1 |
7609
|
tao
|
/**
|
2 |
|
|
* '$RCSfile$'
|
3 |
|
|
* Copyright: 2000-2011 Regents of the University of California and the
|
4 |
|
|
* National Center for Ecological Analysis and Synthesis
|
5 |
|
|
*
|
6 |
|
|
* '$Author: leinfelder $'
|
7 |
|
|
* '$Date: 2012-11-29 16:52:29 -0800 (Thu, 29 Nov 2012) $'
|
8 |
|
|
*
|
9 |
|
|
* This program is free software; you can redistribute it and/or modify
|
10 |
|
|
* it under the terms of the GNU General Public License as published by
|
11 |
|
|
* the Free Software Foundation; either version 2 of the License, or
|
12 |
|
|
* (at your option) any later version.
|
13 |
|
|
*
|
14 |
|
|
* This program is distributed in the hope that it will be useful,
|
15 |
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
16 |
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
17 |
|
|
* GNU General Public License for more details.
|
18 |
|
|
*
|
19 |
|
|
* You should have received a copy of the GNU General Public License
|
20 |
|
|
* along with this program; if not, write to the Free Software
|
21 |
|
|
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
|
22 |
|
|
*/
|
23 |
|
|
package edu.ucsb.nceas.metacat.index;
|
24 |
|
|
|
25 |
|
|
import java.io.ByteArrayInputStream;
|
26 |
7647
|
leinfelder
|
import java.io.ByteArrayOutputStream;
|
27 |
7609
|
tao
|
import java.io.IOException;
|
28 |
|
|
import java.io.InputStream;
|
29 |
7647
|
leinfelder
|
import java.io.OutputStreamWriter;
|
30 |
7609
|
tao
|
import java.io.Writer;
|
31 |
7647
|
leinfelder
|
import java.sql.SQLException;
|
32 |
|
|
import java.util.Hashtable;
|
33 |
8464
|
leinfelder
|
import java.util.List;
|
34 |
|
|
import java.util.Map;
|
35 |
7725
|
tao
|
|
36 |
7680
|
tao
|
import java.util.Set;
|
37 |
7609
|
tao
|
|
38 |
7620
|
tao
|
import javax.xml.parsers.ParserConfigurationException;
|
39 |
|
|
|
40 |
7647
|
leinfelder
|
import org.apache.commons.io.IOUtils;
|
41 |
7609
|
tao
|
import org.apache.commons.logging.Log;
|
42 |
|
|
import org.apache.commons.logging.LogFactory;
|
43 |
|
|
import org.apache.solr.client.solrj.SolrServerException;
|
44 |
7647
|
leinfelder
|
import org.apache.solr.common.params.ModifiableSolrParams;
|
45 |
7609
|
tao
|
import org.apache.solr.common.params.SolrParams;
|
46 |
7725
|
tao
|
|
47 |
7609
|
tao
|
import org.apache.solr.servlet.SolrRequestParsers;
|
48 |
7718
|
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.types.v1.Identifier;
|
52 |
7680
|
tao
|
import org.dataone.service.types.v1.Subject;
|
53 |
8464
|
leinfelder
|
import org.dataone.service.types.v1.SystemMetadata;
|
54 |
7620
|
tao
|
import org.xml.sax.SAXException;
|
55 |
7609
|
tao
|
|
56 |
7647
|
leinfelder
|
import edu.ucsb.nceas.metacat.DBTransform;
|
57 |
8464
|
leinfelder
|
import edu.ucsb.nceas.metacat.common.index.IndexTask;
|
58 |
7725
|
tao
|
import edu.ucsb.nceas.metacat.common.query.SolrQueryResponseWriterFactory;
|
59 |
7718
|
tao
|
import edu.ucsb.nceas.metacat.common.query.SolrQueryService;
|
60 |
|
|
import edu.ucsb.nceas.metacat.common.query.SolrQueryServiceController;
|
61 |
7757
|
leinfelder
|
import edu.ucsb.nceas.metacat.common.query.stream.ContentTypeByteArrayInputStream;
|
62 |
8464
|
leinfelder
|
import edu.ucsb.nceas.metacat.dataone.hazelcast.HazelcastService;
|
63 |
7647
|
leinfelder
|
import edu.ucsb.nceas.utilities.PropertyNotFoundException;
|
64 |
7609
|
tao
|
|
65 |
|
|
|
66 |
|
|
/**
|
67 |
|
|
* This class will query the solr server and return the result.
|
68 |
|
|
* @author tao
|
69 |
|
|
*
|
70 |
|
|
*/
|
71 |
|
|
public class MetacatSolrIndex {
|
72 |
|
|
|
73 |
7634
|
tao
|
|
74 |
7772
|
tao
|
//public static final String SOLRQUERY = "solr";
|
75 |
|
|
//public static final String SOLR_HOME_PROPERTY_NAME = "solr.homeDir";
|
76 |
|
|
//public static final String SOLR_CONFIG_FILE_NAME_PROPERTY_NAME = "solr.configFileName";
|
77 |
|
|
//public static final String SOLR_COLLECTION_NAME_PROPERTY_NAME = "solr.collectionName";
|
78 |
|
|
//public static final String SOLR_SERVER_CLASSNAME_PROPERTY_NAME = "solr.server.classname";
|
79 |
7718
|
tao
|
|
80 |
7707
|
tao
|
|
81 |
7609
|
tao
|
private static Log log = LogFactory.getLog(MetacatSolrIndex.class);
|
82 |
7634
|
tao
|
private static MetacatSolrIndex solrIndex = null;
|
83 |
7609
|
tao
|
|
84 |
7664
|
tao
|
public static MetacatSolrIndex getInstance() throws Exception {
|
85 |
7634
|
tao
|
if (solrIndex == null) {
|
86 |
|
|
solrIndex = new MetacatSolrIndex();
|
87 |
|
|
}
|
88 |
|
|
return solrIndex;
|
89 |
|
|
}
|
90 |
|
|
|
91 |
7609
|
tao
|
/**
|
92 |
|
|
* Constructor
|
93 |
7620
|
tao
|
* @throws SAXException
|
94 |
|
|
* @throws IOException
|
95 |
|
|
* @throws ParserConfigurationException
|
96 |
7609
|
tao
|
*/
|
97 |
7664
|
tao
|
private MetacatSolrIndex() throws Exception {
|
98 |
7718
|
tao
|
|
99 |
7609
|
tao
|
}
|
100 |
|
|
|
101 |
|
|
|
102 |
|
|
|
103 |
7718
|
tao
|
|
104 |
7609
|
tao
|
/**
|
105 |
|
|
* Query the solr server
|
106 |
|
|
* @param query the solr query
|
107 |
7680
|
tao
|
* @param authorizedSubjects the authorized subjects in this query session
|
108 |
7609
|
tao
|
* @return the result as the InputStream
|
109 |
|
|
* @throws SolrServerException
|
110 |
7647
|
leinfelder
|
* @throws ClassNotFoundException
|
111 |
|
|
* @throws SQLException
|
112 |
|
|
* @throws PropertyNotFoundException
|
113 |
7685
|
tao
|
* @throws SAXException
|
114 |
|
|
* @throws ParserConfigurationException
|
115 |
7718
|
tao
|
* @throws UnsupportedType
|
116 |
|
|
* @throws NotFound
|
117 |
|
|
* @throws NotImplemented
|
118 |
7609
|
tao
|
*/
|
119 |
7718
|
tao
|
public InputStream query(String query, Set<Subject>authorizedSubjects) throws SolrServerException, IOException, PropertyNotFoundException, SQLException,
|
120 |
|
|
ClassNotFoundException, ParserConfigurationException, SAXException, NotImplemented, NotFound, UnsupportedType {
|
121 |
7680
|
tao
|
if(authorizedSubjects == null || authorizedSubjects.isEmpty()) {
|
122 |
|
|
throw new SolrServerException("MetacatSolrIndex.query - There is no any authorized subjects(even the public user) in this query session.");
|
123 |
|
|
}
|
124 |
7609
|
tao
|
InputStream inputStream = null;
|
125 |
|
|
SolrParams solrParams = SolrRequestParsers.parseQueryString(query);
|
126 |
7718
|
tao
|
String wt = solrParams.get(SolrQueryService.WT);
|
127 |
7647
|
leinfelder
|
// handle normal and skin-based queries
|
128 |
7718
|
tao
|
if (SolrQueryService.isSupportedWT(wt)) {
|
129 |
|
|
// just handle as normal solr query
|
130 |
|
|
|
131 |
7720
|
tao
|
inputStream = SolrQueryServiceController.getInstance().query(solrParams, authorizedSubjects);
|
132 |
7647
|
leinfelder
|
}
|
133 |
|
|
else {
|
134 |
7718
|
tao
|
// assume it is a skin name
|
135 |
|
|
String qformat = wt;
|
136 |
|
|
|
137 |
|
|
// perform the solr query using wt=XML
|
138 |
|
|
wt = SolrQueryResponseWriterFactory.XML;
|
139 |
|
|
ModifiableSolrParams msp = new ModifiableSolrParams(solrParams);
|
140 |
|
|
msp.set(SolrQueryService.WT, wt);
|
141 |
7720
|
tao
|
inputStream = SolrQueryServiceController.getInstance().query(msp, authorizedSubjects);
|
142 |
7718
|
tao
|
|
143 |
|
|
// apply the stylesheet (XML->HTML)
|
144 |
|
|
DBTransform transformer = new DBTransform();
|
145 |
|
|
String documentContent = IOUtils.toString(inputStream, "UTF-8");
|
146 |
|
|
String sourceType = "solr";
|
147 |
|
|
String targetType = "-//W3C//HTML//EN";
|
148 |
|
|
ByteArrayOutputStream baos = new ByteArrayOutputStream();
|
149 |
|
|
Writer writer = new OutputStreamWriter(baos , "UTF-8");
|
150 |
|
|
// TODO: include more params?
|
151 |
|
|
Hashtable<String, String[]> params = new Hashtable<String, String[]>();
|
152 |
|
|
params.put("qformat", new String[] {qformat});
|
153 |
|
|
transformer.transformXMLDocument(
|
154 |
|
|
documentContent ,
|
155 |
|
|
sourceType,
|
156 |
|
|
targetType ,
|
157 |
|
|
qformat,
|
158 |
|
|
writer,
|
159 |
|
|
params,
|
160 |
|
|
null //sessionid
|
161 |
|
|
);
|
162 |
|
|
|
163 |
|
|
// finally, get the HTML back
|
164 |
7757
|
leinfelder
|
inputStream = new ContentTypeByteArrayInputStream(baos.toByteArray());
|
165 |
|
|
((ContentTypeByteArrayInputStream) inputStream).setContentType("text/html");
|
166 |
7647
|
leinfelder
|
}
|
167 |
|
|
|
168 |
7609
|
tao
|
return inputStream;
|
169 |
7718
|
tao
|
|
170 |
7609
|
tao
|
}
|
171 |
8464
|
leinfelder
|
|
172 |
|
|
public void submit(Identifier pid, SystemMetadata systemMetadata, Map<String, List<Object>> fields) {
|
173 |
|
|
IndexTask task = new IndexTask();
|
174 |
|
|
task.setSystemMetadata(systemMetadata);
|
175 |
|
|
task.setFields(fields);
|
176 |
|
|
HazelcastService.getInstance().getIndexQueue().put(pid, task );
|
177 |
|
|
}
|
178 |
|
|
|
179 |
7718
|
tao
|
|
180 |
7609
|
tao
|
}
|