1
|
/**
|
2
|
* Copyright: 2013 Regents of the University of California and the
|
3
|
* National Center for Ecological Analysis and Synthesis
|
4
|
*
|
5
|
* This program is free software; you can redistribute it and/or modify
|
6
|
* it under the terms of the GNU General Public License as published by
|
7
|
* the Free Software Foundation; either version 2 of the License, or
|
8
|
* (at your option) any later version.
|
9
|
*
|
10
|
* This program is distributed in the hope that it will be useful,
|
11
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
12
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
13
|
* GNU General Public License for more details.
|
14
|
*
|
15
|
* You should have received a copy of the GNU General Public License
|
16
|
* along with this program; if not, write to the Free Software
|
17
|
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
|
18
|
*/
|
19
|
package edu.ucsb.nceas.metacat.index;
|
20
|
|
21
|
import java.io.ByteArrayInputStream;
|
22
|
import java.io.FileInputStream;
|
23
|
import java.io.FileNotFoundException;
|
24
|
import java.io.IOException;
|
25
|
import java.io.InputStream;
|
26
|
import java.net.MalformedURLException;
|
27
|
import java.util.ArrayList;
|
28
|
import java.util.Calendar;
|
29
|
import java.util.Collection;
|
30
|
import java.util.Date;
|
31
|
import java.util.HashMap;
|
32
|
import java.util.Iterator;
|
33
|
import java.util.List;
|
34
|
import java.util.Map;
|
35
|
import java.util.Set;
|
36
|
|
37
|
import javax.xml.parsers.DocumentBuilder;
|
38
|
import javax.xml.parsers.DocumentBuilderFactory;
|
39
|
import javax.xml.parsers.ParserConfigurationException;
|
40
|
import javax.xml.xpath.XPath;
|
41
|
import javax.xml.xpath.XPathExpressionException;
|
42
|
import javax.xml.xpath.XPathFactory;
|
43
|
|
44
|
import org.apache.commons.codec.EncoderException;
|
45
|
import org.apache.commons.collections.CollectionUtils;
|
46
|
import org.apache.commons.io.output.ByteArrayOutputStream;
|
47
|
import org.apache.commons.lang.StringUtils;
|
48
|
import org.apache.commons.logging.Log;
|
49
|
import org.apache.commons.logging.LogFactory;
|
50
|
import org.apache.solr.client.solrj.SolrQuery;
|
51
|
import org.apache.solr.client.solrj.SolrServer;
|
52
|
import org.apache.solr.client.solrj.SolrServerException;
|
53
|
import org.apache.solr.client.solrj.response.QueryResponse;
|
54
|
import org.apache.solr.client.solrj.response.UpdateResponse;
|
55
|
import org.apache.solr.common.SolrDocument;
|
56
|
import org.apache.solr.common.SolrDocumentList;
|
57
|
import org.apache.solr.common.SolrInputDocument;
|
58
|
import org.apache.solr.schema.IndexSchema;
|
59
|
import org.dataone.cn.indexer.XMLNamespaceConfig;
|
60
|
import org.dataone.cn.indexer.convert.SolrDateConverter;
|
61
|
import org.dataone.cn.indexer.parser.BaseXPathDocumentSubprocessor;
|
62
|
import org.dataone.cn.indexer.parser.IDocumentDeleteSubprocessor;
|
63
|
import org.dataone.cn.indexer.parser.IDocumentSubprocessor;
|
64
|
import org.dataone.cn.indexer.parser.SolrField;
|
65
|
import org.dataone.cn.indexer.solrhttp.SolrDoc;
|
66
|
import org.dataone.cn.indexer.solrhttp.SolrElementField;
|
67
|
import org.dataone.exceptions.MarshallingException;
|
68
|
import org.dataone.service.exceptions.NotFound;
|
69
|
import org.dataone.service.exceptions.NotImplemented;
|
70
|
import org.dataone.service.exceptions.ServiceFailure;
|
71
|
import org.dataone.service.exceptions.UnsupportedType;
|
72
|
import org.dataone.service.types.v1.Event;
|
73
|
import org.dataone.service.types.v1.Identifier;
|
74
|
import org.dataone.service.types.v2.SystemMetadata;
|
75
|
import org.dataone.service.util.DateTimeMarshaller;
|
76
|
import org.dataone.service.util.TypeMarshaller;
|
77
|
import org.dspace.foresite.OREParserException;
|
78
|
import org.jibx.runtime.JiBXException;
|
79
|
import org.w3c.dom.Document;
|
80
|
import org.xml.sax.SAXException;
|
81
|
|
82
|
import edu.ucsb.nceas.metacat.common.index.event.IndexEvent;
|
83
|
import edu.ucsb.nceas.metacat.common.query.SolrQueryServiceController;
|
84
|
import edu.ucsb.nceas.metacat.index.event.EventlogFactory;
|
85
|
import edu.ucsb.nceas.metacat.index.resourcemap.ResourceMapSubprocessor;
|
86
|
|
87
|
/**
|
88
|
* A class does insert, update and remove indexes to a SOLR server
|
89
|
* @author tao
|
90
|
*
|
91
|
*/
|
92
|
public class SolrIndex {
|
93
|
|
94
|
public static final String ID = "id";
|
95
|
private static final String IDQUERY = ID+":*";
|
96
|
private List<IDocumentSubprocessor> subprocessors = null;
|
97
|
private List<IDocumentDeleteSubprocessor> deleteSubprocessors = null;
|
98
|
|
99
|
private SolrServer solrServer = null;
|
100
|
private XMLNamespaceConfig xmlNamespaceConfig = null;
|
101
|
private List<SolrField> sysmetaSolrFields = null;
|
102
|
|
103
|
private static DocumentBuilderFactory documentBuilderFactory = null;
|
104
|
private static DocumentBuilder builder = null;
|
105
|
|
106
|
private static XPathFactory xpathFactory = null;
|
107
|
private static XPath xpath = null;
|
108
|
Log log = LogFactory.getLog(SolrIndex.class);
|
109
|
|
110
|
static {
|
111
|
documentBuilderFactory = DocumentBuilderFactory.newInstance();
|
112
|
documentBuilderFactory.setNamespaceAware(true);
|
113
|
try {
|
114
|
builder = documentBuilderFactory.newDocumentBuilder();
|
115
|
} catch (ParserConfigurationException e) {
|
116
|
e.printStackTrace();
|
117
|
}
|
118
|
xpathFactory = XPathFactory.newInstance();
|
119
|
xpath = xpathFactory.newXPath();
|
120
|
}
|
121
|
|
122
|
/**
|
123
|
* Constructor
|
124
|
* @throws SAXException
|
125
|
* @throws IOException
|
126
|
*/
|
127
|
public SolrIndex(XMLNamespaceConfig xmlNamespaceConfig, List<SolrField> sysmetaSolrFields)
|
128
|
throws XPathExpressionException, ParserConfigurationException, IOException, SAXException {
|
129
|
this.xmlNamespaceConfig = xmlNamespaceConfig;
|
130
|
this.sysmetaSolrFields = sysmetaSolrFields;
|
131
|
init();
|
132
|
}
|
133
|
|
134
|
private void init() throws ParserConfigurationException, XPathExpressionException {
|
135
|
xpath.setNamespaceContext(xmlNamespaceConfig);
|
136
|
initExpressions();
|
137
|
}
|
138
|
|
139
|
private void initExpressions() throws XPathExpressionException {
|
140
|
for (SolrField field : sysmetaSolrFields) {
|
141
|
field.initExpression(xpath);
|
142
|
}
|
143
|
|
144
|
}
|
145
|
|
146
|
|
147
|
/**
|
148
|
* Get the list of the Subprocessors in this index.
|
149
|
* @return the list of the Subprocessors.
|
150
|
*/
|
151
|
public List<IDocumentSubprocessor> getSubprocessors() {
|
152
|
return subprocessors;
|
153
|
}
|
154
|
|
155
|
/**
|
156
|
* Set the list of Subprocessors.
|
157
|
* @param subprocessorList the list will be set.
|
158
|
*/
|
159
|
public void setSubprocessors(List<IDocumentSubprocessor> subprocessorList) {
|
160
|
for (IDocumentSubprocessor subprocessor : subprocessorList) {
|
161
|
if (subprocessor instanceof BaseXPathDocumentSubprocessor) {
|
162
|
((BaseXPathDocumentSubprocessor)subprocessor).initExpression(xpath);
|
163
|
}
|
164
|
}
|
165
|
this.subprocessors = subprocessorList;
|
166
|
}
|
167
|
|
168
|
public List<IDocumentDeleteSubprocessor> getDeleteSubprocessors() {
|
169
|
return deleteSubprocessors;
|
170
|
}
|
171
|
|
172
|
public void setDeleteSubprocessors(
|
173
|
List<IDocumentDeleteSubprocessor> deleteSubprocessors) {
|
174
|
this.deleteSubprocessors = deleteSubprocessors;
|
175
|
}
|
176
|
|
177
|
/**
|
178
|
* Generate the index for the given information
|
179
|
* @param id
|
180
|
* @param systemMetadata
|
181
|
* @param dataStream
|
182
|
* @return
|
183
|
* @throws IOException
|
184
|
* @throws SAXException
|
185
|
* @throws ParserConfigurationException
|
186
|
* @throws XPathExpressionException
|
187
|
* @throws JiBXException
|
188
|
* @throws SolrServerException
|
189
|
* @throws EncoderException
|
190
|
* @throws UnsupportedType
|
191
|
* @throws NotFound
|
192
|
* @throws NotImplemented
|
193
|
*/
|
194
|
private Map<String, SolrDoc> process(String id, SystemMetadata systemMetadata, String objectPath)
|
195
|
throws IOException, SAXException, ParserConfigurationException,
|
196
|
XPathExpressionException, MarshallingException, EncoderException, SolrServerException, NotImplemented, NotFound, UnsupportedType{
|
197
|
|
198
|
// Load the System Metadata document
|
199
|
ByteArrayOutputStream systemMetadataOutputStream = new ByteArrayOutputStream();
|
200
|
TypeMarshaller.marshalTypeToOutputStream(systemMetadata, systemMetadataOutputStream);
|
201
|
ByteArrayInputStream systemMetadataStream = new ByteArrayInputStream(systemMetadataOutputStream.toByteArray());
|
202
|
Document sysMetaDoc = generateXmlDocument(systemMetadataStream);
|
203
|
if (sysMetaDoc == null) {
|
204
|
log.error("Could not load System metadata for ID: " + id);
|
205
|
return null;
|
206
|
}
|
207
|
|
208
|
// Extract the field values from the System Metadata
|
209
|
List<SolrElementField> sysSolrFields = processSysmetaFields(sysMetaDoc, id);
|
210
|
SolrDoc indexDocument = new SolrDoc(sysSolrFields);
|
211
|
Map<String, SolrDoc> docs = new HashMap<String, SolrDoc>();
|
212
|
docs.put(id, indexDocument);
|
213
|
|
214
|
// get the format id for this object
|
215
|
String formatId = indexDocument.getFirstFieldValue(SolrElementField.FIELD_OBJECTFORMAT);
|
216
|
|
217
|
// Determine if subprocessors are available for this ID
|
218
|
if (subprocessors != null) {
|
219
|
// for each subprocessor loaded from the spring config
|
220
|
for (IDocumentSubprocessor subprocessor : subprocessors) {
|
221
|
// Does this subprocessor apply?
|
222
|
if (subprocessor.canProcess(formatId)) {
|
223
|
// if so, then extract the additional information from the
|
224
|
// document.
|
225
|
try {
|
226
|
// docObject = the resource map document or science
|
227
|
// metadata document.
|
228
|
// note that resource map processing touches all objects
|
229
|
// referenced by the resource map.
|
230
|
FileInputStream dataStream = new FileInputStream(objectPath);
|
231
|
if (!dataStream.getFD().valid()) {
|
232
|
log.error("Could not load OBJECT file for ID,Path=" + id + ", "
|
233
|
+ objectPath);
|
234
|
//throw new Exception("Could not load OBJECT for ID " + id );
|
235
|
} else {
|
236
|
docs = subprocessor.processDocument(id, docs, dataStream);
|
237
|
}
|
238
|
} catch (Exception e) {
|
239
|
e.printStackTrace();
|
240
|
log.error(e.getMessage(), e);
|
241
|
throw new SolrServerException(e.getMessage());
|
242
|
}
|
243
|
}
|
244
|
}
|
245
|
}
|
246
|
|
247
|
// TODO: in the XPathDocumentParser class in d1_cn_index_process module,
|
248
|
// merge is only for resource map. We need more work here.
|
249
|
for (SolrDoc mergeDoc : docs.values()) {
|
250
|
if (!mergeDoc.isMerged()) {
|
251
|
mergeWithIndexedDocument(mergeDoc);
|
252
|
}
|
253
|
}
|
254
|
|
255
|
//SolrElementAdd addCommand = getAddCommand(new ArrayList<SolrDoc>(docs.values()));
|
256
|
|
257
|
return docs;
|
258
|
}
|
259
|
|
260
|
/**
|
261
|
* Merge updates with existing solr documents
|
262
|
*
|
263
|
* This method appears to re-set the data package field data into the
|
264
|
* document about to be updated in the solr index. Since packaging
|
265
|
* information is derived from the package document (resource map), this
|
266
|
* information is not present when processing a document contained in a data
|
267
|
* package. This method replaces those values from the existing solr index
|
268
|
* record for the document being processed. -- sroseboo, 1-18-12
|
269
|
*
|
270
|
* @param indexDocument
|
271
|
* @return
|
272
|
* @throws IOException
|
273
|
* @throws EncoderException
|
274
|
* @throws XPathExpressionException
|
275
|
* @throws SAXException
|
276
|
* @throws ParserConfigurationException
|
277
|
* @throws SolrServerException
|
278
|
* @throws UnsupportedType
|
279
|
* @throws NotFound
|
280
|
* @throws NotImplemented
|
281
|
*/
|
282
|
// TODO:combine merge function with resourcemap merge function
|
283
|
|
284
|
private SolrDoc mergeWithIndexedDocument(SolrDoc indexDocument) throws IOException,
|
285
|
EncoderException, XPathExpressionException, SolrServerException, ParserConfigurationException, SAXException, NotImplemented, NotFound, UnsupportedType {
|
286
|
List<String> ids = new ArrayList<String>();
|
287
|
ids.add(indexDocument.getIdentifier());
|
288
|
List<SolrDoc> indexedDocuments = ResourceMapSubprocessor.getSolrDocs(ids);
|
289
|
SolrDoc indexedDocument = indexedDocuments == null || indexedDocuments.size() <= 0 ? null
|
290
|
: indexedDocuments.get(0);
|
291
|
|
292
|
IndexSchema indexSchema = SolrQueryServiceController.getInstance().getSchema();
|
293
|
|
294
|
if (indexedDocument == null || indexedDocument.getFieldList().size() <= 0) {
|
295
|
return indexDocument;
|
296
|
} else {
|
297
|
for (SolrElementField field : indexedDocument.getFieldList()) {
|
298
|
if ((field.getName().equals(SolrElementField.FIELD_ISDOCUMENTEDBY)
|
299
|
|| field.getName().equals(SolrElementField.FIELD_DOCUMENTS) || field
|
300
|
.getName().equals(SolrElementField.FIELD_RESOURCEMAP))
|
301
|
&& !indexDocument.hasFieldWithValue(field.getName(), field.getValue())) {
|
302
|
indexDocument.addField(field);
|
303
|
} else if (!indexSchema.isCopyFieldTarget(indexSchema.getField(field.getName())) && !indexDocument.hasField(field.getName())) {
|
304
|
indexDocument.addField(field);
|
305
|
}
|
306
|
}
|
307
|
|
308
|
indexDocument.setMerged(true);
|
309
|
return indexDocument;
|
310
|
}
|
311
|
}
|
312
|
|
313
|
/*
|
314
|
* Generate a Document from the InputStream
|
315
|
*/
|
316
|
private Document generateXmlDocument(InputStream smdStream) throws SAXException {
|
317
|
Document doc = null;
|
318
|
|
319
|
try {
|
320
|
doc = builder.parse(smdStream);
|
321
|
} catch (IOException e) {
|
322
|
log.error(e.getMessage(), e);
|
323
|
}
|
324
|
|
325
|
return doc;
|
326
|
}
|
327
|
|
328
|
/*
|
329
|
* Index the fields of the system metadata
|
330
|
*/
|
331
|
private List<SolrElementField> processSysmetaFields(Document doc, String identifier) {
|
332
|
|
333
|
List<SolrElementField> fieldList = new ArrayList<SolrElementField>();
|
334
|
// solrFields is the list of fields defined in the application context
|
335
|
|
336
|
for (SolrField field : sysmetaSolrFields) {
|
337
|
try {
|
338
|
// the field.getFields method can return a single value or
|
339
|
// multiple values for multi-valued fields
|
340
|
// or can return multiple SOLR document fields.
|
341
|
fieldList.addAll(field.getFields(doc, identifier));
|
342
|
} catch (Exception e) {
|
343
|
e.printStackTrace();
|
344
|
}
|
345
|
}
|
346
|
return fieldList;
|
347
|
|
348
|
}
|
349
|
|
350
|
/**
|
351
|
* Check the parameters of the insert or update methods.
|
352
|
* @param pid
|
353
|
* @param systemMetadata
|
354
|
* @param data
|
355
|
* @throws SolrServerException
|
356
|
*/
|
357
|
private void checkParams(Identifier pid, SystemMetadata systemMetadata, String objectPath) throws SolrServerException {
|
358
|
if(pid == null || pid.getValue() == null || pid.getValue().trim().equals("")) {
|
359
|
throw new SolrServerException("The identifier of the indexed document should not be null or blank.");
|
360
|
}
|
361
|
if(systemMetadata == null) {
|
362
|
throw new SolrServerException("The system metadata of the indexed document "+pid.getValue()+ " should not be null.");
|
363
|
}
|
364
|
if(objectPath == null) {
|
365
|
throw new SolrServerException("The indexed document itself for pid "+pid.getValue()+" should not be null.");
|
366
|
}
|
367
|
}
|
368
|
|
369
|
/**
|
370
|
* Insert the indexes for a document.
|
371
|
* @param pid the id of this document
|
372
|
* @param systemMetadata the system metadata associated with the data object
|
373
|
* @param data the path to the object file itself
|
374
|
* @throws SolrServerException
|
375
|
* @throws JiBXException
|
376
|
* @throws EncoderException
|
377
|
* @throws UnsupportedType
|
378
|
* @throws NotFound
|
379
|
* @throws NotImplemented
|
380
|
*/
|
381
|
private synchronized void insert(Identifier pid, SystemMetadata systemMetadata, String objectPath)
|
382
|
throws IOException, SAXException, ParserConfigurationException,
|
383
|
XPathExpressionException, SolrServerException, MarshallingException, EncoderException, NotImplemented, NotFound, UnsupportedType {
|
384
|
checkParams(pid, systemMetadata, objectPath);
|
385
|
Map<String, SolrDoc> docs = process(pid.getValue(), systemMetadata, objectPath);
|
386
|
|
387
|
//transform the Map to the SolrInputDocument which can be used by the solr server
|
388
|
if(docs != null) {
|
389
|
Set<String> ids = docs.keySet();
|
390
|
for(String id : ids) {
|
391
|
if(id != null) {
|
392
|
SolrDoc doc = docs.get(id);
|
393
|
insertToIndex(doc);
|
394
|
}
|
395
|
|
396
|
}
|
397
|
}
|
398
|
}
|
399
|
|
400
|
/**
|
401
|
* Adds the given fields to the solr index for the given pid, preserving the index values
|
402
|
* that previously existed
|
403
|
* @param pid
|
404
|
* @param fields
|
405
|
*/
|
406
|
public void insertFields(Identifier pid, Map<String, List<Object>> fields) {
|
407
|
|
408
|
try {
|
409
|
// copy the original values already indexed for this document
|
410
|
SolrQuery query = new SolrQuery("id:\"" + pid.getValue() + "\"");
|
411
|
QueryResponse res = solrServer.query(query);
|
412
|
SolrDoc doc = new SolrDoc();
|
413
|
|
414
|
// include existing values if they exist
|
415
|
IndexSchema indexSchema = SolrQueryServiceController.getInstance().getSchema();
|
416
|
|
417
|
if (res.getResults().size() > 0) {
|
418
|
SolrDocument orig = res.getResults().get(0);
|
419
|
for (String fieldName: orig.getFieldNames()) {
|
420
|
// don't transfer the copyTo fields, otherwise there are errors
|
421
|
if (indexSchema.isCopyFieldTarget(indexSchema.getField(fieldName))) {
|
422
|
continue;
|
423
|
}
|
424
|
for (Object value: orig.getFieldValues(fieldName)) {
|
425
|
String stringValue = value.toString();
|
426
|
// special handling for dates in ISO 8601
|
427
|
if (value instanceof Date) {
|
428
|
stringValue = DateTimeMarshaller.serializeDateToUTC((Date)value);
|
429
|
SolrDateConverter converter = new SolrDateConverter();
|
430
|
stringValue = converter.convert(stringValue);
|
431
|
}
|
432
|
SolrElementField field = new SolrElementField(fieldName, stringValue);
|
433
|
log.debug("Adding field: " + fieldName);
|
434
|
doc.addField(field);
|
435
|
}
|
436
|
}
|
437
|
}
|
438
|
|
439
|
// add the additional fields we are trying to include in the index
|
440
|
for (String fieldName: fields.keySet()) {
|
441
|
List<Object> values = fields.get(fieldName);
|
442
|
for (Object value: values) {
|
443
|
if (!doc.hasFieldWithValue(fieldName, value.toString())) {
|
444
|
if (indexSchema.getField(fieldName).multiValued()) {
|
445
|
doc.addField(new SolrElementField(fieldName, value.toString()));
|
446
|
} else {
|
447
|
doc.updateOrAddField(fieldName, value.toString());
|
448
|
}
|
449
|
}
|
450
|
}
|
451
|
}
|
452
|
|
453
|
// make sure there is an id in the solrdoc so it is added to the index
|
454
|
if (!doc.hasField(ID)) {
|
455
|
doc.updateOrAddField(ID, pid.getValue());
|
456
|
}
|
457
|
|
458
|
// insert the whole thing
|
459
|
insertToIndex(doc);
|
460
|
} catch (Exception e) {
|
461
|
String error = "SolrIndex.insetFields - could not update the solr index: " + e.getMessage();
|
462
|
writeEventLog(null, pid, error);
|
463
|
log.error(error, e);
|
464
|
}
|
465
|
|
466
|
}
|
467
|
|
468
|
/*
|
469
|
* Insert a SolrDoc to the solr server.
|
470
|
*/
|
471
|
private synchronized void insertToIndex(SolrDoc doc) throws SolrServerException, IOException {
|
472
|
if(doc != null ) {
|
473
|
SolrInputDocument solrDoc = new SolrInputDocument();
|
474
|
List<SolrElementField> list = doc.getFieldList();
|
475
|
if(list != null) {
|
476
|
//solrDoc.addField(METACATPIDFIELD, pid);
|
477
|
Iterator<SolrElementField> iterator = list.iterator();
|
478
|
while (iterator.hasNext()) {
|
479
|
SolrElementField field = iterator.next();
|
480
|
if(field != null) {
|
481
|
String value = field.getValue();
|
482
|
String name = field.getName();
|
483
|
//System.out.println("add name/value pair - "+name+"/"+value);
|
484
|
solrDoc.addField(name, value);
|
485
|
}
|
486
|
}
|
487
|
}
|
488
|
if(!solrDoc.isEmpty()) {
|
489
|
/*IndexEvent event = new IndexEvent();
|
490
|
event.setDate(Calendar.getInstance().getTime());
|
491
|
Identifier pid = new Identifier();
|
492
|
pid.setValue(doc.getIdentifier());
|
493
|
event.setIdentifier(pid);*/
|
494
|
try {
|
495
|
UpdateResponse response = solrServer.add(solrDoc);
|
496
|
solrServer.commit();
|
497
|
/*event.setType(IndexEvent.SUCCESSINSERT);
|
498
|
event.setDescription("Successfully insert the solr index for the id "+pid.getValue());
|
499
|
try {
|
500
|
EventlogFactory.createIndexEventLog().write(event);
|
501
|
} catch (Exception e) {
|
502
|
log.error("SolrIndex.insertToIndex - IndexEventLog can't log the index inserting event :"+e.getMessage());
|
503
|
}*/
|
504
|
} catch (SolrServerException e) {
|
505
|
/*event.setAction(Event.CREATE);
|
506
|
event.setDescription("Failed to insert the solr index for the id "+pid.getValue()+" since "+e.getMessage());
|
507
|
try {
|
508
|
EventlogFactory.createIndexEventLog().write(event);
|
509
|
} catch (Exception ee) {
|
510
|
log.error("SolrIndex.insertToIndex - IndexEventLog can't log the index inserting event :"+ee.getMessage());
|
511
|
}*/
|
512
|
throw e;
|
513
|
} catch (IOException e) {
|
514
|
/*event.setAction(Event.CREATE);
|
515
|
event.setDescription("Failed to insert the solr index for the id "+pid.getValue()+" since "+e.getMessage());
|
516
|
try {
|
517
|
EventlogFactory.createIndexEventLog().write(event);
|
518
|
} catch (Exception ee) {
|
519
|
log.error("SolrIndex.insertToIndex - IndexEventLog can't log the index inserting event :"+ee.getMessage());
|
520
|
}*/
|
521
|
throw e;
|
522
|
|
523
|
}
|
524
|
//System.out.println("=================the response is:\n"+response.toString());
|
525
|
}
|
526
|
}
|
527
|
}
|
528
|
|
529
|
/**
|
530
|
* Update the solr index. This method handles the three scenarios:
|
531
|
* 1. Remove an existing doc - if the the system metadata shows the value of the archive is true,
|
532
|
* remove the index for the previous version(s) and generate new index for the doc.
|
533
|
* 2. Add a new doc - if the system metadata shows the value of the archive is false, generate the
|
534
|
* index for the doc.
|
535
|
*/
|
536
|
public void update(Identifier pid, SystemMetadata systemMetadata) {
|
537
|
if(systemMetadata==null || pid==null) {
|
538
|
log.error("SolrIndex.update - the systemMetadata or pid is null. So nothing will be indexed.");
|
539
|
return;
|
540
|
}
|
541
|
String objectPath = null;
|
542
|
try {
|
543
|
if (systemMetadata.getArchived() == null || !systemMetadata.getArchived()) {
|
544
|
objectPath = DistributedMapsFactory.getObjectPathMap().get(pid);
|
545
|
}
|
546
|
update(pid, systemMetadata, objectPath);
|
547
|
EventlogFactory.createIndexEventLog().remove(pid);
|
548
|
} catch (Exception e) {
|
549
|
String error = "SolrIndex.update - could not update the solr index since " + e.getMessage();
|
550
|
writeEventLog(systemMetadata, pid, error);
|
551
|
log.error(error, e);
|
552
|
}
|
553
|
}
|
554
|
|
555
|
|
556
|
/**
|
557
|
* Update the solr index. This method handles the three scenarios:
|
558
|
* 1. Remove an existing doc - if the the system metadata shows the value of the archive is true,
|
559
|
* remove the index for the previous version(s) and generate new index for the doc.
|
560
|
* 2. Add a new doc - if the system metadata shows the value of the archive is false, generate the
|
561
|
* index for the doc.
|
562
|
* @param pid
|
563
|
* @param systemMetadata
|
564
|
* @param data
|
565
|
* @throws SolrServerException
|
566
|
* @throws ServiceFailure
|
567
|
* @throws XPathExpressionException
|
568
|
* @throws NotImplemented
|
569
|
* @throws NotFound
|
570
|
* @throws UnsupportedType
|
571
|
* @throws IOException
|
572
|
* @throws SAXException
|
573
|
* @throws ParserConfigurationException
|
574
|
* @throws OREParserException
|
575
|
* @throws JiBXException
|
576
|
* @throws EncoderException
|
577
|
*/
|
578
|
void update(Identifier pid, SystemMetadata systemMetadata, String objectPath) throws Exception {
|
579
|
//checkParams(pid, systemMetadata, objectPath);
|
580
|
if(systemMetadata==null || pid==null) {
|
581
|
log.error("SolrIndex.update - the systemMetadata or pid is null. So nothing will be indexed.");
|
582
|
return;
|
583
|
}
|
584
|
boolean isArchive = systemMetadata.getArchived() != null && systemMetadata.getArchived();
|
585
|
if(isArchive ) {
|
586
|
//delete the index for the archived objects
|
587
|
remove(pid.getValue(), systemMetadata);
|
588
|
log.info("SolrIndex.update============================= archive the idex for the identifier "+pid);
|
589
|
} else {
|
590
|
//generate index for either add or update.
|
591
|
insert(pid, systemMetadata, objectPath);
|
592
|
log.info("SolrIndex.update============================= insert index for the identifier "+pid);
|
593
|
}
|
594
|
}
|
595
|
|
596
|
|
597
|
|
598
|
/*
|
599
|
* Is the pid a resource map
|
600
|
*/
|
601
|
private boolean isDataPackage(String pid, SystemMetadata sysmeta) throws FileNotFoundException, ServiceFailure {
|
602
|
boolean isDataPackage = false;
|
603
|
//SystemMetadata sysmeta = DistributedMapsFactory.getSystemMetadata(pid);
|
604
|
if(sysmeta != null) {
|
605
|
isDataPackage = IndexGeneratorTimerTask.isResourceMap(sysmeta.getFormatId());
|
606
|
}
|
607
|
return isDataPackage;
|
608
|
}
|
609
|
|
610
|
private boolean isPartOfDataPackage(String pid) throws XPathExpressionException, NotImplemented, NotFound, UnsupportedType, SolrServerException, IOException, ParserConfigurationException, SAXException {
|
611
|
SolrDoc dataPackageIndexDoc = ResourceMapSubprocessor.getSolrDoc(pid);
|
612
|
if (dataPackageIndexDoc != null) {
|
613
|
String resourceMapId = dataPackageIndexDoc
|
614
|
.getFirstFieldValue(SolrElementField.FIELD_RESOURCEMAP);
|
615
|
return StringUtils.isNotEmpty(resourceMapId);
|
616
|
} else {
|
617
|
return false;
|
618
|
}
|
619
|
}
|
620
|
/**
|
621
|
* Remove the indexed associated with specified pid.
|
622
|
* @param pid the pid which the indexes are associated with
|
623
|
* @throws IOException
|
624
|
* @throws SolrServerException
|
625
|
* @throws ParserConfigurationException
|
626
|
* @throws SAXException
|
627
|
* @throws UnsupportedType
|
628
|
* @throws NotFound
|
629
|
* @throws NotImplemented
|
630
|
* @throws XPathExpressionException
|
631
|
* @throws ServiceFailure
|
632
|
* @throws OREParserException
|
633
|
*/
|
634
|
private void remove(String pid, SystemMetadata sysmeta) throws Exception {
|
635
|
if (isDataPackage(pid, sysmeta)) {
|
636
|
removeDataPackage(pid);
|
637
|
} else if (isPartOfDataPackage(pid)) {
|
638
|
removeFromDataPackage(pid);
|
639
|
} else {
|
640
|
removeFromIndex(pid);
|
641
|
}
|
642
|
}
|
643
|
|
644
|
/*
|
645
|
* Remove the resource map from the solr index. It doesn't only remove the index for itself and also
|
646
|
* remove the relationship for the related metadata and data objects.
|
647
|
*/
|
648
|
private void removeDataPackage(String pid) throws Exception {
|
649
|
removeFromIndex(pid);
|
650
|
List<SolrDoc> docsToUpdate = getUpdatedSolrDocsByRemovingResourceMap(pid);
|
651
|
if (docsToUpdate != null && !docsToUpdate.isEmpty()) {
|
652
|
//SolrElementAdd addCommand = new SolrElementAdd(docsToUpdate);
|
653
|
//httpService.sendUpdate(solrIndexUri, addCommand);
|
654
|
for(SolrDoc doc : docsToUpdate) {
|
655
|
removeFromIndex(doc.getIdentifier());
|
656
|
insertToIndex(doc);
|
657
|
}
|
658
|
}
|
659
|
|
660
|
}
|
661
|
|
662
|
/*
|
663
|
* Get the list of the solr doc which need to be updated because the removal of the resource map
|
664
|
*/
|
665
|
private List<SolrDoc> getUpdatedSolrDocsByRemovingResourceMap(String resourceMapId)
|
666
|
throws UnsupportedType, NotFound, SolrServerException, ParserConfigurationException, SAXException, MalformedURLException, IOException, XPathExpressionException {
|
667
|
List<SolrDoc> updatedSolrDocs = null;
|
668
|
if (resourceMapId != null && !resourceMapId.trim().equals("")) {
|
669
|
/*List<SolrDoc> docsContainResourceMap = httpService.getDocumentsByResourceMap(
|
670
|
solrQueryUri, resourceMapId);*/
|
671
|
List<SolrDoc> docsContainResourceMap = ResourceMapSubprocessor.getDocumentsByResourceMap(resourceMapId);
|
672
|
updatedSolrDocs = removeResourceMapRelationship(docsContainResourceMap,
|
673
|
resourceMapId);
|
674
|
}
|
675
|
return updatedSolrDocs;
|
676
|
}
|
677
|
|
678
|
/*
|
679
|
* Get the list of the solr doc which need to be updated because the removal of the resource map
|
680
|
*/
|
681
|
private List<SolrDoc> removeResourceMapRelationship(List<SolrDoc> docsContainResourceMap,
|
682
|
String resourceMapId) throws XPathExpressionException, IOException {
|
683
|
List<SolrDoc> totalUpdatedSolrDocs = new ArrayList<SolrDoc>();
|
684
|
if (docsContainResourceMap != null && !docsContainResourceMap.isEmpty()) {
|
685
|
for (SolrDoc doc : docsContainResourceMap) {
|
686
|
List<SolrDoc> updatedSolrDocs = new ArrayList<SolrDoc>();
|
687
|
List<String> resourceMapIdStrs = doc
|
688
|
.getAllFieldValues(SolrElementField.FIELD_RESOURCEMAP);
|
689
|
List<String> dataIdStrs = doc
|
690
|
.getAllFieldValues(SolrElementField.FIELD_DOCUMENTS);
|
691
|
List<String> metadataIdStrs = doc
|
692
|
.getAllFieldValues(SolrElementField.FIELD_ISDOCUMENTEDBY);
|
693
|
if ((dataIdStrs == null || dataIdStrs.isEmpty())
|
694
|
&& (metadataIdStrs == null || metadataIdStrs.isEmpty())) {
|
695
|
// only has resourceMap field, doesn't have either documentBy or documents fields.
|
696
|
// so we only remove the resource map field.
|
697
|
doc.removeFieldsWithValue(SolrElementField.FIELD_RESOURCEMAP, resourceMapId);
|
698
|
updatedSolrDocs.add(doc);
|
699
|
} else if ((dataIdStrs != null && !dataIdStrs.isEmpty())
|
700
|
&& (metadataIdStrs == null || metadataIdStrs.isEmpty())) {
|
701
|
//The solr doc is for a metadata object since the solr doc documents data files
|
702
|
updatedSolrDocs = removeAggregatedItems(resourceMapId, doc, resourceMapIdStrs,
|
703
|
dataIdStrs, SolrElementField.FIELD_DOCUMENTS);
|
704
|
} else if ((dataIdStrs == null || dataIdStrs.isEmpty())
|
705
|
&& (metadataIdStrs != null && !metadataIdStrs.isEmpty())) {
|
706
|
//The solr doc is for a data object since it documentedBy elements.
|
707
|
updatedSolrDocs = removeAggregatedItems(resourceMapId, doc, resourceMapIdStrs,
|
708
|
metadataIdStrs, SolrElementField.FIELD_ISDOCUMENTEDBY);
|
709
|
} else if ((dataIdStrs != null && !dataIdStrs.isEmpty())
|
710
|
&& (metadataIdStrs != null && !metadataIdStrs.isEmpty())){
|
711
|
// both metadata and data for one object
|
712
|
List<SolrDoc> solrDocsRemovedDocuments = removeAggregatedItems(resourceMapId, doc, resourceMapIdStrs,
|
713
|
dataIdStrs, SolrElementField.FIELD_DOCUMENTS);
|
714
|
List<SolrDoc> solrDocsRemovedDocumentBy = removeAggregatedItems(resourceMapId, doc, resourceMapIdStrs,
|
715
|
metadataIdStrs, SolrElementField.FIELD_ISDOCUMENTEDBY);
|
716
|
updatedSolrDocs = mergeUpdatedSolrDocs(solrDocsRemovedDocumentBy, solrDocsRemovedDocuments);
|
717
|
}
|
718
|
//move them to the final result
|
719
|
if(updatedSolrDocs != null) {
|
720
|
for(SolrDoc updatedDoc: updatedSolrDocs) {
|
721
|
totalUpdatedSolrDocs.add(updatedDoc);
|
722
|
}
|
723
|
}
|
724
|
|
725
|
}
|
726
|
|
727
|
}
|
728
|
return totalUpdatedSolrDocs;
|
729
|
}
|
730
|
|
731
|
/*
|
732
|
* Process the list of ids of the documentBy/documents in a slor doc.
|
733
|
*/
|
734
|
private List<SolrDoc> removeAggregatedItems(String targetResourceMapId, SolrDoc doc,
|
735
|
List<String> resourceMapIdsInDoc, List<String> aggregatedItemsInDoc, String fieldNameRemoved) {
|
736
|
List<SolrDoc> updatedSolrDocs = new ArrayList<SolrDoc>();
|
737
|
if (doc != null && resourceMapIdsInDoc != null && aggregatedItemsInDoc != null
|
738
|
&& fieldNameRemoved != null) {
|
739
|
if (resourceMapIdsInDoc.size() == 1) {
|
740
|
//only has one resource map. remove the resource map. also remove the documentBy
|
741
|
doc.removeFieldsWithValue(SolrElementField.FIELD_RESOURCEMAP, targetResourceMapId);
|
742
|
doc.removeAllFields(fieldNameRemoved);
|
743
|
updatedSolrDocs.add(doc);
|
744
|
} else if (resourceMapIdsInDoc.size() > 1) {
|
745
|
//we have multiple resource maps. We should match them.
|
746
|
Map<String, String> ids = matchResourceMapsAndItems(doc.getIdentifier(),
|
747
|
targetResourceMapId, resourceMapIdsInDoc, aggregatedItemsInDoc, fieldNameRemoved);
|
748
|
if (ids != null) {
|
749
|
for (String id : ids.keySet()) {
|
750
|
doc.removeFieldsWithValue(fieldNameRemoved, id);
|
751
|
}
|
752
|
}
|
753
|
doc.removeFieldsWithValue(SolrElementField.FIELD_RESOURCEMAP,
|
754
|
targetResourceMapId);
|
755
|
updatedSolrDocs.add(doc);
|
756
|
/*if (aggregatedItemsInDoc.size() > 1) {
|
757
|
|
758
|
|
759
|
} else {
|
760
|
//multiple resource map aggregate same metadata and data. Just remove the resource map
|
761
|
doc.removeFieldsWithValue(SolrElementField.FIELD_RESOURCEMAP,
|
762
|
targetResourceMapId);
|
763
|
updatedSolrDocs.add(doc);
|
764
|
}*/
|
765
|
}
|
766
|
}
|
767
|
return updatedSolrDocs;
|
768
|
}
|
769
|
|
770
|
/*
|
771
|
* Return a map of mapping aggregation id map the target resourceMapId.
|
772
|
* This will look the aggregation information in another side - If the targetId
|
773
|
* is a metadata object, we will look the data objects which it describes; If
|
774
|
* the targetId is a data object, we will look the metadata object which documents it.
|
775
|
*/
|
776
|
private Map<String, String> matchResourceMapsAndItems(String targetId,
|
777
|
String targetResourceMapId, List<String> originalResourceMaps, List<String> aggregatedItems, String fieldName) {
|
778
|
Map<String, String> map = new HashMap<String, String>();
|
779
|
if (targetId != null && targetResourceMapId != null && aggregatedItems != null
|
780
|
&& fieldName != null) {
|
781
|
String newFieldName = null;
|
782
|
if (fieldName.equals(SolrElementField.FIELD_ISDOCUMENTEDBY)) {
|
783
|
newFieldName = SolrElementField.FIELD_DOCUMENTS;
|
784
|
} else if (fieldName.equals(SolrElementField.FIELD_DOCUMENTS)) {
|
785
|
newFieldName = SolrElementField.FIELD_ISDOCUMENTEDBY;
|
786
|
}
|
787
|
if (newFieldName != null) {
|
788
|
for (String item : aggregatedItems) {
|
789
|
SolrDoc doc = null;
|
790
|
try {
|
791
|
doc = getDocumentById(item);
|
792
|
List<String> fieldValues = doc.getAllFieldValues(newFieldName);
|
793
|
List<String> resourceMapIds = doc
|
794
|
.getAllFieldValues(SolrElementField.FIELD_RESOURCEMAP);
|
795
|
if ((fieldValues != null && fieldValues.contains(targetId))
|
796
|
&& (resourceMapIds != null && resourceMapIds
|
797
|
.contains(targetResourceMapId))) {
|
798
|
//okay, we found the target aggregation item id and the resource map id
|
799
|
//in this solr doc. However, we need check if another resource map with different
|
800
|
//id but specify the same relationship. If we have the id(s), we should not
|
801
|
// remove the documents( or documentBy) element since we need to preserve the
|
802
|
// relationship for the remain resource map.
|
803
|
boolean hasDuplicateIds = false;
|
804
|
if(originalResourceMaps != null) {
|
805
|
for(String id :resourceMapIds) {
|
806
|
if (originalResourceMaps.contains(id) && !id.equals(targetResourceMapId)) {
|
807
|
hasDuplicateIds = true;
|
808
|
break;
|
809
|
}
|
810
|
}
|
811
|
}
|
812
|
if(!hasDuplicateIds) {
|
813
|
map.put(item, targetResourceMapId);
|
814
|
}
|
815
|
|
816
|
}
|
817
|
} catch (Exception e) {
|
818
|
log.warn("SolrIndex.matchResourceMapsAndItems - can't get the solrdoc for the id "
|
819
|
+ item + " since " + e.getMessage());
|
820
|
}
|
821
|
}
|
822
|
}
|
823
|
}
|
824
|
return map;
|
825
|
}
|
826
|
|
827
|
/*
|
828
|
* Get the solr index doc from the index server for the given id.
|
829
|
*/
|
830
|
private SolrDoc getDocumentById(String id) throws NotImplemented, NotFound, UnsupportedType,
|
831
|
SolrServerException, ParserConfigurationException, SAXException, XPathExpressionException, IOException {
|
832
|
SolrDoc doc = ResourceMapSubprocessor.getSolrDoc(id);
|
833
|
return doc;
|
834
|
}
|
835
|
|
836
|
/*
|
837
|
* Merge two list of updated solr docs. removedDocumentBy has the correct information about documentBy element.
|
838
|
* removedDocuments has the correct information about the documents element.
|
839
|
* So we go through the two list and found the two docs having the same identifier.
|
840
|
* Get the list of the documents value from the one in the removedDoucments (1).
|
841
|
* Remove all values of documents from the one in the removedDocumentBy.
|
842
|
* Then copy the list of documents value from (1) to to the one in the removedDocumentBy.
|
843
|
*/
|
844
|
private List<SolrDoc> mergeUpdatedSolrDocs(List<SolrDoc>removedDocumentBy, List<SolrDoc>removedDocuments) {
|
845
|
List<SolrDoc> mergedDocuments = new ArrayList<SolrDoc>();
|
846
|
if(removedDocumentBy == null || removedDocumentBy.isEmpty()) {
|
847
|
mergedDocuments = removedDocuments;
|
848
|
} else if (removedDocuments == null || removedDocuments.isEmpty()) {
|
849
|
mergedDocuments = removedDocumentBy;
|
850
|
} else {
|
851
|
int sizeOfDocBy = removedDocumentBy.size();
|
852
|
int sizeOfDocs = removedDocuments.size();
|
853
|
for(int i=sizeOfDocBy-1; i>= 0; i--) {
|
854
|
SolrDoc docInRemovedDocBy = removedDocumentBy.get(i);
|
855
|
for(int j= sizeOfDocs-1; j>=0; j--) {
|
856
|
SolrDoc docInRemovedDocs = removedDocuments.get(j);
|
857
|
if(docInRemovedDocBy.getIdentifier().equals(docInRemovedDocs.getIdentifier())) {
|
858
|
//find the same doc in both list. let's merge them.
|
859
|
//first get all the documents element from the docWithDocs(it has the correct information about the documents element)
|
860
|
List<String> idsInDocuments = docInRemovedDocs.getAllFieldValues(SolrElementField.FIELD_DOCUMENTS);
|
861
|
docInRemovedDocBy.removeAllFields(SolrElementField.FIELD_DOCUMENTS);//clear out any documents element in docInRemovedDocBy
|
862
|
//add the Documents element from the docInRemovedDocs if it has any.
|
863
|
// The docInRemovedDocs has the correct information about the documentBy. Now it copied the correct information of the documents element.
|
864
|
// So docInRemovedDocs has both correct information about the documentBy and documents elements.
|
865
|
if(idsInDocuments != null) {
|
866
|
for(String id : idsInDocuments) {
|
867
|
if(id != null && !id.trim().equals("")) {
|
868
|
docInRemovedDocBy.addField(new SolrElementField(SolrElementField.FIELD_DOCUMENTS, id));
|
869
|
}
|
870
|
|
871
|
}
|
872
|
}
|
873
|
//intersect the resource map ids.
|
874
|
List<String> resourceMapIdsInWithDocs = docInRemovedDocs.getAllFieldValues(SolrElementField.FIELD_RESOURCEMAP);
|
875
|
List<String> resourceMapIdsInWithDocBy = docInRemovedDocBy.getAllFieldValues(SolrElementField.FIELD_RESOURCEMAP);
|
876
|
docInRemovedDocBy.removeAllFields(SolrElementField.FIELD_RESOURCEMAP);
|
877
|
Collection resourceMapIds = CollectionUtils.union(resourceMapIdsInWithDocs, resourceMapIdsInWithDocBy);
|
878
|
if(resourceMapIds != null) {
|
879
|
for(Object idObj : resourceMapIds) {
|
880
|
String id = (String)idObj;
|
881
|
docInRemovedDocBy.addField(new SolrElementField(SolrElementField.FIELD_RESOURCEMAP, id));
|
882
|
}
|
883
|
}
|
884
|
//we don't need do anything about the documentBy elements since the docInRemovedDocBy has the correct information.
|
885
|
mergedDocuments.add(docInRemovedDocBy);
|
886
|
//delete the two documents from the list
|
887
|
removedDocumentBy.remove(i);
|
888
|
removedDocuments.remove(j);
|
889
|
break;
|
890
|
}
|
891
|
|
892
|
}
|
893
|
}
|
894
|
// when we get there, if the two lists are empty, this will be a perfect merge. However, if something are left. we
|
895
|
//just put them in.
|
896
|
for(SolrDoc doc: removedDocumentBy) {
|
897
|
mergedDocuments.add(doc);
|
898
|
}
|
899
|
for(SolrDoc doc: removedDocuments) {
|
900
|
mergedDocuments.add(doc);
|
901
|
}
|
902
|
}
|
903
|
return mergedDocuments;
|
904
|
}
|
905
|
|
906
|
|
907
|
/*
|
908
|
* Remove a pid which is part of resource map.
|
909
|
*/
|
910
|
private void removeFromDataPackage(String pid) throws Exception {
|
911
|
SolrDoc indexedDoc = ResourceMapSubprocessor.getSolrDoc(pid);
|
912
|
removeFromIndex(pid);
|
913
|
List<SolrDoc> docsToUpdate = new ArrayList<SolrDoc>();
|
914
|
|
915
|
List<String> documents = indexedDoc.getAllFieldValues(SolrElementField.FIELD_DOCUMENTS);
|
916
|
for (String documentsValue : documents) {
|
917
|
SolrDoc solrDoc = ResourceMapSubprocessor.getSolrDoc(documentsValue);
|
918
|
solrDoc.removeFieldsWithValue(SolrElementField.FIELD_ISDOCUMENTEDBY, pid);
|
919
|
removeFromIndex(documentsValue);
|
920
|
insertToIndex(solrDoc);
|
921
|
}
|
922
|
|
923
|
List<String> documentedBy = indexedDoc
|
924
|
.getAllFieldValues(SolrElementField.FIELD_ISDOCUMENTEDBY);
|
925
|
for (String documentedByValue : documentedBy) {
|
926
|
SolrDoc solrDoc = ResourceMapSubprocessor.getSolrDoc(documentedByValue);
|
927
|
solrDoc.removeFieldsWithValue(SolrElementField.FIELD_DOCUMENTS, pid);
|
928
|
//docsToUpdate.add(solrDoc);
|
929
|
removeFromIndex(documentedByValue);
|
930
|
insertToIndex(solrDoc);
|
931
|
}
|
932
|
|
933
|
//SolrElementAdd addCommand = new SolrElementAdd(docsToUpdate);
|
934
|
//httpService.sendUpdate(solrIndexUri, addCommand);
|
935
|
}
|
936
|
|
937
|
/*
|
938
|
* Remove a pid from the solr index
|
939
|
*/
|
940
|
private synchronized void removeFromIndex(String identifier) throws Exception {
|
941
|
|
942
|
|
943
|
Map<String, SolrDoc> docs = new HashMap<String, SolrDoc>();
|
944
|
|
945
|
for (IDocumentDeleteSubprocessor deleteSubprocessor : deleteSubprocessors) {
|
946
|
docs.putAll(deleteSubprocessor.processDocForDelete(identifier, docs));
|
947
|
}
|
948
|
List<SolrDoc> docsToUpdate = new ArrayList<SolrDoc>();
|
949
|
List<String> idsToIndex = new ArrayList<String>();
|
950
|
for (String idToUpdate : docs.keySet()) {
|
951
|
if (docs.get(idToUpdate) != null) {
|
952
|
docsToUpdate.add(docs.get(idToUpdate));
|
953
|
} else {
|
954
|
idsToIndex.add(idToUpdate);
|
955
|
}
|
956
|
}
|
957
|
|
958
|
// update the docs we have
|
959
|
for (SolrDoc docToUpdate : docsToUpdate) {
|
960
|
insertToIndex(docToUpdate);
|
961
|
}
|
962
|
|
963
|
// delete this one
|
964
|
deleteDocFromIndex(identifier);
|
965
|
|
966
|
// index the rest
|
967
|
for (String idToIndex : idsToIndex) {
|
968
|
Identifier pid = new Identifier();
|
969
|
pid.setValue(idToIndex);
|
970
|
SystemMetadata sysMeta = DistributedMapsFactory.getSystemMetadata(idToIndex);
|
971
|
if (SolrDoc.visibleInIndex(sysMeta)) {
|
972
|
String objectPath = DistributedMapsFactory.getObjectPathMap().get(pid);
|
973
|
insert(pid, sysMeta, objectPath);
|
974
|
}
|
975
|
}
|
976
|
|
977
|
}
|
978
|
|
979
|
private void deleteDocFromIndex(String pid) throws Exception {
|
980
|
if (pid != null && !pid.trim().equals("")) {
|
981
|
/*IndexEvent event = new IndexEvent();
|
982
|
event.setDate(Calendar.getInstance().getTime());
|
983
|
Identifier identifier = new Identifier();
|
984
|
identifier.setValue(pid);
|
985
|
event.setIdentifier(identifier);*/
|
986
|
try {
|
987
|
solrServer.deleteById(pid);
|
988
|
solrServer.commit();
|
989
|
/*event.setType(IndexEvent.SUCCESSDELETE);
|
990
|
event.setDescription("Successfully remove the solr index for the id "+identifier.getValue());
|
991
|
try {
|
992
|
EventlogFactory.createIndexEventLog().write(event);
|
993
|
} catch (Exception e) {
|
994
|
log.error("SolrIndex.removeFromIndex - IndexEventLog can't log the index deleting event :"+e.getMessage());
|
995
|
}*/
|
996
|
} catch (SolrServerException e) {
|
997
|
/*event.setAction(Event.DELETE);
|
998
|
event.setDescription("Failurely remove the solr index for the id "+identifier.getValue()+" since "+e.getMessage());
|
999
|
try {
|
1000
|
EventlogFactory.createIndexEventLog().write(event);
|
1001
|
} catch (Exception ee) {
|
1002
|
log.error("SolrIndex.removeFromIndex - IndexEventLog can't log the index deleting event :"+ee.getMessage());
|
1003
|
}*/
|
1004
|
throw e;
|
1005
|
|
1006
|
} catch (IOException e) {
|
1007
|
/*event.setAction(Event.DELETE);
|
1008
|
event.setDescription("Failurely remove the solr index for the id "+identifier.getValue()+" since "+e.getMessage());
|
1009
|
try {
|
1010
|
EventlogFactory.createIndexEventLog().write(event);
|
1011
|
} catch (Exception ee) {
|
1012
|
log.error("SolrIndex.removeFromIndex - IndexEventLog can't log the index deleting event :"+ee.getMessage());
|
1013
|
}*/
|
1014
|
throw e;
|
1015
|
}
|
1016
|
|
1017
|
}
|
1018
|
|
1019
|
}
|
1020
|
|
1021
|
/**
|
1022
|
* Get the solrServer
|
1023
|
* @return
|
1024
|
*/
|
1025
|
public SolrServer getSolrServer() {
|
1026
|
return solrServer;
|
1027
|
}
|
1028
|
|
1029
|
/**
|
1030
|
* Set the solrServer.
|
1031
|
* @param solrServer
|
1032
|
*/
|
1033
|
public void setSolrServer(SolrServer solrServer) {
|
1034
|
this.solrServer = solrServer;
|
1035
|
}
|
1036
|
|
1037
|
/**
|
1038
|
* Get all indexed ids in the solr server.
|
1039
|
* @return an empty list if there is no index.
|
1040
|
* @throws SolrServerException
|
1041
|
*/
|
1042
|
public List<String> getSolrIds() throws SolrServerException {
|
1043
|
List<String> list = new ArrayList<String>();
|
1044
|
SolrQuery query = new SolrQuery(IDQUERY);
|
1045
|
query.setRows(Integer.MAX_VALUE);
|
1046
|
query.setFields(ID);
|
1047
|
QueryResponse response = solrServer.query(query);
|
1048
|
SolrDocumentList docs = response.getResults();
|
1049
|
if(docs != null) {
|
1050
|
for(SolrDocument doc :docs) {
|
1051
|
String identifier = (String)doc.getFieldValue(ID);
|
1052
|
//System.out.println("======================== "+identifier);
|
1053
|
list.add(identifier);
|
1054
|
}
|
1055
|
}
|
1056
|
return list;
|
1057
|
}
|
1058
|
|
1059
|
private void writeEventLog(SystemMetadata systemMetadata, Identifier pid, String error) {
|
1060
|
IndexEvent event = new IndexEvent();
|
1061
|
event.setIdentifier(pid);
|
1062
|
event.setDate(Calendar.getInstance().getTime());
|
1063
|
String action = null;
|
1064
|
if (systemMetadata == null ) {
|
1065
|
action = Event.CREATE.xmlValue();
|
1066
|
event.setAction(Event.CREATE);
|
1067
|
}
|
1068
|
else if(systemMetadata.getArchived() != null && systemMetadata.getArchived()) {
|
1069
|
action = Event.DELETE.xmlValue();
|
1070
|
event.setAction(Event.DELETE);
|
1071
|
} else {
|
1072
|
action = Event.CREATE.xmlValue();
|
1073
|
event.setAction(Event.CREATE);
|
1074
|
}
|
1075
|
event.setDescription("Failed to "+action+"the solr index for the id "+pid.getValue()+" since "+error);
|
1076
|
try {
|
1077
|
EventlogFactory.createIndexEventLog().write(event);
|
1078
|
} catch (Exception ee) {
|
1079
|
log.error("SolrIndex.insertToIndex - IndexEventLog can't log the index inserting event :"+ee.getMessage());
|
1080
|
}
|
1081
|
}
|
1082
|
}
|