1 | /*
|
---|
2 | * AbstractSearch.java
|
---|
3 | * Copyright (C) 2005 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 | */
|
---|
19 | package org.greenstone.gsdl3.service;
|
---|
20 |
|
---|
21 | // Greenstone classes
|
---|
22 | import java.util.ArrayList;
|
---|
23 | import java.util.HashMap;
|
---|
24 |
|
---|
25 | import org.apache.log4j.Logger;
|
---|
26 | import org.greenstone.gsdl3.util.AbstractBasicDocument;
|
---|
27 | import org.greenstone.gsdl3.util.BasicDocument;
|
---|
28 | import org.greenstone.gsdl3.util.GSPath;
|
---|
29 | import org.greenstone.gsdl3.util.GSXML;
|
---|
30 | import org.greenstone.gsdl3.util.XMLConverter;
|
---|
31 | import org.w3c.dom.Document;
|
---|
32 | import org.w3c.dom.Element;
|
---|
33 | import org.w3c.dom.NodeList;
|
---|
34 |
|
---|
35 | /**
|
---|
36 | * Partially implements a generic search service
|
---|
37 | *
|
---|
38 | */
|
---|
39 |
|
---|
40 | public abstract class AbstractSearch extends ServiceRack
|
---|
41 | {
|
---|
42 |
|
---|
43 | static Logger logger = Logger.getLogger(org.greenstone.gsdl3.service.AbstractSearch.class.getName());
|
---|
44 |
|
---|
45 | // the search service
|
---|
46 | protected String QUERY_SERVICE = null; // set by subclass
|
---|
47 |
|
---|
48 | // compulsory params
|
---|
49 | protected static final String INDEX_PARAM = "index";
|
---|
50 | protected static final String QUERY_PARAM = "query";
|
---|
51 | protected static final String RAW_PARAM = "rawquery";
|
---|
52 |
|
---|
53 | // optional standard params - some of these have to be implemented
|
---|
54 | protected static final String MAXDOCS_PARAM = "maxDocs";
|
---|
55 | protected static final String HITS_PER_PAGE_PARAM = "hitsPerPage";
|
---|
56 | protected static final String START_PAGE_PARAM = "startPage";
|
---|
57 |
|
---|
58 | protected AbstractBasicDocument gs_doc = null;
|
---|
59 |
|
---|
60 | /** can more than one index be searched at the same time? */
|
---|
61 | protected boolean does_multi_index_search = false;
|
---|
62 | /** does this service support paging of results? */
|
---|
63 | protected boolean does_paging = false;
|
---|
64 | /** does this service support asking for a subset of results? */
|
---|
65 | protected boolean does_chunking = false;
|
---|
66 | /** does this service support faceting search results */
|
---|
67 | protected boolean does_faceting = false;
|
---|
68 | /**
|
---|
69 | * the default document type - use if all documents are the same type
|
---|
70 | */
|
---|
71 | protected String default_document_type = null;
|
---|
72 | /**
|
---|
73 | * the default index, or comma separated list if more than one is the
|
---|
74 | * default (with start and end commas, eg ,TI,SU,). Should be set by
|
---|
75 | * configure()
|
---|
76 | */
|
---|
77 | protected String default_index = "";
|
---|
78 |
|
---|
79 | protected Element service_metadata_list = null;
|
---|
80 | protected HashMap<String, String> paramDefaults = null;
|
---|
81 |
|
---|
82 | public AbstractSearch()
|
---|
83 | {
|
---|
84 | paramDefaults = new HashMap<String, String>();
|
---|
85 | }
|
---|
86 |
|
---|
87 | /**
|
---|
88 | * Sets up the short service info for service by QUERY_SERVICE (e.g.
|
---|
89 | * TextQuery or AudioQuery) If other services will be provided, should be
|
---|
90 | * added in the subclass configure also looks for search format info, and
|
---|
91 | * document format info
|
---|
92 | */
|
---|
93 | public boolean configure(Element info, Element extra_info)
|
---|
94 | {
|
---|
95 | if (!super.configure(info, extra_info))
|
---|
96 | {
|
---|
97 | return false;
|
---|
98 | }
|
---|
99 |
|
---|
100 | logger.info("Configuring AbstractSearch...");
|
---|
101 |
|
---|
102 | this.config_info = info;
|
---|
103 |
|
---|
104 | // set up short_service_info_
|
---|
105 | // => for now just has id and type. the name (lang dependent)
|
---|
106 | // will be added in if the list is requested.
|
---|
107 |
|
---|
108 | Element tq_service = this.desc_doc.createElement(GSXML.SERVICE_ELEM);
|
---|
109 | tq_service.setAttribute(GSXML.TYPE_ATT, GSXML.SERVICE_TYPE_QUERY);
|
---|
110 | tq_service.setAttribute(GSXML.NAME_ATT, QUERY_SERVICE);
|
---|
111 | this.short_service_info.appendChild(tq_service);
|
---|
112 |
|
---|
113 | // load up paging defaults
|
---|
114 | if (does_chunking) {
|
---|
115 | paramDefaults.put(MAXDOCS_PARAM, "100");
|
---|
116 | }
|
---|
117 | if (does_paging) {
|
---|
118 | paramDefaults.put(HITS_PER_PAGE_PARAM, "20");
|
---|
119 | paramDefaults.put(START_PAGE_PARAM, "1");
|
---|
120 | }
|
---|
121 |
|
---|
122 | // load up any search param defaults
|
---|
123 | Element search_elem = (Element) GSXML.getChildByTagName(extra_info, GSXML.SEARCH_ELEM);
|
---|
124 | if (search_elem != null) {
|
---|
125 | getSearchParamDefaults(search_elem);
|
---|
126 | }
|
---|
127 | // add some format info to service map if there is any
|
---|
128 | // => lookin extra info first look in buildConfig
|
---|
129 |
|
---|
130 | Element format = (Element) GSXML.getChildByTagName(info, GSXML.FORMAT_ELEM);
|
---|
131 |
|
---|
132 | if (format == null)
|
---|
133 | {
|
---|
134 | // try to find a format element inside <search> that contains a gsf:template. Note what if we have only xsl:templates??
|
---|
135 |
|
---|
136 | NodeList format_elems = null;
|
---|
137 | if (search_elem != null)
|
---|
138 | {
|
---|
139 | format_elems = search_elem.getElementsByTagName(GSXML.FORMAT_ELEM);
|
---|
140 | }
|
---|
141 | for (int i = 0; i < format_elems.getLength(); i++)
|
---|
142 | {
|
---|
143 | format = (Element) format_elems.item(i);
|
---|
144 | if (format.getElementsByTagName("gsf:template").getLength() != 0)
|
---|
145 | {
|
---|
146 | break;
|
---|
147 | }
|
---|
148 | }
|
---|
149 | }//end of if(format==null)
|
---|
150 | //
|
---|
151 | if (format != null)
|
---|
152 | {
|
---|
153 | this.format_info_map.put(QUERY_SERVICE, this.desc_doc.importNode(format, true));
|
---|
154 | }
|
---|
155 |
|
---|
156 | // look for document display format - for documentType
|
---|
157 | String path = GSPath.appendLink(GSXML.DISPLAY_ELEM, GSXML.FORMAT_ELEM);
|
---|
158 | Element display_format = (Element) GSXML.getNodeByPath(extra_info, path);
|
---|
159 | if (display_format != null)
|
---|
160 | {
|
---|
161 | // check for docType option.
|
---|
162 | Element doc_type_opt = GSXML.getNamedElement(display_format, "gsf:option", GSXML.NAME_ATT, "documentType");
|
---|
163 | if (doc_type_opt != null)
|
---|
164 | {
|
---|
165 | String value = doc_type_opt.getAttribute(GSXML.VALUE_ATT);
|
---|
166 | if (!value.equals(""))
|
---|
167 | {
|
---|
168 | this.default_document_type = value;
|
---|
169 | }
|
---|
170 | }
|
---|
171 | }
|
---|
172 |
|
---|
173 | // Base line for document (might be overriden by sub-classes)
|
---|
174 | gs_doc = new BasicDocument(this.default_document_type);
|
---|
175 |
|
---|
176 | return true;
|
---|
177 | }
|
---|
178 |
|
---|
179 | protected void getSearchParamDefaults(Element search_elem) {
|
---|
180 |
|
---|
181 | NodeList param_defaults_list = GSXML.getChildrenByTagName(search_elem, GSXML.PARAM_DEFAULT_ELEM);
|
---|
182 | for (int i=0; i<param_defaults_list.getLength(); i++) {
|
---|
183 | Element paramdef = (Element)param_defaults_list.item(i);
|
---|
184 | String name = paramdef.getAttribute(GSXML.NAME_ATT);
|
---|
185 | String val = paramdef.getAttribute(GSXML.VALUE_ATT);
|
---|
186 | if (!name.equals("") && !val.equals("")) {
|
---|
187 | paramDefaults.put(name, val);
|
---|
188 | }
|
---|
189 | }
|
---|
190 | }
|
---|
191 | /**
|
---|
192 | * returns a basic description for QUERY_SERVICE. If a subclass provides
|
---|
193 | * other services they need to provide their own descriptions
|
---|
194 | */
|
---|
195 | protected Element getServiceDescription(Document doc, String service, String lang, String subset)
|
---|
196 | {
|
---|
197 | if (!service.equals(QUERY_SERVICE))
|
---|
198 | {
|
---|
199 | return null;
|
---|
200 | }
|
---|
201 |
|
---|
202 | Element tq_service = doc.createElement(GSXML.SERVICE_ELEM);
|
---|
203 | tq_service.setAttribute(GSXML.TYPE_ATT, GSXML.SERVICE_TYPE_QUERY);
|
---|
204 | tq_service.setAttribute(GSXML.NAME_ATT, QUERY_SERVICE);
|
---|
205 | if (subset == null || subset.equals(GSXML.DISPLAY_TEXT_ELEM + GSXML.LIST_MODIFIER))
|
---|
206 | {
|
---|
207 | tq_service.appendChild(GSXML.createDisplayTextElement(doc, GSXML.DISPLAY_TEXT_NAME, getServiceName(QUERY_SERVICE, lang)));
|
---|
208 | tq_service.appendChild(GSXML.createDisplayTextElement(doc, GSXML.DISPLAY_TEXT_SUBMIT, getServiceSubmit(QUERY_SERVICE, lang)));
|
---|
209 | tq_service.appendChild(GSXML.createDisplayTextElement(doc, GSXML.DISPLAY_TEXT_DESCRIPTION, getServiceDescription(QUERY_SERVICE, lang)));
|
---|
210 | }
|
---|
211 | if (subset == null || subset.equals(GSXML.PARAM_ELEM + GSXML.LIST_MODIFIER))
|
---|
212 | {
|
---|
213 | Element param_list = doc.createElement(GSXML.PARAM_ELEM + GSXML.LIST_MODIFIER);
|
---|
214 | addCustomQueryParams(param_list, lang);
|
---|
215 | addStandardQueryParams(param_list, lang);
|
---|
216 | tq_service.appendChild(param_list);
|
---|
217 | }
|
---|
218 | if (subset == null || subset.equals(GSXML.METADATA_ELEM + GSXML.LIST_MODIFIER)) {
|
---|
219 |
|
---|
220 | if (service_metadata_list == null) {
|
---|
221 | Document ml_doc = XMLConverter.newDOM();
|
---|
222 | service_metadata_list = ml_doc.createElement(GSXML.METADATA_ELEM+GSXML.LIST_MODIFIER);
|
---|
223 | if (does_paging) {
|
---|
224 | service_metadata_list.appendChild(GSXML.createMetadataElement(ml_doc, "does_paging", "true"));
|
---|
225 | }
|
---|
226 | }
|
---|
227 | tq_service.appendChild(doc.importNode(service_metadata_list, true));
|
---|
228 | }
|
---|
229 | return tq_service;
|
---|
230 |
|
---|
231 | }
|
---|
232 |
|
---|
233 | // perhaps these should be changed to search down the class hierarchy for
|
---|
234 | // values - do can just put the info in the resource bundle to use it
|
---|
235 | /** returns the default name for the TextQuery service */
|
---|
236 | protected String getServiceName(String service_id, String lang)
|
---|
237 | {
|
---|
238 | return getTextString(service_id + ".name", lang);
|
---|
239 | }
|
---|
240 |
|
---|
241 | /** returns the default description for the TextQuery service */
|
---|
242 | protected String getServiceDescription(String service_id, String lang)
|
---|
243 | {
|
---|
244 | return getTextString(service_id + ".description", lang);
|
---|
245 | }
|
---|
246 |
|
---|
247 | /** returns the default submit button text for the TextQuery service */
|
---|
248 | protected String getServiceSubmit(String service_id, String lang)
|
---|
249 | {
|
---|
250 | return getTextString(service_id + ".submit", lang);
|
---|
251 |
|
---|
252 | }
|
---|
253 |
|
---|
254 | /** adds the standard query params into the service description */
|
---|
255 | protected void addStandardQueryParams(Element param_list, String lang)
|
---|
256 | {
|
---|
257 | // this test is not so good. here we are using absence of default index
|
---|
258 | // to determine whether we have indexes or not. But in other places,
|
---|
259 | // absence of default index just means to use the first one as default.
|
---|
260 | if (!default_index.equals(""))
|
---|
261 | {
|
---|
262 | createParameter(INDEX_PARAM, param_list, lang);
|
---|
263 | }
|
---|
264 | if (does_chunking)
|
---|
265 | {
|
---|
266 | createParameter(MAXDOCS_PARAM, param_list, lang);
|
---|
267 | }
|
---|
268 | if (does_paging)
|
---|
269 | {
|
---|
270 | createParameter(HITS_PER_PAGE_PARAM, param_list, lang);
|
---|
271 | createParameter(START_PAGE_PARAM, param_list, lang);
|
---|
272 | }
|
---|
273 | createParameter(QUERY_PARAM, param_list, lang);
|
---|
274 | }
|
---|
275 |
|
---|
276 | /**
|
---|
277 | * adds any service specific query params into the service default
|
---|
278 | * implementation: add nothing. subclasses may need to override this to add
|
---|
279 | * in their specific parameters
|
---|
280 | */
|
---|
281 | protected void addCustomQueryParams(Element param_list, String lang)
|
---|
282 | {
|
---|
283 | // default behaviour, do nothing
|
---|
284 | }
|
---|
285 |
|
---|
286 | protected void createParameter(String name, Element param_list, String lang)
|
---|
287 | {
|
---|
288 | createParameter(name, param_list, lang, null);
|
---|
289 | }
|
---|
290 |
|
---|
291 | protected void createParameter(String name, Element param_list, String lang, String default_value)
|
---|
292 | {
|
---|
293 | // at this level, not interested in boolean return type
|
---|
294 | createParameterChain(name, param_list, lang, default_value);
|
---|
295 | }
|
---|
296 |
|
---|
297 | /**
|
---|
298 | * default implementations for the standard parameters plus some other
|
---|
299 | * common ones index, maxDocs, hitsPerPage, startPage
|
---|
300 | */
|
---|
301 |
|
---|
302 | protected boolean createParameterChain(String name, Element param_list, String lang, String default_value)
|
---|
303 | {
|
---|
304 | Document doc = param_list.getOwnerDocument();
|
---|
305 | Element param = null;
|
---|
306 | String param_default = default_value;
|
---|
307 | if (default_value == null) {
|
---|
308 | // have we got a stored up default? will be null if not there
|
---|
309 | param_default = paramDefaults.get(name);
|
---|
310 | }
|
---|
311 | if (name.equals(QUERY_PARAM) || name.equals(RAW_PARAM))
|
---|
312 | {
|
---|
313 | param = GSXML.createParameterDescription(doc, name, getTextString("param." + name, lang), GSXML.PARAM_TYPE_STRING, param_default, null, null);
|
---|
314 | param_list.appendChild(param);
|
---|
315 | return true;
|
---|
316 | }
|
---|
317 | else if (name.equals(INDEX_PARAM))
|
---|
318 | {
|
---|
319 | // should we make these class fields?
|
---|
320 | ArrayList<String> index_ids = new ArrayList<String>();
|
---|
321 | ArrayList<String> index_names = new ArrayList<String>();
|
---|
322 | getIndexData(index_ids, index_names, lang);
|
---|
323 | String param_type = GSXML.PARAM_TYPE_ENUM_SINGLE;
|
---|
324 | if (does_multi_index_search)
|
---|
325 | {
|
---|
326 | param_type = GSXML.PARAM_TYPE_ENUM_MULTI;
|
---|
327 | }
|
---|
328 | if (param_default == null)
|
---|
329 | {
|
---|
330 | param_default = this.default_index;
|
---|
331 | }
|
---|
332 | param = GSXML.createParameterDescription2(doc, INDEX_PARAM, getTextString("param." + INDEX_PARAM, lang), param_type, param_default, index_ids, index_names);
|
---|
333 | param_list.appendChild(param);
|
---|
334 | return true;
|
---|
335 | }
|
---|
336 | else if (name.equals(MAXDOCS_PARAM))
|
---|
337 | {
|
---|
338 | param = GSXML.createParameterDescription(doc, name, getTextString("param." + name, lang), GSXML.PARAM_TYPE_INTEGER, param_default, null, null);
|
---|
339 | param_list.appendChild(param);
|
---|
340 | return true;
|
---|
341 | }
|
---|
342 | else if (name.equals(HITS_PER_PAGE_PARAM))
|
---|
343 | {
|
---|
344 | param = GSXML.createParameterDescription(doc, name, getTextString("param." + name, lang), GSXML.PARAM_TYPE_INTEGER, param_default, null, null);
|
---|
345 | param_list.appendChild(param);
|
---|
346 | return true;
|
---|
347 | }
|
---|
348 | else if (name.equals(START_PAGE_PARAM))
|
---|
349 | {
|
---|
350 | // start page - set to 1 for the search page
|
---|
351 | param = GSXML.createParameterDescription(doc, START_PAGE_PARAM, "", GSXML.PARAM_TYPE_INVISIBLE, param_default, null, null);
|
---|
352 | param_list.appendChild(param);
|
---|
353 | return true;
|
---|
354 | }
|
---|
355 |
|
---|
356 | // Get to there then none of the above params matched
|
---|
357 | // => return false so the chain can continue
|
---|
358 | return false;
|
---|
359 | }
|
---|
360 |
|
---|
361 | /**
|
---|
362 | * create an element to go into the search results list. A node element has
|
---|
363 | * the form <docNode nodeId='xxx' nodeType='leaf' docType='hierarchy'
|
---|
364 | * rank='0.23'/>
|
---|
365 | */
|
---|
366 | protected Element createDocNode(Document doc, String node_id, String rank)
|
---|
367 | {
|
---|
368 | return this.gs_doc.createDocNode(doc, node_id, rank);
|
---|
369 | }
|
---|
370 |
|
---|
371 | /**
|
---|
372 | * returns the document type of the doc that the specified node belongs to.
|
---|
373 | * should be one of GSXML.DOC_TYPE_SIMPLE, GSXML.DOC_TYPE_PAGED,
|
---|
374 | * GSXML.DOC_TYPE_HIERARCHY
|
---|
375 | */
|
---|
376 | protected String getDocType(String node_id)
|
---|
377 | {
|
---|
378 | return this.gs_doc.getDocType(node_id);
|
---|
379 | }
|
---|
380 |
|
---|
381 | /**
|
---|
382 | * returns the node type of the specified node. should be one of
|
---|
383 | * GSXML.NODE_TYPE_LEAF, GSXML.NODE_TYPE_INTERNAL, GSXML.NODE_TYPE_ROOT
|
---|
384 | */
|
---|
385 | protected String getNodeType(String node_id, String doc_type)
|
---|
386 | {
|
---|
387 | return this.gs_doc.getNodeType(node_id, doc_type);
|
---|
388 | }
|
---|
389 |
|
---|
390 | /** returns true if the node has child nodes */
|
---|
391 | protected boolean hasChildren(String node_id)
|
---|
392 | {
|
---|
393 | return this.gs_doc.hasChildren(node_id);
|
---|
394 | }
|
---|
395 |
|
---|
396 | /** returns true if the node has a parent */
|
---|
397 | protected boolean hasParent(String node_id)
|
---|
398 | {
|
---|
399 | return this.gs_doc.hasParent(node_id);
|
---|
400 | }
|
---|
401 |
|
---|
402 | /**
|
---|
403 | * get the details about the indexes available must be implemented by
|
---|
404 | * subclass there must be at least one index
|
---|
405 | */
|
---|
406 | abstract protected void getIndexData(ArrayList<String> index_ids, ArrayList<String> index_names, String lang);
|
---|
407 |
|
---|
408 | }
|
---|