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: jones $'
|
12
|
* '$Date: 2000-06-26 03:35:05 -0700 (Mon, 26 Jun 2000) $'
|
13
|
* '$Revision: 203 $'
|
14
|
*/
|
15
|
|
16
|
package edu.ucsb.nceas.metacat;
|
17
|
|
18
|
import java.io.*;
|
19
|
import java.util.Stack;
|
20
|
import java.util.Vector;
|
21
|
import java.util.Enumeration;
|
22
|
|
23
|
import org.xml.sax.Attributes;
|
24
|
import org.xml.sax.InputSource;
|
25
|
import org.xml.sax.SAXException;
|
26
|
import org.xml.sax.SAXParseException;
|
27
|
import org.xml.sax.XMLReader;
|
28
|
import org.xml.sax.helpers.XMLReaderFactory;
|
29
|
import org.xml.sax.helpers.DefaultHandler;
|
30
|
|
31
|
/**
|
32
|
* A Class that represents a structured query, and can be
|
33
|
* constructed from an XML serialization conforming to @see pathquery.dtd.
|
34
|
* The printSQL() method can be used to print a SQL serialization of the query.
|
35
|
*/
|
36
|
public class QuerySpecification extends DefaultHandler {
|
37
|
|
38
|
// Query data structures
|
39
|
private String meta_file_id;
|
40
|
private String querytitle;
|
41
|
private Vector doctypeList;
|
42
|
private QueryGroup query = null;
|
43
|
|
44
|
private Stack elementStack;
|
45
|
private Stack queryStack;
|
46
|
private String currentValue;
|
47
|
private String currentPathexpr;
|
48
|
private String parserName = null;
|
49
|
|
50
|
/**
|
51
|
* construct an instance of the QuerySpecification class
|
52
|
*
|
53
|
* @param queryspec the XML representation of the query (should conform
|
54
|
* to pathquery.dtd) as a Reader
|
55
|
* @param parserName the fully qualified name of a Java Class implementing
|
56
|
* the org.xml.sax.XMLReader interface
|
57
|
*/
|
58
|
public QuerySpecification( Reader queryspec, String parserName )
|
59
|
throws IOException {
|
60
|
super();
|
61
|
|
62
|
// Initialize the class variables
|
63
|
doctypeList = new Vector();
|
64
|
elementStack = new Stack();
|
65
|
queryStack = new Stack();
|
66
|
this.parserName = parserName;
|
67
|
|
68
|
// Initialize the parser and read the queryspec
|
69
|
XMLReader parser = initializeParser();
|
70
|
if (parser == null) {
|
71
|
System.err.println("SAX parser not instantiated properly.");
|
72
|
}
|
73
|
try {
|
74
|
parser.parse(new InputSource(queryspec));
|
75
|
} catch (SAXException e) {
|
76
|
System.err.println("error parsing data");
|
77
|
System.err.println(e.getMessage());
|
78
|
}
|
79
|
}
|
80
|
|
81
|
/**
|
82
|
* construct an instance of the QuerySpecification class
|
83
|
*
|
84
|
* @param queryspec the XML representation of the query (should conform
|
85
|
* to pathquery.dtd) as a String
|
86
|
* @param parserName the fully qualified name of a Java Class implementing
|
87
|
* the org.xml.sax.Parser interface
|
88
|
*/
|
89
|
public QuerySpecification( String queryspec, String parserName )
|
90
|
throws IOException {
|
91
|
this(new StringReader(queryspec), parserName);
|
92
|
}
|
93
|
|
94
|
/** Main routine for testing */
|
95
|
static public void main(String[] args) {
|
96
|
|
97
|
if (args.length < 1) {
|
98
|
System.err.println("Wrong number of arguments!!!");
|
99
|
System.err.println("USAGE: java QuerySpecification <xmlfile>");
|
100
|
return;
|
101
|
} else {
|
102
|
String xmlfile = args[0];
|
103
|
|
104
|
try {
|
105
|
MetaCatUtil util = new MetaCatUtil();
|
106
|
FileReader xml = new FileReader(new File(xmlfile));
|
107
|
QuerySpecification qspec =
|
108
|
new QuerySpecification(xml, util.getOption("saxparser"));
|
109
|
System.out.println(qspec.printSQL());
|
110
|
|
111
|
} catch (IOException e) {
|
112
|
System.err.println(e.getMessage());
|
113
|
}
|
114
|
|
115
|
}
|
116
|
}
|
117
|
|
118
|
/**
|
119
|
* Set up the SAX parser for reading the XML serialized query
|
120
|
*/
|
121
|
private XMLReader initializeParser() {
|
122
|
XMLReader parser = null;
|
123
|
|
124
|
// Set up the SAX document handlers for parsing
|
125
|
try {
|
126
|
|
127
|
// Get an instance of the parser
|
128
|
parser = XMLReaderFactory.createXMLReader(parserName);
|
129
|
|
130
|
// Set the ContentHandler to this instance
|
131
|
parser.setContentHandler(this);
|
132
|
|
133
|
// Set the error Handler to this instance
|
134
|
parser.setErrorHandler(this);
|
135
|
|
136
|
} catch (Exception e) {
|
137
|
System.err.println(e.toString());
|
138
|
}
|
139
|
|
140
|
return parser;
|
141
|
}
|
142
|
|
143
|
/**
|
144
|
* callback method used by the SAX Parser when the start tag of an
|
145
|
* element is detected. Used in this context to parse and store
|
146
|
* the query information in class variables.
|
147
|
*/
|
148
|
public void startElement (String uri, String localName,
|
149
|
String qName, Attributes atts)
|
150
|
throws SAXException {
|
151
|
BasicNode currentNode = new BasicNode(localName);
|
152
|
// add attributes to BasicNode here
|
153
|
if (atts != null) {
|
154
|
int len = atts.getLength();
|
155
|
for (int i = 0; i < len; i++) {
|
156
|
currentNode.setAttribute(atts.getLocalName(i), atts.getValue(i));
|
157
|
}
|
158
|
}
|
159
|
|
160
|
elementStack.push(currentNode);
|
161
|
if (currentNode.getTagName().equals("querygroup")) {
|
162
|
QueryGroup currentGroup = new QueryGroup(
|
163
|
currentNode.getAttribute("operator"));
|
164
|
if (query == null) {
|
165
|
query = currentGroup;
|
166
|
} else {
|
167
|
QueryGroup parentGroup = (QueryGroup)queryStack.peek();
|
168
|
parentGroup.addChild(currentGroup);
|
169
|
}
|
170
|
queryStack.push(currentGroup);
|
171
|
}
|
172
|
}
|
173
|
|
174
|
/**
|
175
|
* callback method used by the SAX Parser when the end tag of an
|
176
|
* element is detected. Used in this context to parse and store
|
177
|
* the query information in class variables.
|
178
|
*/
|
179
|
public void endElement (String uri, String localName,
|
180
|
String qName) throws SAXException {
|
181
|
BasicNode leaving = (BasicNode)elementStack.pop();
|
182
|
if (leaving.getTagName().equals("queryterm")) {
|
183
|
boolean isCaseSensitive = (new Boolean(
|
184
|
leaving.getAttribute("casesensitive"))).booleanValue();
|
185
|
QueryTerm currentTerm = null;
|
186
|
if (currentPathexpr == null) {
|
187
|
currentTerm = new QueryTerm(isCaseSensitive,
|
188
|
leaving.getAttribute("searchmode"),currentValue);
|
189
|
} else {
|
190
|
currentTerm = new QueryTerm(isCaseSensitive,
|
191
|
leaving.getAttribute("searchmode"),currentValue,
|
192
|
currentPathexpr);
|
193
|
}
|
194
|
QueryGroup currentGroup = (QueryGroup)queryStack.peek();
|
195
|
currentGroup.addChild(currentTerm);
|
196
|
currentValue = null;
|
197
|
currentPathexpr = null;
|
198
|
} else if (leaving.getTagName().equals("querygroup")) {
|
199
|
QueryGroup leavingGroup = (QueryGroup)queryStack.pop();
|
200
|
}
|
201
|
}
|
202
|
|
203
|
/**
|
204
|
* callback method used by the SAX Parser when the text sequences of an
|
205
|
* xml stream are detected. Used in this context to parse and store
|
206
|
* the query information in class variables.
|
207
|
*/
|
208
|
public void characters(char ch[], int start, int length) {
|
209
|
|
210
|
String inputString = new String(ch, start, length);
|
211
|
BasicNode currentNode = (BasicNode)elementStack.peek();
|
212
|
String currentTag = currentNode.getTagName();
|
213
|
if (currentTag.equals("meta_file_id")) {
|
214
|
meta_file_id = inputString;
|
215
|
} else if (currentTag.equals("querytitle")) {
|
216
|
querytitle = inputString;
|
217
|
} else if (currentTag.equals("value")) {
|
218
|
currentValue = inputString;
|
219
|
} else if (currentTag.equals("pathexpr")) {
|
220
|
currentPathexpr = inputString;
|
221
|
} else if (currentTag.equals("returndoctype")) {
|
222
|
doctypeList.add(inputString);
|
223
|
}
|
224
|
}
|
225
|
|
226
|
|
227
|
/**
|
228
|
* create a SQL serialization of the query that this instance represents
|
229
|
*/
|
230
|
public String printSQL() {
|
231
|
StringBuffer self = new StringBuffer();
|
232
|
|
233
|
// This determines the returned results
|
234
|
self.append("SELECT docid,docname,doctype,doctitle ");
|
235
|
self.append("FROM xml_documents WHERE docid IN (");
|
236
|
|
237
|
// This determines the documents that meet the query conditions
|
238
|
self.append(query.printSQL());
|
239
|
|
240
|
self.append(") ");
|
241
|
|
242
|
// Add SQL to filter for doctypes requested in the query
|
243
|
if (!doctypeList.isEmpty()) {
|
244
|
boolean firstdoctype = true;
|
245
|
self.append(" AND (");
|
246
|
Enumeration en = doctypeList.elements();
|
247
|
while (en.hasMoreElements()) {
|
248
|
String currentDoctype = (String)en.nextElement();
|
249
|
if (firstdoctype) {
|
250
|
firstdoctype = false;
|
251
|
self.append(" doctype = '" + currentDoctype + "'");
|
252
|
} else {
|
253
|
self.append(" OR doctype = '" + currentDoctype + "'");
|
254
|
}
|
255
|
}
|
256
|
self.append(") ");
|
257
|
}
|
258
|
|
259
|
return self.toString();
|
260
|
}
|
261
|
|
262
|
/**
|
263
|
* create a String description of the query that this instance represents.
|
264
|
* This should become a way to get the XML serialization of the query.
|
265
|
*/
|
266
|
public String toString() {
|
267
|
return "meta_file_id=" + meta_file_id + "\n" +
|
268
|
"querytitle=" + querytitle + "\n" + query;
|
269
|
}
|
270
|
|
271
|
/** a utility class that represents a group of terms in a query */
|
272
|
private class QueryGroup {
|
273
|
private String operator = null; // indicates how query terms are combined
|
274
|
private Vector children = null; // the list of query terms and groups
|
275
|
|
276
|
/**
|
277
|
* construct a new QueryGroup
|
278
|
*
|
279
|
* @param operator the boolean conector used to connect query terms
|
280
|
* in this query group
|
281
|
*/
|
282
|
public QueryGroup(String operator) {
|
283
|
this.operator = operator;
|
284
|
children = new Vector();
|
285
|
}
|
286
|
|
287
|
/**
|
288
|
* Add a child QueryGroup to this QueryGroup
|
289
|
*
|
290
|
* @param qgroup the query group to be added to the list of terms
|
291
|
*/
|
292
|
public void addChild(QueryGroup qgroup) {
|
293
|
children.add((Object)qgroup);
|
294
|
}
|
295
|
|
296
|
/**
|
297
|
* Add a child QueryTerm to this QueryGroup
|
298
|
*
|
299
|
* @param qterm the query term to be added to the list of terms
|
300
|
*/
|
301
|
public void addChild(QueryTerm qterm) {
|
302
|
children.add((Object)qterm);
|
303
|
}
|
304
|
|
305
|
/**
|
306
|
* Retrieve an Enumeration of query terms for this QueryGroup
|
307
|
*/
|
308
|
public Enumeration getChildren() {
|
309
|
return children.elements();
|
310
|
}
|
311
|
|
312
|
/**
|
313
|
* create a SQL serialization of the query that this instance represents
|
314
|
*/
|
315
|
public String printSQL() {
|
316
|
StringBuffer self = new StringBuffer();
|
317
|
boolean first = true;
|
318
|
|
319
|
self.append("(");
|
320
|
|
321
|
Enumeration en= getChildren();
|
322
|
while (en.hasMoreElements()) {
|
323
|
Object qobject = en.nextElement();
|
324
|
if (first) {
|
325
|
first = false;
|
326
|
} else {
|
327
|
self.append(" " + operator + " ");
|
328
|
}
|
329
|
if (qobject instanceof QueryGroup) {
|
330
|
QueryGroup qg = (QueryGroup)qobject;
|
331
|
self.append(qg.printSQL());
|
332
|
} else if (qobject instanceof QueryTerm) {
|
333
|
QueryTerm qt = (QueryTerm)qobject;
|
334
|
self.append(qt.printSQL());
|
335
|
} else {
|
336
|
System.err.println("qobject wrong type: fatal error");
|
337
|
}
|
338
|
}
|
339
|
self.append(") \n");
|
340
|
return self.toString();
|
341
|
}
|
342
|
|
343
|
/**
|
344
|
* create a String description of the query that this instance represents.
|
345
|
* This should become a way to get the XML serialization of the query.
|
346
|
*/
|
347
|
public String toString() {
|
348
|
StringBuffer self = new StringBuffer();
|
349
|
|
350
|
self.append(" (Query group operator=" + operator + "\n");
|
351
|
Enumeration en= getChildren();
|
352
|
while (en.hasMoreElements()) {
|
353
|
Object qobject = en.nextElement();
|
354
|
self.append(qobject);
|
355
|
}
|
356
|
self.append(" )\n");
|
357
|
return self.toString();
|
358
|
}
|
359
|
}
|
360
|
|
361
|
/** a utility class that represents a single term in a query */
|
362
|
private class QueryTerm {
|
363
|
private boolean casesensitive = false;
|
364
|
private String searchmode = null;
|
365
|
private String value = null;
|
366
|
private String pathexpr = null;
|
367
|
|
368
|
/**
|
369
|
* Construct a new instance of a query term for a free text search
|
370
|
* (using the value only)
|
371
|
*
|
372
|
* @param casesensitive flag indicating whether case is used to match
|
373
|
* @param searchmode determines what kind of substring match is performed
|
374
|
* (one of starts-with|ends-with|contains|matches-exactly)
|
375
|
* @param value the text value to match
|
376
|
*/
|
377
|
public QueryTerm(boolean casesensitive, String searchmode,
|
378
|
String value) {
|
379
|
this.casesensitive = casesensitive;
|
380
|
this.searchmode = searchmode;
|
381
|
this.value = value;
|
382
|
}
|
383
|
|
384
|
/**
|
385
|
* Construct a new instance of a query term for a structured search
|
386
|
* (matching the value only for those nodes in the pathexpr)
|
387
|
*
|
388
|
* @param casesensitive flag indicating whether case is used to match
|
389
|
* @param searchmode determines what kind of substring match is performed
|
390
|
* (one of starts-with|ends-with|contains|matches-exactly)
|
391
|
* @param value the text value to match
|
392
|
* @param pathexpr the hierarchical path to the nodes to be searched
|
393
|
*/
|
394
|
public QueryTerm(boolean casesensitive, String searchmode,
|
395
|
String value, String pathexpr) {
|
396
|
this(casesensitive, searchmode, value);
|
397
|
this.pathexpr = pathexpr;
|
398
|
}
|
399
|
|
400
|
/** determine if the QueryTerm is case sensitive */
|
401
|
public boolean isCaseSensitive() {
|
402
|
return casesensitive;
|
403
|
}
|
404
|
|
405
|
/** get the searchmode parameter */
|
406
|
public String getSearchMode() {
|
407
|
return searchmode;
|
408
|
}
|
409
|
|
410
|
/** get the Value parameter */
|
411
|
public String getValue() {
|
412
|
return value;
|
413
|
}
|
414
|
|
415
|
/** get the path expression parameter */
|
416
|
public String getPathExpression() {
|
417
|
return pathexpr;
|
418
|
}
|
419
|
|
420
|
/**
|
421
|
* create a SQL serialization of the query that this instance represents
|
422
|
*/
|
423
|
public String printSQL() {
|
424
|
StringBuffer self = new StringBuffer();
|
425
|
|
426
|
// Uppercase the search string if case match is not important
|
427
|
String casevalue = null;
|
428
|
String nodedataterm = null;
|
429
|
|
430
|
if (casesensitive) {
|
431
|
nodedataterm = "nodedata";
|
432
|
casevalue = value;
|
433
|
} else {
|
434
|
nodedataterm = "UPPER(nodedata)";
|
435
|
casevalue = value.toUpperCase();
|
436
|
}
|
437
|
|
438
|
// Add appropriate wildcards to search string
|
439
|
String searchvalue = null;
|
440
|
if (searchmode.equals("starts-with")) {
|
441
|
searchvalue = casevalue + "%";
|
442
|
} else if (searchmode.equals("ends-with")) {
|
443
|
searchvalue = "%" + casevalue;
|
444
|
} else if (searchmode.equals("contains")) {
|
445
|
searchvalue = "%" + casevalue + "%";
|
446
|
} else {
|
447
|
searchvalue = casevalue;
|
448
|
}
|
449
|
|
450
|
self.append("SELECT DISTINCT docid FROM xml_nodes WHERE \n");
|
451
|
|
452
|
if (pathexpr != null) {
|
453
|
self.append(nodedataterm + " LIKE " + "'" + searchvalue + "' ");
|
454
|
self.append("AND parentnodeid IN ");
|
455
|
self.append("(SELECT nodeid FROM xml_index WHERE path LIKE " +
|
456
|
"'" + pathexpr + "') " );
|
457
|
} else {
|
458
|
self.append(nodedataterm + " LIKE " + "'" + searchvalue + "' ");
|
459
|
}
|
460
|
|
461
|
return self.toString();
|
462
|
}
|
463
|
|
464
|
/**
|
465
|
* create a String description of the query that this instance represents.
|
466
|
* This should become a way to get the XML serialization of the query.
|
467
|
*/
|
468
|
public String toString() {
|
469
|
StringBuffer self = new StringBuffer();
|
470
|
|
471
|
self.append(" Query Term iscasesensitive=" + casesensitive + "\n");
|
472
|
self.append(" searchmode=" + searchmode + "\n");
|
473
|
self.append(" value=" + value + "\n");
|
474
|
if (pathexpr != null) {
|
475
|
self.append(" pathexpr=" + pathexpr + "\n");
|
476
|
}
|
477
|
|
478
|
return self.toString();
|
479
|
}
|
480
|
}
|
481
|
}
|
482
|
|
483
|
/**
|
484
|
* '$Log$
|
485
|
* 'Revision 1.9.2.3 2000/06/25 23:38:17 jones
|
486
|
* 'Added RCSfile keyword
|
487
|
* '
|
488
|
* 'Revision 1.9.2.2 2000/06/25 23:34:18 jones
|
489
|
* 'Changed documentation formatting, added log entries at bottom of source files
|
490
|
* ''
|
491
|
*/
|