1
|
/**
|
2
|
* '$RCSfile$'
|
3
|
* Purpose: A class that gets Accession Number, check for uniqueness
|
4
|
* and register it into db
|
5
|
* Copyright: 2000 Regents of the University of California and the
|
6
|
* National Center for Ecological Analysis and Synthesis
|
7
|
* Authors: Jivka Bojilova, Matt Jones
|
8
|
*
|
9
|
* '$Author: leinfelder $'
|
10
|
* '$Date: 2011-11-02 20:40:12 -0700 (Wed, 02 Nov 2011) $'
|
11
|
* '$Revision: 6595 $'
|
12
|
*
|
13
|
* This program is free software; you can redistribute it and/or modify
|
14
|
* it under the terms of the GNU General Public License as published by
|
15
|
* the Free Software Foundation; either version 2 of the License, or
|
16
|
* (at your option) any later version.
|
17
|
*
|
18
|
* This program is distributed in the hope that it will be useful,
|
19
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
20
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
21
|
* GNU General Public License for more details.
|
22
|
*
|
23
|
* You should have received a copy of the GNU General Public License
|
24
|
* along with this program; if not, write to the Free Software
|
25
|
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
|
26
|
*/
|
27
|
package edu.ucsb.nceas.metacat.index;
|
28
|
|
29
|
import java.io.FileInputStream;
|
30
|
import java.io.FileNotFoundException;
|
31
|
import java.io.InputStream;
|
32
|
import java.util.ArrayList;
|
33
|
import java.util.Date;
|
34
|
import java.util.List;
|
35
|
import java.util.TimerTask;
|
36
|
import java.util.Vector;
|
37
|
|
38
|
import org.apache.commons.logging.Log;
|
39
|
import org.apache.commons.logging.LogFactory;
|
40
|
import org.apache.solr.client.solrj.SolrServerException;
|
41
|
import org.dataone.client.MNode;
|
42
|
import org.dataone.configuration.Settings;
|
43
|
import org.dataone.service.exceptions.InvalidRequest;
|
44
|
import org.dataone.service.exceptions.InvalidToken;
|
45
|
import org.dataone.service.exceptions.NotAuthorized;
|
46
|
import org.dataone.service.exceptions.NotImplemented;
|
47
|
import org.dataone.service.exceptions.ServiceFailure;
|
48
|
import org.dataone.service.types.v1.Identifier;
|
49
|
import org.dataone.service.types.v1.ObjectFormatIdentifier;
|
50
|
import org.dataone.service.types.v1.ObjectInfo;
|
51
|
import org.dataone.service.types.v1.ObjectList;
|
52
|
import org.dataone.service.types.v1.SystemMetadata;
|
53
|
|
54
|
import com.hazelcast.core.IMap;
|
55
|
import com.hazelcast.core.ISet;
|
56
|
|
57
|
|
58
|
/**
|
59
|
* A class represents the object to generate massive solr indexes.
|
60
|
* This can happen during an update of Metacat (generating index for all existing documents)
|
61
|
* or regenerate index for those documents
|
62
|
* failing to build index during the insert or update.
|
63
|
*
|
64
|
* @author tao
|
65
|
*
|
66
|
*/
|
67
|
public class IndexGenerator extends TimerTask {
|
68
|
|
69
|
private static final int FIRST =0;
|
70
|
private static final int SECOND =1;
|
71
|
public static final int WAITTIME = 10000;
|
72
|
public static final int MAXWAITNUMBER = 180;
|
73
|
private static final String HTTP = "http://";
|
74
|
private static final String MNAPPENDIX = "/d1/mn";
|
75
|
private static final String RESOURCEMAPPROPERYNAME = "index.resourcemap.namespace";
|
76
|
public static final String WAITIMEPOPERTYNAME = "index.regenerate.start.waitingtime";
|
77
|
public static final String MAXATTEMPTSPROPERTYNAME = "index.regenerate.start.maxattempts";
|
78
|
|
79
|
private static int waitingTime = WAITTIME;
|
80
|
private static int maxAttempts = MAXWAITNUMBER;
|
81
|
|
82
|
private SolrIndex solrIndex = null;
|
83
|
//private SystemMetadataEventListener systemMetadataListener = null;
|
84
|
private IMap<Identifier, SystemMetadata> systemMetadataMap;
|
85
|
private IMap<Identifier, String> objectPathMap;
|
86
|
private Log log = LogFactory.getLog(IndexGenerator.class);
|
87
|
//private MNode mNode = null;
|
88
|
private static List<String> resourceMapNamespaces = null;
|
89
|
|
90
|
/**
|
91
|
* Constructor
|
92
|
* @param solrIndex
|
93
|
* @param systemMetadataListener
|
94
|
*/
|
95
|
public IndexGenerator(SolrIndex solrIndex) {
|
96
|
this.solrIndex = solrIndex;
|
97
|
resourceMapNamespaces = Settings.getConfiguration().getList(RESOURCEMAPPROPERYNAME);
|
98
|
//this.systemMetadataListener = systemMetadataListener;
|
99
|
//this.mNode = new MNode(buildMNBaseURL());
|
100
|
try {
|
101
|
waitingTime = Settings.getConfiguration().getInt(WAITIMEPOPERTYNAME);
|
102
|
maxAttempts = Settings.getConfiguration().getInt(MAXATTEMPTSPROPERTYNAME);
|
103
|
} catch (Exception e) {
|
104
|
log.warn("IndexGenerator.constructor - couldn't read the waiting time or maxattempts from the metacat.properties file since : "+e.getMessage()+". Default values will be used");
|
105
|
waitingTime = WAITTIME;
|
106
|
maxAttempts = MAXWAITNUMBER;
|
107
|
}
|
108
|
}
|
109
|
|
110
|
/**
|
111
|
* Build the index for all documents in Metacat without overwriting.
|
112
|
* @throws SolrServerException
|
113
|
* @throws ServiceFailure
|
114
|
* @throws NotImplemented
|
115
|
* @throws NotAuthorized
|
116
|
* @throws InvalidToken
|
117
|
* @throws InvalidRequest
|
118
|
*/
|
119
|
public void indexAll() throws InvalidRequest, InvalidToken, NotAuthorized,
|
120
|
NotImplemented, ServiceFailure, SolrServerException, FileNotFoundException {
|
121
|
boolean force = false;
|
122
|
indexAll(force);
|
123
|
}
|
124
|
|
125
|
/**
|
126
|
* Build the index for all documents. If force is true, the existed index for documents
|
127
|
* will be overwritten.
|
128
|
* @param force
|
129
|
* @throws SolrServerException
|
130
|
* @throws ServiceFailure
|
131
|
* @throws NotImplemented
|
132
|
* @throws NotAuthorized
|
133
|
* @throws InvalidToken
|
134
|
* @throws InvalidRequest
|
135
|
*/
|
136
|
public void indexAll(boolean force) throws InvalidRequest, InvalidToken,
|
137
|
NotAuthorized, NotImplemented, ServiceFailure, SolrServerException, FileNotFoundException {
|
138
|
Date since = null;
|
139
|
Date until = null;
|
140
|
index(since, until, force);
|
141
|
}
|
142
|
|
143
|
/**
|
144
|
* Build the index for the docs which have been modified since the specified date.
|
145
|
* @param since
|
146
|
* @param force
|
147
|
* @throws SolrServerException
|
148
|
* @throws ServiceFailure
|
149
|
* @throws NotImplemented
|
150
|
* @throws NotAuthorized
|
151
|
* @throws InvalidToken
|
152
|
* @throws InvalidRequest
|
153
|
*/
|
154
|
public void index(Date since, boolean force) throws InvalidRequest, InvalidToken,
|
155
|
NotAuthorized, NotImplemented, ServiceFailure, SolrServerException, FileNotFoundException {
|
156
|
Date until = null;
|
157
|
index(since, until, force);
|
158
|
}
|
159
|
|
160
|
/**
|
161
|
* Build the index for the docs which have been modified between the specified date.s
|
162
|
* @param since
|
163
|
* @param until
|
164
|
* @param force
|
165
|
* @throws SolrServerException
|
166
|
* @throws ServiceFailure
|
167
|
* @throws NotImplemented
|
168
|
* @throws NotAuthorized
|
169
|
* @throws InvalidToken
|
170
|
* @throws InvalidRequest
|
171
|
* @throws FileNotFoundException
|
172
|
*/
|
173
|
public void index(Date since, Date until, boolean force) throws SolrServerException, InvalidRequest,
|
174
|
InvalidToken, NotAuthorized, NotImplemented, ServiceFailure, FileNotFoundException {
|
175
|
List<String> solrIds = null;
|
176
|
initSystemMetadataMap();
|
177
|
initObjectPathMap();
|
178
|
List[] allMetacatIds = getMetacatIds(since, until);
|
179
|
List<String> otherMetacatIds = allMetacatIds[FIRST];
|
180
|
List<String> resourceMapIds = allMetacatIds[SECOND];
|
181
|
log.info("the metacat ids (exception resource map -----------------------------"+otherMetacatIds);
|
182
|
log.info("the metacat resroucemap ids -----------------------------"+resourceMapIds);
|
183
|
if(!force) {
|
184
|
solrIds = getSolrDocIds();
|
185
|
}
|
186
|
log.info("the solr ids -----------------------------"+solrIds);
|
187
|
index(otherMetacatIds, solrIds, force);
|
188
|
index(resourceMapIds, solrIds, force);
|
189
|
|
190
|
|
191
|
}
|
192
|
|
193
|
/*
|
194
|
* Doing index when it is necessary when compare metacatids and solrids.
|
195
|
*/
|
196
|
private void index(List<String> metacatIds, List<String>solrIds, boolean force) {
|
197
|
if(metacatIds != null) {
|
198
|
for(String metacatId : metacatIds) {
|
199
|
if(metacatId != null) {
|
200
|
boolean buildIndex = true;
|
201
|
if(!force && solrIds != null && solrIds.contains(metacatId)) {
|
202
|
//solr already indexs the id and we don't force it to rebuild it, so set the buildIndex to be false
|
203
|
buildIndex = false;
|
204
|
}
|
205
|
if(buildIndex) {
|
206
|
try {
|
207
|
generateIndex(metacatId);
|
208
|
} catch (Exception e) {
|
209
|
log.error("IndexGenerator.index - Metacat Index couldn't generate the index for the id - "+metacatId+" because "+e.getMessage());
|
210
|
}
|
211
|
|
212
|
}
|
213
|
}
|
214
|
}
|
215
|
}
|
216
|
}
|
217
|
|
218
|
public void run() {
|
219
|
try {
|
220
|
indexAll();
|
221
|
} catch (InvalidRequest e) {
|
222
|
// TODO Auto-generated catch block
|
223
|
//e.printStackTrace();
|
224
|
log.error("IndexGenerator.run - Metadata-Index couldn't generate indexes for those documents which haven't been indexed : "+e.getMessage());
|
225
|
} catch (InvalidToken e) {
|
226
|
// TODO Auto-generated catch block
|
227
|
//e.printStackTrace();
|
228
|
log.error("IndexGenerator.run - Metadata-Index couldn't generate indexes for those documents which haven't been indexed : "+e.getMessage());
|
229
|
} catch (NotAuthorized e) {
|
230
|
// TODO Auto-generated catch block
|
231
|
//e.printStackTrace();
|
232
|
} catch (NotImplemented e) {
|
233
|
// TODO Auto-generated catch block
|
234
|
//e.printStackTrace();
|
235
|
log.error("IndexGenerator.run - Metadata-Index couldn't generate indexes for those documents which haven't been indexed : "+e.getMessage());
|
236
|
} catch (ServiceFailure e) {
|
237
|
// TODO Auto-generated catch block
|
238
|
//e.printStackTrace();
|
239
|
log.error("IndexGenerator.run - Metadata-Index couldn't generate indexes for those documents which haven't been indexed : "+e.getMessage());
|
240
|
} catch (SolrServerException e) {
|
241
|
// TODO Auto-generated catch block
|
242
|
//e.printStackTrace();
|
243
|
log.error("IndexGenerator.run - Metadata-Index couldn't generate indexes for those documents which haven't been indexed : "+e.getMessage());
|
244
|
} catch (FileNotFoundException e) {
|
245
|
log.error("IndexGenerator.run - Metadata-Index couldn't generate indexes for those documents which haven't been indexed : "+e.getMessage());
|
246
|
}
|
247
|
}
|
248
|
|
249
|
/*
|
250
|
* Get the indexed ids list from the solr server.
|
251
|
* An empty list will be returned if there is no ids.
|
252
|
*/
|
253
|
private List<String> getSolrDocIds() throws SolrServerException {
|
254
|
List<String> ids = solrIndex.getSolrIds();
|
255
|
return ids;
|
256
|
}
|
257
|
|
258
|
/*
|
259
|
* Get an array of the list of ids of the metacat. If since and util are null, it will return all of them.
|
260
|
* The first element of the list is the ids except the resource map. The second elements of the list is the ids of the resource map.
|
261
|
* The reason to split them is when we index the resource map, we need the index of the document in the resource map ready.
|
262
|
*/
|
263
|
private List[] getMetacatIds(Date since, Date until) throws InvalidRequest,
|
264
|
InvalidToken, NotAuthorized, NotImplemented, ServiceFailure, FileNotFoundException {
|
265
|
|
266
|
List<String> resourceMapIds = new ArrayList();
|
267
|
List<String> otherIds = new ArrayList();
|
268
|
List[] ids = new List[2];
|
269
|
ids[FIRST]= otherIds;
|
270
|
ids[SECOND] = resourceMapIds;
|
271
|
ISet<Identifier> metacatIds = DistributedMapsFactory.getIdentifiersSet();
|
272
|
if(metacatIds != null) {
|
273
|
for(Identifier identifier : metacatIds) {
|
274
|
if(identifier != null && identifier.getValue() != null && !identifier.getValue().equals("")) {
|
275
|
SystemMetadata sysmeta = getSystemMetadata(identifier.getValue());
|
276
|
if(sysmeta != null && !sysmeta.getArchived()) {
|
277
|
ObjectFormatIdentifier formatId =sysmeta.getFormatId();
|
278
|
//System.out.println("the object format id is "+formatId.getValue());
|
279
|
//System.out.println("the ============ resourcMapNamespaces"+resourceMapNamespaces);
|
280
|
boolean correctTimeRange = false;
|
281
|
Date sysDate = sysmeta.getDateSysMetadataModified();
|
282
|
if(since == null && until == null) {
|
283
|
correctTimeRange = true;
|
284
|
} else if (since != null && until == null) {
|
285
|
if(sysDate.getTime() >= since.getTime()) {
|
286
|
correctTimeRange = true;
|
287
|
}
|
288
|
} else if (since == null && until != null) {
|
289
|
if(sysDate.getTime() <= until.getTime()) {
|
290
|
correctTimeRange = true;
|
291
|
}
|
292
|
} else if (since != null && until != null) {
|
293
|
if(sysDate.getTime() >= since.getTime() && sysDate.getTime() <= until.getTime()) {
|
294
|
correctTimeRange = true;
|
295
|
}
|
296
|
}
|
297
|
if(correctTimeRange && formatId != null && formatId.getValue() != null && resourceMapNamespaces != null && isResourceMap(formatId)) {
|
298
|
resourceMapIds.add(identifier.getValue());
|
299
|
} else {
|
300
|
otherIds.add(identifier.getValue());
|
301
|
}
|
302
|
}
|
303
|
}
|
304
|
}
|
305
|
}
|
306
|
/*if(objects != null) {
|
307
|
List<ObjectInfo> objectInfoList = objects.getObjectInfoList();
|
308
|
if(objectInfoList != null) {
|
309
|
for(ObjectInfo info : objectInfoList) {
|
310
|
if(info != null) {
|
311
|
Identifier identifier = info.getIdentifier();
|
312
|
if(identifier != null && identifier.getValue() != null && !identifier.getValue().equals("")) {
|
313
|
SystemMetadata sysmeta = getSystemMetadata(identifier.getValue());
|
314
|
if(sysmeta != null && !sysmeta.getArchived()) {
|
315
|
ObjectFormatIdentifier formatId =sysmeta.getFormatId();
|
316
|
//System.out.println("the object format id is "+formatId.getValue());
|
317
|
//System.out.println("the ============ resourcMapNamespaces"+resourceMapNamespaces);
|
318
|
if(formatId != null && formatId.getValue() != null && resourceMapNamespaces != null && isResourceMap(formatId)) {
|
319
|
resourceMapIds.add(identifier.getValue());
|
320
|
} else {
|
321
|
otherIds.add(identifier.getValue());
|
322
|
}
|
323
|
}
|
324
|
//ids.add(identifier.getValue());
|
325
|
}
|
326
|
}
|
327
|
}
|
328
|
}
|
329
|
}*/
|
330
|
return ids;
|
331
|
}
|
332
|
|
333
|
/*
|
334
|
* If the specified ObjectFormatIdentifier is a resrouce map namespace.
|
335
|
*/
|
336
|
public static boolean isResourceMap(ObjectFormatIdentifier formatId) {
|
337
|
boolean isResourceMap = false;
|
338
|
if(formatId != null && resourceMapNamespaces != null) {
|
339
|
for(String namespace : resourceMapNamespaces) {
|
340
|
if(namespace != null && formatId.getValue() != null && !formatId.getValue().trim().equals("") && formatId.getValue().equals(namespace)) {
|
341
|
isResourceMap = true;
|
342
|
break;
|
343
|
}
|
344
|
}
|
345
|
}
|
346
|
return isResourceMap;
|
347
|
}
|
348
|
|
349
|
|
350
|
|
351
|
/*
|
352
|
* Generate index for the id.
|
353
|
*/
|
354
|
private void generateIndex(String id) throws Exception {
|
355
|
if(id != null) {
|
356
|
SystemMetadata sysmeta = getSystemMetadata(id);
|
357
|
//only update none-archived id.
|
358
|
if(sysmeta != null && !sysmeta.getArchived()) {
|
359
|
InputStream data = getDataObject(id);
|
360
|
Identifier obsolete = sysmeta.getObsoletes();
|
361
|
List<String> obsoleteChain = null;
|
362
|
if(obsolete != null) {
|
363
|
obsoleteChain = getObsoletes(id);
|
364
|
}
|
365
|
solrIndex.update(id, obsoleteChain, sysmeta, data);
|
366
|
} else {
|
367
|
throw new Exception("IndexGenerator.generate - there is no found SystemMetadata associated with the id "+id);
|
368
|
}
|
369
|
|
370
|
}
|
371
|
}
|
372
|
|
373
|
/*
|
374
|
* Initialize the system metadata map
|
375
|
*/
|
376
|
private void initSystemMetadataMap() throws FileNotFoundException, ServiceFailure{
|
377
|
int times = 0;
|
378
|
if(systemMetadataMap == null) {
|
379
|
systemMetadataMap = DistributedMapsFactory.getSystemMetadataMap();
|
380
|
/*while(true) {
|
381
|
try {
|
382
|
systemMetadataMap = DistributedMapsFactory.getSystemMetadataMap();
|
383
|
break;
|
384
|
} catch (FileNotFoundException e) {
|
385
|
throw e;
|
386
|
} catch (ServiceFailure e) {
|
387
|
if(times <= maxAttempts) {
|
388
|
log.warn("IndexGenerator.initSystemMetadataMap - the hazelcast service is not ready : "
|
389
|
+e.getMessage()+"\nWe will try to access it "+waitingTime/1000+" seconds later ");
|
390
|
try {
|
391
|
Thread.sleep(waitingTime);
|
392
|
} catch (Exception ee) {
|
393
|
log.warn("IndexGenerator.initSystemMetadataMap - the thread can't sleep for "+waitingTime/1000+" seconds to wait the hazelcast service");
|
394
|
}
|
395
|
|
396
|
} else {
|
397
|
throw new ServiceFailure("0000", "IndexGenerator.initSystemMetadataMap - the hazelcast service is not ready even though Metacat-index wailted for "+maxAttempts*waitingTime/1000+" seconds. We can't get the system metadata from it and the building index can't happen this time");
|
398
|
}
|
399
|
}
|
400
|
times++;
|
401
|
}*/
|
402
|
}
|
403
|
}
|
404
|
|
405
|
/*
|
406
|
* We should call this method after calling initSystemMetadataMap since this method doesn't have the mechanism to wait the readiness of the hazelcast service
|
407
|
*/
|
408
|
private void initObjectPathMap() throws FileNotFoundException, ServiceFailure {
|
409
|
if(objectPathMap == null) {
|
410
|
objectPathMap = DistributedMapsFactory.getObjectPathMap();
|
411
|
}
|
412
|
}
|
413
|
/**
|
414
|
* Get an InputStream as the data object for the specific pid.
|
415
|
* @param pid
|
416
|
* @return
|
417
|
* @throws FileNotFoundException
|
418
|
*/
|
419
|
private InputStream getDataObject(String pid) throws FileNotFoundException {
|
420
|
Identifier identifier = new Identifier();
|
421
|
identifier.setValue(pid);
|
422
|
String objectPath = objectPathMap.get(identifier);
|
423
|
InputStream data = null;
|
424
|
data = new FileInputStream(objectPath);
|
425
|
return data;
|
426
|
|
427
|
}
|
428
|
|
429
|
/**
|
430
|
* Get the SystemMetadata for the specified id from the distributed Map.
|
431
|
* The null maybe is returned if there is no system metadata found.
|
432
|
* @param id the specified id.
|
433
|
* @return the SystemMetadata associated with the id.
|
434
|
*/
|
435
|
private SystemMetadata getSystemMetadata(String id) {
|
436
|
SystemMetadata metadata = null;
|
437
|
if(systemMetadataMap != null && id != null) {
|
438
|
Identifier identifier = new Identifier();
|
439
|
identifier.setValue(id);
|
440
|
metadata = systemMetadataMap.get(identifier);
|
441
|
}
|
442
|
return metadata;
|
443
|
}
|
444
|
|
445
|
/**
|
446
|
* Get the obsoletes chain of the specified id. The returned list doesn't include
|
447
|
* the specified id itself. The newer version has the lower index number in the list.
|
448
|
* Empty list will be returned if there is no document to be obsoleted by this id.
|
449
|
* @param id
|
450
|
* @return
|
451
|
*/
|
452
|
private List<String> getObsoletes(String id) {
|
453
|
List<String> obsoletes = new ArrayList<String>();
|
454
|
while (id != null) {
|
455
|
SystemMetadata metadata = getSystemMetadata(id);
|
456
|
id = null;//set it to be null in order to stop the while loop if the id can't be assinged to a new value in the following code.
|
457
|
if(metadata != null) {
|
458
|
Identifier identifier = metadata.getObsoletes();
|
459
|
if(identifier != null && identifier.getValue() != null && !identifier.getValue().trim().equals("")) {
|
460
|
obsoletes.add(identifier.getValue());
|
461
|
id = identifier.getValue();
|
462
|
}
|
463
|
}
|
464
|
}
|
465
|
return obsoletes;
|
466
|
}
|
467
|
|
468
|
/**
|
469
|
* Overwrite and do nothing
|
470
|
*/
|
471
|
public boolean cancel() {
|
472
|
return true;
|
473
|
}
|
474
|
|
475
|
}
|