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.Serializable; 021import java.util.Map; 022 023import org.apache.commons.logging.Log; 024import org.apache.commons.logging.LogFactory; 025import org.nuxeo.ecm.core.api.Blobs; 026import org.nuxeo.ecm.core.api.blobholder.BlobHolder; 027import org.nuxeo.ecm.core.api.blobholder.SimpleBlobHolder; 028import org.nuxeo.ecm.core.convert.api.ConversionException; 029import org.nuxeo.ecm.core.convert.api.ConversionService; 030import org.nuxeo.ecm.core.convert.extension.Converter; 031import org.nuxeo.ecm.core.convert.extension.ConverterDescriptor; 032import org.nuxeo.runtime.api.Framework; 033 034/** 035 * Converter that tries to find a way to extract full text content according to input mime-type. 036 * 037 * @author tiry 038 */ 039public class FullTextConverter implements Converter { 040 041 private static final String TEXT_PLAIN_MT = "text/plain"; 042 043 private static final Log log = LogFactory.getLog(FullTextConverter.class); 044 045 protected ConverterDescriptor descriptor; 046 047 @Override 048 public BlobHolder convert(BlobHolder blobHolder, Map<String, Serializable> parameters) throws ConversionException { 049 050 String srcMT = blobHolder.getBlob().getMimeType(); 051 052 if (TEXT_PLAIN_MT.equals(srcMT)) { 053 // no need to convert ! 054 return blobHolder; 055 } 056 057 ConversionService cs = Framework.getLocalService(ConversionService.class); 058 059 String converterName = cs.getConverterName(srcMT, TEXT_PLAIN_MT); 060 061 if (converterName != null) { 062 if (converterName.equals(descriptor.getConverterName())) { 063 // Should never happen ! 064 log.debug("Existing from converter to avoid a loop"); 065 return new SimpleBlobHolder(Blobs.createBlob("")); 066 } 067 return cs.convert(converterName, blobHolder, parameters); 068 } else { 069 log.debug("Unable to find full text extractor for source mime type" + srcMT); 070 return new SimpleBlobHolder(Blobs.createBlob("")); 071 } 072 } 073 074 @Override 075 public void init(ConverterDescriptor descriptor) { 076 this.descriptor = descriptor; 077 } 078 079}