import org.codelibs.fess.crawler.exception.SitemapsException;
import org.xml.sax.Attributes;
import org.xml.sax.SAXNotRecognizedException;
import org.xml.sax.SAXNotSupportedException;
import org.xml.sax.helpers.DefaultHandler;

/**
 * Helper class for parsing and validating sitemaps.
 * It supports XML sitemaps, XML sitemap indexes, and text sitemaps,
 * and can handle GZIP compressed sitemaps.

            final ExtractData data = new ExtractData(content != null ? content : StringUtil.EMPTY);
            final Enumeration<Header> headers = message.getAllHeaders();
            while (headers.hasMoreElements()) {
                final Header header = headers.nextElement();
                data.putValue(header.getName(), header.getValue());
            }
            putValue(data, "Content-ID", message.getContentID());

import org.codelibs.fess.crawler.helper.ContentLengthHelper;
import org.codelibs.fess.crawler.helper.EncodingHelper;
import org.codelibs.fess.crawler.helper.MemoryDataHelper;
import org.codelibs.fess.crawler.helper.RobotsTxtHelper;
import org.codelibs.fess.crawler.helper.SitemapsHelper;
import org.codelibs.fess.crawler.helper.UrlConvertHelper;
import org.codelibs.fess.crawler.helper.impl.LogHelperImpl;

import org.codelibs.fess.crawler.exception.CrawlingAccessException;
import org.codelibs.fess.crawler.exception.MaxLengthExceededException;
import org.codelibs.fess.crawler.helper.ContentLengthHelper;
import org.codelibs.fess.crawler.helper.MimeTypeHelper;

import io.minio.BucketExistsArgs;
import io.minio.GetObjectArgs;
import io.minio.GetObjectTagsArgs;
import io.minio.ListObjectsArgs;
import io.minio.MinioClient;

import org.codelibs.fess.crawler.exception.CrawlerSystemException;
import org.codelibs.fess.crawler.exception.CrawlingAccessException;
import org.codelibs.fess.crawler.helper.EncodingHelper;
import org.codelibs.fess.crawler.helper.UrlConvertHelper;
import org.codelibs.fess.crawler.util.CharUtil;
import org.codelibs.fess.crawler.util.XPathAPI;
import org.codelibs.nekohtml.parsers.DOMParser;
import org.w3c.dom.Document;

import org.codelibs.fess.crawler.filter.UrlFilter;
import org.codelibs.fess.crawler.filter.impl.UrlFilterImpl;
import org.codelibs.fess.crawler.helper.MemoryDataHelper;
import org.codelibs.fess.crawler.helper.RobotsTxtHelper;
import org.codelibs.fess.crawler.helper.impl.MimeTypeHelperImpl;
import org.codelibs.fess.crawler.service.impl.UrlFilterServiceImpl;
import org.codelibs.fess.crawler.util.CrawlerWebServer;

 * 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.helper.impl;

import java.util.Set;

import org.apache.logging.log4j.LogManager;
import org.apache.logging.log4j.Logger;
import org.codelibs.fess.crawler.entity.RequestData;

import org.codelibs.fess.crawler.exception.CrawlingAccessException;
import org.codelibs.fess.crawler.exception.MaxLengthExceededException;
import org.codelibs.fess.crawler.helper.ContentLengthHelper;
import org.codelibs.fess.crawler.helper.MimeTypeHelper;

import jakarta.annotation.Resource;

/**
 * FileSystemClient is CrawlerClient implementation to crawl files on a file
 * system.
 *
 * @author shinsuke
 *

                return null;
            }

            data.setContent(EntityUtils.toString(response.getEntity(), Charsets.UTF_8));
            final Header[] headers = response.getAllHeaders();
            for (final Header header : headers) {
                data.putValue(header.getName(), header.getValue());
            }
        } catch (final IOException e) {
            throw new ExtractException(e);

import org.codelibs.fess.crawler.exception.CrawlingAccessException;
import org.codelibs.fess.crawler.exception.MaxLengthExceededException;
import org.codelibs.fess.crawler.helper.ContentLengthHelper;
import org.codelibs.fess.crawler.helper.MimeTypeHelper;

import jakarta.annotation.Resource;
import org.codelibs.jcifs.smb1.Config;
import org.codelibs.jcifs.smb1.ACE;
import org.codelibs.jcifs.smb1.NtlmPasswordAuthentication;