1 | /*
|
---|
2 | * AbstractGS2AudioSearch.java
|
---|
3 | * Copyright (C) 2011 New Zealand Digital Library, http://www.nzdl.org
|
---|
4 | *
|
---|
5 | * This program is free software; you can redistribute it and/or modify
|
---|
6 | * the Free Software Foundation; either version 2 of the License, or
|
---|
7 | * (at your option) any later version.
|
---|
8 | *
|
---|
9 | * This program is distributed in the hope that it will be useful,
|
---|
10 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
11 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
12 | * GNU General Public License for more details.
|
---|
13 | *
|
---|
14 | * You should have received a copy of the GNU General Public License
|
---|
15 | * along with this program; if not, write to the Free Software
|
---|
16 | * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
17 | */
|
---|
18 | package org.greenstone.gsdl3.service;
|
---|
19 |
|
---|
20 |
|
---|
21 | // Greenstone classes
|
---|
22 | import org.greenstone.gsdl3.util.OID;
|
---|
23 | import org.greenstone.gsdl3.util.DBInfo;
|
---|
24 | import org.greenstone.gsdl3.util.GSXML;
|
---|
25 | import org.greenstone.gsdl3.util.BasicDocumentDatabase;
|
---|
26 | import org.greenstone.gsdl3.util.GSFile;
|
---|
27 |
|
---|
28 | // XML classes
|
---|
29 | import org.w3c.dom.Document;
|
---|
30 | import org.w3c.dom.Element;
|
---|
31 | import org.w3c.dom.NodeList;
|
---|
32 |
|
---|
33 | // java
|
---|
34 | import java.util.Vector;
|
---|
35 | import java.util.ArrayList;
|
---|
36 | import java.util.HashMap;
|
---|
37 | import java.util.Map;
|
---|
38 | import java.util.Set;
|
---|
39 | import java.util.Iterator;
|
---|
40 | import java.io.File;
|
---|
41 |
|
---|
42 | import org.apache.log4j.*;
|
---|
43 |
|
---|
44 | public abstract class AbstractGS2AudioSearch
|
---|
45 | extends AbstractAudioSearch
|
---|
46 | {
|
---|
47 |
|
---|
48 | protected static final String EQUIV_TERM_ELEM = "equivTerm";
|
---|
49 |
|
---|
50 | protected static final String STEM_ATT = "stem";
|
---|
51 | protected static final String NUM_DOCS_MATCH_ATT = "numDocsMatch";
|
---|
52 | protected static final String FREQ_ATT = "freq";
|
---|
53 |
|
---|
54 | // Elements used in the config file that are specific to this class
|
---|
55 | protected static final String DEFAULT_INDEX_ELEM = "defaultIndex";
|
---|
56 | protected static final String INDEX_STEM_ELEM = "indexStem";
|
---|
57 | protected static final String INDEX_ELEM = "index";
|
---|
58 | protected static final String DEFAULT_INDEX_SUBCOLLECTION_ELEM = "defaultIndexSubcollection";
|
---|
59 | protected static final String DEFAULT_INDEX_LANGUAGE_ELEM = "defaultIndexLanguage";
|
---|
60 | protected static final String INDEX_SUBCOLLECTION_ELEM = "indexSubcollection";
|
---|
61 | protected static final String INDEX_LANGUAGE_ELEM = "indexLanguage";
|
---|
62 |
|
---|
63 |
|
---|
64 | // Some indexing options
|
---|
65 | protected static final String STEMINDEX_OPTION = "stemIndexes";
|
---|
66 | protected static final String MAXNUMERIC_OPTION = "maxnumeric";
|
---|
67 |
|
---|
68 | /** the stem used for the index files */
|
---|
69 | protected String index_stem = null;
|
---|
70 |
|
---|
71 | // stem indexes available
|
---|
72 | protected boolean does_case=true;
|
---|
73 | protected boolean does_stem=true;
|
---|
74 | protected boolean does_accent=false;
|
---|
75 |
|
---|
76 | // maxnumeric -
|
---|
77 | protected int maxnumeric = 4;
|
---|
78 |
|
---|
79 | BasicDocumentDatabase gs_doc_db = null;
|
---|
80 |
|
---|
81 | static Logger logger = Logger.getLogger(org.greenstone.gsdl3.service.AbstractGS2AudioSearch.class.getName());
|
---|
82 |
|
---|
83 |
|
---|
84 | /** constructor */
|
---|
85 | public AbstractGS2AudioSearch()
|
---|
86 | {
|
---|
87 |
|
---|
88 | }
|
---|
89 | public void cleanUp() {
|
---|
90 | super.cleanUp();
|
---|
91 | this.gs_doc_db.cleanUp();
|
---|
92 | }
|
---|
93 |
|
---|
94 | /** configure this service */
|
---|
95 | public boolean configure(Element info, Element extra_info)
|
---|
96 | {
|
---|
97 | if (!super.configure(info,extra_info)) {
|
---|
98 | return false;
|
---|
99 | }
|
---|
100 |
|
---|
101 | // find out what kind of database we have
|
---|
102 | Element database_type_elem = (Element) GSXML.getChildByTagName(info, GSXML.DATABASE_TYPE_ELEM);
|
---|
103 | String database_type = null;
|
---|
104 | if (database_type_elem != null) {
|
---|
105 | database_type = database_type_elem.getAttribute(GSXML.NAME_ATT);
|
---|
106 | }
|
---|
107 | if (database_type == null || database_type.equals("")) {
|
---|
108 | database_type = "gdbm"; // the default
|
---|
109 | }
|
---|
110 |
|
---|
111 | // the index stem is either the collection name or is specified in the config file
|
---|
112 | Element index_stem_elem = (Element) GSXML.getChildByTagName(info, INDEX_STEM_ELEM);
|
---|
113 | if (index_stem_elem != null) {
|
---|
114 | this.index_stem = index_stem_elem.getAttribute(GSXML.NAME_ATT);
|
---|
115 | }
|
---|
116 | if (this.index_stem == null || this.index_stem.equals("")) {
|
---|
117 | logger.warn("indexStem element not found, stem will default to collection name");
|
---|
118 | this.index_stem = this.cluster_name;
|
---|
119 | }
|
---|
120 |
|
---|
121 | // replaces default AbstractSearch version with one tied to database
|
---|
122 | gs_doc_db = new BasicDocumentDatabase(database_type,this.site_home,
|
---|
123 | this.cluster_name,
|
---|
124 | this.index_stem);
|
---|
125 | if (!gs_doc_db.isValid()) {
|
---|
126 | logger.error("Failed to open Document Database.");
|
---|
127 | return false;
|
---|
128 | }
|
---|
129 | this.gs_doc = gs_doc_db;
|
---|
130 |
|
---|
131 | // do we support any of the extended features?
|
---|
132 | does_chunking = true;
|
---|
133 |
|
---|
134 | // Get the default index out of <defaultIndex> (buildConfig.xml)
|
---|
135 | Element def = (Element) GSXML.getChildByTagName(info, DEFAULT_INDEX_ELEM);
|
---|
136 | if (def != null) {
|
---|
137 | this.default_index = def.getAttribute(GSXML.SHORTNAME_ATT);
|
---|
138 | } // otherwise will be "", and the first one will be the default
|
---|
139 |
|
---|
140 |
|
---|
141 | // Get index options ...
|
---|
142 | // ... but there are currently no index options supported for audio
|
---|
143 | // so nothing to do
|
---|
144 |
|
---|
145 | // Similarly the following will not currently do anything, but leave
|
---|
146 | // in for the time we do start supporting different indexes
|
---|
147 | //
|
---|
148 | // *** Also, it is identical to that in AbstractGS2TextSearch, so
|
---|
149 | // consider putting into supporting routine and sharing
|
---|
150 | // (realted in 'info' object
|
---|
151 |
|
---|
152 | // get display info from extra info
|
---|
153 | if (extra_info !=null) {
|
---|
154 | Document owner = info.getOwnerDocument();
|
---|
155 | // so far we have index specific display elements, and global format elements
|
---|
156 | NodeList indexes = info.getElementsByTagName(GSXML.INDEX_ELEM);
|
---|
157 | Element config_search = (Element)GSXML.getChildByTagName(extra_info, GSXML.SEARCH_ELEM);
|
---|
158 |
|
---|
159 | for (int i=0; i<indexes.getLength();i++) {
|
---|
160 | Element ind = (Element)indexes.item(i);
|
---|
161 | String name = ind.getAttribute(GSXML.NAME_ATT);
|
---|
162 | Element node_extra = GSXML.getNamedElement(config_search,
|
---|
163 | GSXML.INDEX_ELEM,
|
---|
164 | GSXML.NAME_ATT,
|
---|
165 | name);
|
---|
166 | if (node_extra == null) {
|
---|
167 | logger.error("haven't found extra info for index named "+name);
|
---|
168 | continue;
|
---|
169 | }
|
---|
170 |
|
---|
171 | // get the display elements if any - displayName
|
---|
172 | NodeList display_names = node_extra.getElementsByTagName(GSXML.DISPLAY_TEXT_ELEM);
|
---|
173 | if (display_names !=null) {
|
---|
174 | for (int j=0; j<display_names.getLength(); j++) {
|
---|
175 | Element e = (Element)display_names.item(j);
|
---|
176 | ind.appendChild(owner.importNode(e, true));
|
---|
177 | }
|
---|
178 | }
|
---|
179 | } // for each index
|
---|
180 | }
|
---|
181 |
|
---|
182 | return true;
|
---|
183 | }
|
---|
184 |
|
---|
185 | protected void getIndexData(ArrayList index_ids, ArrayList index_names, String lang)
|
---|
186 | {
|
---|
187 | logger.info("Trivial index support for audio for now. Adding hardwired 'audioDB' to index");
|
---|
188 | index_names.add("audioDB");
|
---|
189 | }
|
---|
190 | }
|
---|
191 |
|
---|
192 |
|
---|