KickJava   Java API By Example, From Geeks To Geeks.

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


1 /*
2  * File : $Source: /usr/local/cvs/opencms/src/org/opencms/search/documents/CmsDocumentHtml.java,v $
3  * Date : $Date: 2005/06/23 11:11:29 $
4  * Version: $Revision: 1.8 $
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.CmsProperty;
37 import org.opencms.file.CmsPropertyDefinition;
38 import org.opencms.file.CmsResource;
39 import org.opencms.main.CmsException;
40 import org.opencms.main.OpenCms;
41 import org.opencms.search.A_CmsIndexResource;
42 import org.opencms.search.CmsIndexException;
43 import org.opencms.search.extractors.CmsExtractorHtml;
44 import org.opencms.search.extractors.I_CmsExtractionResult;
45
46 /**
47  * Lucene document factory class to extract index data from a cms resource
48  * containing plain html data.<p>
49  *
50  * @author Alexander Kandzior
51  *
52  * @version $Revision: 1.8 $
53  *
54  * @since 6.0.0
55  */
56 public class CmsDocumentHtml extends A_CmsVfsDocument {
57
58     /**
59      * Creates a new instance of this lucene document factory.<p>
60      *
61      * @param name name of the documenttype
62      */
63     public CmsDocumentHtml(String JavaDoc name) {
64
65         super(name);
66     }
67
68     /**
69      * Returns the raw text content of a given vfs resource containing HTML data.<p>
70      *
71      * @see org.opencms.search.documents.A_CmsVfsDocument#extractContent(org.opencms.file.CmsObject, org.opencms.search.A_CmsIndexResource, java.lang.String)
72      */
73     public I_CmsExtractionResult extractContent(CmsObject cms, A_CmsIndexResource indexResource, String JavaDoc language)
74     throws CmsIndexException, CmsException {
75
76         CmsResource resource = (CmsResource)indexResource.getData();
77         CmsFile file = readFile(cms, resource);
78
79         try {
80             String JavaDoc path = cms.getRequestContext().removeSiteRoot(resource.getRootPath());
81             CmsProperty encProp = cms.readPropertyObject(path, CmsPropertyDefinition.PROPERTY_CONTENT_ENCODING, true);
82             String JavaDoc encoding = encProp.getValue(OpenCms.getSystemInfo().getDefaultEncoding());
83
84             return CmsExtractorHtml.getExtractor().extractText(file.getContents(), encoding);
85         } catch (Exception JavaDoc e) {
86             throw new CmsIndexException(
87                 Messages.get().container(Messages.ERR_TEXT_EXTRACTION_1, resource.getRootPath()),
88                 e);
89         }
90     }
91 }
Popular Tags