001: //gzipParser.java
002: //------------------------
003: //part of YaCy
004: //(C) by Michael Peter Christen; mc@anomic.de
005: //first published on http://www.anomic.de
006: //Frankfurt, Germany, 2005
007: //
008: //this file is contributed by Martin Thelian
009: //last major change: 24.04.2005
010: //
011: //This program is free software; you can redistribute it and/or modify
012: //it under the terms of the GNU General Public License as published by
013: //the Free Software Foundation; either version 2 of the License, or
014: //(at your option) any later version.
015: //
016: //This program is distributed in the hope that it will be useful,
017: //but WITHOUT ANY WARRANTY; without even the implied warranty of
018: //MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
019: //GNU General Public License for more details.
020: //
021: //You should have received a copy of the GNU General Public License
022: //along with this program; if not, write to the Free Software
023: //Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
024: //
025: //Using this software in any meaning (reading, learning, copying, compiling,
026: //running) means that you agree that the Author(s) is (are) not responsible
027: //for cost, loss of data or any harm that may be caused directly or indirectly
028: //by usage of this softare or this documentation. The usage of this software
029: //is on your own risk. The installation and usage (starting/running) of this
030: //software may allow other people or application to access your computer and
031: //any attached devices and is highly dependent on the configuration of the
032: //software which must be done by the user of the software; the author(s) is
033: //(are) also not responsible for proper configuration and usage of the
034: //software, even if provoked by documentation provided together with
035: //the software.
036: //
037: //Any changes to this file according to the GPL as documented in the file
038: //gpl.txt aside this file in the shipment you received can be done to the
039: //lines that follows this copyright notice here, but changes must not be
040: //done inside the copyright notive above. A re-distribution must contain
041: //the intact and unchanged copyright notice.
042: //Contributions and changes to the program code must be marked as such.
043:
044: package de.anomic.plasma.parser.gzip;
045:
046: import java.io.File;
047: import java.io.FileOutputStream;
048: import java.io.InputStream;
049: import java.util.Hashtable;
050: import java.util.zip.GZIPInputStream;
051:
052: import de.anomic.plasma.plasmaParser;
053: import de.anomic.plasma.plasmaParserDocument;
054: import de.anomic.plasma.parser.AbstractParser;
055: import de.anomic.plasma.parser.Parser;
056: import de.anomic.plasma.parser.ParserException;
057: import de.anomic.yacy.yacyURL;
058:
059: public class gzipParser extends AbstractParser implements Parser {
060:
061: /**
062: * a list of mime types that are supported by this parser class
063: * @see #getSupportedMimeTypes()
064: */
065: public static final Hashtable<String, String> SUPPORTED_MIME_TYPES = new Hashtable<String, String>();
066: static {
067: SUPPORTED_MIME_TYPES.put("application/x-gzip", "gz,tgz");
068: SUPPORTED_MIME_TYPES.put("application/gzip", "gz,tgz");
069: }
070:
071: /**
072: * a list of library names that are needed by this parser
073: * @see Parser#getLibxDependences()
074: */
075: private static final String[] LIBX_DEPENDENCIES = new String[] {};
076:
077: public gzipParser() {
078: super (LIBX_DEPENDENCIES);
079: this .parserName = "GNU Zip Compressed Archive Parser";
080: }
081:
082: public Hashtable<String, String> getSupportedMimeTypes() {
083: return SUPPORTED_MIME_TYPES;
084: }
085:
086: public plasmaParserDocument parse(yacyURL location,
087: String mimeType, String charset, InputStream source)
088: throws ParserException, InterruptedException {
089:
090: File tempFile = null;
091: try {
092: int read = 0;
093: byte[] data = new byte[1024];
094:
095: GZIPInputStream zippedContent = new GZIPInputStream(source);
096:
097: tempFile = File.createTempFile("gunzip", "tmp");
098: tempFile.deleteOnExit();
099:
100: // creating a temp file to store the uncompressed data
101: FileOutputStream out = new FileOutputStream(tempFile);
102:
103: // reading gzip file and store it uncompressed
104: while ((read = zippedContent.read(data, 0, 1024)) != -1) {
105: out.write(data, 0, read);
106: }
107: zippedContent.close();
108: out.close();
109:
110: // check for interruption
111: checkInterruption();
112:
113: // creating a new parser class to parse the unzipped content
114: plasmaParser theParser = new plasmaParser();
115: return theParser
116: .parseSource(location, null, null, tempFile);
117: } catch (Exception e) {
118: if (e instanceof InterruptedException)
119: throw (InterruptedException) e;
120: if (e instanceof ParserException)
121: throw (ParserException) e;
122:
123: throw new ParserException(
124: "Unexpected error while parsing gzip file. "
125: + e.getMessage(), location);
126: } finally {
127: if (tempFile != null)
128: tempFile.delete();
129: }
130: }
131:
132: public void reset() {
133: // Nothing todo here at the moment
134: super.reset();
135: }
136: }
|