1 | /*
|
---|
2 | * GS2MGPPSearch.java
|
---|
3 | * Copyright (C) 2002 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.mgpp.*;
|
---|
23 | import org.greenstone.gsdl3.util.*;
|
---|
24 |
|
---|
25 | // XML classes
|
---|
26 | import org.w3c.dom.Document;
|
---|
27 | import org.w3c.dom.Element;
|
---|
28 | import org.w3c.dom.NodeList;
|
---|
29 |
|
---|
30 | // java classes
|
---|
31 | import java.util.Iterator;
|
---|
32 | import java.util.Set;
|
---|
33 | import java.util.HashMap;
|
---|
34 | import java.util.Map;
|
---|
35 | import java.util.ArrayList;
|
---|
36 | import java.util.Vector;
|
---|
37 | import java.io.File;
|
---|
38 |
|
---|
39 | import org.apache.log4j.*;
|
---|
40 |
|
---|
41 | /**
|
---|
42 | *
|
---|
43 | * @author <a href="mailto:[email protected]">Katherine Don</a>
|
---|
44 | * @author <a href="mailto:[email protected]">Michael Dewsnip</a>
|
---|
45 | */
|
---|
46 |
|
---|
47 | public class GS2MGPPSearch
|
---|
48 | extends AbstractGS2FieldSearch {
|
---|
49 | private static MGPPSearchWrapper mgpp_src=null;
|
---|
50 |
|
---|
51 | private String physical_index_name = "idx";
|
---|
52 |
|
---|
53 | static Logger logger = Logger.getLogger(org.greenstone.gsdl3.service.GS2MGPPSearch.class.getName());
|
---|
54 |
|
---|
55 | /** constructor */
|
---|
56 | public GS2MGPPSearch() {
|
---|
57 | if(mgpp_src == null) {
|
---|
58 | mgpp_src = new MGPPSearchWrapper();
|
---|
59 | }
|
---|
60 | }
|
---|
61 |
|
---|
62 | public void cleanUp() {
|
---|
63 | super.cleanUp();
|
---|
64 | mgpp_src.unloadIndexData();
|
---|
65 | }
|
---|
66 | /** process a query */
|
---|
67 | protected Element processAnyQuery(Element request, int query_type) {
|
---|
68 | synchronized (mgpp_src) {
|
---|
69 | return super.processAnyQuery(request, query_type);
|
---|
70 | }
|
---|
71 | }
|
---|
72 | /** configure this service */
|
---|
73 | public boolean configure(Element info, Element extra_info) {
|
---|
74 | if (!super.configure(info, extra_info)){
|
---|
75 | return false;
|
---|
76 | }
|
---|
77 |
|
---|
78 | // the default level is also the level which gdbm is expecting
|
---|
79 | // this must not be overwritten
|
---|
80 | mgpp_src.setReturnLevel(this.default_gdbm_level);
|
---|
81 | // return term info
|
---|
82 | mgpp_src.setReturnTerms(true);
|
---|
83 | // set the default - this may be overwritten by query params
|
---|
84 | mgpp_src.setQueryLevel(this.default_level);
|
---|
85 | mgpp_src.setMaxNumeric(this.maxnumeric);
|
---|
86 |
|
---|
87 | return true;
|
---|
88 | }
|
---|
89 |
|
---|
90 | /** add in the mgpp specific params to TextQuery */
|
---|
91 | protected void addCustomQueryParams(Element param_list, String lang) {
|
---|
92 | super.addCustomQueryParams(param_list, lang);
|
---|
93 | createParameter(RANK_PARAM, param_list, lang);
|
---|
94 | }
|
---|
95 |
|
---|
96 | protected boolean setUpQueryer(HashMap params) {
|
---|
97 |
|
---|
98 | // set up the query params
|
---|
99 | Set entries = params.entrySet();
|
---|
100 | Iterator i = entries.iterator();
|
---|
101 | String physical_sub_index_name= this.default_index_subcollection;
|
---|
102 | String physical_index_language_name= this.default_index_language;
|
---|
103 | while (i.hasNext()) {
|
---|
104 | Map.Entry m = (Map.Entry)i.next();
|
---|
105 | String name = (String)m.getKey();
|
---|
106 | String value = (String)m.getValue();
|
---|
107 |
|
---|
108 | if (name.equals(CASE_PARAM)) {
|
---|
109 | boolean val = (value.equals(BOOLEAN_PARAM_ON)?true:false);
|
---|
110 | mgpp_src.setCase(val);
|
---|
111 | } else if (name.equals(STEM_PARAM)) {
|
---|
112 | boolean val = (value.equals(BOOLEAN_PARAM_ON)?true:false);
|
---|
113 | mgpp_src.setStem(val);
|
---|
114 | } else if (name.equals(ACCENT_PARAM)) {
|
---|
115 | boolean val = (value.equals(BOOLEAN_PARAM_ON)?true:false);
|
---|
116 | mgpp_src.setAccentFold(val);
|
---|
117 | } else if (name.equals(MAXDOCS_PARAM)&& !value.equals("")) {
|
---|
118 | int docs = Integer.parseInt(value);
|
---|
119 | mgpp_src.setMaxDocs(docs);
|
---|
120 | } else if (name.equals(LEVEL_PARAM)) {
|
---|
121 | mgpp_src.setQueryLevel(value);
|
---|
122 | } else if (name.equals(MATCH_PARAM)) {
|
---|
123 | int mode;
|
---|
124 | if (value.equals(MATCH_PARAM_ALL)) mode=1;
|
---|
125 | else mode=0;
|
---|
126 | mgpp_src.setMatchMode(mode);
|
---|
127 | } else if (name.equals(RANK_PARAM)) {
|
---|
128 | if (value.equals(RANK_PARAM_RANK)) {
|
---|
129 | mgpp_src.setSortByRank(true);
|
---|
130 | } else if (value.equals(RANK_PARAM_NONE)) {
|
---|
131 | mgpp_src.setSortByRank(false);
|
---|
132 | }
|
---|
133 | } else if (name.equals(INDEX_SUBCOLLECTION_PARAM)) {
|
---|
134 | physical_sub_index_name=value;
|
---|
135 | }else if (name.equals(INDEX_LANGUAGE_PARAM)){
|
---|
136 | physical_index_language_name=value;
|
---|
137 | } // ignore any others
|
---|
138 | }
|
---|
139 |
|
---|
140 | if (physical_index_name.equals("idx")){
|
---|
141 | if (physical_sub_index_name!=null) {
|
---|
142 | physical_index_name+=physical_sub_index_name;
|
---|
143 | }
|
---|
144 | if (physical_index_language_name!=null){
|
---|
145 | physical_index_name+=physical_index_language_name;
|
---|
146 | }
|
---|
147 | }
|
---|
148 |
|
---|
149 | // set up mgpp_src
|
---|
150 | String indexdir = GSFile.collectionBaseDir(this.site_home, this.cluster_name) + File.separatorChar + GSFile.collectionIndexPath(this.index_stem, physical_index_name);
|
---|
151 | mgpp_src.loadIndexData(indexdir);
|
---|
152 |
|
---|
153 | return true;
|
---|
154 | }
|
---|
155 |
|
---|
156 | protected Object runQuery(String query) {
|
---|
157 | mgpp_src.runQuery(query);
|
---|
158 | MGPPQueryResult mqr= mgpp_src.getQueryResult();
|
---|
159 | return mqr;
|
---|
160 |
|
---|
161 | }
|
---|
162 |
|
---|
163 | protected long numDocsMatched(Object query_result) {
|
---|
164 | return ((MGPPQueryResult)query_result).getTotalDocs();
|
---|
165 | }
|
---|
166 |
|
---|
167 | protected String [] getDocIDs(Object query_result) {
|
---|
168 |
|
---|
169 | Vector docs = ((MGPPQueryResult)query_result).getDocs();
|
---|
170 | String [] doc_nums = new String [docs.size()];
|
---|
171 | for (int d = 0; d < docs.size(); d++) {
|
---|
172 | doc_nums[d] = Long.toString((((MGPPDocInfo) docs.elementAt(d)).num_));
|
---|
173 | }
|
---|
174 | return doc_nums;
|
---|
175 | }
|
---|
176 |
|
---|
177 | protected String [] getDocRanks(Object query_result) {
|
---|
178 |
|
---|
179 | Vector docs = ((MGPPQueryResult)query_result).getDocs();
|
---|
180 | String [] doc_ranks = new String [docs.size()];
|
---|
181 | for (int d = 0; d < docs.size(); d++) {
|
---|
182 | doc_ranks[d] = Float.toString(((MGPPDocInfo) docs.elementAt(d)).rank_);
|
---|
183 | }
|
---|
184 | return doc_ranks;
|
---|
185 | }
|
---|
186 |
|
---|
187 | protected boolean addTermInfo(Element term_list, HashMap params,
|
---|
188 | Object query_result) {
|
---|
189 |
|
---|
190 | String query_level = (String)params.get(LEVEL_PARAM); // the current query level
|
---|
191 |
|
---|
192 | Vector terms = ((MGPPQueryResult)query_result).getTerms();
|
---|
193 | for (int t = 0; t < terms.size(); t++) {
|
---|
194 | MGPPTermInfo term_info = (MGPPTermInfo) terms.get(t);
|
---|
195 |
|
---|
196 | Element term_elem = this.doc.createElement(GSXML.TERM_ELEM);
|
---|
197 | term_elem.setAttribute(GSXML.NAME_ATT, term_info.term_);
|
---|
198 | term_elem.setAttribute(STEM_ATT, "" + term_info.stem_method_);
|
---|
199 | term_elem.setAttribute(FREQ_ATT, "" + term_info.term_freq_);
|
---|
200 | term_elem.setAttribute(NUM_DOCS_MATCH_ATT, "" + term_info.match_docs_);
|
---|
201 | String field = term_info.tag_;
|
---|
202 | if (field.equals(query_level)) {
|
---|
203 | // ignore
|
---|
204 | field = "";
|
---|
205 | }
|
---|
206 | term_elem.setAttribute(FIELD_ATT, field);
|
---|
207 |
|
---|
208 | Vector equiv_terms = term_info.equiv_terms_;
|
---|
209 | Element equiv_term_list = this.doc.createElement(EQUIV_TERM_ELEM+GSXML.LIST_MODIFIER);
|
---|
210 | term_elem.appendChild(equiv_term_list);
|
---|
211 |
|
---|
212 | for (int et = 0; et < equiv_terms.size(); et++) {
|
---|
213 | String equiv_term = (String) equiv_terms.get(et);
|
---|
214 |
|
---|
215 | Element equiv_term_elem = this.doc.createElement(GSXML.TERM_ELEM);
|
---|
216 | equiv_term_elem.setAttribute(GSXML.NAME_ATT, equiv_term);
|
---|
217 | equiv_term_elem.setAttribute(NUM_DOCS_MATCH_ATT, "");
|
---|
218 | equiv_term_elem.setAttribute(FREQ_ATT, "");
|
---|
219 | equiv_term_list.appendChild(equiv_term_elem);
|
---|
220 | }
|
---|
221 |
|
---|
222 | term_list.appendChild(term_elem);
|
---|
223 | }
|
---|
224 | return true;
|
---|
225 | }
|
---|
226 |
|
---|
227 |
|
---|
228 | protected String addFieldInfo(String query, String field) {
|
---|
229 | if (field.equals("") || field.equals("ZZ")) {
|
---|
230 | return query;
|
---|
231 | }
|
---|
232 | return "["+query+"]:"+field;
|
---|
233 | }
|
---|
234 | protected void addQueryElem(StringBuffer final_query, String query,
|
---|
235 | String field, String combine) {
|
---|
236 |
|
---|
237 | String comb="";
|
---|
238 | if (final_query.length()>0) {
|
---|
239 | comb = " "+combine+" ";
|
---|
240 | }
|
---|
241 | final_query.append(comb+addFieldInfo(query,field));
|
---|
242 | }
|
---|
243 |
|
---|
244 | protected String addStemOptions(String query, String stem,
|
---|
245 | String casef, String accent) {
|
---|
246 | String mods = "#";
|
---|
247 | if (casef != null) {
|
---|
248 | if (casef.equals("1")) {
|
---|
249 | mods += "i";
|
---|
250 | } else {
|
---|
251 | mods += "c";
|
---|
252 | }
|
---|
253 | }
|
---|
254 | if (stem != null) {
|
---|
255 | if (stem.equals("1")) {
|
---|
256 | mods += "s";
|
---|
257 | } else {
|
---|
258 | mods+= "u";
|
---|
259 | }
|
---|
260 | }
|
---|
261 | if (accent != null) {
|
---|
262 | if (accent.equals("1")) {
|
---|
263 | mods += "f";
|
---|
264 | } else {
|
---|
265 | mods += "a";
|
---|
266 | }
|
---|
267 | }
|
---|
268 |
|
---|
269 | StringBuffer temp = new StringBuffer();
|
---|
270 | String [] terms = query.split(" ");
|
---|
271 | for (int i=0; i<terms.length; i++) {
|
---|
272 | String t = terms[i].trim();
|
---|
273 | // what is the TX bit about???
|
---|
274 | if (!t.equals("") && !t.equals("TX")) {
|
---|
275 | temp.append(" "+t+mods);
|
---|
276 | }
|
---|
277 | }
|
---|
278 | return temp.toString();
|
---|
279 | }
|
---|
280 |
|
---|
281 | }
|
---|
282 |
|
---|
283 |
|
---|