001/* 002 * (C) Copyright 2006-2016 Nuxeo SA (http://nuxeo.com/) and others. 003 * 004 * Licensed under the Apache License, Version 2.0 (the "License"); 005 * you may not use this file except in compliance with the License. 006 * You may obtain a copy of the License at 007 * 008 * http://www.apache.org/licenses/LICENSE-2.0 009 * 010 * Unless required by applicable law or agreed to in writing, software 011 * distributed under the License is distributed on an "AS IS" BASIS, 012 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 013 * See the License for the specific language governing permissions and 014 * limitations under the License. 015 * 016 * Contributors: 017 * Nuxeo - initial API and implementation 018 * 019 */ 020 021package org.nuxeo.ecm.platform.mimetype.detectors; 022 023import static java.nio.charset.StandardCharsets.UTF_8; 024 025import java.io.File; 026import java.io.IOException; 027import java.util.Map; 028 029import org.apache.commons.io.FileUtils; 030import org.apache.commons.logging.Log; 031import org.apache.commons.logging.LogFactory; 032import org.nuxeo.runtime.api.Framework; 033 034import net.sf.jmimemagic.MagicDetector; 035 036public class MsoXmlMimetypeSniffer implements MagicDetector { 037 038 private static final Log log = LogFactory.getLog(MsoXmlMimetypeSniffer.class); 039 040 @Override 041 public String getDisplayName() { 042 return "XML Microsoft 2003 MimeType Detector"; 043 } 044 045 @Override 046 public String[] getHandledExtensions() { 047 return new String[] { "xml" }; 048 } 049 050 @Override 051 public String[] getHandledTypes() { 052 return new String[] { "application/vnd.ms-excel", "application/msword", }; 053 } 054 055 @Override 056 public String getName() { 057 return "msoxml2003detector"; 058 } 059 060 @Override 061 public String getVersion() { 062 return "0.1"; 063 } 064 065 @Override 066 public String[] process(byte[] data, int offset, int length, long bitmask, char comparator, String mimeType, 067 @SuppressWarnings("rawtypes") Map params) { 068 String[] mimetypes = {}; 069 File file = null; 070 try { 071 file = Framework.createTempFile("magicdetector", ".xml"); 072 FileUtils.writeByteArrayToFile(file, data); 073 mimetypes = guessMsoXml(file); 074 } catch (IOException e) { 075 log.error(e); 076 } finally { 077 if (file != null) { 078 file.delete(); 079 } 080 } 081 return mimetypes; 082 } 083 084 @Override 085 public String[] process(File file, int offset, int length, long bitmask, char comparator, String mimeType, 086 @SuppressWarnings("rawtypes") Map params) { 087 return guessMsoXml(file); 088 } 089 090 public String[] guessMsoXml(File file) { 091 String[] mimetype = {}; 092 try { 093 String content = FileUtils.readFileToString(file, UTF_8); 094 if (content.contains("<?mso-application progid=\"Word.Document\"?>")) { 095 String[] type = { getHandledTypes()[1] }; 096 mimetype = type; 097 } else { 098 if (content.contains("<?mso-application progid=\"Excel.Sheet\"?>")) { 099 String[] type = { getHandledTypes()[0] }; 100 mimetype = type; 101 } 102 } 103 } catch (IOException e) { 104 log.error(e); 105 } 106 return mimetype; 107 } 108 109}