source: trunk/gsdl3/src/java/org/greenstone/gsdl3/service/PhindService.java@ 3471

Last change on this file since 3471 was 3471, checked in by kjdon, 22 years ago

service modules now belong to a serviceCluster or colleciton - collection_name has been changed to the more general cluster_name. service module cant configure itself from a file - we no longer know where the appropriate file is. so must be configured by passing the xml node to the configure method
  • Property svn:keywords set to Author Date Id Revision
File size: 15.9 KB
Line 
1/*
2 * PhindService.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 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 2 of the License, or
8 * (at your option) any later version.
9 *
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
14 *
15 * You should have received a copy of the GNU General Public License
16 * along with this program; if not, write to the Free Software
17 * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
18 */
19package org.greenstone.gsdl3.service;
20
21import org.greenstone.gsdl3.util.*;
22
23import org.greenstone.mgpp.*;
24import org.w3c.dom.Document;
25import org.w3c.dom.Node;
26import org.w3c.dom.Element;
27import org.w3c.dom.Text;
28
29import java.util.Vector;
30import java.util.HashMap;
31/**
32 * PhindService - the phind phrase browsing service
33 *
34 * @author <a href="mailto:[email protected]">Katherine Don</a>
35 * @version $Revision: 3471 $
36 */
37public class PhindService
38 extends ServiceModule {
39
40 private MGPPWrapper mgpp_src_=null;
41 private String basepath_ = null;
42 public PhindService() {
43 mgpp_src_ = new MGPPWrapper();
44 // set up the default params
45 mgpp_src_.setQueryLevel("Document");
46 mgpp_src_.setReturnLevel("Document");
47 mgpp_src_.setMaxDocs(5);
48 mgpp_src_.setStem(false);
49 mgpp_src_.setCase(true);
50 }
51 /** configure the service module
52 *
53 * @param info a DOM Element containing any config info for the service
54 * @return true if configured
55 */
56 public boolean configure(Element info) {
57
58 System.out.println("configuring PhindService");
59
60 // set up short_service_info_ - for now just has name and type
61 Element e = doc_.createElement("service");
62 e.setAttribute("type", "query");
63 e.setAttribute("name", "PhindApplet");
64 short_service_info_.appendChild(e);
65
66 // set up service_info_map_ - we only have one element, and it has
67 // no extra info yet - we are not processing the config info
68 Element f = doc_.createElement("service");
69 f.setAttribute("type", "query");
70 f.setAttribute("name", "PhindApplet");
71
72 // add in the applet info for the phind applet
73 // need to make this dynamic - library names etc
74 // change the applet params - have a single param with the library name
75 // this is left blank at this end, and must be filled in by applet action - if the library name is not needed, this param is left out
76 // phindcgi param now is not complete - library must be prepended to it.
77 String app_info = "<applet CODEBASE='lib/java' CODE='org.greenstone.applet.phind.Phind.class' ARCHIVE='phind.jar, xercesImpl.jar, gsdl3.jar, jaxp.jar, xml-apis.jar' WIDTH='500' HEIGHT='400'><PARAM NAME='library' VALUE=''/> <PARAM NAME='phindcgi' VALUE='?a=a&amp;sa=r&amp;sn=Phind'/>";
78 app_info +="<PARAM NAME='collection' VALUE='";
79 app_info += cluster_name_;
80 app_info += "'/> <PARAM NAME='classifier' VALUE='1'/> <PARAM NAME='orientation' VALUE='vertical'/> <PARAM NAME='depth' VALUE='2'/> <PARAM NAME='resultorder' VALUE='L,l,E,e,D,d'/> <PARAM NAME='backdrop' VALUE='interfaces/default/images/phindbg1.jpg'/><PARAM NAME='fontsize' VALUE='10'/> <PARAM NAME='blocksize' VALUE='10'/>The Phind java applet.</applet>";
81
82 Document dom = converter_.getDOM(app_info);
83 Element app_elem = dom.getDocumentElement();
84 f.appendChild(doc_.importNode(app_elem, true));
85
86 service_info_map_.put("PhindApplet", f);
87
88 return true;
89 }
90
91 protected Element processService(String name, Element request) {
92
93 if (!name.equals("PhindApplet")) {
94 System.err.println("PhindService:you have asked for a non-existant service - "+name+"!");
95 return null;
96 }
97 Element param_elem = (Element)GSXML.getChildByTagName(request, "paramList");
98 HashMap params = GSXML.extractParams(param_elem);
99
100 long first_e = Long.parseLong((String)params.get("pfe"));
101 long last_e = Long.parseLong((String)params.get("ple"));
102 long first_l = Long.parseLong((String)params.get("pfl"));
103 long last_l = Long.parseLong((String)params.get("pll"));
104 long first_d = Long.parseLong((String)params.get("pfd"));
105 long last_d = Long.parseLong((String)params.get("pld"));
106
107 long phrase;
108 String phrase_str = (String)params.get("ppnum");
109 if (phrase_str == null || phrase_str.equals("")) {
110 phrase=0;
111 } else {
112 phrase = Long.parseLong(phrase_str);
113 }
114 String word = (String)params.get("pptext");
115 String phind_index = (String)params.get("pc");
116 // the location of the mgpp database files
117 basepath_ = GSFile.phindBaseDir(site_home_, cluster_name_, phind_index);
118
119 // the result element
120 Element result = doc_.createElement("response");
121 String from = GSPath.appendLink(cluster_name_, "PhindApplet");
122 result.setAttribute("from", from);
123 result.setAttribute("type", "query");
124
125 // applet result info must be in appletInfo element
126 Element applet_data = doc_.createElement("appletData");
127 result.appendChild(applet_data);
128 Element phind_data = doc_.createElement("phindData");
129 applet_data.appendChild(phind_data);
130
131
132 // if we dont know the phrase number, look it up
133 if (phrase == 0) {
134 if (word==null || word.equals("")) {
135 Element error = phindError("no word or phrase");
136 phind_data.appendChild(error);
137 return result;
138 }
139 phrase = findPhraseNumberFromWord( word);
140 System.out.println("phind, term number for "+word+" is "+phrase);
141 }
142 if (phrase==0) {
143 // the word is not in the collection
144 // return a phind error string
145 Element error = phindError("the term "+word+" is not in the collection");
146 phind_data.appendChild(error);
147 return result;
148 }
149
150 // get the phrase data into the phind_data node
151 getPhraseData(phind_data, phrase, first_l, last_l,
152 first_e, last_e, first_d, last_d);
153 return result;
154
155
156 }// processService
157
158 protected long findPhraseNumberFromWord(String word) {
159
160 // set the mgpp index data - we are looking up pword
161 mgpp_src_.loadIndexData(basepath_, "pword");
162
163 mgpp_src_.runQuery(word);
164
165 MGPPQueryResult res = mgpp_src_.getQueryResult();
166 Vector docs = res.getDocs();
167 if (docs.size()==0) {
168 // phrase not found
169 return 0;
170 }
171 MGPPDocInfo doc = (MGPPDocInfo)docs.firstElement();
172 return doc.num_;
173 }
174
175 protected boolean getPhraseData(Element phind_data,
176 long phrase, long first_l, long last_l,
177 long first_e, long last_e, long first_d,
178 long last_d) {
179
180 String record = mgpp_src_.getDocument(basepath_, "pdata", "Document",
181 phrase);
182 if (record.equals("")) {
183 Element error = phindError("somethings gone wrong - we haven't got a record for phrase number "+phrase);
184 phind_data.appendChild(error);
185 return false;
186 }
187
188 System.out.println("record="+record);
189 // parse the record - its in gordons cryptic form
190 // ":word:tf:ef:df:el:dl:lf:ll"
191 // el: e,e,e
192 // dl: d;f,d;f,
193 // lf and ll may be null
194 // l: type,dest, dest; type,dest,dest
195
196 // ignore everything up to and including first colon (has
197 // <Document>3505: at the start)
198 record = record.substring(record.indexOf(':')+1);
199
200 // split on ':'
201 String [] fields = record.split(":");
202 String word = fields[0];
203 String tf = fields[1];
204 String ef = fields[2];
205 String df = fields[3];
206
207
208 String expansions = fields[4];
209 String documents = fields[5];
210 String lf = "0";
211 String linklist = "";
212 if (fields.length > 7) {// have thesaurus stuff
213 lf =fields[6];
214 linklist = fields[7];
215 }
216
217 // the phindData attributes and phrase
218 phind_data.setAttribute("id", Long.toString(phrase));
219 phind_data.setAttribute("df", df);
220 phind_data.setAttribute("ef", ef);
221 phind_data.setAttribute("lf", lf);
222 phind_data.setAttribute("tf", tf);
223 GSXML.createTextElement(doc_, "phrase", word);
224
225 addExpansionList(phind_data, expansions, word, ef, first_e, last_e);
226 addDocumentList(phind_data, documents, word, df, first_d, last_d);
227 if (!lf.equals("0")) {
228 System.out.println("adding thesaurus stuff");
229 addThesaurusList(phind_data, linklist, word, lf, first_l, last_l);
230 }
231 return true;
232 }
233
234 protected boolean addExpansionList( Element phind_data, String record,
235 String word,
236 String freq,
237 long first, long last) {
238
239 Element expansion_list = doc_.createElement("expansionList");
240 phind_data.appendChild(expansion_list);
241 expansion_list.setAttribute("length", freq);
242 expansion_list.setAttribute("start", Long.toString(first));
243 expansion_list.setAttribute("end", Long.toString(last));
244
245 // get the list of strings
246 String [] expansions = record.split(",");
247 int length = expansions.length;
248 if (length < last) last = length;
249 for (long i = first; i < last; i++) {
250 long num = Long.parseLong(expansions[(int)i]);
251 Element expansion = getExpansion( num, word);
252 expansion.setAttribute("num", Long.toString(i));
253 expansion_list.appendChild(expansion);
254 }
255 return true;
256 }
257
258 protected Element getExpansion(long phrase_num,
259 String orig_phrase) {
260
261 // look up the phrase in the pdata thingy
262 String record = mgpp_src_.getDocument(basepath_, "pdata", "Document",
263 phrase_num);
264
265 if (record ==null || record.equals("")) return null;
266
267 // ignore everything up to and including first colon
268 record = record.substring(record.indexOf(':')+1);
269
270 String [] fields = record.split(":");
271 String phrase = fields[0];
272 String tf = fields[1];
273 //String ef = fields[2]; dont use this
274 String df = fields[3];
275
276 Element expansion = doc_.createElement("expansion");
277 expansion.setAttribute("tf", tf);
278 expansion.setAttribute("df", df);
279 expansion.setAttribute("id", Long.toString(phrase_num));
280
281 // get teh suffix and prefix
282 String [] ends = splitPhraseOnWord(phrase, orig_phrase);
283 if (!ends[0].equals("")) {
284 expansion.appendChild(GSXML.createTextElement(doc_, "prefix", ends[0]));
285 }
286 if (!ends[1].equals("")) {
287 expansion.appendChild(GSXML.createTextElement(doc_, "suffix", ends[1]));
288 }
289
290 return expansion;
291
292 }
293
294 protected boolean addDocumentList(Element phind_data, String record,
295 String word,
296 String freq,
297 long first, long last) {
298
299 Element document_list = doc_.createElement("documentList");
300 phind_data.appendChild(document_list);
301 document_list.setAttribute("length", freq);
302 document_list.setAttribute("start", Long.toString(first));
303 document_list.setAttribute("end", Long.toString(last));
304
305 // get the list of doc,freq
306 String [] doc_freqs = record.split(";");
307 int length = doc_freqs.length;
308 if (length<last) last=length;
309
310 for (long i = first; i < last; i++) {
311 String doc_elem = doc_freqs[(int)i];
312 int p = doc_elem.indexOf(',');
313 long doc_num;
314 String doc_freq;
315 if (p == -1) { // there is no freq in the record
316 doc_num =Long.parseLong(doc_elem);
317 doc_freq = "1";
318 } else {
319 doc_num = Long.parseLong(doc_elem.substring(0,p));
320 doc_freq = doc_elem.substring(p+1);
321 }
322 Element document = getDocument( doc_num);
323 document.setAttribute("freq", doc_freq);
324 document.setAttribute("num", Long.toString(i));
325 document_list.appendChild(document);
326 }
327
328
329 return true;
330 }
331
332
333 protected Element getDocument(long doc_num) {
334
335 // look up the phrase in the docs thingy
336 String record = mgpp_src_.getDocument(basepath_, "docs", "Document",
337 doc_num);
338
339 if (record ==null || record.equals("")) return null;
340 System.out.println("doc record:"+record);
341
342 // ignore everything up to and including first \t
343 record = record.substring(record.indexOf('\t')+1);
344
345 String [] fields = record.split("\t");
346 String hash = fields[0];
347 String title = fields[1];
348
349 Element d = doc_.createElement("document");
350 d.setAttribute("hash", hash);
351 d.appendChild(GSXML.createTextElement(doc_, "title", title));
352
353 return d;
354
355 }
356 protected boolean addThesaurusList(Element phind_data, String record,
357 String word,
358 String freq,
359 long first, long last) {
360
361
362 Element thesaurus_list = doc_.createElement("thesaurusList");
363 phind_data.appendChild(thesaurus_list);
364 thesaurus_list.setAttribute("length", freq);
365 thesaurus_list.setAttribute("start", Long.toString(first));
366 thesaurus_list.setAttribute("end", Long.toString(last));
367
368 System.out.println("record for thesaurus="+record);
369
370 // get the list of type,dest,dest
371 String [] links = record.split(";");
372 int length = links.length;
373 long index = 0;
374 for (int i = 0; i < length; i++) { // go through the entries
375 String link_info = links[(int)i];
376 String [] items = link_info.split(",");
377 // the first entry is teh type
378 String type = items[0];
379 for (int j = 1; j<items.length; j++, index++) {
380 if (index >= first && index < last) { // only output the ones we want
381 long phrase = Long.parseLong(items[j]);
382 Element t = getThesaurus(phrase);
383 t.setAttribute("type", type);
384 thesaurus_list.appendChild(t);
385 }
386 }
387 }
388
389 return true;
390 }
391
392 protected Element getThesaurus(long phrase_num) {
393
394 // look up the phrase in the pdata thingy
395 String record = mgpp_src_.getDocument(basepath_, "pdata", "Document",
396 phrase_num);
397
398 if (record ==null || record.equals("")) return null;
399
400 // ignore everything up to and including first colon
401 record = record.substring(record.indexOf(':')+1);
402
403 String [] fields = record.split(":");
404 String phrase = fields[0];
405 String tf = fields[1];
406 //String ef = fields[2]; dont use this
407 String df = fields[3];
408
409 Element thesaurus = doc_.createElement("thesaurus");
410 thesaurus.setAttribute("tf", tf);
411 thesaurus.setAttribute("df", df);
412 thesaurus.setAttribute("id", Long.toString(phrase_num));
413 thesaurus.appendChild(GSXML.createTextElement(doc_, "phrase", phrase));
414 return thesaurus;
415
416 }
417
418 /** returns an array of two elements - the prefix and the suffix*/
419 protected String [] splitPhraseOnWord(String phrase, String word) {
420
421 if (word.equals("")) {
422
423 String [] res = {phrase, ""};
424 return res;
425 }
426 // use 2 so that we only split on the first occurrance. trailing empty strings should be included
427 String [] result = phrase.split(word, 2);
428 if (result.length !=2) {
429 System.out.println("didn't get two substrings!!");
430 }
431 return result;
432
433 }
434
435 protected Element phindError(String message) {
436 Element e = doc_.createElement("phindError");
437 Text t = doc_.createTextNode(message);
438 e.appendChild(t);
439 return e;
440 }
441
442}
443
444
445 /*
446 // CREATE dummy response
447 Element res = doc_.createElement("response");
448 res.setAttribute("from", "PhindApplet");
449 Element data = doc_.createElement("service");
450 Element app_data = doc_.createElement("appletData");
451 data.appendChild(app_data);
452 String phind_info ="<phindData id='2507' tf='19424' ef='1632' df='1843' lf='0'><phrase>FOREST</phrase><expansionList length='1632' start='0' end='10'><expansion num='0' id='177648' tf='2162' df='519'><suffix>MANAGEMENT</suffix></expansion> <expansion num='1' id='177531' tf='1958' df='566'><suffix>PRODUCTS</suffix></expansion> <expansion num='2' id='177469' tf='1328' df='532'><suffix>RESOURCES</suffix></expansion> <expansion num='3' id='177773' tf='943' df='177'><suffix>GENETIC</suffix></expansion> <expansion num='4' id='177335' tf='736' df='258'><prefix>SUSTAINABLE</prefix></expansion> </expansionList><documentList length='1843' start='0' end='10'><document num='0' hash='HASH011fb8a7d8bf781ab3cbb087' freq='363'><title>FO-edu List of Countries 0</title></document><document num='1' hash='HASH27ae41229eb0636849a5be' freq='344' ><title>FO-edu List of Countries 1</title></document><document num='2' hash='HASH0187ef85c9dbf5bf132ea1d1' freq='263'><title>FO-edu List of Countries 2</title></document><document num='3' hash='HASH0125ec9ef67960446f471280' freq='238'><title>FO-edu List of Countries 3</title></document><document num='4' hash='HASH67087f7717eb35050ce1ac' freq='213'><title>FO-edu List of Countries 4</title></document></documentList><thesaurusList><thesaurus num='3' id='36506' tf='0' df='0' type='RT'><phrase>FRANCOPHONE</phrase></thesaurus></thesaurusList></phindData>";
453
454 Node t = converter_.getDOM(phind_info).getDocumentElement();
455 app_data.appendChild(doc_.importNode(t, true));
456
457 res.appendChild(data);
458
459 return res;
460 */
Note: See TracBrowser for help on using the repository browser.