1
|
/**
|
2
|
* '$RCSfile$'
|
3
|
* Purpose: A Class that represents a structured query, and can be
|
4
|
* constructed from an XML serialization conforming to
|
5
|
* pathquery.dtd. The printSQL() method can be used to print
|
6
|
* a SQL serialization of the query.
|
7
|
* Copyright: 2000 Regents of the University of California and the
|
8
|
* National Center for Ecological Analysis and Synthesis
|
9
|
* Authors: Matt Jones
|
10
|
*
|
11
|
* '$Author: leinfelder $'
|
12
|
* '$Date: 2012-10-18 15:14:22 -0700 (Thu, 18 Oct 2012) $'
|
13
|
* '$Revision: 7407 $'
|
14
|
*
|
15
|
* This program is free software; you can redistribute it and/or modify
|
16
|
* it under the terms of the GNU General Public License as published by
|
17
|
* the Free Software Foundation; either version 2 of the License, or
|
18
|
* (at your option) any later version.
|
19
|
*
|
20
|
* This program is distributed in the hope that it will be useful,
|
21
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
22
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
23
|
* GNU General Public License for more details.
|
24
|
*
|
25
|
* You should have received a copy of the GNU General Public License
|
26
|
* along with this program; if not, write to the Free Software
|
27
|
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
|
28
|
*/
|
29
|
|
30
|
package edu.ucsb.nceas.metacat;
|
31
|
|
32
|
import java.util.ArrayList;
|
33
|
import java.util.Enumeration;
|
34
|
import java.util.List;
|
35
|
import java.util.Vector;
|
36
|
|
37
|
import org.apache.log4j.Logger;
|
38
|
|
39
|
import edu.ucsb.nceas.metacat.shared.MetacatUtilException;
|
40
|
import edu.ucsb.nceas.metacat.util.SystemUtil;
|
41
|
|
42
|
/** a utility class that represents a group of terms in a query */
|
43
|
public class QueryGroup {
|
44
|
private String operator = null; // indicates how query terms are combined
|
45
|
private Vector children = null; // the list of query terms and groups
|
46
|
private int countPercentageSearchItem = 0;
|
47
|
private Vector queryTermsWithSameValue = null;//this two dimension vectors.
|
48
|
//will hold query terms which has same search value.
|
49
|
private Vector queryTermsInPathIndex = null; //this vector holds query terms without same value
|
50
|
// and search path is in path index.
|
51
|
private Vector<QueryTerm> queryTerms = null;//this vector only holds query terms without same search value.
|
52
|
// and search path is NOT in path index.
|
53
|
private Vector queryGroupsChildren = null;
|
54
|
private static Logger logMetacat = Logger.getLogger(QueryGroup.class);
|
55
|
public static String UNION = "UNION";
|
56
|
public static String INTERSECT = "INTERSECT";
|
57
|
|
58
|
|
59
|
/**
|
60
|
* construct a new QueryGroup
|
61
|
*
|
62
|
* @param operator the boolean conector used to connect query terms
|
63
|
* in this query group
|
64
|
*/
|
65
|
public QueryGroup(String operator) {
|
66
|
this.operator = operator;
|
67
|
children = new Vector();
|
68
|
queryTermsWithSameValue = new Vector();
|
69
|
queryTermsInPathIndex = new Vector();
|
70
|
queryTerms = new Vector<QueryTerm>();
|
71
|
queryGroupsChildren = new Vector();
|
72
|
}
|
73
|
|
74
|
/**
|
75
|
* Add a child QueryGroup to this QueryGroup
|
76
|
*
|
77
|
* @param qgroup the query group to be added to the list of terms
|
78
|
*/
|
79
|
public void addChild(QueryGroup qgroup) {
|
80
|
children.add((Object)qgroup);
|
81
|
queryGroupsChildren.add(qgroup);
|
82
|
}
|
83
|
|
84
|
/**
|
85
|
* Add a child QueryTerm to this QueryGroup
|
86
|
*
|
87
|
* @param qterm the query term to be added to the list of terms
|
88
|
*/
|
89
|
public void addChild(QueryTerm qterm) {
|
90
|
children.add((Object)qterm);
|
91
|
handleNewQueryTerms(qterm);
|
92
|
|
93
|
}
|
94
|
|
95
|
/*
|
96
|
* Retrieve an Enumeration of query terms for this QueryGroup
|
97
|
*/
|
98
|
private Enumeration getChildren() {
|
99
|
return children.elements();
|
100
|
}
|
101
|
|
102
|
public int getPercentageSymbolCount()
|
103
|
{
|
104
|
return countPercentageSearchItem;
|
105
|
}
|
106
|
|
107
|
/**
|
108
|
* create a SQL serialization of the query that this instance represents
|
109
|
*/
|
110
|
public String printSQL(boolean useXMLIndex, List<Object> parameterValues) {
|
111
|
|
112
|
StringBuffer self = new StringBuffer();
|
113
|
StringBuffer queryString = new StringBuffer();
|
114
|
|
115
|
boolean first = true;
|
116
|
|
117
|
if (!queryTermsWithSameValue.isEmpty() || !queryTermsInPathIndex.isEmpty())
|
118
|
{
|
119
|
// keep track of the values we add as prepared statement question marks (?)
|
120
|
List<Object> groupValues = new ArrayList<Object>();
|
121
|
String pathIndexQueryString = printSQLStringInPathIndex(groupValues);
|
122
|
parameterValues.addAll(groupValues);
|
123
|
queryString.append(pathIndexQueryString);
|
124
|
if (queryString != null)
|
125
|
{
|
126
|
first = false;
|
127
|
}
|
128
|
}
|
129
|
|
130
|
for (int i=0; i<queryGroupsChildren.size(); i++)
|
131
|
{
|
132
|
|
133
|
// keep track of the values we add as prepared statement question marks (?)
|
134
|
List<Object> childrenValues = new ArrayList<Object>();
|
135
|
// get the group
|
136
|
QueryGroup qg = (QueryGroup) queryGroupsChildren.elementAt(i);
|
137
|
String queryGroupSQL = qg.printSQL(useXMLIndex, childrenValues);
|
138
|
logMetacat.info("In QueryGroup.printSQL.. found a QueryGroup: " + queryGroupSQL);
|
139
|
if (first) {
|
140
|
first = false;
|
141
|
} else {
|
142
|
if(!queryString.toString().equals("") && queryGroupSQL != null &&!queryGroupSQL.equals("")){
|
143
|
queryString.append(" " + operator + " ");
|
144
|
}
|
145
|
}
|
146
|
// add the sql
|
147
|
queryString.append(queryGroupSQL);
|
148
|
// add the parameter values
|
149
|
parameterValues.addAll(childrenValues);
|
150
|
|
151
|
// count percentage number
|
152
|
int count = qg.getPercentageSymbolCount();
|
153
|
countPercentageSearchItem = countPercentageSearchItem + count;
|
154
|
}
|
155
|
|
156
|
for (int i=0; i<queryTerms.size(); i++)
|
157
|
{
|
158
|
// keep track of the values we add as prepared statement question marks (?)
|
159
|
List<Object> termValues = new ArrayList<Object>();
|
160
|
// get the term
|
161
|
QueryTerm qt = (QueryTerm)queryTerms.elementAt(i);
|
162
|
String termQueryString = qt.printSQL(useXMLIndex, termValues);
|
163
|
logMetacat.info("In QueryGroup.printSQL.. found a QueryGroup: " + termQueryString);
|
164
|
if (!(qt.getSearchMode().equals("contains") && qt.getValue().equals("%"))){
|
165
|
if (first) {
|
166
|
first = false;
|
167
|
} else {
|
168
|
if(!queryString.toString().equals("")){
|
169
|
queryString.append(" " + operator + " ");
|
170
|
}
|
171
|
}
|
172
|
// include the sql
|
173
|
queryString.append(termQueryString);
|
174
|
// include the parameter values
|
175
|
parameterValues.addAll(termValues);
|
176
|
|
177
|
// count percerntage number
|
178
|
int count = qt.getPercentageSymbolCount();
|
179
|
countPercentageSearchItem = countPercentageSearchItem + count;
|
180
|
}
|
181
|
}
|
182
|
|
183
|
if(!queryString.toString().equals("")){
|
184
|
self.append("(");
|
185
|
self.append(queryString.toString());
|
186
|
self.append(")");
|
187
|
}
|
188
|
|
189
|
logMetacat.info("In QueryGroup.printSQL.. final query returned is: "
|
190
|
+ self.toString());
|
191
|
return self.toString();
|
192
|
}
|
193
|
|
194
|
|
195
|
|
196
|
|
197
|
/*
|
198
|
* If every query term in a queryGroup share a search value and search path
|
199
|
* is in xml_path_index, we should use a new query to replace the original query term query in order to
|
200
|
* improve performance. Also if even the term doesn't share any value with other term
|
201
|
* we still use "OR" to replace UNION action (we only handle union operator in the query group).
|
202
|
*
|
203
|
*/
|
204
|
private String printSQLStringInPathIndex(List<Object> parameterValues)
|
205
|
{
|
206
|
String sql ="";
|
207
|
String value ="";
|
208
|
StringBuffer sqlBuff = new StringBuffer();
|
209
|
int index =0;
|
210
|
if (queryTermsWithSameValue != null && queryTermsInPathIndex != null)
|
211
|
{
|
212
|
|
213
|
sqlBuff.append("SELECT DISTINCT docid FROM xml_path_index WHERE ");
|
214
|
if (!queryTermsWithSameValue.isEmpty())
|
215
|
{
|
216
|
boolean firstVector = true;
|
217
|
for (int j=0; j<queryTermsWithSameValue.size(); j++)
|
218
|
{
|
219
|
Vector queryTermVector = (Vector)queryTermsWithSameValue.elementAt(j);
|
220
|
QueryTerm term1 = (QueryTerm)queryTermVector.elementAt(0);
|
221
|
value = term1.getValue();
|
222
|
boolean first = true;
|
223
|
if (firstVector)
|
224
|
{
|
225
|
firstVector = false;
|
226
|
}
|
227
|
else
|
228
|
{
|
229
|
sqlBuff.append(" "+"OR"+" ");
|
230
|
}
|
231
|
|
232
|
sqlBuff.append(" (");
|
233
|
|
234
|
// keep track of parameter values
|
235
|
List<Object> searchValues = new ArrayList<Object>();
|
236
|
|
237
|
// get the general search criteria (no path info)
|
238
|
String searchTermSQL = term1.printSearchExprSQL(searchValues);
|
239
|
|
240
|
// add the SQL
|
241
|
sqlBuff.append(searchTermSQL);
|
242
|
|
243
|
// add parameter values
|
244
|
parameterValues.addAll(searchValues);
|
245
|
|
246
|
sqlBuff.append("AND path IN ( ");
|
247
|
|
248
|
//gets every path in query term object
|
249
|
for (int i=0; i<queryTermVector.size(); i++)
|
250
|
{
|
251
|
QueryTerm term = (QueryTerm)queryTermVector.elementAt(i);
|
252
|
value = term.getValue();
|
253
|
String path = term.getPathExpression();
|
254
|
if (path != null && !path.equals(""))
|
255
|
{
|
256
|
if (first)
|
257
|
{
|
258
|
first = false;
|
259
|
sqlBuff.append("?");
|
260
|
parameterValues.add(path);
|
261
|
}
|
262
|
else
|
263
|
{
|
264
|
sqlBuff.append(", ?");
|
265
|
parameterValues.add(path);
|
266
|
}
|
267
|
index++;
|
268
|
if (value != null && (value.equals("%") || value.equals("%%%")))
|
269
|
{
|
270
|
countPercentageSearchItem++;
|
271
|
}
|
272
|
}
|
273
|
}
|
274
|
sqlBuff.append(" ))");
|
275
|
|
276
|
}
|
277
|
}
|
278
|
if (!queryTermsInPathIndex.isEmpty())
|
279
|
{
|
280
|
for (int j=0; j<queryTermsInPathIndex.size(); j++)
|
281
|
{
|
282
|
QueryTerm term = (QueryTerm)queryTermsInPathIndex.elementAt(j);
|
283
|
if (term != null)
|
284
|
{
|
285
|
term.setInUnionGroup(true);
|
286
|
if (index > 0)
|
287
|
{
|
288
|
sqlBuff.append(" "+"OR"+" ");
|
289
|
}
|
290
|
sqlBuff.append("(");
|
291
|
// keep track of the parameter values for this sql
|
292
|
List<Object> termParameterValues = new ArrayList<Object>();
|
293
|
String termSQL = term.printSQL(true, termParameterValues);
|
294
|
sqlBuff.append(termSQL);
|
295
|
sqlBuff.append(")");
|
296
|
// add the param values
|
297
|
parameterValues.addAll(termParameterValues);
|
298
|
index++;
|
299
|
}
|
300
|
}
|
301
|
}
|
302
|
}
|
303
|
if (index >0)
|
304
|
{
|
305
|
sql = sqlBuff.toString();
|
306
|
}
|
307
|
return sql;
|
308
|
}
|
309
|
|
310
|
/**
|
311
|
* create a String description of the query that this instance represents.
|
312
|
* This should become a way to get the XML serialization of the query.
|
313
|
*/
|
314
|
public String toString() {
|
315
|
StringBuffer self = new StringBuffer();
|
316
|
|
317
|
self.append(" (Query group operator=" + operator + "\n");
|
318
|
Enumeration en= getChildren();
|
319
|
while (en.hasMoreElements()) {
|
320
|
Object qobject = en.nextElement();
|
321
|
self.append(qobject);
|
322
|
}
|
323
|
self.append(" )\n");
|
324
|
return self.toString();
|
325
|
}
|
326
|
|
327
|
/*
|
328
|
* When a new QueryTerm come, first we need to compare it to
|
329
|
* the queryTerm vector, which contains queryTerm that doesn't
|
330
|
* have same search value to any other queryTerm. Here is algorithm.
|
331
|
* 1) If new QueryTerm find a QueryTerm in queryTerms which has same search value,
|
332
|
* them create a new vector which contain both QueryTerms and add the new vector
|
333
|
* to two-dimention vector queryTermsWithSameValue, and remove the QueryTerm which
|
334
|
* was in queryTerm.
|
335
|
* 2) If new QueryTerm couldn't find a QueryTerm in queryTerms which has same search value,
|
336
|
* then search queryTermsWithSameValue, to see if this vector already has the search value.
|
337
|
* 2.1) if has the search value, add the new QueryTerm to the queryTermsWithSameValue.
|
338
|
* 2.2) if hasn't, add the new QueryTerm to queryTerms vector.
|
339
|
*/
|
340
|
private void handleNewQueryTerms(QueryTerm newTerm)
|
341
|
{
|
342
|
// currently we only handle UNION group
|
343
|
if (newTerm != null )
|
344
|
{
|
345
|
//System.out.println("new term is not null branch in handle new query term");
|
346
|
//we only handle union operator now.
|
347
|
try {
|
348
|
if (operator != null
|
349
|
&& operator.equalsIgnoreCase(UNION)
|
350
|
&& SystemUtil.getPathsForIndexing().contains(
|
351
|
newTerm.getPathExpression())) {
|
352
|
// System.out.println("in only union branch in handle new
|
353
|
// query term");
|
354
|
for (int i = 0; i < queryTermsInPathIndex.size(); i++) {
|
355
|
QueryTerm term = (QueryTerm) queryTermsInPathIndex.elementAt(i);
|
356
|
if (term != null && term.hasSameSearchValue(newTerm)) {
|
357
|
// System.out.println("1Move a query term and add a
|
358
|
// new query term into search value in handle new
|
359
|
// query term");
|
360
|
// find a target which has same search value
|
361
|
Vector<QueryTerm> newSameValueVector = new Vector<QueryTerm>();
|
362
|
newSameValueVector.add(term);
|
363
|
newSameValueVector.addElement(newTerm);
|
364
|
queryTermsWithSameValue.add(newSameValueVector);
|
365
|
queryTermsInPathIndex.remove(i);
|
366
|
return;
|
367
|
}
|
368
|
}
|
369
|
// no same search value was found in queryTerms.
|
370
|
// then we need search queryTermsWithSameValue
|
371
|
for (int i = 0; i < queryTermsWithSameValue.size(); i++) {
|
372
|
Vector sameValueVec = (Vector) queryTermsWithSameValue.elementAt(i);
|
373
|
// we only compare the first query term
|
374
|
QueryTerm term = (QueryTerm) sameValueVec.elementAt(0);
|
375
|
if (term != null && term.hasSameSearchValue(newTerm)) {
|
376
|
// System.out.println("2add a new query term into
|
377
|
// search value in handle new query term");
|
378
|
sameValueVec.add(newTerm);
|
379
|
return;
|
380
|
}
|
381
|
}
|
382
|
// nothing found, but the search path is still in
|
383
|
// xml_path_index,
|
384
|
// save it into queryTermsInPathIndex vector
|
385
|
queryTermsInPathIndex.add(newTerm);
|
386
|
return;
|
387
|
}
|
388
|
} catch (MetacatUtilException ue) {
|
389
|
logMetacat.warn("Could not get index paths: " + ue.getMessage());
|
390
|
}
|
391
|
|
392
|
// add this newTerm to queryTerms since we couldn't find it in xml_path_index
|
393
|
queryTerms.add(newTerm);
|
394
|
}
|
395
|
|
396
|
}
|
397
|
}
|