1 |
7696
|
tao
|
/**
|
2 |
|
|
* This program is free software; you can redistribute it and/or modify
|
3 |
|
|
* it under the terms of the GNU General Public License as published by
|
4 |
|
|
* the Free Software Foundation; either version 2 of the License, or
|
5 |
|
|
* (at your option) any later version.
|
6 |
|
|
*
|
7 |
|
|
* This program is distributed in the hope that it will be useful,
|
8 |
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
9 |
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
10 |
|
|
* GNU General Public License for more details.
|
11 |
|
|
*
|
12 |
|
|
* You should have received a copy of the GNU General Public License
|
13 |
|
|
* along with this program; if not, write to the Free Software
|
14 |
|
|
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
|
15 |
|
|
*/
|
16 |
|
|
package edu.ucsb.nceas.metacat.index.resourcemap;
|
17 |
|
|
|
18 |
|
|
import java.io.IOException;
|
19 |
9018
|
leinfelder
|
import java.io.InputStream;
|
20 |
8893
|
tao
|
import java.net.MalformedURLException;
|
21 |
7696
|
tao
|
import java.util.ArrayList;
|
22 |
7710
|
tao
|
import java.util.Date;
|
23 |
7696
|
tao
|
import java.util.HashMap;
|
24 |
|
|
import java.util.List;
|
25 |
|
|
import java.util.Map;
|
26 |
|
|
|
27 |
|
|
import javax.xml.parsers.ParserConfigurationException;
|
28 |
|
|
import javax.xml.xpath.XPathExpressionException;
|
29 |
|
|
|
30 |
|
|
import org.apache.commons.codec.EncoderException;
|
31 |
|
|
import org.apache.commons.logging.Log;
|
32 |
|
|
import org.apache.commons.logging.LogFactory;
|
33 |
|
|
import org.apache.solr.client.solrj.SolrServer;
|
34 |
|
|
import org.apache.solr.client.solrj.SolrServerException;
|
35 |
|
|
import org.apache.solr.client.solrj.response.QueryResponse;
|
36 |
|
|
import org.apache.solr.common.SolrDocument;
|
37 |
|
|
import org.apache.solr.common.params.SolrParams;
|
38 |
7710
|
tao
|
import org.apache.solr.schema.IndexSchema;
|
39 |
7696
|
tao
|
import org.apache.solr.servlet.SolrRequestParsers;
|
40 |
9028
|
leinfelder
|
import org.dataone.cn.indexer.XmlDocumentUtility;
|
41 |
8464
|
leinfelder
|
import org.dataone.cn.indexer.convert.SolrDateConverter;
|
42 |
9028
|
leinfelder
|
import org.dataone.cn.indexer.parser.BaseXPathDocumentSubprocessor;
|
43 |
7696
|
tao
|
import org.dataone.cn.indexer.parser.IDocumentSubprocessor;
|
44 |
|
|
import org.dataone.cn.indexer.resourcemap.ResourceMap;
|
45 |
8296
|
tao
|
import org.dataone.cn.indexer.resourcemap.ResourceMapFactory;
|
46 |
7696
|
tao
|
import org.dataone.cn.indexer.solrhttp.SolrDoc;
|
47 |
|
|
import org.dataone.cn.indexer.solrhttp.SolrElementField;
|
48 |
7733
|
tao
|
import org.dataone.service.exceptions.NotFound;
|
49 |
|
|
import org.dataone.service.exceptions.NotImplemented;
|
50 |
|
|
import org.dataone.service.exceptions.UnsupportedType;
|
51 |
8464
|
leinfelder
|
import org.dataone.service.util.DateTimeMarshaller;
|
52 |
8023
|
tao
|
import org.dspace.foresite.OREParserException;
|
53 |
7696
|
tao
|
import org.w3c.dom.Document;
|
54 |
|
|
import org.xml.sax.SAXException;
|
55 |
|
|
|
56 |
8464
|
leinfelder
|
import edu.ucsb.nceas.metacat.common.SolrServerFactory;
|
57 |
7733
|
tao
|
import edu.ucsb.nceas.metacat.common.query.SolrQueryServiceController;
|
58 |
7696
|
tao
|
import edu.ucsb.nceas.metacat.index.SolrIndex;
|
59 |
|
|
|
60 |
7710
|
tao
|
|
61 |
7696
|
tao
|
/**
|
62 |
|
|
* A solr index parser for the ResourceMap file.
|
63 |
|
|
* The solr doc of the ResourceMap self only has the system metadata information.
|
64 |
|
|
* The solr docs of the science metadata doc and data file have the resource map package information.
|
65 |
|
|
*/
|
66 |
9028
|
leinfelder
|
public class ResourceMapSubprocessor extends BaseXPathDocumentSubprocessor implements IDocumentSubprocessor {
|
67 |
7696
|
tao
|
|
68 |
|
|
private static final String QUERY ="q=id:";
|
69 |
8893
|
tao
|
private static final String QUERY2="q="+SolrElementField.FIELD_RESOURCEMAP+":";
|
70 |
7696
|
tao
|
private static Log log = LogFactory.getLog(SolrIndex.class);
|
71 |
|
|
private static SolrServer solrServer = null;
|
72 |
|
|
static {
|
73 |
|
|
try {
|
74 |
|
|
solrServer = SolrServerFactory.createSolrServer();
|
75 |
|
|
} catch (Exception e) {
|
76 |
|
|
log.error("ResourceMapSubprocessor - can't generate the SolrServer since - "+e.getMessage());
|
77 |
|
|
}
|
78 |
|
|
}
|
79 |
|
|
|
80 |
|
|
@Override
|
81 |
|
|
public Map<String, SolrDoc> processDocument(String identifier, Map<String, SolrDoc> docs,
|
82 |
9018
|
leinfelder
|
InputStream is) throws IOException, EncoderException, SAXException,
|
83 |
8134
|
tao
|
XPathExpressionException, ParserConfigurationException, SolrServerException, NotImplemented, NotFound, UnsupportedType, OREParserException, ResourceMapException {
|
84 |
7696
|
tao
|
SolrDoc resourceMapDoc = docs.get(identifier);
|
85 |
9028
|
leinfelder
|
Document doc = XmlDocumentUtility.generateXmlDocument(is);
|
86 |
9018
|
leinfelder
|
List<SolrDoc> processedDocs = processResourceMap(resourceMapDoc, doc );
|
87 |
7696
|
tao
|
Map<String, SolrDoc> processedDocsMap = new HashMap<String, SolrDoc>();
|
88 |
|
|
for (SolrDoc processedDoc : processedDocs) {
|
89 |
|
|
processedDocsMap.put(processedDoc.getIdentifier(), processedDoc);
|
90 |
|
|
}
|
91 |
|
|
return processedDocsMap;
|
92 |
|
|
}
|
93 |
|
|
|
94 |
|
|
private List<SolrDoc> processResourceMap(SolrDoc indexDocument, Document resourceMapDocument)
|
95 |
8134
|
tao
|
throws XPathExpressionException, IOException, SAXException, ParserConfigurationException, EncoderException, SolrServerException, NotImplemented, NotFound, UnsupportedType, OREParserException, ResourceMapException{
|
96 |
8023
|
tao
|
//ResourceMap resourceMap = new ResourceMap(resourceMapDocument);
|
97 |
8296
|
tao
|
ResourceMap resourceMap = ResourceMapFactory.buildResourceMap(resourceMapDocument);
|
98 |
7851
|
tao
|
List<String> documentIds = resourceMap.getAllDocumentIDs();//this list includes the resourceMap id itself.
|
99 |
7696
|
tao
|
//List<SolrDoc> updateDocuments = getHttpService().getDocuments(getSolrQueryUri(), documentIds);
|
100 |
7851
|
tao
|
List<SolrDoc> updateDocuments = getSolrDocs(resourceMap.getIdentifier(), documentIds);
|
101 |
7696
|
tao
|
List<SolrDoc> mergedDocuments = resourceMap.mergeIndexedDocuments(updateDocuments);
|
102 |
7710
|
tao
|
/*if(mergedDocuments != null) {
|
103 |
|
|
for(SolrDoc doc : mergedDocuments) {
|
104 |
|
|
ByteArrayOutputStream out = new ByteArrayOutputStream();
|
105 |
|
|
doc.serialize(out, "UTF-8");
|
106 |
|
|
String result = new String(out.toByteArray(), "UTF-8");
|
107 |
|
|
System.out.println("after updated document===========================");
|
108 |
|
|
System.out.println(result);
|
109 |
|
|
}
|
110 |
|
|
}*/
|
111 |
7696
|
tao
|
mergedDocuments.add(indexDocument);
|
112 |
|
|
return mergedDocuments;
|
113 |
|
|
}
|
114 |
|
|
|
115 |
8134
|
tao
|
private List<SolrDoc> getSolrDocs(String resourceMapId, List<String> ids) throws SolrServerException, IOException, ParserConfigurationException, SAXException, XPathExpressionException, NotImplemented, NotFound, UnsupportedType, ResourceMapException {
|
116 |
7851
|
tao
|
List<SolrDoc> list = new ArrayList<SolrDoc>();
|
117 |
|
|
if(ids != null) {
|
118 |
|
|
for(String id : ids) {
|
119 |
8464
|
leinfelder
|
SolrDoc doc = getSolrDoc(id);
|
120 |
7851
|
tao
|
if(doc != null) {
|
121 |
|
|
list.add(doc);
|
122 |
|
|
} else if ( !id.equals(resourceMapId)) {
|
123 |
8134
|
tao
|
throw new ResourceMapException("Solr index doesn't have the information about the id "+id+" which is a component in the resource map "+resourceMapId+". Metacat-Index can't process the resource map prior to its components.");
|
124 |
7851
|
tao
|
}
|
125 |
|
|
}
|
126 |
|
|
}
|
127 |
|
|
return list;
|
128 |
|
|
}
|
129 |
|
|
|
130 |
7696
|
tao
|
/*
|
131 |
|
|
* Get the SolrDoc list for the list of the ids.
|
132 |
|
|
*/
|
133 |
7733
|
tao
|
public static List<SolrDoc> getSolrDocs(List<String> ids) throws SolrServerException, IOException, ParserConfigurationException, SAXException, XPathExpressionException, NotImplemented, NotFound, UnsupportedType {
|
134 |
7696
|
tao
|
List<SolrDoc> list = new ArrayList<SolrDoc>();
|
135 |
|
|
if(ids != null) {
|
136 |
|
|
for(String id : ids) {
|
137 |
8464
|
leinfelder
|
SolrDoc doc = getSolrDoc(id);
|
138 |
7696
|
tao
|
if(doc != null) {
|
139 |
|
|
list.add(doc);
|
140 |
|
|
}
|
141 |
|
|
}
|
142 |
|
|
}
|
143 |
|
|
return list;
|
144 |
|
|
}
|
145 |
|
|
|
146 |
8464
|
leinfelder
|
/*
|
147 |
|
|
* Get the SolrDoc for the specified id
|
148 |
|
|
*/
|
149 |
|
|
public static SolrDoc getSolrDoc(String id) throws SolrServerException,
|
150 |
|
|
IOException, ParserConfigurationException, SAXException,
|
151 |
|
|
XPathExpressionException, NotImplemented, NotFound, UnsupportedType {
|
152 |
8893
|
tao
|
int targetIndex = 0;
|
153 |
8552
|
leinfelder
|
SolrDoc doc = null;
|
154 |
8893
|
tao
|
String query = QUERY + "\"" + id + "\"";
|
155 |
|
|
List<SolrDoc> list = getDocumentsByQuery(query);
|
156 |
|
|
if(list != null && !list.isEmpty()) {
|
157 |
|
|
doc = list.get(targetIndex);
|
158 |
|
|
}
|
159 |
8464
|
leinfelder
|
return doc;
|
160 |
|
|
}
|
161 |
8893
|
tao
|
|
162 |
|
|
/**
|
163 |
|
|
* Get a list of solr documents which's resourcemap field matches the given value.
|
164 |
|
|
* @param resourceMapId - the target resource map id
|
165 |
|
|
* @return the list of solr document
|
166 |
|
|
* @throws MalformedURLException
|
167 |
|
|
* @throws UnsupportedType
|
168 |
|
|
* @throws NotFound
|
169 |
|
|
* @throws SolrServerException
|
170 |
|
|
* @throws ParserConfigurationException
|
171 |
|
|
* @throws IOException
|
172 |
|
|
* @throws SAXException
|
173 |
|
|
*/
|
174 |
|
|
public static List<SolrDoc> getDocumentsByResourceMap(String resourceMapId) throws MalformedURLException,
|
175 |
|
|
UnsupportedType, NotFound, SolrServerException, ParserConfigurationException, IOException, SAXException {
|
176 |
|
|
String query = QUERY2 + "\"" + resourceMapId + "\"";
|
177 |
|
|
return getDocumentsByQuery(query);
|
178 |
|
|
}
|
179 |
|
|
|
180 |
|
|
/**
|
181 |
|
|
* Get a list of slor docs which match the query.
|
182 |
|
|
* @param query - a string of a query
|
183 |
|
|
* @return
|
184 |
|
|
* @throws SolrServerException
|
185 |
|
|
* @throws MalformedURLException
|
186 |
|
|
* @throws UnsupportedType
|
187 |
|
|
* @throws NotFound
|
188 |
|
|
* @throws ParserConfigurationException
|
189 |
|
|
* @throws IOException
|
190 |
|
|
* @throws SAXException
|
191 |
|
|
*/
|
192 |
|
|
public static List<SolrDoc> getDocumentsByQuery(String query) throws SolrServerException, MalformedURLException, UnsupportedType,
|
193 |
|
|
NotFound, ParserConfigurationException, IOException, SAXException {
|
194 |
|
|
List<SolrDoc> docs = new ArrayList<SolrDoc>();
|
195 |
|
|
if (solrServer != null && query != null && !query.trim().equals("")) {
|
196 |
|
|
SolrParams solrParams = SolrRequestParsers.parseQueryString(query);
|
197 |
|
|
QueryResponse qr = solrServer.query(solrParams);
|
198 |
|
|
if (qr != null && qr.getResults() != null) {
|
199 |
|
|
for(int i=0; i<qr.getResults().size(); i++) {
|
200 |
|
|
SolrDocument orig = qr.getResults().get(i);
|
201 |
|
|
SolrDoc doc = new SolrDoc();
|
202 |
|
|
IndexSchema indexSchema = SolrQueryServiceController.getInstance().getSchema();
|
203 |
|
|
for (String fieldName : orig.getFieldNames()) {
|
204 |
|
|
// don't transfer the copyTo fields, otherwise there are errors
|
205 |
|
|
if (indexSchema.isCopyFieldTarget(indexSchema.getField(fieldName))) {
|
206 |
|
|
continue;
|
207 |
|
|
}
|
208 |
|
|
for (Object value : orig.getFieldValues(fieldName)) {
|
209 |
|
|
String stringValue = value.toString();
|
210 |
|
|
// special handling for dates in ISO 8601
|
211 |
|
|
if (value instanceof Date) {
|
212 |
|
|
stringValue = DateTimeMarshaller.serializeDateToUTC((Date) value);
|
213 |
|
|
SolrDateConverter converter = new SolrDateConverter();
|
214 |
|
|
stringValue = converter.convert(stringValue);
|
215 |
|
|
}
|
216 |
|
|
SolrElementField field = new SolrElementField(fieldName, stringValue);
|
217 |
|
|
log.debug("Adding field: " + fieldName);
|
218 |
|
|
doc.addField(field);
|
219 |
|
|
}
|
220 |
|
|
}
|
221 |
|
|
docs.add(doc);
|
222 |
|
|
}
|
223 |
|
|
|
224 |
|
|
}
|
225 |
|
|
}
|
226 |
|
|
return docs;
|
227 |
|
|
}
|
228 |
8464
|
leinfelder
|
|
229 |
|
|
|
230 |
7696
|
tao
|
}
|