1
|
package edu.ucsb.nceas.metacat.admin.upgrade.dataone;
|
2
|
/**
|
3
|
* '$RCSfile$'
|
4
|
* Purpose: A Class for upgrading the database to version 1.5
|
5
|
* Copyright: 2000 Regents of the University of California and the
|
6
|
* National Center for Ecological Analysis and Synthesis
|
7
|
* Authors: Saurabh Garg
|
8
|
*
|
9
|
* '$Author: leinfelder $'
|
10
|
* '$Date: 2014-07-23 16:19:48 -0700 (Wed, 23 Jul 2014) $'
|
11
|
* '$Revision: 8810 $'
|
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
|
|
28
|
|
29
|
import java.io.InputStream;
|
30
|
import java.math.BigInteger;
|
31
|
import java.util.ArrayList;
|
32
|
import java.util.Calendar;
|
33
|
import java.util.Collections;
|
34
|
import java.util.Date;
|
35
|
import java.util.List;
|
36
|
import java.util.Map;
|
37
|
|
38
|
import org.apache.commons.beanutils.BeanUtils;
|
39
|
import org.apache.commons.io.IOUtils;
|
40
|
import org.apache.commons.logging.Log;
|
41
|
import org.apache.commons.logging.LogFactory;
|
42
|
import org.dataone.client.v2.itk.D1Client;
|
43
|
import org.dataone.client.v2.MNode;
|
44
|
import org.dataone.client.v2.formats.ObjectFormatCache;
|
45
|
import org.dataone.ore.ResourceMapFactory;
|
46
|
import org.dataone.service.types.v1.Checksum;
|
47
|
import org.dataone.service.types.v1.Identifier;
|
48
|
import org.dataone.service.types.v1.ObjectFormatIdentifier;
|
49
|
import org.dataone.service.types.v1.ObjectInfo;
|
50
|
import org.dataone.service.types.v1.ObjectList;
|
51
|
import org.dataone.service.types.v2.SystemMetadata;
|
52
|
import org.dataone.service.types.v1.util.ChecksumUtil;
|
53
|
import org.dspace.foresite.ResourceMap;
|
54
|
|
55
|
import edu.ucsb.nceas.metacat.DBUtil;
|
56
|
import edu.ucsb.nceas.metacat.DocumentImpl;
|
57
|
import edu.ucsb.nceas.metacat.MetaCatServlet;
|
58
|
import edu.ucsb.nceas.metacat.admin.AdminException;
|
59
|
import edu.ucsb.nceas.metacat.admin.upgrade.UpgradeUtilityInterface;
|
60
|
import edu.ucsb.nceas.metacat.dataone.SystemMetadataFactory;
|
61
|
import edu.ucsb.nceas.metacat.properties.PropertyService;
|
62
|
import edu.ucsb.nceas.utilities.PropertyNotFoundException;
|
63
|
import edu.ucsb.nceas.utilities.SortedProperties;
|
64
|
|
65
|
public class GenerateORE implements UpgradeUtilityInterface {
|
66
|
|
67
|
private static Log log = LogFactory.getLog(GenerateORE.class);
|
68
|
|
69
|
private int serverLocation = 1;
|
70
|
|
71
|
public boolean upgrade() throws AdminException {
|
72
|
boolean success = true;
|
73
|
|
74
|
// include ORE, data, for this server only
|
75
|
boolean includeOre = true;
|
76
|
boolean downloadData = false;
|
77
|
try {
|
78
|
downloadData = Boolean.parseBoolean(PropertyService.getProperty("dataone.ore.downloaddata"));
|
79
|
} catch (PropertyNotFoundException e) {
|
80
|
// ignore, default to false
|
81
|
log.warn("Could not find ORE 'dataone.ore.downloaddata' property, defaulting to false", e);
|
82
|
}
|
83
|
|
84
|
|
85
|
try {
|
86
|
// get only local ids for this server
|
87
|
List<String> idList = null;
|
88
|
|
89
|
idList = DBUtil.getAllDocidsByType(DocumentImpl.EML2_0_0NAMESPACE, true, serverLocation);
|
90
|
filterOutExisting(idList);
|
91
|
Collections.sort(idList);
|
92
|
SystemMetadataFactory.generateSystemMetadata(idList, includeOre, downloadData);
|
93
|
|
94
|
idList = DBUtil.getAllDocidsByType(DocumentImpl.EML2_0_1NAMESPACE, true, serverLocation);
|
95
|
filterOutExisting(idList);
|
96
|
Collections.sort(idList);
|
97
|
SystemMetadataFactory.generateSystemMetadata(idList, includeOre, downloadData);
|
98
|
|
99
|
idList = DBUtil.getAllDocidsByType(DocumentImpl.EML2_1_0NAMESPACE, true, serverLocation);
|
100
|
filterOutExisting(idList);
|
101
|
Collections.sort(idList);
|
102
|
SystemMetadataFactory.generateSystemMetadata(idList, includeOre, downloadData);
|
103
|
|
104
|
idList = DBUtil.getAllDocidsByType(DocumentImpl.EML2_1_1NAMESPACE, true, serverLocation);
|
105
|
filterOutExisting(idList);
|
106
|
Collections.sort(idList);
|
107
|
SystemMetadataFactory.generateSystemMetadata(idList, includeOre, downloadData);
|
108
|
|
109
|
} catch (Exception e) {
|
110
|
String msg = "Problem generating missing system metadata: " + e.getMessage();
|
111
|
log.error(msg, e);
|
112
|
success = false;
|
113
|
throw new AdminException(msg);
|
114
|
}
|
115
|
return success;
|
116
|
}
|
117
|
|
118
|
private List<String> filterOutExisting(List<String> idList) {
|
119
|
List<String> toRemove = new ArrayList<String>();
|
120
|
for (String id: idList) {
|
121
|
Identifier identifier = new Identifier();
|
122
|
identifier.setValue(id);
|
123
|
boolean exists = SystemMetadataFactory.oreExistsFor(identifier);
|
124
|
if (exists) {
|
125
|
toRemove.add(id);
|
126
|
}
|
127
|
}
|
128
|
for (String id: toRemove) {
|
129
|
idList.remove(id);
|
130
|
}
|
131
|
return idList;
|
132
|
}
|
133
|
|
134
|
public int getServerLocation() {
|
135
|
return serverLocation;
|
136
|
}
|
137
|
|
138
|
public void setServerLocation(int serverLocation) {
|
139
|
this.serverLocation = serverLocation;
|
140
|
}
|
141
|
|
142
|
/**
|
143
|
* Need to update the existing ORE maps to have correct dateTime serializations
|
144
|
* see: https://redmine.dataone.org/issues/3046
|
145
|
* @param mnBaseUrl
|
146
|
*/
|
147
|
public static void updateOREdateFormat(String mnBaseUrl) {
|
148
|
|
149
|
List<Identifier> orePids = getAllOREpids(mnBaseUrl);
|
150
|
updateOREs(orePids, "b", mnBaseUrl);
|
151
|
}
|
152
|
|
153
|
/**
|
154
|
* Retrieves a list of all ORE objects on the given MN
|
155
|
* @param mnBaseUrl
|
156
|
* @return
|
157
|
*/
|
158
|
public static List<Identifier> getAllOREpids(String mnBaseUrl) {
|
159
|
|
160
|
MNode mn = null;
|
161
|
ObjectFormatIdentifier formatId = null;
|
162
|
List<Identifier> pids = null;
|
163
|
try {
|
164
|
|
165
|
// get the MN
|
166
|
mn = D1Client.getMN(mnBaseUrl);
|
167
|
|
168
|
// get the ORE format id
|
169
|
formatId = ObjectFormatCache.getInstance().getFormat("http://www.openarchives.org/ore/terms").getFormatId();
|
170
|
|
171
|
// get the objects that match
|
172
|
ObjectList objectList = mn.listObjects(null, null, null, formatId, null, null, 0, Integer.MAX_VALUE);
|
173
|
pids = new ArrayList<Identifier>();
|
174
|
for (ObjectInfo o: objectList.getObjectInfoList()) {
|
175
|
pids.add(o.getIdentifier());
|
176
|
}
|
177
|
|
178
|
} catch (Exception e) {
|
179
|
log.error("Could not get MN list of ORE pids", e);
|
180
|
}
|
181
|
|
182
|
return pids;
|
183
|
|
184
|
}
|
185
|
|
186
|
/**
|
187
|
* Updates the given OREs by regenerating and reserializing the RDF using the updated foresite library
|
188
|
* Only non-obsolete, non-archived ORE objects are updated and their SystemMetadata is based on the original version.
|
189
|
* see: https://redmine.dataone.org/issues/3046
|
190
|
* @param orePids
|
191
|
* @param pidSuffix
|
192
|
* @param mnBaseUrl
|
193
|
*/
|
194
|
public static void updateOREs(List<Identifier> orePids, String pidSuffix, String mnBaseUrl) {
|
195
|
|
196
|
// get a MN client for this local node, or use the given baseUrl
|
197
|
MNode mn = null;
|
198
|
try {
|
199
|
mn = D1Client.getMN(mnBaseUrl);
|
200
|
} catch (Exception e) {
|
201
|
log.error("Could not get MN client", e);
|
202
|
// nothing more we can do here
|
203
|
return;
|
204
|
}
|
205
|
|
206
|
// make sure we have something for the suffix
|
207
|
if (pidSuffix == null) {
|
208
|
pidSuffix = "b";
|
209
|
}
|
210
|
|
211
|
for (Identifier orePid: orePids) {
|
212
|
try {
|
213
|
|
214
|
log.debug("processing ORE pid: " + orePid.getValue());
|
215
|
|
216
|
// get original SystemMetadata
|
217
|
SystemMetadata originalOreSysMeta = mn.getSystemMetadata(null, orePid);
|
218
|
|
219
|
// only update the CURRENT revision of the ORE
|
220
|
if (originalOreSysMeta.getObsoletedBy() != null || (originalOreSysMeta.getArchived() != null && originalOreSysMeta.getArchived())) {
|
221
|
log.debug("ORE pid is obsolete or archived, skipping: " + orePid.getValue());
|
222
|
continue;
|
223
|
}
|
224
|
|
225
|
// get the original ORE map
|
226
|
InputStream originalOreStream = mn.get(null, orePid);
|
227
|
Map<Identifier, Map<Identifier, List<Identifier>>> originalOre = ResourceMapFactory.getInstance().parseResourceMap(originalOreStream);
|
228
|
|
229
|
// generate the updated ORE map, in this case we aren't changing any values, just altering the serialization using a newer foresite library
|
230
|
Identifier updatedOrePid = new Identifier();
|
231
|
updatedOrePid.setValue(orePid.getValue() + pidSuffix);
|
232
|
ResourceMap updatedOre = ResourceMapFactory.getInstance().createResourceMap(updatedOrePid , originalOre.entrySet().iterator().next().getValue());
|
233
|
String resourceMapXML = ResourceMapFactory.getInstance().serializeResourceMap(updatedOre);
|
234
|
Checksum oreChecksum = ChecksumUtil.checksum(IOUtils.toInputStream(resourceMapXML, MetaCatServlet.DEFAULT_ENCODING), "MD5");
|
235
|
Date today = Calendar.getInstance().getTime();
|
236
|
|
237
|
// copy the existing SystemMetada into the new SystemMetadata
|
238
|
SystemMetadata updatedOreSysMeta = new SystemMetadata();
|
239
|
BeanUtils.copyProperties(updatedOreSysMeta, originalOreSysMeta);
|
240
|
|
241
|
// set the new SystemMetadata values
|
242
|
updatedOreSysMeta.setIdentifier(updatedOrePid);
|
243
|
updatedOreSysMeta.setObsoletes(orePid);
|
244
|
updatedOreSysMeta.setObsoletedBy(null);
|
245
|
updatedOreSysMeta.setArchived(false);
|
246
|
updatedOreSysMeta.setChecksum(oreChecksum);
|
247
|
updatedOreSysMeta.setSize(BigInteger.valueOf(SystemMetadataFactory.sizeOfStream(IOUtils.toInputStream(resourceMapXML, MetaCatServlet.DEFAULT_ENCODING))));
|
248
|
updatedOreSysMeta.setDateSysMetadataModified(today);
|
249
|
updatedOreSysMeta.setDateUploaded(today);
|
250
|
updatedOreSysMeta.setReplicaList(null);
|
251
|
|
252
|
// save the updated ORE to the MN
|
253
|
InputStream updatedOreStream = IOUtils.toInputStream(resourceMapXML, MetaCatServlet.DEFAULT_ENCODING);
|
254
|
mn.update(null, orePid, updatedOreStream, updatedOrePid, updatedOreSysMeta);
|
255
|
|
256
|
|
257
|
} catch (Exception e) {
|
258
|
log.error("Could not update ORE map: " + orePid, e);
|
259
|
|
260
|
// go to the next record, there's nothing else to do here
|
261
|
continue;
|
262
|
}
|
263
|
|
264
|
}
|
265
|
}
|
266
|
|
267
|
public static void main(String [] ags){
|
268
|
|
269
|
try {
|
270
|
// set up the properties based on the test/deployed configuration of the workspace
|
271
|
SortedProperties testProperties =
|
272
|
new SortedProperties("test/test.properties");
|
273
|
testProperties.load();
|
274
|
String metacatContextDir = testProperties.getProperty("metacat.contextDir");
|
275
|
PropertyService.getInstance(metacatContextDir + "/WEB-INF");
|
276
|
// now run it
|
277
|
GenerateORE upgrader = new GenerateORE();
|
278
|
upgrader.upgrade();
|
279
|
|
280
|
} catch (Exception ex) {
|
281
|
System.out.println("Exception:" + ex.getMessage());
|
282
|
ex.printStackTrace();
|
283
|
}
|
284
|
}
|
285
|
}
|