1
|
/**
|
2
|
* '$RCSfile$'
|
3
|
* Purpose: A Class that handles the SAX XML events as they
|
4
|
* are generated from XML documents
|
5
|
* Copyright: 2000 Regents of the University of California and the
|
6
|
* National Center for Ecological Analysis and Synthesis
|
7
|
* Authors: Matt Jones, Jivka Bojilova
|
8
|
* Release: @release@
|
9
|
*
|
10
|
* '$Author: jones $'
|
11
|
* '$Date: 2001-01-18 11:52:00 -0800 (Thu, 18 Jan 2001) $'
|
12
|
* '$Revision: 669 $'
|
13
|
*
|
14
|
* This program is free software; you can redistribute it and/or modify
|
15
|
* it under the terms of the GNU General Public License as published by
|
16
|
* the Free Software Foundation; either version 2 of the License, or
|
17
|
* (at your option) any later version.
|
18
|
*
|
19
|
* This program is distributed in the hope that it will be useful,
|
20
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
21
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
22
|
* GNU General Public License for more details.
|
23
|
*
|
24
|
* You should have received a copy of the GNU General Public License
|
25
|
* along with this program; if not, write to the Free Software
|
26
|
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
|
27
|
*/
|
28
|
|
29
|
package edu.ucsb.nceas.metacat;
|
30
|
|
31
|
import java.sql.*;
|
32
|
import java.io.StringReader;
|
33
|
import java.util.Stack;
|
34
|
import java.util.Vector;
|
35
|
import java.util.Enumeration;
|
36
|
import java.util.EmptyStackException;
|
37
|
|
38
|
import org.xml.sax.Attributes;
|
39
|
import org.xml.sax.SAXException;
|
40
|
import org.xml.sax.SAXParseException;
|
41
|
import org.xml.sax.ext.DeclHandler;
|
42
|
import org.xml.sax.ext.LexicalHandler;
|
43
|
import org.xml.sax.helpers.DefaultHandler;
|
44
|
|
45
|
/**
|
46
|
* A database aware Class implementing callback bethods for the SAX parser to
|
47
|
* call when processing the XML stream and generating events
|
48
|
*/
|
49
|
public class DBSAXHandler extends DefaultHandler
|
50
|
implements LexicalHandler, DeclHandler, Runnable {
|
51
|
|
52
|
private boolean atFirstElement;
|
53
|
private boolean processingDTD;
|
54
|
private String docname = null;
|
55
|
private String doctype;
|
56
|
private String systemid;
|
57
|
private boolean stackCreated = false;
|
58
|
private Stack nodeStack;
|
59
|
private Vector nodeIndex;
|
60
|
private Connection conn = null;
|
61
|
private DocumentImpl currentDocument;
|
62
|
private DBSAXNode rootNode;
|
63
|
private String action = null;
|
64
|
private String docid = null;
|
65
|
private String user = null;
|
66
|
private String group = null;
|
67
|
private Thread xmlIndex;
|
68
|
private boolean endDocument = false;
|
69
|
private int serverCode = 1;
|
70
|
|
71
|
private static final int MAXDATACHARS = 4000;
|
72
|
private static final int MAXTITLELEN = 1000;
|
73
|
|
74
|
/** Construct an instance of the handler class
|
75
|
*
|
76
|
* @param conn the JDBC connection to which information is written
|
77
|
*/
|
78
|
public DBSAXHandler(Connection conn) {
|
79
|
this.conn = conn;
|
80
|
this.atFirstElement = true;
|
81
|
this.processingDTD = false;
|
82
|
|
83
|
// Create the stack for keeping track of node context
|
84
|
// if it doesn't already exist
|
85
|
if (!stackCreated) {
|
86
|
nodeStack = new Stack();
|
87
|
nodeIndex = new Vector();
|
88
|
stackCreated = true;
|
89
|
}
|
90
|
}
|
91
|
|
92
|
public DBSAXHandler(Connection conn, String action, String docid,
|
93
|
String user, String group, int serverCode)
|
94
|
{
|
95
|
this(conn);
|
96
|
this.action = action;
|
97
|
this.docid = docid;
|
98
|
this.user = user;
|
99
|
this.group = group;
|
100
|
this.xmlIndex = new Thread(this);
|
101
|
this.serverCode = serverCode;
|
102
|
}
|
103
|
|
104
|
/** Construct an instance of the handler class
|
105
|
*
|
106
|
* @param conn the JDBC connection to which information is written
|
107
|
* @param action - "INSERT" or "UPDATE"
|
108
|
* @param docid to be inserted or updated into JDBC connection
|
109
|
* @param user the user connected to MetaCat servlet
|
110
|
*/
|
111
|
public DBSAXHandler(Connection conn,String action,String docid,
|
112
|
String user, String group)
|
113
|
{
|
114
|
this(conn);
|
115
|
this.action = action;
|
116
|
this.docid = docid;
|
117
|
this.user = user;
|
118
|
this.group = group;
|
119
|
this.xmlIndex = new Thread(this);
|
120
|
//this.xmlIndex.setPriority(Thread.MIN_PRIORITY);
|
121
|
}
|
122
|
|
123
|
/** SAX Handler that receives notification of beginning of the document */
|
124
|
public void startDocument() throws SAXException {
|
125
|
MetaCatUtil.debugMessage("start Document");
|
126
|
|
127
|
// Create the document node representation as root
|
128
|
rootNode = new DBSAXNode(conn, this.docid);
|
129
|
// Add the node to the stack, so that any text data can be
|
130
|
// added as it is encountered
|
131
|
nodeStack.push(rootNode);
|
132
|
}
|
133
|
|
134
|
/** SAX Handler that receives notification of end of the document */
|
135
|
public void endDocument() throws SAXException {
|
136
|
MetaCatUtil.debugMessage("end Document");
|
137
|
// Starting new thread for writing XML Index.
|
138
|
// It calls the run method of the thread.
|
139
|
try {
|
140
|
xmlIndex.start();
|
141
|
} catch (NullPointerException e) {
|
142
|
xmlIndex = null;
|
143
|
throw new
|
144
|
SAXException("Problem with starting thread for writing XML Index. " +
|
145
|
e.getMessage());
|
146
|
}
|
147
|
}
|
148
|
|
149
|
/** SAX Handler that is called at the start of each XML element */
|
150
|
public void startElement(String uri, String localName,
|
151
|
String qName, Attributes atts)
|
152
|
throws SAXException {
|
153
|
MetaCatUtil.debugMessage("Start ELEMENT " + localName);
|
154
|
|
155
|
DBSAXNode parentNode = null;
|
156
|
DBSAXNode currentNode = null;
|
157
|
|
158
|
// Get a reference to the parent node for the id
|
159
|
try {
|
160
|
parentNode = (DBSAXNode)nodeStack.peek();
|
161
|
} catch (EmptyStackException e) {
|
162
|
parentNode = null;
|
163
|
}
|
164
|
|
165
|
// Document representation that points to the root document node
|
166
|
if (atFirstElement) {
|
167
|
atFirstElement = false;
|
168
|
// If no DOCTYPE declaration: docname = root element name
|
169
|
if (docname == null) {
|
170
|
docname = localName;
|
171
|
doctype = docname;
|
172
|
MetaCatUtil.debugMessage("DOCNAME-a: " + docname);
|
173
|
MetaCatUtil.debugMessage("DOCTYPE-a: " + doctype);
|
174
|
} else if (doctype == null) {
|
175
|
doctype = docname;
|
176
|
MetaCatUtil.debugMessage("DOCTYPE-b: " + doctype);
|
177
|
}
|
178
|
rootNode.writeNodename(docname);
|
179
|
try {
|
180
|
currentDocument = new DocumentImpl(conn, rootNode.getNodeID(),
|
181
|
docname, doctype, docid, action, user,
|
182
|
this.serverCode);
|
183
|
} catch (Exception ane) {
|
184
|
throw (new SAXException("Error with " + action, ane));
|
185
|
}
|
186
|
// not needed any more
|
187
|
//rootNode.writeDocID(currentDocument.getDocID());
|
188
|
}
|
189
|
|
190
|
// Create the current node representation
|
191
|
currentNode = new DBSAXNode(conn, localName, parentNode,
|
192
|
currentDocument.getRootNodeID(),docid,
|
193
|
currentDocument.getDoctype());
|
194
|
|
195
|
// Add all of the attributes
|
196
|
for (int i=0; i<atts.getLength(); i++) {
|
197
|
currentNode.setAttribute(atts.getLocalName(i), atts.getValue(i), docid);
|
198
|
}
|
199
|
|
200
|
// Add the node to the stack, so that any text data can be
|
201
|
// added as it is encountered
|
202
|
nodeStack.push(currentNode);
|
203
|
// Add the node to the vector used by thread for writing XML Index
|
204
|
nodeIndex.addElement(currentNode);
|
205
|
|
206
|
}
|
207
|
|
208
|
// The run method of xmlIndex thread. It writes XML Index for the document.
|
209
|
public void run () {
|
210
|
DBSAXNode currNode = null;
|
211
|
DBSAXNode prevNode = null;
|
212
|
Connection dbconn = null;
|
213
|
String doctype = currentDocument.getDoctype();
|
214
|
int step = 0;
|
215
|
int counter = 0;
|
216
|
|
217
|
try {
|
218
|
// Opening separate db connection for writing XML Index
|
219
|
MetaCatUtil util = new MetaCatUtil();
|
220
|
dbconn = util.openDBConnection();
|
221
|
dbconn.setAutoCommit(false);
|
222
|
|
223
|
// Going through the elements of the document and writing its Index
|
224
|
Enumeration nodes = nodeIndex.elements();
|
225
|
while ( nodes.hasMoreElements() ) {
|
226
|
currNode = (DBSAXNode)nodes.nextElement();
|
227
|
currNode.updateNodeIndex(dbconn, docid, doctype);
|
228
|
}
|
229
|
|
230
|
dbconn.commit();
|
231
|
|
232
|
//if this is a package file then write the package info to
|
233
|
//the xml_relation table. RelationHandler checks to see
|
234
|
//if it is a package file so you don't have to do it here.
|
235
|
if ( doctype.equals(util.getOption("packagedoctype")) )
|
236
|
{
|
237
|
DocumentImpl xmldoc = new DocumentImpl(dbconn, docid);
|
238
|
RelationHandler rth = new RelationHandler(xmldoc, dbconn);
|
239
|
}
|
240
|
else if ( doctype.equals(util.getOption("accessdoctype")) )
|
241
|
{
|
242
|
DocumentImpl xmldoc = new DocumentImpl(dbconn, docid);
|
243
|
String xml = xmldoc.toString();
|
244
|
try {
|
245
|
AccessControlList aclobj =
|
246
|
new AccessControlList(dbconn, docid, new StringReader(xml),
|
247
|
user, group);
|
248
|
dbconn.commit();
|
249
|
} catch (SAXException e) {
|
250
|
try {
|
251
|
dbconn.rollback();
|
252
|
dbconn.close();
|
253
|
} catch (SQLException sqle) {}
|
254
|
System.out.println("Error writing ACL. " + e.getMessage());
|
255
|
}
|
256
|
}
|
257
|
|
258
|
|
259
|
dbconn.close();
|
260
|
|
261
|
} catch (Exception e) {
|
262
|
try {
|
263
|
dbconn.rollback();
|
264
|
dbconn.close();
|
265
|
} catch (SQLException sqle) {}
|
266
|
System.out.println("Error writing XML Index. " + e.getMessage());
|
267
|
}
|
268
|
}
|
269
|
|
270
|
/** SAX Handler that is called for each XML text node */
|
271
|
public void characters(char[] cbuf, int start, int len) throws SAXException {
|
272
|
MetaCatUtil.debugMessage("CHARACTERS");
|
273
|
DBSAXNode currentNode = (DBSAXNode)nodeStack.peek();
|
274
|
String data = null;
|
275
|
int leftover = len;
|
276
|
int offset = start;
|
277
|
boolean moredata = true;
|
278
|
|
279
|
// This loop deals with the case where there are more characters
|
280
|
// than can fit in a single database text field (limit is
|
281
|
// MAXDATACHARS). If the text to be inserted exceeds MAXDATACHARS,
|
282
|
// write a series of nodes that are MAXDATACHARS long, and then the
|
283
|
// final node contains the remainder
|
284
|
while (moredata) {
|
285
|
if (leftover > MAXDATACHARS) {
|
286
|
data = new String(cbuf, offset, MAXDATACHARS);
|
287
|
leftover -= MAXDATACHARS;
|
288
|
offset += MAXDATACHARS;
|
289
|
} else {
|
290
|
data = new String(cbuf, offset, leftover);
|
291
|
moredata = false;
|
292
|
}
|
293
|
|
294
|
// Write the content of the node to the database
|
295
|
currentNode.writeChildNodeToDB("TEXT", null, data, docid);
|
296
|
}
|
297
|
|
298
|
// write the title of document if there are tag <title>
|
299
|
if ( currentNode.getTagName().equals("title") ) {
|
300
|
if ( leftover > MAXTITLELEN )
|
301
|
currentDocument.setTitle(new String(cbuf, start, MAXTITLELEN));
|
302
|
else
|
303
|
currentDocument.setTitle(new String(cbuf, start, leftover));
|
304
|
}
|
305
|
}
|
306
|
|
307
|
/**
|
308
|
* SAX Handler that is called for each XML text node that is
|
309
|
* Ignorable white space
|
310
|
*/
|
311
|
public void ignorableWhitespace(char[] cbuf, int start, int len)
|
312
|
throws SAXException {
|
313
|
// When validation is turned "on", white spaces are reported here
|
314
|
// When validation is turned "off" white spaces are not reported here,
|
315
|
// but through characters() callback
|
316
|
MetaCatUtil.debugMessage("IGNORABLEWHITESPACE");
|
317
|
//System.out.println("Whitespace:" + len + "x" + new String(cbuf, start, len) + "x");
|
318
|
|
319
|
DBSAXNode currentNode = (DBSAXNode)nodeStack.peek();
|
320
|
String data = null;
|
321
|
int leftover = len;
|
322
|
int offset = start;
|
323
|
boolean moredata = true;
|
324
|
|
325
|
// This loop deals with the case where there are more characters
|
326
|
// than can fit in a single database text field (limit is
|
327
|
// MAXDATACHARS). If the text to be inserted exceeds MAXDATACHARS,
|
328
|
// write a series of nodes that are MAXDATACHARS long, and then the
|
329
|
// final node contains the remainder
|
330
|
while (moredata) {
|
331
|
if (leftover > MAXDATACHARS) {
|
332
|
data = new String(cbuf, offset, MAXDATACHARS);
|
333
|
leftover -= MAXDATACHARS;
|
334
|
offset += MAXDATACHARS;
|
335
|
} else {
|
336
|
data = new String(cbuf, offset, leftover);
|
337
|
moredata = false;
|
338
|
}
|
339
|
|
340
|
// Write the content of the node to the database
|
341
|
currentNode.writeChildNodeToDB("TEXT", null, data, docid);
|
342
|
}
|
343
|
}
|
344
|
|
345
|
/**
|
346
|
* SAX Handler called once for each processing instruction found:
|
347
|
* node that PI may occur before or after the root element.
|
348
|
*/
|
349
|
public void processingInstruction(String target, String data)
|
350
|
throws SAXException {
|
351
|
MetaCatUtil.debugMessage("PI");
|
352
|
DBSAXNode currentNode = (DBSAXNode)nodeStack.peek();
|
353
|
currentNode.writeChildNodeToDB("PI", target, data, docid);
|
354
|
}
|
355
|
|
356
|
/** SAX Handler that is called at the end of each XML element */
|
357
|
public void endElement(String uri, String localName,
|
358
|
String qName) throws SAXException {
|
359
|
MetaCatUtil.debugMessage("End ELEMENT " + localName);
|
360
|
|
361
|
// Get the node from the stack
|
362
|
DBSAXNode currentNode = (DBSAXNode)nodeStack.pop();
|
363
|
}
|
364
|
|
365
|
//
|
366
|
// the next section implements the LexicalHandler interface
|
367
|
//
|
368
|
|
369
|
/** SAX Handler that receives notification of DOCTYPE. Sets the DTD */
|
370
|
public void startDTD(String name, String publicId, String systemId)
|
371
|
throws SAXException {
|
372
|
docname = name;
|
373
|
doctype = publicId;
|
374
|
systemid = systemId;
|
375
|
|
376
|
MetaCatUtil.debugMessage("Start DTD");
|
377
|
MetaCatUtil.debugMessage("DOCNAME: " + docname);
|
378
|
MetaCatUtil.debugMessage("DOCTYPE: " + doctype);
|
379
|
MetaCatUtil.debugMessage(" SYSID: " + systemid);
|
380
|
}
|
381
|
|
382
|
/**
|
383
|
* SAX Handler that receives notification of end of DTD
|
384
|
*/
|
385
|
public void endDTD() throws SAXException {
|
386
|
MetaCatUtil.debugMessage("end DTD");
|
387
|
}
|
388
|
|
389
|
/**
|
390
|
* SAX Handler that receives notification of comments in the DTD
|
391
|
*/
|
392
|
public void comment(char[] ch, int start, int length) throws SAXException {
|
393
|
MetaCatUtil.debugMessage("COMMENT");
|
394
|
if ( !processingDTD ) {
|
395
|
DBSAXNode currentNode = (DBSAXNode)nodeStack.peek();
|
396
|
currentNode.writeChildNodeToDB("COMMENT", null, new String(ch), docid);
|
397
|
}
|
398
|
}
|
399
|
|
400
|
/**
|
401
|
* SAX Handler that receives notification of the start of CDATA sections
|
402
|
*/
|
403
|
public void startCDATA() throws SAXException {
|
404
|
MetaCatUtil.debugMessage("start CDATA");
|
405
|
}
|
406
|
|
407
|
/**
|
408
|
* SAX Handler that receives notification of the end of CDATA sections
|
409
|
*/
|
410
|
public void endCDATA() throws SAXException {
|
411
|
MetaCatUtil.debugMessage("end CDATA");
|
412
|
}
|
413
|
|
414
|
/**
|
415
|
* SAX Handler that receives notification of the start of entities
|
416
|
*/
|
417
|
public void startEntity(String name) throws SAXException {
|
418
|
MetaCatUtil.debugMessage("start ENTITY: " + name);
|
419
|
if (name.equals("[dtd]")) {
|
420
|
processingDTD = true;
|
421
|
}
|
422
|
}
|
423
|
|
424
|
/**
|
425
|
* SAX Handler that receives notification of the end of entities
|
426
|
*/
|
427
|
public void endEntity(String name) throws SAXException {
|
428
|
MetaCatUtil.debugMessage("end ENTITY: " + name);
|
429
|
if (name.equals("[dtd]")) {
|
430
|
processingDTD = false;
|
431
|
}
|
432
|
}
|
433
|
|
434
|
/**
|
435
|
* SAX Handler that receives notification of element declarations
|
436
|
*/
|
437
|
public void elementDecl(String name, String model)
|
438
|
throws org.xml.sax.SAXException {
|
439
|
MetaCatUtil.debugMessage("ELEMENTDECL: " + name + " " + model);
|
440
|
}
|
441
|
|
442
|
/**
|
443
|
* SAX Handler that receives notification of attribute declarations
|
444
|
*/
|
445
|
public void attributeDecl(String eName, String aName,
|
446
|
String type, String valueDefault, String value)
|
447
|
throws org.xml.sax.SAXException {
|
448
|
MetaCatUtil.debugMessage("ATTRIBUTEDECL: " + eName + " "
|
449
|
+ aName + " " + type + " " + valueDefault + " "
|
450
|
+ value);
|
451
|
}
|
452
|
|
453
|
/**
|
454
|
* SAX Handler that receives notification of internal entity declarations
|
455
|
*/
|
456
|
public void internalEntityDecl(String name, String value)
|
457
|
throws org.xml.sax.SAXException {
|
458
|
MetaCatUtil.debugMessage("INTERNENTITYDECL: " + name + " " + value);
|
459
|
}
|
460
|
|
461
|
/**
|
462
|
* SAX Handler that receives notification of external entity declarations
|
463
|
*/
|
464
|
public void externalEntityDecl(String name, String publicId,
|
465
|
String systemId)
|
466
|
throws org.xml.sax.SAXException {
|
467
|
MetaCatUtil.debugMessage("EXTERNENTITYDECL: " + name + " " + publicId
|
468
|
+ " " + systemId);
|
469
|
}
|
470
|
|
471
|
//
|
472
|
// the next section implements the ErrorHandler interface
|
473
|
//
|
474
|
|
475
|
/**
|
476
|
* SAX Handler that receives notification of fatal parsing errors
|
477
|
*/
|
478
|
public void fatalError(SAXParseException exception) throws SAXException {
|
479
|
MetaCatUtil.debugMessage("FATALERROR");
|
480
|
throw (new SAXException("Fatal processing error.", exception));
|
481
|
}
|
482
|
|
483
|
/**
|
484
|
* SAX Handler that receives notification of recoverable parsing errors
|
485
|
*/
|
486
|
public void error(SAXParseException exception) throws SAXException {
|
487
|
MetaCatUtil.debugMessage("ERROR");
|
488
|
throw (new SAXException("Processing error.", exception));
|
489
|
}
|
490
|
|
491
|
/**
|
492
|
* SAX Handler that receives notification of warnings
|
493
|
*/
|
494
|
public void warning(SAXParseException exception) throws SAXException {
|
495
|
MetaCatUtil.debugMessage("WARNING");
|
496
|
throw (new SAXException("Warning.", exception));
|
497
|
}
|
498
|
|
499
|
//
|
500
|
// Helper, getter and setter methods
|
501
|
//
|
502
|
|
503
|
/**
|
504
|
* get the document name
|
505
|
*/
|
506
|
public String getDocname() {
|
507
|
return docname;
|
508
|
}
|
509
|
|
510
|
/**
|
511
|
* get the document processing state
|
512
|
*/
|
513
|
public boolean processingDTD() {
|
514
|
return processingDTD;
|
515
|
}
|
516
|
}
|