001/*
002 * (C) Copyright 2002-2007 Nuxeo SAS (http://nuxeo.com/) and contributors.
003 *
004 * All rights reserved. This program and the accompanying materials
005 * are made available under the terms of the GNU Lesser General Public License
006 * (LGPL) version 2.1 which accompanies this distribution, and is available at
007 * http://www.gnu.org/licenses/lgpl.html
008 *
009 * This library is distributed in the hope that it will be useful,
010 * but WITHOUT ANY WARRANTY; without even the implied warranty of
011 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
012 * Lesser General Public License for more details.
013 *
014 * Contributors:
015 *     Nuxeo - initial API and implementation
016 *
017 */
018package org.nuxeo.ecm.core.convert.plugins.text.extractors;
019
020import java.io.File;
021import java.io.FileInputStream;
022import java.io.FileOutputStream;
023import java.io.IOException;
024import java.io.InputStream;
025import java.io.OutputStream;
026import java.io.Serializable;
027import java.util.Map;
028
029import org.apache.commons.logging.Log;
030import org.apache.commons.logging.LogFactory;
031import org.apache.poi.hslf.extractor.PowerPointExtractor;
032import org.nuxeo.ecm.core.api.Blob;
033import org.nuxeo.ecm.core.api.Blobs;
034import org.nuxeo.ecm.core.api.blobholder.BlobHolder;
035import org.nuxeo.ecm.core.convert.api.ConversionException;
036import org.nuxeo.ecm.core.convert.cache.SimpleCachableBlobHolder;
037import org.nuxeo.ecm.core.convert.extension.Converter;
038import org.nuxeo.ecm.core.convert.extension.ConverterDescriptor;
039
040/**
041 * @deprecated subsumed by MSOffice2TextConverter
042 */
043@Deprecated
044public class PPT2TextConverter implements Converter {
045
046    private static final Log log = LogFactory.getLog(PPT2TextConverter.class);
047
048    @Override
049    public BlobHolder convert(BlobHolder blobHolder, Map<String, Serializable> parameters) throws ConversionException {
050        File f = null;
051        OutputStream fas = null;
052
053        PowerPointExtractor extractor = null;
054        try {
055            extractor = new PowerPointExtractor(blobHolder.getBlob().getStream());
056
057            byte[] bytes = extractor.getText().getBytes();
058            f = File.createTempFile("po-ppt2text", ".txt");
059            fas = new FileOutputStream(f);
060            fas.write(bytes);
061
062            Blob blob;
063            try (InputStream in = new FileInputStream(f)) {
064                blob = Blobs.createBlob(in);
065            }
066            blob.setMimeType("text/plain");
067            return new SimpleCachableBlobHolder(blob);
068        } catch (IOException e) {
069            throw new ConversionException("Error during PPT2Text conversion", e);
070        } finally {
071            if (extractor != null) {
072                try {
073                    extractor.close();
074                } catch (IOException e) {
075                    log.error(e, e);
076                }
077            }
078            if (fas != null) {
079                try {
080                    fas.close();
081                } catch (IOException e) {
082                    log.error(e, e);
083                }
084            }
085            if (f != null) {
086                f.delete();
087            }
088        }
089    }
090
091    @Override
092    public void init(ConverterDescriptor descriptor) {
093    }
094
095}