KickJava   Java API By Example, From Geeks To Geeks.

Java > Open Source Codes > org > opencms > search > documents > CmsDocumentMsExcel


1 /*
2  * File : $Source: /usr/local/cvs/opencms/src/org/opencms/search/documents/CmsDocumentMsExcel.java,v $
3  * Date : $Date: 2005/06/23 11:11:29 $
4  * Version: $Revision: 1.9 $
5  *
6  * This library is part of OpenCms -
7  * the Open Source Content Mananagement System
8  *
9  * Copyright (c) 2005 Alkacon Software GmbH (http://www.alkacon.com)
10  *
11  * This library is free software; you can redistribute it and/or
12  * modify it under the terms of the GNU Lesser General Public
13  * License as published by the Free Software Foundation; either
14  * version 2.1 of the License, or (at your option) any later version.
15  *
16  * This library is distributed in the hope that it will be useful,
17  * but WITHOUT ANY WARRANTY; without even the implied warranty of
18  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
19  * Lesser General Public License for more details.
20  *
21  * For further information about Alkacon Software GmbH, please see the
22  * company website: http://www.alkacon.com
23  *
24  * For further information about OpenCms, please see the
25  * project website: http://www.opencms.org
26  *
27  * You should have received a copy of the GNU Lesser General Public
28  * License along with this library; if not, write to the Free Software
29  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
30  */
31
32 package org.opencms.search.documents;
33
34 import org.opencms.file.CmsFile;
35 import org.opencms.file.CmsObject;
36 import org.opencms.file.CmsResource;
37 import org.opencms.main.CmsException;
38 import org.opencms.search.A_CmsIndexResource;
39 import org.opencms.search.CmsIndexException;
40 import org.opencms.search.extractors.CmsExtractorMsExcel;
41 import org.opencms.search.extractors.I_CmsExtractionResult;
42
43 import java.io.FileNotFoundException JavaDoc;
44
45 /**
46  * Lucene document factory class to extract index data from a cms resource
47  * containing MS Excel data.<p>
48  *
49  * @author Carsten Weinholz
50  *
51  * @version $Revision: 1.9 $
52  *
53  * @since 6.0.0
54  */
55 public class CmsDocumentMsExcel extends A_CmsVfsDocument {
56
57     /**
58      * Creates a new instance of this lucene document factory.<p>
59      *
60      * @param name name of the documenttype
61      */
62     public CmsDocumentMsExcel(String JavaDoc name) {
63
64         super(name);
65     }
66
67     /**
68      * Returns the raw text content of a given vfs resource containing MS Excel data.<p>
69      *
70      * @see org.opencms.search.documents.A_CmsVfsDocument#extractContent(org.opencms.file.CmsObject, org.opencms.search.A_CmsIndexResource, java.lang.String)
71      */
72     public I_CmsExtractionResult extractContent(CmsObject cms, A_CmsIndexResource indexResource, String JavaDoc language)
73     throws CmsIndexException, CmsException {
74
75         CmsResource resource = (CmsResource)indexResource.getData();
76         CmsFile file = readFile(cms, resource);
77
78         try {
79             return CmsExtractorMsExcel.getExtractor().extractText(file.getContents());
80         } catch (Exception JavaDoc e) {
81             if (e instanceof FileNotFoundException JavaDoc) {
82                 if (e.getMessage() != null && e.getMessage().indexOf("Workbook") > 0) {
83                     // special case: catch Excel95 format error
84 throw new CmsIndexException(Messages.get().container(
85                         Messages.ERR_NO_EXCEL_FORMAT_1,
86                         resource.getRootPath()), e);
87                 }
88             }
89             throw new CmsIndexException(
90                 Messages.get().container(Messages.ERR_TEXT_EXTRACTION_1, resource.getRootPath()),
91                 e);
92         }
93     }
94 }
Popular Tags