//gzipParser.java //------------------------ //part of YaCy //(C) by Michael Peter Christen; mc@yacy.net //first published on http://www.anomic.de //Frankfurt, Germany, 2005 // //this file is contributed by Martin Thelian // // $LastChangedDate$ // $LastChangedRevision$ // $LastChangedBy$ // //This program is free software; you can redistribute it and/or modify //it under the terms of the GNU General Public License as published by //the Free Software Foundation; either version 2 of the License, or //(at your option) any later version. // //This program is distributed in the hope that it will be useful, //but WITHOUT ANY WARRANTY; without even the implied warranty of //MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the //GNU General Public License for more details. // //You should have received a copy of the GNU General Public License //along with this program; if not, write to the Free Software //Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA package net.yacy.document.parser; import java.io.File; import java.io.FileOutputStream; import java.io.InputStream; import java.util.Date; import java.util.zip.GZIPInputStream; import net.yacy.cora.document.id.DigestURL; import net.yacy.cora.document.id.MultiProtocolURL; import net.yacy.document.AbstractParser; import net.yacy.document.Document; import net.yacy.document.Parser; import net.yacy.document.TextParser; import net.yacy.document.VocabularyScraper; import net.yacy.kelondro.util.FileUtils; import org.apache.commons.compress.compressors.gzip.GzipUtils; /** * Parses a gz archive. * Unzips and parses the content and adds it to the created main document */ public class gzipParser extends AbstractParser implements Parser { public gzipParser() { super("GNU Zip Compressed Archive Parser"); this.SUPPORTED_EXTENSIONS.add("gz"); this.SUPPORTED_EXTENSIONS.add("tgz"); this.SUPPORTED_MIME_TYPES.add("application/x-gzip"); this.SUPPORTED_MIME_TYPES.add("application/gzip"); this.SUPPORTED_MIME_TYPES.add("application/x-gunzip"); this.SUPPORTED_MIME_TYPES.add("application/gzipped"); this.SUPPORTED_MIME_TYPES.add("application/gzip-compressed"); this.SUPPORTED_MIME_TYPES.add("gzip/document"); } @Override public Document[] parse( final DigestURL location, final String mimeType, final String charset, final VocabularyScraper scraper, final int timezoneOffset, final InputStream source) throws Parser.Failure, InterruptedException { File tempFile = null; Document maindoc = null; try { int read = 0; final byte[] data = new byte[1024]; final GZIPInputStream zippedContent = new GZIPInputStream(source); tempFile = File.createTempFile("gunzip","tmp"); // creating a temp file to store the uncompressed data final FileOutputStream out = new FileOutputStream(tempFile); // reading gzip file and store it uncompressed while ((read = zippedContent.read(data, 0, 1024)) != -1) { out.write(data, 0, read); } zippedContent.close(); out.close(); final String filename = location.getFileName(); // create maindoc for this gzip container, register with supplied url & mime maindoc = new Document( location, mimeType, charset, this, null, null, AbstractParser.singleList(filename.isEmpty() ? location.toTokens() : MultiProtocolURL.unescape(filename)), // title null, null, null, null, 0.0d, 0.0d, (Object) null, null, null, null, false, new Date()); // creating a new parser class to parse the unzipped content final String contentfilename = GzipUtils.getUncompressedFilename(location.getFileName()); final String mime = TextParser.mimeOf(DigestURL.getFileExtension(contentfilename)); Document[] docs = TextParser.parseSource(location, mime, null, scraper, timezoneOffset, 999, tempFile); if (docs != null) maindoc.addSubDocuments(docs); } catch (final Exception e) { if (e instanceof InterruptedException) throw (InterruptedException) e; if (e instanceof Parser.Failure) throw (Parser.Failure) e; throw new Parser.Failure("Unexpected error while parsing gzip file. " + e.getMessage(),location); } finally { if (tempFile != null) FileUtils.deletedelete(tempFile); } return maindoc == null ? null : new Document[]{maindoc}; } }