1 |
7606
|
tao
|
/**
|
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.InputStream;
|
30 |
|
|
import java.util.ArrayList;
|
31 |
|
|
import java.util.Date;
|
32 |
|
|
import java.util.List;
|
33 |
|
|
import java.util.Vector;
|
34 |
|
|
|
35 |
|
|
import org.apache.commons.logging.Log;
|
36 |
|
|
import org.apache.commons.logging.LogFactory;
|
37 |
|
|
import org.apache.solr.client.solrj.SolrServerException;
|
38 |
|
|
import org.dataone.client.MNode;
|
39 |
|
|
import org.dataone.configuration.Settings;
|
40 |
|
|
import org.dataone.service.exceptions.InvalidRequest;
|
41 |
|
|
import org.dataone.service.exceptions.InvalidToken;
|
42 |
|
|
import org.dataone.service.exceptions.NotAuthorized;
|
43 |
|
|
import org.dataone.service.exceptions.NotImplemented;
|
44 |
|
|
import org.dataone.service.exceptions.ServiceFailure;
|
45 |
|
|
import org.dataone.service.types.v1.Identifier;
|
46 |
7607
|
tao
|
import org.dataone.service.types.v1.ObjectInfo;
|
47 |
7606
|
tao
|
import org.dataone.service.types.v1.ObjectList;
|
48 |
|
|
import org.dataone.service.types.v1.SystemMetadata;
|
49 |
|
|
|
50 |
|
|
|
51 |
|
|
/**
|
52 |
|
|
* A class represents the object to generate massive solr indexes.
|
53 |
|
|
* This can happen during an update of Metacat (generating index for all existing documents)
|
54 |
|
|
* or regenerate index for those documents
|
55 |
|
|
* failing to build index during the insert or update.
|
56 |
|
|
*
|
57 |
|
|
* @author tao
|
58 |
|
|
*
|
59 |
|
|
*/
|
60 |
7613
|
tao
|
public class IndexGenerator implements Runnable {
|
61 |
7606
|
tao
|
|
62 |
7687
|
tao
|
private static final int WAITTIME = 10000;
|
63 |
|
|
private static final int MAXWAITNUMBER = 180;
|
64 |
7606
|
tao
|
private static final String HTTP = "http://";
|
65 |
|
|
private static final String MNAPPENDIX = "/d1/mn";
|
66 |
|
|
private SolrIndex solrIndex = null;
|
67 |
|
|
private SystemMetadataEventListener systemMetadataListener = null;
|
68 |
|
|
private Log log = LogFactory.getLog(IndexGenerator.class);
|
69 |
|
|
private MNode mNode = null;
|
70 |
|
|
|
71 |
|
|
/**
|
72 |
|
|
* Constructor
|
73 |
|
|
* @param solrIndex
|
74 |
|
|
* @param systemMetadataListener
|
75 |
|
|
*/
|
76 |
|
|
public IndexGenerator(SolrIndex solrIndex, SystemMetadataEventListener systemMetadataListener) {
|
77 |
|
|
this.solrIndex = solrIndex;
|
78 |
|
|
this.systemMetadataListener = systemMetadataListener;
|
79 |
|
|
this.mNode = new MNode(buildMNBaseURL());
|
80 |
|
|
}
|
81 |
|
|
|
82 |
|
|
/**
|
83 |
|
|
* Build the index for all documents in Metacat without overwriting.
|
84 |
|
|
* @throws SolrServerException
|
85 |
|
|
* @throws ServiceFailure
|
86 |
|
|
* @throws NotImplemented
|
87 |
|
|
* @throws NotAuthorized
|
88 |
|
|
* @throws InvalidToken
|
89 |
|
|
* @throws InvalidRequest
|
90 |
|
|
*/
|
91 |
|
|
public void indexAll() throws InvalidRequest, InvalidToken, NotAuthorized,
|
92 |
|
|
NotImplemented, ServiceFailure, SolrServerException {
|
93 |
|
|
boolean force = false;
|
94 |
|
|
indexAll(force);
|
95 |
|
|
}
|
96 |
|
|
|
97 |
|
|
/**
|
98 |
|
|
* Build the index for all documents. If force is true, the existed index for documents
|
99 |
|
|
* will be overwritten.
|
100 |
|
|
* @param force
|
101 |
|
|
* @throws SolrServerException
|
102 |
|
|
* @throws ServiceFailure
|
103 |
|
|
* @throws NotImplemented
|
104 |
|
|
* @throws NotAuthorized
|
105 |
|
|
* @throws InvalidToken
|
106 |
|
|
* @throws InvalidRequest
|
107 |
|
|
*/
|
108 |
|
|
public void indexAll(boolean force) throws InvalidRequest, InvalidToken,
|
109 |
|
|
NotAuthorized, NotImplemented, ServiceFailure, SolrServerException {
|
110 |
|
|
Date since = null;
|
111 |
|
|
Date until = null;
|
112 |
|
|
index(since, until, force);
|
113 |
|
|
}
|
114 |
|
|
|
115 |
|
|
/**
|
116 |
|
|
* Build the index for the docs which have been modified since the specified date.
|
117 |
|
|
* @param since
|
118 |
|
|
* @param force
|
119 |
|
|
* @throws SolrServerException
|
120 |
|
|
* @throws ServiceFailure
|
121 |
|
|
* @throws NotImplemented
|
122 |
|
|
* @throws NotAuthorized
|
123 |
|
|
* @throws InvalidToken
|
124 |
|
|
* @throws InvalidRequest
|
125 |
|
|
*/
|
126 |
|
|
public void index(Date since, boolean force) throws InvalidRequest, InvalidToken,
|
127 |
|
|
NotAuthorized, NotImplemented, ServiceFailure, SolrServerException {
|
128 |
|
|
Date until = null;
|
129 |
|
|
index(since, until, force);
|
130 |
|
|
}
|
131 |
|
|
|
132 |
|
|
/**
|
133 |
|
|
* Build the index for the docs which have been modified between the specified date.s
|
134 |
|
|
* @param since
|
135 |
|
|
* @param until
|
136 |
|
|
* @param force
|
137 |
|
|
* @throws SolrServerException
|
138 |
|
|
* @throws ServiceFailure
|
139 |
|
|
* @throws NotImplemented
|
140 |
|
|
* @throws NotAuthorized
|
141 |
|
|
* @throws InvalidToken
|
142 |
|
|
* @throws InvalidRequest
|
143 |
|
|
*/
|
144 |
|
|
public void index(Date since, Date until, boolean force) throws SolrServerException, InvalidRequest,
|
145 |
|
|
InvalidToken, NotAuthorized, NotImplemented, ServiceFailure {
|
146 |
|
|
List<String> solrIds = null;
|
147 |
|
|
List<String> metacatIds = null;
|
148 |
|
|
if(!force) {
|
149 |
|
|
solrIds = getSolrDocIds();
|
150 |
|
|
}
|
151 |
7607
|
tao
|
log.info("the solr ids -----------------------------"+solrIds);
|
152 |
7606
|
tao
|
metacatIds = getMetadataIds(since, until);
|
153 |
7607
|
tao
|
log.info("the metacat ids -----------------------------"+metacatIds);
|
154 |
7688
|
tao
|
if(metacatIds != null) {
|
155 |
7606
|
tao
|
for(String metacatId : metacatIds) {
|
156 |
|
|
if(metacatId != null) {
|
157 |
|
|
boolean buildIndex = true;
|
158 |
|
|
if(!force && solrIds != null && solrIds.contains(metacatId)) {
|
159 |
|
|
//solr already indexs the id and we don't force it to rebuild it, so set the buildIndex to be false
|
160 |
|
|
buildIndex = false;
|
161 |
|
|
}
|
162 |
|
|
if(buildIndex) {
|
163 |
|
|
try {
|
164 |
|
|
generateIndex(metacatId);
|
165 |
|
|
} catch (Exception e) {
|
166 |
|
|
log.error("IndexGenerator.index - Metacat Index couldn't generate the index for the id - "+metacatId+" because "+e.getMessage());
|
167 |
|
|
}
|
168 |
|
|
|
169 |
|
|
}
|
170 |
|
|
}
|
171 |
|
|
}
|
172 |
7688
|
tao
|
}
|
173 |
7606
|
tao
|
}
|
174 |
|
|
|
175 |
7613
|
tao
|
public void run() {
|
176 |
|
|
try {
|
177 |
|
|
indexAll();
|
178 |
|
|
} catch (InvalidRequest e) {
|
179 |
|
|
// TODO Auto-generated catch block
|
180 |
|
|
//e.printStackTrace();
|
181 |
|
|
log.error("IndexGenerator.run - Metadata-Index couldn't generate indexes for those documents which haven't been indexed : "+e.getMessage());
|
182 |
|
|
} catch (InvalidToken e) {
|
183 |
|
|
// TODO Auto-generated catch block
|
184 |
|
|
//e.printStackTrace();
|
185 |
|
|
log.error("IndexGenerator.run - Metadata-Index couldn't generate indexes for those documents which haven't been indexed : "+e.getMessage());
|
186 |
|
|
} catch (NotAuthorized e) {
|
187 |
|
|
// TODO Auto-generated catch block
|
188 |
|
|
//e.printStackTrace();
|
189 |
|
|
} catch (NotImplemented e) {
|
190 |
|
|
// TODO Auto-generated catch block
|
191 |
|
|
//e.printStackTrace();
|
192 |
|
|
log.error("IndexGenerator.run - Metadata-Index couldn't generate indexes for those documents which haven't been indexed : "+e.getMessage());
|
193 |
|
|
} catch (ServiceFailure e) {
|
194 |
|
|
// TODO Auto-generated catch block
|
195 |
|
|
//e.printStackTrace();
|
196 |
|
|
log.error("IndexGenerator.run - Metadata-Index couldn't generate indexes for those documents which haven't been indexed : "+e.getMessage());
|
197 |
|
|
} catch (SolrServerException e) {
|
198 |
|
|
// TODO Auto-generated catch block
|
199 |
|
|
//e.printStackTrace();
|
200 |
|
|
log.error("IndexGenerator.run - Metadata-Index couldn't generate indexes for those documents which haven't been indexed : "+e.getMessage());
|
201 |
|
|
}
|
202 |
|
|
}
|
203 |
|
|
|
204 |
7606
|
tao
|
/*
|
205 |
|
|
* Get the indexed ids list from the solr server.
|
206 |
|
|
* An empty list will be returned if there is no ids.
|
207 |
|
|
*/
|
208 |
|
|
private List<String> getSolrDocIds() throws SolrServerException {
|
209 |
|
|
List<String> ids = solrIndex.getSolrIds();
|
210 |
|
|
return ids;
|
211 |
|
|
}
|
212 |
|
|
|
213 |
|
|
/*
|
214 |
|
|
* Get the ids of the metacat. If since and util are null, it will return all of them
|
215 |
|
|
*/
|
216 |
|
|
private List<String> getMetadataIds(Date since, Date until) throws InvalidRequest,
|
217 |
|
|
InvalidToken, NotAuthorized, NotImplemented, ServiceFailure {
|
218 |
7607
|
tao
|
List<String> ids = new ArrayList();
|
219 |
7606
|
tao
|
ObjectList objects = null;
|
220 |
7687
|
tao
|
int times = 0;
|
221 |
|
|
while (true) {
|
222 |
|
|
try {
|
223 |
|
|
mNode.ping();
|
224 |
|
|
break;
|
225 |
|
|
} catch (Exception e) {
|
226 |
|
|
if(times <= MAXWAITNUMBER) {
|
227 |
|
|
log.warn("IndexGenerator.getMetadataIds - the mnode "+ mNode.getNodeBaseServiceUrl()+
|
228 |
|
|
" is not ready :" +e.getMessage()+"\nWe will try to access it 10 seconds later ");
|
229 |
|
|
try {
|
230 |
|
|
Thread.sleep(WAITTIME);
|
231 |
|
|
} catch (Exception ee) {
|
232 |
|
|
log.warn("IndexGenerator.getMetadataIds - the thread can't sleep for 10 seconds to wait the MNode");
|
233 |
|
|
}
|
234 |
|
|
|
235 |
|
|
} else {
|
236 |
|
|
throw new ServiceFailure("503", "IndexGenerator.getMetadataIds - the mnode "+ mNode.getNodeBaseServiceUrl()+
|
237 |
|
|
" is not ready even though Metacat-index wailted for 30 minutes. We can't get the objects list from it and the building index can't happen this time");
|
238 |
|
|
}
|
239 |
|
|
|
240 |
|
|
}
|
241 |
|
|
times++;
|
242 |
|
|
}
|
243 |
7606
|
tao
|
if(since == null && until == null) {
|
244 |
|
|
objects = mNode.listObjects();
|
245 |
|
|
} else {
|
246 |
|
|
objects = mNode.listObjects(since, until, null, true, 0, Integer.MAX_VALUE);
|
247 |
|
|
}
|
248 |
7607
|
tao
|
if(objects != null) {
|
249 |
|
|
List<ObjectInfo> objectInfoList = objects.getObjectInfoList();
|
250 |
|
|
if(objectInfoList != null) {
|
251 |
|
|
for(ObjectInfo info : objectInfoList) {
|
252 |
|
|
if(info != null) {
|
253 |
|
|
Identifier identifier = info.getIdentifier();
|
254 |
|
|
if(identifier != null && identifier.getValue() != null && !identifier.getValue().equals("")) {
|
255 |
|
|
ids.add(identifier.getValue());
|
256 |
|
|
}
|
257 |
|
|
}
|
258 |
|
|
}
|
259 |
|
|
}
|
260 |
|
|
}
|
261 |
7606
|
tao
|
return ids;
|
262 |
|
|
}
|
263 |
|
|
|
264 |
|
|
/*
|
265 |
|
|
* Build up the mn base url
|
266 |
|
|
*/
|
267 |
|
|
private String buildMNBaseURL() {
|
268 |
7607
|
tao
|
String url = HTTP+Settings.getConfiguration().getString("server.name")+":"+
|
269 |
7606
|
tao
|
Settings.getConfiguration().getString("server.httpPort")+"/"+
|
270 |
7607
|
tao
|
Settings.getConfiguration().getString("application.context")+MNAPPENDIX;
|
271 |
7606
|
tao
|
log.info("IndexGenerator.buildMNBaseURL - the base url of MNode is "+url);
|
272 |
|
|
return url;
|
273 |
|
|
}
|
274 |
|
|
|
275 |
|
|
|
276 |
|
|
/*
|
277 |
|
|
* Generate index for the id.
|
278 |
|
|
*/
|
279 |
|
|
private void generateIndex(String id) throws Exception {
|
280 |
|
|
if(id != null) {
|
281 |
|
|
SystemMetadata sysmeta = systemMetadataListener.getSystemMetadata(id);
|
282 |
|
|
if(sysmeta != null) {
|
283 |
|
|
InputStream data = systemMetadataListener.getDataObject(id);
|
284 |
|
|
Identifier obsolete = sysmeta.getObsoletes();
|
285 |
7627
|
tao
|
List<String> obsoleteChain = null;
|
286 |
7606
|
tao
|
if(obsolete != null) {
|
287 |
7627
|
tao
|
obsoleteChain = systemMetadataListener.getObsoletes(id);
|
288 |
|
|
}
|
289 |
|
|
solrIndex.update(id, obsoleteChain, sysmeta, data);
|
290 |
7606
|
tao
|
} else {
|
291 |
|
|
throw new Exception("IndexGenerator.generate - there is no found SystemMetadata associated with the id "+id);
|
292 |
|
|
}
|
293 |
|
|
|
294 |
|
|
}
|
295 |
|
|
}
|
296 |
|
|
}
|