1 | /*
|
---|
2 | * SharedSoleneGS2FieldSearch.java -- shared base code for Solr and Lucene
|
---|
3 | * Copyright (C) 2006 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 |
|
---|
19 | package org.greenstone.gsdl3.service;
|
---|
20 |
|
---|
21 | // Greenstone classes
|
---|
22 | import java.util.ArrayList;
|
---|
23 |
|
---|
24 | import org.apache.log4j.Logger;
|
---|
25 | import org.greenstone.LuceneWrapper4.SharedSoleneQuery;
|
---|
26 | import org.greenstone.gsdl3.util.GSXML;
|
---|
27 | import org.greenstone.gsdl3.util.XMLConverter;
|
---|
28 |
|
---|
29 | import org.w3c.dom.Document;
|
---|
30 | import org.w3c.dom.Element;
|
---|
31 | import org.w3c.dom.NodeList;
|
---|
32 |
|
---|
33 | // Shared code for Solr and Lucene GS2FieldSearch
|
---|
34 |
|
---|
35 | public abstract class SharedSoleneGS2FieldSearch extends AbstractGS2FieldSearch
|
---|
36 | {
|
---|
37 |
|
---|
38 | protected static String RANK_PARAM_RANK = "rank";
|
---|
39 | protected static String RANK_PARAM_NONE = "none";
|
---|
40 |
|
---|
41 | protected static final String SORT_ELEM = "sort";
|
---|
42 | protected static final String DEFAULT_SORT_ELEM = "defaultSort";
|
---|
43 |
|
---|
44 | static Logger logger = Logger.getLogger(org.greenstone.gsdl3.service.SharedSoleneGS2FieldSearch.class.getName());
|
---|
45 |
|
---|
46 | protected SharedSoleneQuery solene_src = null;
|
---|
47 | protected String default_sort = "";
|
---|
48 |
|
---|
49 | public SharedSoleneGS2FieldSearch()
|
---|
50 | {
|
---|
51 | does_paging = true;
|
---|
52 | does_chunking = false; // actually it can but we don't want to do a max docs, as we are already paging results
|
---|
53 |
|
---|
54 | // Lucene/Solr uses double operators, not single
|
---|
55 | //AND_OPERATOR = "&&";
|
---|
56 | //OR_OPERATOR = "||";
|
---|
57 | AND_OPERATOR = "AND";
|
---|
58 | OR_OPERATOR = "OR";
|
---|
59 | NOT_OPERATOR = "NOT";
|
---|
60 |
|
---|
61 | }
|
---|
62 |
|
---|
63 | /** configure this service */
|
---|
64 | public boolean configure(Element info, Element extra_info)
|
---|
65 | {
|
---|
66 | if (!super.configure(info, extra_info))
|
---|
67 | {
|
---|
68 | return false;
|
---|
69 | }
|
---|
70 |
|
---|
71 | // the search element
|
---|
72 | Element config_search = (Element) GSXML.getChildByTagName(extra_info, GSXML.SEARCH_ELEM);
|
---|
73 | Document owner = info.getOwnerDocument();
|
---|
74 | // find the sort fields in serviceRack xml, and add in the deisplayItems if any
|
---|
75 | NodeList sort_nodes = info.getElementsByTagName(SORT_ELEM);
|
---|
76 |
|
---|
77 | for (int i = 0; i < sort_nodes.getLength(); i++)
|
---|
78 | {
|
---|
79 | Element sort = (Element) sort_nodes.item(i);
|
---|
80 | String name = sort.getAttribute(GSXML.NAME_ATT);
|
---|
81 | Element node_extra = GSXML.getNamedElement(config_search, SORT_ELEM, GSXML.NAME_ATT, name);
|
---|
82 | if (node_extra == null)
|
---|
83 | {
|
---|
84 | logger.error("haven't found extra info for sort field named " + name);
|
---|
85 | continue;
|
---|
86 | }
|
---|
87 |
|
---|
88 | // get the display elements if any - displayName
|
---|
89 | NodeList display_names = node_extra.getElementsByTagName(GSXML.DISPLAY_TEXT_ELEM);
|
---|
90 | if (display_names != null)
|
---|
91 | {
|
---|
92 | for (int j = 0; j < display_names.getLength(); j++)
|
---|
93 | {
|
---|
94 | Element e = (Element) display_names.item(j);
|
---|
95 | sort.appendChild(owner.importNode(e, true));
|
---|
96 | }
|
---|
97 | }
|
---|
98 | } // for each sortfield
|
---|
99 |
|
---|
100 | // get the default sort field
|
---|
101 | Element def = (Element) GSXML.getChildByTagName(info, DEFAULT_SORT_ELEM);
|
---|
102 | if (def != null)
|
---|
103 | {
|
---|
104 | this.default_sort = def.getAttribute(GSXML.SHORTNAME_ATT);
|
---|
105 | }
|
---|
106 |
|
---|
107 | // Lucene/Solr doesn't do case folding or stemming or accent folding at the
|
---|
108 | // moment
|
---|
109 | does_case = false;
|
---|
110 | does_stem = false;
|
---|
111 | does_accent = false;
|
---|
112 |
|
---|
113 | return true;
|
---|
114 | }
|
---|
115 |
|
---|
116 | /** add in the Lucene/Solr specific params to TextQuery */
|
---|
117 | protected void addCustomQueryParams(Element param_list, String lang)
|
---|
118 | {
|
---|
119 | super.addCustomQueryParams(param_list, lang);
|
---|
120 | /** Lucene's/Solr's rank (sort) param is based on sort fields, not ranked/not */
|
---|
121 | createParameter(RANK_PARAM, param_list, lang);
|
---|
122 | }
|
---|
123 | /** add in Lucene/SOLR specific params for AdvancedFieldQuery */
|
---|
124 | protected void addCustomQueryParamsAdvField(Element param_list, String lang)
|
---|
125 | {
|
---|
126 | super.addCustomQueryParamsAdvField(param_list, lang);
|
---|
127 | createParameter(RANK_PARAM, param_list, lang);
|
---|
128 |
|
---|
129 | }
|
---|
130 |
|
---|
131 | /** create a param and add to the list */
|
---|
132 | /** we override this to do a special rank param */
|
---|
133 | protected void createParameter(String name, Element param_list, String lang)
|
---|
134 | {
|
---|
135 | Document doc = param_list.getOwnerDocument();
|
---|
136 | Element param = null;
|
---|
137 | String param_default = paramDefaults.get(name);
|
---|
138 | if (name.equals(RANK_PARAM))
|
---|
139 | {
|
---|
140 | // get the fields
|
---|
141 | ArrayList<String> fields = new ArrayList<String>();
|
---|
142 | ArrayList<String> field_names = new ArrayList<String>();
|
---|
143 | param_default = default_sort;
|
---|
144 | if (!getSortData(fields, field_names, lang)) {
|
---|
145 | fields.add(RANK_PARAM_RANK);
|
---|
146 | fields.add(RANK_PARAM_NONE);
|
---|
147 | field_names.add(getTextString("param." + RANK_PARAM + "." + RANK_PARAM_RANK, lang));
|
---|
148 | field_names.add(getTextString("param." + RANK_PARAM + "." + RANK_PARAM_NONE, lang));
|
---|
149 | param_default = RANK_PARAM_RANK;
|
---|
150 | }
|
---|
151 | else {
|
---|
152 | if (param_default == null) {
|
---|
153 | param_default = fields.get(0);
|
---|
154 | }
|
---|
155 | }
|
---|
156 | param = GSXML.createParameterDescription2(doc, name, getTextString("param." + name, lang), GSXML.PARAM_TYPE_ENUM_SINGLE, param_default, fields, field_names);
|
---|
157 |
|
---|
158 | }
|
---|
159 |
|
---|
160 | if (param != null)
|
---|
161 | {
|
---|
162 | param_list.appendChild(param);
|
---|
163 | }
|
---|
164 | else
|
---|
165 | {
|
---|
166 | super.createParameter(name, param_list, lang);
|
---|
167 | }
|
---|
168 |
|
---|
169 | }
|
---|
170 |
|
---|
171 | protected boolean getSortData(ArrayList<String> sort_ids, ArrayList<String> sort_names, String lang) {
|
---|
172 |
|
---|
173 | Element sort_list = (Element) GSXML.getChildByTagName(this.config_info, SORT_ELEM + GSXML.LIST_MODIFIER);
|
---|
174 | if (sort_list == null) return false;
|
---|
175 | NodeList sorts = sort_list.getElementsByTagName(SORT_ELEM);
|
---|
176 | int len = sorts.getLength();
|
---|
177 | if (len == 0) return false;
|
---|
178 | for (int i = 0; i < len; i++)
|
---|
179 | {
|
---|
180 | Element sort = (Element) sorts.item(i);
|
---|
181 | String shortname = sort.getAttribute(GSXML.SHORTNAME_ATT);
|
---|
182 | sort_ids.add(shortname);
|
---|
183 | String display_name = getDisplayText(sort, GSXML.DISPLAY_TEXT_NAME, lang, "en", "metadata_names");
|
---|
184 | if (display_name.equals(""))
|
---|
185 | {
|
---|
186 | // use the sort name and look up in dictionary
|
---|
187 | display_name = sort.getAttribute(GSXML.NAME_ATT);
|
---|
188 | if (display_name.equals(""))
|
---|
189 | {
|
---|
190 | display_name = shortname;
|
---|
191 | }
|
---|
192 | else {
|
---|
193 | display_name = getTextString("param.sortBy." + display_name, lang);
|
---|
194 | }
|
---|
195 | }
|
---|
196 | sort_names.add(display_name);
|
---|
197 |
|
---|
198 | }
|
---|
199 | return true;
|
---|
200 | }
|
---|
201 |
|
---|
202 | protected String addFieldInfo(String query, String field)
|
---|
203 | {
|
---|
204 | // currently, allfields (ZZ) is stored as a extra field for Lucene
|
---|
205 | if (field.equals(""))
|
---|
206 | { // || field.equals("ZZ")) {
|
---|
207 | return query;
|
---|
208 | }
|
---|
209 | return field + ":(" + query + ")";
|
---|
210 | }
|
---|
211 |
|
---|
212 | protected void addQueryElem(StringBuffer s, String q, String f, String c)
|
---|
213 | {
|
---|
214 |
|
---|
215 | String combine = "";
|
---|
216 | if (s.length() > 0)
|
---|
217 | {
|
---|
218 | combine = " " + c + " ";
|
---|
219 | }
|
---|
220 | s.append(combine + addFieldInfo(q, f));
|
---|
221 | }
|
---|
222 |
|
---|
223 | /** Lucene/Solr doesn't use these options at the moment */
|
---|
224 | protected String addStemOptions(String query, String stem, String casef, String accent)
|
---|
225 | {
|
---|
226 | return query;
|
---|
227 | }
|
---|
228 |
|
---|
229 | /**
|
---|
230 | * Lucene/Solr does not use internal ids. It just uses hash ids. So we need
|
---|
231 | * to override these methods so no conversion is done.
|
---|
232 | */
|
---|
233 | /** convert indexer internal id to Greenstone oid */
|
---|
234 | protected String internalNum2OID(long docnum)
|
---|
235 | {
|
---|
236 | return Long.toString(docnum);
|
---|
237 | }
|
---|
238 |
|
---|
239 | protected String internalNum2OID(String docnum)
|
---|
240 | {
|
---|
241 | return docnum;
|
---|
242 |
|
---|
243 | }
|
---|
244 | }
|
---|