler.entity.AccessResultImpl;
import org.codelibs.fess.crawler.entity.UrlQueueImpl;
import org.codelibs.fess.crawler.helper.MemoryDataHelper;
import org.dbflute.utflute.core.PlainTestCase;
import org.mockito.InjectMocks;
import org.mockito.Mock;

import org.codelibs.core.lang.StringUtil;
import org.codelibs.core.misc.Tuple3;
import org.codelibs.fess.crawler.builder.RequestDataBuilder;
import org.codelibs.fess.crawler.client.CrawlerClient;
import org.codelibs.fess.crawler.client.CrawlerClientFactory;
import org.codelibs.fess.crawler.entity.ResponseData;
import org.codelibs.fess.crawler.exception.CrawlingAccessException;
import org.codelibs.fess.exception.ThumbnailGenerationException;

    /**
     * The unique identifier of the crawling information entry being edited.
     * This is a required field for identifying which crawling info to update.
     */
    @Required
    @Size(max = 1000)
    public String id;

    /**
     * The session identifier of the crawling session.
     * This is a required field that identifies the specific crawling session.
     * Maximum length is 20 characters.
     */

        String[] keywords = { "java", "crawler", "testing" };
        data.putValues("keywords", keywords);

        String[] retrievedKeywords = data.getValues("keywords");
        assertNotNull(retrievedKeywords);
        assertEquals(3, retrievedKeywords.length);
        assertEquals("java", retrievedKeywords[0]);
        assertEquals("crawler", retrievedKeywords[1]);

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

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

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.ExtractData;
import org.codelibs.fess.crawler.exception.CrawlerSystemException;

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

import java.io.ByteArrayInputStream;
import java.io.File;
import java.io.FileInputStream;
import java.io.IOException;
import java.io.InputStream;

import org.codelibs.fess.crawler.entity.ResponseData;
import org.codelibs.fess.crawler.exception.CrawlingAccessException;
import org.dbflute.utflute.core.PlainTestCase;

/**

 * either express or implied. See the License for the specific language
 * governing permissions and limitations under the License.
 */
package org.codelibs.fess.crawler.util;

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

/**
 * A callback interface for processing access results.
 *
 * @param <RESULT> the type of access result that extends {@link AccessResult}
 */