1 | /**
|
---|
2 | *#########################################################################
|
---|
3 | * FedoraGS3Connection.java - works with the demo-client for Greenstone 3,
|
---|
4 | * of the Greenstone digital library suite from the New Zealand Digital
|
---|
5 | * Library Project at the * University of Waikato, New Zealand.
|
---|
6 | * <BR><BR>
|
---|
7 | * Copyright (C) 2008 New Zealand Digital Library Project
|
---|
8 | * <BR><BR>
|
---|
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 | * <BR><BR>
|
---|
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 | */
|
---|
20 |
|
---|
21 | package org.greenstone.fedora.services;
|
---|
22 |
|
---|
23 |
|
---|
24 | import java.io.StringReader;
|
---|
25 |
|
---|
26 | import org.apache.log4j.Logger;
|
---|
27 | import org.greenstone.fedora.services.FedoraGS3Exception.AuthenticationFailedException;
|
---|
28 | import org.greenstone.fedora.services.FedoraGS3Exception.CancelledException;
|
---|
29 | import org.greenstone.fedora.services.FedoraGS3Exception.FedoraGS3RunException;
|
---|
30 | import org.greenstone.fedora.services.FedoraGS3Exception.NotAFedoraServerException;
|
---|
31 | import org.greenstone.gsdl3.util.GSXML;
|
---|
32 | import org.w3c.dom.Document;
|
---|
33 | import org.w3c.dom.Element;
|
---|
34 | import org.w3c.dom.Attr;
|
---|
35 | import org.w3c.dom.Text;
|
---|
36 | import org.w3c.dom.NodeList;
|
---|
37 | import org.w3c.dom.Node;
|
---|
38 | import org.xml.sax.InputSource;
|
---|
39 |
|
---|
40 | import java.io.File;
|
---|
41 | import java.util.HashMap;
|
---|
42 | import java.util.Properties;
|
---|
43 | import java.util.Map;
|
---|
44 |
|
---|
45 | import javax.swing.JOptionPane;
|
---|
46 |
|
---|
47 | import org.xml.sax.SAXException;
|
---|
48 | import java.io.UnsupportedEncodingException;
|
---|
49 | import java.io.IOException;
|
---|
50 | import javax.net.ssl.SSLHandshakeException;
|
---|
51 | import java.net.ConnectException;
|
---|
52 | import java.net.MalformedURLException;
|
---|
53 | import java.rmi.RemoteException;
|
---|
54 | import javax.xml.parsers.ParserConfigurationException;
|
---|
55 | import javax.xml.transform.TransformerException;
|
---|
56 |
|
---|
57 | /**
|
---|
58 | * Class that extends FedoraConnection in order to be able to use
|
---|
59 | * Fedora's web services to retrieve the specific datastreams of
|
---|
60 | * Greenstone documents stored in Fedora's repository. This class
|
---|
61 | * provides methods that convert those datastreams into Greenstone3
|
---|
62 | * XML response messages which are returned.
|
---|
63 | * @author ak19
|
---|
64 | */
|
---|
65 | public class FedoraGS3Connection
|
---|
66 | extends FedoraConnection implements FedoraToGS3Interface,
|
---|
67 | FedoraToGS3Interface.Constants
|
---|
68 | {
|
---|
69 | /** The logging instance for this class */
|
---|
70 | private static final Logger LOG = Logger.getLogger(
|
---|
71 | FedoraGS3Connection.class.getName());
|
---|
72 |
|
---|
73 | /** Default name of Fedora index */
|
---|
74 | private static final String DEFAULT_FEDORA_INDEX = "BasicIndex";
|
---|
75 |
|
---|
76 | /** Complete list of services that are supported our FedoraGS3 would
|
---|
77 | * support if everything goes well. If a connection to FedoraGSearch
|
---|
78 | * cannot be established, the query services will no longer be
|
---|
79 | * available. The actual services supported are given by member
|
---|
80 | * variable serviceNames. */
|
---|
81 | protected static final String[] SERVICES = {
|
---|
82 | "DocumentContentRetrieve", "DocumentMetadataRetrieve",
|
---|
83 | "DocumentStructureRetrieve",
|
---|
84 | "TextQuery", "FieldQuery",
|
---|
85 | "ClassifierBrowse", "ClassifierBrowseMetadataRetrieve"
|
---|
86 | };
|
---|
87 |
|
---|
88 | /** List of services actually supported by our FedoraGS3 repository
|
---|
89 | * after construction. If FedoraGenericSearch can't be connected to,
|
---|
90 | * then query services will not be offered */
|
---|
91 | protected String[] serviceNames;
|
---|
92 |
|
---|
93 | /** The object used to connect to FedoraGenericSearch, which is used
|
---|
94 | * for doing full-text searching */
|
---|
95 | protected GSearchConnection fedoraGSearch;
|
---|
96 |
|
---|
97 | /** The url for the wsdl file of FedoraGSearch's web services
|
---|
98 | * by default this will be the Fedora server's base URL
|
---|
99 | * concatenated to "gsearch/services/FgsOperations?wsdl" */
|
---|
100 | protected String gSearchWSDLURL;
|
---|
101 |
|
---|
102 | /** The last part of the gSearchWSDL URL. The first part is
|
---|
103 | * the same as the fedora server's base url. */
|
---|
104 | protected String gSearchWSDLSuffix;
|
---|
105 |
|
---|
106 | /** The name of the index that FedoraGSearch will index the GS3
|
---|
107 | * documents into. If no name is specified in the properties file,
|
---|
108 | * this will default to FedoraIndex. */
|
---|
109 | protected String gSearchIndexName;
|
---|
110 |
|
---|
111 | /** 5 argument constructor is the same as that of superclass FedoraConnection:
|
---|
112 | * @param protocol can be either http or https
|
---|
113 | * @param host is the host where the fedora server is listening
|
---|
114 | * @param port is the port where the fedora server is listening
|
---|
115 | * @param fedoraServerUsername is the username for administrative
|
---|
116 | * authentication required to access the fedora server.
|
---|
117 | * @param fedoraServerPassword is the password for administrative
|
---|
118 | * authentication required to access the fedora server. If no password was set
|
---|
119 | * when installing Fedora, leave the field "".
|
---|
120 | * Instantiates a FedoraGS3Connection object which connects to Fedora's
|
---|
121 | * web services through stub classes and tries to connect to FedoraGSearch's
|
---|
122 | * web services through the default WSDL location for it
|
---|
123 | * ("gsearch/services/FgsOperations?wsdl"). If another url is to be used,
|
---|
124 | * call setGSearchWSDLURL(url) after the constructor instead.
|
---|
125 | */
|
---|
126 | public FedoraGS3Connection(String protocol, String host, int port,
|
---|
127 | String fedoraServerUsername, String fedoraServerPassword)
|
---|
128 | throws ParserConfigurationException, MalformedURLException,
|
---|
129 | SSLHandshakeException, RemoteException, AuthenticationFailedException,
|
---|
130 | NotAFedoraServerException, ConnectException, Exception
|
---|
131 | {
|
---|
132 | super(protocol, host, port, fedoraServerUsername, fedoraServerPassword);
|
---|
133 | // super() will call setInitialisationProperties(properties)
|
---|
134 | // And that will try to instantiate the GSearchConnection.
|
---|
135 | }
|
---|
136 |
|
---|
137 | /** No-argument constructor which is the same as that of superclass
|
---|
138 | * FedoraConnection: it displays a small dialog requesting input for the
|
---|
139 | * host, port, administrative password and username of the fedora server.
|
---|
140 | * If no password was set on the fedora repository when installing it,
|
---|
141 | * the user can leave the password field blank. */
|
---|
142 | public FedoraGS3Connection()
|
---|
143 | throws ParserConfigurationException, MalformedURLException,
|
---|
144 | CancelledException, ConnectException, RemoteException,
|
---|
145 | SSLHandshakeException, Exception
|
---|
146 | {
|
---|
147 | super();
|
---|
148 | // super() will call setInitialisationProperties(properties)
|
---|
149 | // And that will try to instantiate the GSearchConnection.
|
---|
150 | }
|
---|
151 |
|
---|
152 | /** Single-argument constructor which is the same as that of superclass
|
---|
153 | * FedoraConnection: it takes the name of the properties file where
|
---|
154 | * connection initialisation values may already be provided and then
|
---|
155 | * displays a small dialog requesting input for the host, port,
|
---|
156 | * administrative password and username of the fedora server showing
|
---|
157 | * the values in the properties file as default. If the necessary
|
---|
158 | * initialisation are not present in the file, the corresponding fields
|
---|
159 | * in the dialog will be blank.
|
---|
160 | * If no password was set on the fedora repository when installing it,
|
---|
161 | * the user can leave the password field blank. */
|
---|
162 | public FedoraGS3Connection(File propertiesFilename)
|
---|
163 | throws ParserConfigurationException, MalformedURLException,
|
---|
164 | CancelledException, ConnectException, RemoteException,
|
---|
165 | SSLHandshakeException, Exception
|
---|
166 | {
|
---|
167 | super(propertiesFilename);
|
---|
168 | // super() will call setInitialisationProperties(properties)
|
---|
169 | // And that will try to instantiate the GSearchConnection.
|
---|
170 | }
|
---|
171 |
|
---|
172 | /** The superclass constructor calls this method passing any preset
|
---|
173 | * properties loaded from a propertiesFile. This method is overridden
|
---|
174 | * here in order to instantiate the gSearchConnection based on the
|
---|
175 | * - gSearchWSDLSuffix that will be appended to the fedora base url.
|
---|
176 | * (If one was not provided in the properties file, gSearchWSDLURL defaults
|
---|
177 | * to something of the form
|
---|
178 | * "http://<fedorahost:port>/fedoragsearch/services/FgsOperations?wsdl"
|
---|
179 | * which is the baseURL of fedora concatenated to the default gSearchWSDLSuffix,
|
---|
180 | * "gsearch/services/FgsOperations?wsdl".
|
---|
181 | * - name of the index into which the GS3 documents have been indexed
|
---|
182 | * and which FedoraGenericSearch should use to perform searches. If none is
|
---|
183 | * given in the properties file, then the index name defaults to "FedoraIndex".
|
---|
184 | * @param properties is the Properties Map loaded from a properties file
|
---|
185 | * (if there was any) which specifies such things as host and port of the
|
---|
186 | * FedoraServer, but can also specify the property "gsearch.wsdlURL.suffix".
|
---|
187 | * At the end of this method, properties' "gsearch.wsdlURL.suffix" will be set
|
---|
188 | * to whatever the final value of this.gSearchWSDLURL' suffix is, and
|
---|
189 | * "gsearch.indexName" will be set to to whatever the final value of
|
---|
190 | * this.gSearchIndexName is.
|
---|
191 | */
|
---|
192 | protected void setInitialisationProperties(Properties properties)
|
---|
193 | throws ParserConfigurationException, MalformedURLException,
|
---|
194 | CancelledException, ConnectException, RemoteException,
|
---|
195 | SSLHandshakeException, Exception
|
---|
196 | {
|
---|
197 | super.setInitialisationProperties(properties);
|
---|
198 | // gsearchWSDL URL suffix, if not specified, defaults to
|
---|
199 | // "fedoragsearch/services/FgsOperations?wsdl" which is
|
---|
200 | // concatenated to the baseURL of fedora to give the gsearchWSDLURL.
|
---|
201 | this.gSearchWSDLSuffix = properties.getProperty(
|
---|
202 | "gsearch.wsdlURL.suffix", "gsearch/services/FgsOperations?wsdl");
|
---|
203 | this.gSearchWSDLURL = this.baseURL + this.gSearchWSDLSuffix;
|
---|
204 | // Set the property to whatever this.gSearchWSDLURL is now,
|
---|
205 | // so that it will be written out to the properties file again
|
---|
206 | properties.setProperty("gsearch.wsdlURL.suffix", this.gSearchWSDLSuffix);
|
---|
207 |
|
---|
208 | // Similarly for the name of the index FedoraGenericSearch should use
|
---|
209 | // when performing searches for GS3 docs stored in Fedora's repository.
|
---|
210 | this.gSearchIndexName = properties.getProperty(
|
---|
211 | "gsearch.indexName", DEFAULT_FEDORA_INDEX); // default to FedoraIndex
|
---|
212 | properties.setProperty("gsearch.indexName", this.gSearchIndexName);
|
---|
213 | // Create a connection to FedoraGSearch's web services:
|
---|
214 | initSearchFunctionality();
|
---|
215 | }
|
---|
216 |
|
---|
217 | /** Overridden init method to work with the 5 argument constructor, so that we can
|
---|
218 | * bypass using setInitialisationProperties() which works with a Properties map.
|
---|
219 | */
|
---|
220 | protected void init(String protocol, String host, String port,
|
---|
221 | String fedoraServerUsername, String fedoraServerPassword)
|
---|
222 | throws ParserConfigurationException, MalformedURLException,
|
---|
223 | AuthenticationFailedException, RemoteException, Exception
|
---|
224 | {
|
---|
225 | super.init(protocol, host, port, fedoraServerUsername, fedoraServerPassword);
|
---|
226 | this.gSearchWSDLSuffix = "gsearch/services/FgsOperations?wsdl";
|
---|
227 | this.gSearchWSDLURL = this.baseURL + this.gSearchWSDLSuffix;
|
---|
228 | this.gSearchIndexName = DEFAULT_FEDORA_INDEX;
|
---|
229 | initSearchFunctionality();
|
---|
230 | }
|
---|
231 |
|
---|
232 |
|
---|
233 | /** Init method that instantiates a GSearchConnection object used
|
---|
234 | * to work with the separate FedoraGSearch web services.
|
---|
235 | * The url of the WSDL for FedoraGSearch's web services is worked out
|
---|
236 | * from the baseURL of the Fedora server.
|
---|
237 | */
|
---|
238 | protected void initSearchFunctionality()
|
---|
239 | {
|
---|
240 | try {
|
---|
241 | this.fedoraGSearch = null;
|
---|
242 | this.fedoraGSearch = new GSearchConnection(
|
---|
243 | gSearchWSDLURL, gSearchIndexName);
|
---|
244 | this.serviceNames = SERVICES;
|
---|
245 | } catch(Exception e){
|
---|
246 | LOG.error("Cannot connect to FedoraGSearch's web services at "
|
---|
247 | + gSearchWSDLURL + "\nQuery services will not be available.");
|
---|
248 | // If an exception occurs, something has gone wrong when
|
---|
249 | // trying to connect to FedoraGSearch's web services. This
|
---|
250 | // means, we can't offer query services, as that's provided
|
---|
251 | // by FedoraGSearch
|
---|
252 | serviceNames = null;
|
---|
253 | int countOfNonQueryServices = 0;
|
---|
254 | for(int i = 0; i < SERVICES.length; i++) {
|
---|
255 | // do not count query services
|
---|
256 | if(!SERVICES[i].toLowerCase().contains("query")) {
|
---|
257 | countOfNonQueryServices++;
|
---|
258 | }
|
---|
259 | }
|
---|
260 | // Services now supported are everything except Query services
|
---|
261 | serviceNames = new String[countOfNonQueryServices];
|
---|
262 | int j = 0;
|
---|
263 | for(int i = 0; i < SERVICES.length; i++) {
|
---|
264 | if(!SERVICES[i].toLowerCase().contains("query")) {
|
---|
265 | serviceNames[j] = SERVICES[i];
|
---|
266 | j++; // valid serviceName, so increment serviceName counter
|
---|
267 | }
|
---|
268 |
|
---|
269 | }
|
---|
270 | }
|
---|
271 | }
|
---|
272 |
|
---|
273 | /** @return the gSearchWSDLURL, the url of the WSDL for the
|
---|
274 | * FedoraGSearch web services */
|
---|
275 | public String getGSearchWSDLURL() { return gSearchWSDLURL; }
|
---|
276 |
|
---|
277 | /** Sets the member variable gSearchWSDLURL that specify the location of
|
---|
278 | * the WSDL file of FedoraGSearch's web services. Then it attempts
|
---|
279 | * to instantiate a connection to those web services.
|
---|
280 | * @param url is the new url of the GSearch web services WSDL file */
|
---|
281 | public void setGSearchWSDLURL(String url) {
|
---|
282 | this.gSearchWSDLURL = url;
|
---|
283 | initSearchFunctionality();
|
---|
284 | }
|
---|
285 |
|
---|
286 | /** @return the gSearchIndexName, the name of the index Fedora Generic
|
---|
287 | * Search will search in (where GS3 docs have been indexed into). */
|
---|
288 | public String getGSearchIndexName() { return gSearchIndexName; }
|
---|
289 |
|
---|
290 | /** Sets the member variable gSearchIndexName that specifies the name
|
---|
291 | * of the index containing indexed GS3 documents. Then it attempts
|
---|
292 | * to instantiate a connection to the Fedora GSearch web services using
|
---|
293 | * this changed value for indexName.
|
---|
294 | * @param indexName is the new name of the index containing indexed GS3
|
---|
295 | * docs that GSearch should search in. */
|
---|
296 | public void setGSearchIndexName(String indexName) {
|
---|
297 | this.gSearchIndexName = indexName;
|
---|
298 | initSearchFunctionality();
|
---|
299 | }
|
---|
300 |
|
---|
301 | /** @return the array of the services actually supported by FedoraGS3 */
|
---|
302 | protected String[] getServiceNames() { return this.serviceNames;}
|
---|
303 |
|
---|
304 | /**
|
---|
305 | * For finding out if the sectionNumber is given as part of the docID.
|
---|
306 | * @param docID is the String that contains the docPID and may also
|
---|
307 | * contain the section number.
|
---|
308 | * @return true if the document identifier docID contains a section-
|
---|
309 | * number, and false if it consists solely of the docPID.
|
---|
310 | * That is, true is returned if
|
---|
311 | * <pre>docID = "greenstone:colName-<docPID>-<sectionNum>"</pre>
|
---|
312 | * and false is returned if
|
---|
313 | * <pre>docID = "greenstone:colName-<docPID>"</pre>
|
---|
314 | * */
|
---|
315 | protected boolean containsSectionNumber(String docID) {
|
---|
316 | // if there are two hyphens in the docID, then there are sections
|
---|
317 | // (and the section number is appended at end of docID)
|
---|
318 | // docID = "greenstone:colName-<docPID>-<sectionNum>"
|
---|
319 | return (docID.lastIndexOf(HYPHEN) != docID.indexOf(HYPHEN));
|
---|
320 | }
|
---|
321 |
|
---|
322 | /** This method will extract the docPID from docID and return it.
|
---|
323 | * (If a sectionNumber is suffixed to the docID, the docPID which is
|
---|
324 | * the prefix is returned; otherwise the docID is the docPID and is
|
---|
325 | * returned)
|
---|
326 | * @param docID is the String that contains the docPID and may also
|
---|
327 | * contain the section number.
|
---|
328 | * @return only the docPID portion of the docID.
|
---|
329 | */
|
---|
330 | protected String getDocPIDFromDocID(String docID) {
|
---|
331 | if(containsSectionNumber(docID))
|
---|
332 | return docID.substring(0, docID.lastIndexOf(HYPHEN));
|
---|
333 | // else (if there's no sectionNumber), docID is the docPID
|
---|
334 | return docID;
|
---|
335 | }
|
---|
336 |
|
---|
337 | /** This method will return the section Number, if there's any
|
---|
338 | * suffixed to the docID. Otherwise it will return the empty string
|
---|
339 | * @param docID is the String that contains the docPID and may also
|
---|
340 | * contain the section number.
|
---|
341 | * @return only the sectionID portion of the docID - if any, else "".
|
---|
342 | */
|
---|
343 | protected String getSectionIDFromDocID(String docID) {
|
---|
344 | if(containsSectionNumber(docID))
|
---|
345 | return docID.substring(
|
---|
346 | docID.lastIndexOf(HYPHEN)+1, docID.length());
|
---|
347 | return "";
|
---|
348 | }
|
---|
349 |
|
---|
350 | /** Given a list of collectionIDs, returns a GS3 DocumentMetadataRetrieve
|
---|
351 | * response message that gives the metadata for each collection identified
|
---|
352 | * @param collIDs is an array of fedora pids identifying collections in the
|
---|
353 | * fedora repository
|
---|
354 | * @return a GS3 DocumentMetadataRetrieve response message containing the
|
---|
355 | * EX metadata for all the requested collections */
|
---|
356 | public String getCollectionMetadata(String[] collIDs) {
|
---|
357 | return getMetadata(collIDs, new String[] {"all"});
|
---|
358 | }
|
---|
359 |
|
---|
360 | /** Given a list of document identifiers, a GS3 DocumentMetadataRetrieve
|
---|
361 | * response message is returned containing the metadata for each document.
|
---|
362 | * @param docIDs is an array of document identifiers (docID can either be
|
---|
363 | * <pid>s items (documents) in the fedora repository, or
|
---|
364 | * "<pid>-sectionNumber".
|
---|
365 | * @return a GS3 DocumentMetadataRetrieve response message containing the
|
---|
366 | * EX, DC, DLS metadata for all the requested documents
|
---|
367 | * @param metadata is the list of metadata elements to be retrieved for each doc */
|
---|
368 | public String getDocumentMetadata(String[] docIDs, String[] metadata) {
|
---|
369 | return getMetadata(docIDs, metadata);
|
---|
370 | }
|
---|
371 |
|
---|
372 | /** Given a collectionID, returns a GS3 DocumentMetadataRetrieve
|
---|
373 | * response message that gives the metadata for the collection identified
|
---|
374 | * @param collID is a fedora pid identifying a collection in its repository
|
---|
375 | * @return a GS3 DocumentMetadataRetrieve response message containing the
|
---|
376 | * EX metadata for the requested collection
|
---|
377 | * @param metadata is the list of metadata elements to be retrieved for each doc */
|
---|
378 | public String getCollectionMetadata(String collID) {
|
---|
379 | return getMetadata(new String[] {collID}, new String[] {"all"});
|
---|
380 | }
|
---|
381 |
|
---|
382 | /** Given a document identifier, returns a GS3 DocumentMetadataRetrieve
|
---|
383 | * response message containing the metadata for the document.
|
---|
384 | * @param docID is a document identifier (docID can either be a <pid>
|
---|
385 | * of an item (document) in the fedora repository, or it can be
|
---|
386 | * "<pid>-sectionNumber".
|
---|
387 | * @return a GS3 DocumentMetadataRetrieve response message containing the
|
---|
388 | * EX, DC, DLS metadata for the requested document */
|
---|
389 | public String getDocumentMetadata(String docID, String[] metadata) {
|
---|
390 | return getMetadata(new String[] {docID}, metadata);
|
---|
391 | }
|
---|
392 |
|
---|
393 | /** @return a greenstone DocumentMetadataRetrieve response for the
|
---|
394 | * documents or collections indicated by the docIDsOrCollIDs.
|
---|
395 | * @param docIDsOrCollIDs is an array of identifiers which may be either the
|
---|
396 | * fedora pids for collections, or otherwise may be a document identifier.
|
---|
397 | * In the last case, the document ID may consist of either
|
---|
398 | * "documentPID-sectionNumber" or may just be just fedora documentPID
|
---|
399 | * @param metadata is the list of metadata elements to be retrieved for each doc */
|
---|
400 | public String getMetadata(String[] docIDsOrCollIDs, String[] metadata)
|
---|
401 | {
|
---|
402 | Document doc = builder.newDocument();
|
---|
403 | FedoraGS3RunException ex = null;
|
---|
404 |
|
---|
405 | Element docNodeList = doc.createElement(
|
---|
406 | GSXML.DOC_NODE_ELEM+GSXML.LIST_MODIFIER);
|
---|
407 |
|
---|
408 | try{
|
---|
409 | for(int i = 0; i < docIDsOrCollIDs.length; i++) {
|
---|
410 | // create the <documentNode> containing the metadata
|
---|
411 | // for each document docID
|
---|
412 | Element docNode = getMetadata(doc, docIDsOrCollIDs[i], metadata);
|
---|
413 | docNodeList.appendChild(docNode);
|
---|
414 | }
|
---|
415 | } catch(Exception e) {
|
---|
416 | ex = new FedoraGS3RunException(e);
|
---|
417 | ex.setSpecifics("EX (and/or DC, DLS) metadata datastream");
|
---|
418 | }
|
---|
419 |
|
---|
420 | Element responseMsg = createResponseMessage(doc, docNodeList, ex,
|
---|
421 | GSXML.REQUEST_TYPE_PROCESS, "DocumentMetadataRetrieve");
|
---|
422 | try{
|
---|
423 | return FedoraCommons.elementToString(responseMsg);
|
---|
424 | } catch(TransformerException e) {
|
---|
425 | return FedoraGS3RunException.xmlToStringConversionFailureResponseMsg
|
---|
426 | + " " + e;
|
---|
427 | }
|
---|
428 | }
|
---|
429 |
|
---|
430 | /** Method that takes a new DOM document, as well as an identifier of either
|
---|
431 | * a collection or document (which may be a fedora pid for the collection
|
---|
432 | * or document, or may be the documentPid-sectionNumber for a document) and
|
---|
433 | * returns a documentNode element for it:
|
---|
434 | * <documentNode><metadataList>
|
---|
435 | * <metadata name="">value</metadata>
|
---|
436 | * ...
|
---|
437 | * </metadataList></documentNode>
|
---|
438 | * @return documentNode containing the metadata for the collection or
|
---|
439 | * document given by parameter ID
|
---|
440 | * @param id denotes a collection pid, a document pid or a docID of the
|
---|
441 | * form "documentpid-sectionNumber"
|
---|
442 | * @param metadata is the list of metadata elements to be retrieved for each doc */
|
---|
443 | protected Element getMetadata(Document doc, String id, String[] metadata)
|
---|
444 | throws RemoteException, UnsupportedEncodingException,
|
---|
445 | SAXException, IOException
|
---|
446 | {
|
---|
447 | // We're going to create the documentNode nested inside the following
|
---|
448 | // documentNodeList:
|
---|
449 | // <documentNodeList>
|
---|
450 | // <documentNode nodeID=""><metadataList>
|
---|
451 | // <metadata name="">value</metadata>
|
---|
452 | // </metadataList></documentNode>
|
---|
453 | // <documentNode>...</documentNode>
|
---|
454 | // </documentNodeList>
|
---|
455 | // <documentNodeList>
|
---|
456 |
|
---|
457 | // <documentNode nodeID="docID"> - the docNode on which a structure
|
---|
458 | // retrieve is being performed
|
---|
459 | Element docNode = doc.createElement(GSXML.DOC_NODE_ELEM);
|
---|
460 | Attr attribute = doc.createAttribute(GSXML.NODE_ID_ATT);
|
---|
461 | attribute.setValue(id);
|
---|
462 | docNode.setAttributeNode(attribute);
|
---|
463 |
|
---|
464 | // <metadataList>
|
---|
465 | Element metadataList = doc.createElement(
|
---|
466 | GSXML.METADATA_ELEM+GSXML.LIST_MODIFIER);
|
---|
467 |
|
---|
468 | String ex = "";
|
---|
469 | String dc = "";
|
---|
470 | String dls = "";
|
---|
471 | if(id.endsWith(_COLLECTION)) { // docID refers to a collection
|
---|
472 | // Obtain the "EX" datastream (extracted metadata) for the collection
|
---|
473 | ex = this.getEX(id);
|
---|
474 | }
|
---|
475 | else { // docID refers to a document
|
---|
476 | // work out the document's fedora PID and section ID, and then
|
---|
477 | // obtain the EX (extracted metadata) and DC datastreams for the doc
|
---|
478 |
|
---|
479 | // Note that EX/DC for pid="greenstone:<colname>-docPID-1"
|
---|
480 | // is the same as for pid="greenstone:<colname>-docPID"
|
---|
481 | // That is, <Section id="1"> refers to the toplevel document docPID
|
---|
482 | // If requested for top-level document, there may also be DLS meta
|
---|
483 | String sectionID = getSectionIDFromDocID(id);
|
---|
484 | String docPID = getDocPIDFromDocID(id);
|
---|
485 | if(sectionID.equals("") || sectionID.equals("1")) {
|
---|
486 | // metadata of toplevel document is requested
|
---|
487 | ex = this.getEX(docPID); // slightly faster than doing
|
---|
488 | //getSectionEXMetadata(docID, "1")
|
---|
489 | dc = this.getDC(docPID);
|
---|
490 | dls = this.getDLS(docPID);
|
---|
491 | }
|
---|
492 | else {
|
---|
493 | ex = getSectionEXMetadata(docPID, sectionID);
|
---|
494 | dc = getSectionDCMetadata(docPID, sectionID);
|
---|
495 | }
|
---|
496 | }
|
---|
497 |
|
---|
498 | String metafields = "";
|
---|
499 | for(int i = 0; i < metadata.length; i++) {
|
---|
500 | metafields = metafields + metadata[i] + "|";
|
---|
501 | }
|
---|
502 |
|
---|
503 | // Adding in metadata sets in alphabetical order
|
---|
504 | // DC metadata for a top-level document is different from EX, DLS:
|
---|
505 | // only the element's namespace prefix is "dc", the rest of a tagname
|
---|
506 | // is unknown.
|
---|
507 | if(!dc.equals("")) {
|
---|
508 | addMetadataWithNamespacedTagNames(doc, metadataList,
|
---|
509 | dc, DC, metafields);
|
---|
510 | }
|
---|
511 |
|
---|
512 | // Check if we were supposed to process dls and dc metadata
|
---|
513 | // as well. We only ever do this for top-level documents,
|
---|
514 | // in which case, dls and dc will be non-empty strings
|
---|
515 | if(!dls.equals("")) {
|
---|
516 | addMetadataWithFixedTagName(doc, metadataList, dls, DLS, metafields);
|
---|
517 | }
|
---|
518 |
|
---|
519 | // we definitely have an EX metadatastream for each
|
---|
520 | // collection object, top-level document object,
|
---|
521 | // and document section item
|
---|
522 | addMetadataWithFixedTagName(doc, metadataList, ex, EX, metafields);
|
---|
523 |
|
---|
524 | // now the metadataList has been built up
|
---|
525 | docNode.appendChild(metadataList);
|
---|
526 |
|
---|
527 | return docNode; // return <documentNode> containing the metadata
|
---|
528 | }
|
---|
529 |
|
---|
530 | /** This method retrieves all the metadata elements in the metaDataStream
|
---|
531 | * parameter of the form <"metadataSetNS:metadata">"value"</metadata> where
|
---|
532 | * metadataSetNS is the namespace of each tag, and creates a new element of
|
---|
533 | * the form <metadata name="metadataSetNS:metadata">"value"</metadata> for
|
---|
534 | * each. Each of these are then appended to the metadataList parameter.
|
---|
535 | * @param doc is the Document object using which the new metadata Elements
|
---|
536 | * are to be constructed
|
---|
537 | * @param metadataList is the <metadataList> Element to which the new
|
---|
538 | * metadata Elements are to be appended as children.
|
---|
539 | * @param metaDatastream the metadata datastream in string form (e.g. the
|
---|
540 | * Dublin Core metadata stored in the Fedora repository).
|
---|
541 | * @param metadataSet is the constant datastream identifier, e.g. "DC".
|
---|
542 | * At present this method applies to the DC metadata and any others like it
|
---|
543 | * where each tagname is different except for the constant dc: namespace.
|
---|
544 | * @param metafields is a | separated string containing the metadatafields to
|
---|
545 | * extract or "all" if all fields are requested
|
---|
546 | */
|
---|
547 | protected void addMetadataWithNamespacedTagNames(Document doc, Element metadataList,
|
---|
548 | String metaDatastream, String metadataSet, String metafields)
|
---|
549 | throws SAXException, IOException
|
---|
550 | {
|
---|
551 | Document src = builder.parse(
|
---|
552 | new InputSource(new StringReader(metaDatastream)));
|
---|
553 |
|
---|
554 | // The following doesn't work for some reason: to retrieve all elements
|
---|
555 | // whose namespace prefix starts with "dc", we pass "*" for localName
|
---|
556 | //NodeList dcMetaTags = src.getElementsByTagNameNS(metadataSet.toLowerCase(), "*");
|
---|
557 |
|
---|
558 | // Longer way: get the children of the root document
|
---|
559 | NodeList children = src.getDocumentElement().getChildNodes();
|
---|
560 |
|
---|
561 | for(int i = 0; i < children.getLength(); i++) {
|
---|
562 | String nodeName = children.item(i).getNodeName();
|
---|
563 | // check that the nodename starts with the metadataSet ("dc") namespace,
|
---|
564 | // which simultaneously ensures that the node's an element:
|
---|
565 | if(nodeName.toLowerCase().startsWith(metadataSet.toLowerCase())) {
|
---|
566 | // need to have a period for Greenstone instead of Fedora's colon
|
---|
567 | nodeName = nodeName.replace(COLON, PERIOD);
|
---|
568 | if(metadataSet.equals(DC)) { // dc:title -> dc.Title
|
---|
569 | nodeName = "dc" + PERIOD + Character.toString(Character.toUpperCase(nodeName.charAt(3)))
|
---|
570 | + nodeName.substring(4);
|
---|
571 | }
|
---|
572 |
|
---|
573 | // get the requested metadata fields
|
---|
574 | if(metafields.indexOf("all") != -1 || metafields.indexOf(nodeName) != -1) {
|
---|
575 | Element metatag = (Element)children.item(i);
|
---|
576 | String value = FedoraCommons.getValue(metatag);
|
---|
577 | // <dc:tagname>value</dc:tagname>
|
---|
578 | // we're going to put this in our metadata element as
|
---|
579 | // <metadata name="dc.Tagname">value</metadata>
|
---|
580 |
|
---|
581 | // create metadata of (name, value) pairs in target DOM (doc)
|
---|
582 | Element metadata = doc.createElement(GSXML.METADATA_ELEM);
|
---|
583 | Attr attribute = doc.createAttribute(GSXML.NAME_ATT);
|
---|
584 |
|
---|
585 | attribute.setValue(nodeName);
|
---|
586 | metadata.setAttributeNode(attribute);
|
---|
587 | Text content = doc.createTextNode(value);
|
---|
588 | metadata.appendChild(content);
|
---|
589 | metadataList.appendChild(metadata);
|
---|
590 | }
|
---|
591 | }
|
---|
592 | }
|
---|
593 | }
|
---|
594 |
|
---|
595 | /** This method retrieves all the metadata elements in the metaDataStream
|
---|
596 | * of the form <"namespace:"metadata name="metadataName">value</metadata>
|
---|
597 | * where "namespace" is the namespace prefix of each tag, and metadataName
|
---|
598 | * is the name of the metadata (like author, title). For each element
|
---|
599 | * it creates a corresponding new element of the form
|
---|
600 | * <metadata name="namespace:metadataName">value</metadata>.
|
---|
601 | * Each of these are then appended to the metadataList parameter.
|
---|
602 | * @param doc is the Document object using which the new metadata Elements
|
---|
603 | * are to be constructed
|
---|
604 | * @param metadataList is the <metadataList> Element to which the new
|
---|
605 | * metadata Elements are to be appended as children.
|
---|
606 | * @param metaDatastream the metadata datastream in string form (e.g. the
|
---|
607 | * EX/Greenstone extracted metadata or DLS metadata stored in the Fedora
|
---|
608 | * repository).
|
---|
609 | * @param metadataSet is the constant datastream identifier,
|
---|
610 | * e.g. "DLS" or "EX".
|
---|
611 | * At present this method applies to the DLS and EX metadata as they have
|
---|
612 | * constant tagnames throughout.
|
---|
613 | * @param metafields is a | separated string containing the metadatafields to
|
---|
614 | * extract or "all" if all fields are requested.
|
---|
615 | */
|
---|
616 | protected void addMetadataWithFixedTagName(Document doc, Element metadataList,
|
---|
617 | String metaDatastream, String metadataSet, String metafields)
|
---|
618 | throws SAXException, IOException
|
---|
619 | {
|
---|
620 | // Namespace prefix can be "ex:" or "dls:"
|
---|
621 | String namespacePrefix = "";
|
---|
622 | if(!metadataSet.equals(EX)) {
|
---|
623 | // need to have a period for Greenstone instead of Fedora's colon
|
---|
624 | namespacePrefix = metadataSet.toLowerCase() + PERIOD;
|
---|
625 | }
|
---|
626 |
|
---|
627 | Document src = builder.parse(
|
---|
628 | new InputSource(new StringReader(metaDatastream)));
|
---|
629 | NodeList metaTags = src.getElementsByTagName(
|
---|
630 | metadataSet.toLowerCase()+COLON+METADATA);
|
---|
631 | // Looking for tagnames: <ex:metadata> or <dls:metadata>
|
---|
632 |
|
---|
633 | for(int i = 0; i < metaTags.getLength(); i++) {
|
---|
634 | Element metatag = (Element)metaTags.item(i);
|
---|
635 |
|
---|
636 | // extract the metadata of (name, value) pairs from src DOM
|
---|
637 | // look for <metadata name="name">value</metadata>
|
---|
638 | String name = metatag.hasAttribute(NAME) ?
|
---|
639 | metatag.getAttribute(NAME) : "";
|
---|
640 | // sometimes, there are several metadata for the same name, in this
|
---|
641 | // case, look for a qualifier and append its value to the name to
|
---|
642 | // distinguish it uniquely:
|
---|
643 | if(metatag.hasAttribute(QUALIFIER)) {
|
---|
644 | name = name + HYPHEN + metatag.getAttribute(QUALIFIER);
|
---|
645 | }
|
---|
646 | name = namespacePrefix + name; // prefix with namespace, if any
|
---|
647 | if(metafields.indexOf("all") != -1 || metafields.indexOf(name) != -1) {
|
---|
648 | String value = FedoraCommons.getValue(metatag);
|
---|
649 |
|
---|
650 | // create metadata of (name, value) pairs in target DOM (doc)
|
---|
651 | Element metadata = doc.createElement(GSXML.METADATA_ELEM);
|
---|
652 | Attr attribute = doc.createAttribute(GSXML.NAME_ATT);
|
---|
653 | attribute.setValue(name);
|
---|
654 | metadata.setAttributeNode(attribute);
|
---|
655 | Text content = doc.createTextNode(value);
|
---|
656 | metadata.appendChild(content);
|
---|
657 |
|
---|
658 | metadataList.appendChild(metadata);
|
---|
659 | }
|
---|
660 | }
|
---|
661 | }
|
---|
662 |
|
---|
663 | /** Given a document identifier, returns a GS3 DocumentMetadataRetrieve
|
---|
664 | * response message containing ONLY the Title metadata for the document.
|
---|
665 | * @param docID is a document identifier (docID can either be a <pid>
|
---|
666 | * of an item (document) in the fedora repository, or it can be
|
---|
667 | * "<pid>-sectionNumber".
|
---|
668 | * @return a GS3 DocumentMetadataRetrieve response message containing the
|
---|
669 | * Title metadata for the requested document */
|
---|
670 | public String getTitleMetadata(String docID) {
|
---|
671 | return getTitleMetadata(new String[] { docID });
|
---|
672 | }
|
---|
673 |
|
---|
674 | /** Given a document identifier, returns a GS3 DocumentMetadataRetrieve
|
---|
675 | * response message containing ONLY the Title metadata for the documents.
|
---|
676 | * @param docIDs is a list of document identifiers (where docID can either be
|
---|
677 | * a <pid> of an item (document) in the fedora repository, or it can be
|
---|
678 | * "<pid>-sectionNumber".
|
---|
679 | * @return a GS3 DocumentMetadataRetrieve response message containing the
|
---|
680 | * Title metadata for all the requested documents */
|
---|
681 | public String getTitleMetadata(String[] docIDs) {
|
---|
682 | // Must create message of the following form:
|
---|
683 | // <documentNodeList><documentNode nodeID="docID">
|
---|
684 | // <metadataList><metadata name="Title">sometitle</metadata>
|
---|
685 | // </metadataList></documentNode>
|
---|
686 |
|
---|
687 | Document doc = builder.newDocument();
|
---|
688 | FedoraGS3RunException ex = null;
|
---|
689 |
|
---|
690 | Element docNodeList = doc.createElement(
|
---|
691 | GSXML.DOC_NODE_ELEM+GSXML.LIST_MODIFIER);
|
---|
692 | try{
|
---|
693 | for(int i = 0; i < docIDs.length; i++) {
|
---|
694 | Element docNode = getTitleMetadata(doc, docIDs[i]);
|
---|
695 | docNodeList.appendChild(docNode);
|
---|
696 | }
|
---|
697 | }catch(Exception e) {
|
---|
698 | ex = new FedoraGS3RunException(e);
|
---|
699 | //ex.setSpecifics("EX metadata datastream PID: |" + docIDs[i] + "|"); // for debugging PID
|
---|
700 | ex.setSpecifics("EX metadata datastream");
|
---|
701 | }
|
---|
702 |
|
---|
703 | Element responseMsg = createResponseMessage(doc, docNodeList, ex,
|
---|
704 | GSXML.REQUEST_TYPE_PROCESS, "DocumentMetadataRetrieve");
|
---|
705 | try{
|
---|
706 | return FedoraCommons.elementToString(responseMsg);
|
---|
707 | } catch(TransformerException e) {
|
---|
708 | return FedoraGS3RunException.xmlToStringConversionFailureResponseMsg
|
---|
709 | + " " + e;
|
---|
710 | }
|
---|
711 | }
|
---|
712 |
|
---|
713 | /** Method that takes a new DOM document, as well as an identifier of either
|
---|
714 | * a document or document section and returns a documentNode element containing
|
---|
715 | * the title metadata for it:
|
---|
716 | * <documentNode nodeID="docID"><metadataList>
|
---|
717 | * <metadata name="Title">sometitle</metadata>
|
---|
718 | * </metadataList></documentNode>
|
---|
719 | * @return documentNode containing the metadata for the collection or
|
---|
720 | * document given by parameter ID
|
---|
721 | * @param docID denotes the id of a document or a document section, so id
|
---|
722 | * is either a document-pid or it's of the form documentpid-sectionNumber */
|
---|
723 | protected Element getTitleMetadata(Document doc, String docID)
|
---|
724 | throws RemoteException, UnsupportedEncodingException,
|
---|
725 | SAXException, IOException
|
---|
726 | {
|
---|
727 | // Returns a docNode element of the following form:
|
---|
728 | // <documentNode nodeID="docID">
|
---|
729 | // <metadataList><metadata name="Title">sometitle</metadata></metadataList>
|
---|
730 | // </documentNode>
|
---|
731 |
|
---|
732 | // <documentNode nodeID="docID">
|
---|
733 | Element docNode = doc.createElement(GSXML.DOC_NODE_ELEM);
|
---|
734 | Attr attribute = doc.createAttribute(GSXML.NODE_ID_ATT);
|
---|
735 | attribute.setValue(docID);
|
---|
736 | docNode.setAttributeNode(attribute);
|
---|
737 |
|
---|
738 | // <metadataList>
|
---|
739 | Element metaList = doc.createElement(
|
---|
740 | GSXML.METADATA_ELEM+GSXML.LIST_MODIFIER);
|
---|
741 | // <metadata name="Title">
|
---|
742 | Element metadata = doc.createElement(GSXML.METADATA_ELEM);
|
---|
743 | // if we connect it all up (append children), we can immediately add
|
---|
744 | // the name attribute into the metadata element:
|
---|
745 | metaList.appendChild(metadata);
|
---|
746 | docNode.appendChild(metaList);
|
---|
747 | metadata.setAttribute(GSXML.NAME_ATT, TITLE); // immediately add attribute
|
---|
748 |
|
---|
749 | String title = "";
|
---|
750 | String sectionID = getSectionIDFromDocID(docID);
|
---|
751 | String docPID = getDocPIDFromDocID(docID);
|
---|
752 |
|
---|
753 | // check if title of toplevel document is requested
|
---|
754 | if(sectionID.equals(""))
|
---|
755 | title = this.getDocTitle(docPID);
|
---|
756 | else { // title of document section
|
---|
757 | title = this.getSectionTitle(docPID, sectionID);
|
---|
758 | }
|
---|
759 |
|
---|
760 | metadata.appendChild(doc.createTextNode(title));
|
---|
761 |
|
---|
762 | return docNode;
|
---|
763 | }
|
---|
764 |
|
---|
765 | /** @return a String representing Greenstone3 DocumentMetadataRetrieve XML
|
---|
766 | * containing the requested portion of the document structure of the documents
|
---|
767 | * indicated by docIDs:
|
---|
768 | * @param docID is the document identifier of the document whose hierarchical
|
---|
769 | * structure is requested. The name of the collection is already included in the
|
---|
770 | * docID for a Fedora DL.
|
---|
771 | * @param structure - strings specifying the required structure of the document.
|
---|
772 | * It can be a combination of: ancestors, parent, siblings, children, descendants, entire.
|
---|
773 | * @param info - strings specifying the required structural info of the document.
|
---|
774 | * It can be any combination of: siblingPosition, numSiblings, numChildren.
|
---|
775 | */
|
---|
776 | public String getDocumentStructure(String docID, String[] structure, String[] info) {
|
---|
777 | return getStructure(new String[]{docID}, structure, info);
|
---|
778 | }
|
---|
779 |
|
---|
780 |
|
---|
781 | /** @return a String representing Greenstone3 DocumentMetadataRetrieve XML
|
---|
782 | * containing the requested portion of the document structure of the documents
|
---|
783 | * indicated by docIDs:
|
---|
784 | * @param docIDs is an array of document identifiers of documents whose
|
---|
785 | * hierarchical structures are requested. The name of the collection is already
|
---|
786 | * included in the docID for a Fedora DL.
|
---|
787 | * @param structure - strings specifying the required structure of each document.
|
---|
788 | * It can be a combination of: ancestors, parent, siblings, children, descendants, entire.
|
---|
789 | * @param info - strings specifying the required structural info of each document.
|
---|
790 | * It can be any combination of: siblingPosition, numSiblings, numChildren.
|
---|
791 | */
|
---|
792 | public String getDocumentStructure(String[] docIDs, String[] structure, String[] info) {
|
---|
793 | return getStructure(docIDs, structure, info);
|
---|
794 | }
|
---|
795 |
|
---|
796 | /**
|
---|
797 | * Returns a greenstone3 DocumentStructureRetrieve XML response message
|
---|
798 | * containing the document structures for the given docIDs.
|
---|
799 | * Similar to FedoraConnection.getTOC(), but instead of fedora formatted XML,
|
---|
800 | * greenstone formatted XML is returned. The requested section of the table
|
---|
801 | * of contents (TOC) for a document is converted into the greenstone3 xml
|
---|
802 | * format that is returned upon DocumentStructureRetrieve requests.
|
---|
803 | * @param docIDs the documentIDs for which the section's structure is returned;
|
---|
804 | * where a docID is either a fedora pid <docPID> or <docPID>-<sectionNumber>.
|
---|
805 | * @param structure - the structure of the sections to return. Can be any combination of:
|
---|
806 | * ancestors, parent, siblings, children, descendants, entire.
|
---|
807 | * @param infos - strings containing any combination of the values: numChildren, numSiblings,
|
---|
808 | * siblingPosition. The requested info gets added as attributes to the returned root element.
|
---|
809 | * @return a greenstone3 DocumentStructureRetrieve XML response message in
|
---|
810 | * String format with the structure of the docIDs requested.
|
---|
811 | */
|
---|
812 | protected String getStructure(String[] docIDs, String[] structure, String[] infos)
|
---|
813 | {
|
---|
814 | Document doc = builder.newDocument();
|
---|
815 | FedoraGS3RunException ex = null;
|
---|
816 | // <documentNodeList>
|
---|
817 | Element docNodeList = doc.createElement(
|
---|
818 | GSXML.DOC_NODE_ELEM+GSXML.LIST_MODIFIER);
|
---|
819 |
|
---|
820 | try{
|
---|
821 | // append the <documentNodes> for the docIDs
|
---|
822 | // to the docNodeList
|
---|
823 | //getStructureElement(docNodeList, docIDs, levels);
|
---|
824 | getStructureElement(docNodeList, docIDs, structure, infos);
|
---|
825 | } catch(Exception e) {
|
---|
826 | ex = new FedoraGS3RunException(e);
|
---|
827 | ex.setSpecifics("(requested portion of) TOC datastream");
|
---|
828 | }
|
---|
829 | // insert our <documentNodeList> into a GS3 response message
|
---|
830 | Element responseMsg = createResponseMessage(doc, docNodeList, ex,
|
---|
831 | GSXML.REQUEST_TYPE_PROCESS, "DocumentStructureRetrieve");
|
---|
832 | try{
|
---|
833 | return FedoraCommons.elementToString(responseMsg);
|
---|
834 | } catch(TransformerException e) {
|
---|
835 | return FedoraGS3RunException.xmlToStringConversionFailureResponseMsg
|
---|
836 | + " " + e;
|
---|
837 | }
|
---|
838 | }
|
---|
839 |
|
---|
840 |
|
---|
841 | /** Given a <documentNodeList> portion of a greenstone3
|
---|
842 | * DocumentStructureRetrieve XML response message, this method will populate
|
---|
843 | * it with the <documentNodes> that represent the structure of the given docIDs.
|
---|
844 | * @param docNodeList is a <documentNodeList> to which <documentNodes> of
|
---|
845 | * the doc structures are appended.
|
---|
846 | * @param docIDs the documentIDs for which the section's structure is returned;
|
---|
847 | * where a docID is either a fedora pid <docPID> or <docPID>-<sectionNumber>.
|
---|
848 | * @param structures - the structure of the sections to return. Can be any combination of:
|
---|
849 | * ancestors, parent, siblings, children, descendants, entire.
|
---|
850 | * @param infos - a string containing any combination of the values: numChildren, numSiblings,
|
---|
851 | * siblingPosition. The requested info gets added as attributes to the returned root element.
|
---|
852 | */
|
---|
853 | protected void getStructureElement(Element docNodeList, String[] docIDs,
|
---|
854 | String[] structures, String[] infos)
|
---|
855 | throws RemoteException, UnsupportedEncodingException, SAXException,
|
---|
856 | IOException
|
---|
857 | {
|
---|
858 | // Make one string out of requested structure components, and one string from info components
|
---|
859 | String structure = "";
|
---|
860 | String info = "";
|
---|
861 | for(int i = 0; i < structures.length; i++) {
|
---|
862 | structure = structure + structures[i] + "|";
|
---|
863 | }
|
---|
864 | for(int i = 0; i < infos.length; i++) {
|
---|
865 | info = info + infos[i] + "|";
|
---|
866 | }
|
---|
867 |
|
---|
868 | // process each docID
|
---|
869 | for(int i = 0; i < docIDs.length; i++) {
|
---|
870 | // work out the document's fedora PID and section ID
|
---|
871 | String sectionID = getSectionIDFromDocID(docIDs[i]);
|
---|
872 | String docPID = getDocPIDFromDocID(docIDs[i]);
|
---|
873 | if(sectionID.equals("")) {
|
---|
874 | sectionID = "1";
|
---|
875 | }
|
---|
876 |
|
---|
877 | // get the required section, along with children or descendants
|
---|
878 | Element srcDocElement = getSectionStructureXML(docPID, sectionID, structure, info);
|
---|
879 | Document doc = docNodeList.getOwnerDocument();
|
---|
880 |
|
---|
881 | // copy-and-convert that structure into a structure format for GS3
|
---|
882 | Element docNode = getStructure(doc, docIDs[i], docPID, srcDocElement);
|
---|
883 |
|
---|
884 | if(!info.equals("")) {
|
---|
885 | // <nodeStructureInfo>
|
---|
886 | // <info name="" value="" />
|
---|
887 | // <info name="" value="" />
|
---|
888 | // ...
|
---|
889 | // </nodeStructureInfo>
|
---|
890 | Element nodeStructureInfo = doc.createElement(GSXML.NODE_STRUCTURE_ELEM+GSXML.INFO_ATT);
|
---|
891 | Element root = srcDocElement.getOwnerDocument().getDocumentElement();
|
---|
892 |
|
---|
893 | if(root.hasAttribute("numSiblings")) {
|
---|
894 | String numSiblings = root.getAttribute("numSiblings");
|
---|
895 | Element infoEl = doc.createElement(GSXML.INFO_ATT);
|
---|
896 | infoEl.setAttribute(GSXML.NAME_ATT, "numSiblings");
|
---|
897 | infoEl.setAttribute(GSXML.VALUE_ATT, numSiblings);
|
---|
898 | nodeStructureInfo.appendChild(infoEl);
|
---|
899 | }
|
---|
900 |
|
---|
901 | if(root.hasAttribute("siblingPosition")) {
|
---|
902 | String siblingPosition = root.getAttribute("siblingPosition");
|
---|
903 | Element infoEl = doc.createElement(GSXML.INFO_ATT);
|
---|
904 | infoEl.setAttribute(GSXML.NAME_ATT, "siblingPosition");
|
---|
905 | infoEl.setAttribute(GSXML.VALUE_ATT, siblingPosition);
|
---|
906 | nodeStructureInfo.appendChild(infoEl);
|
---|
907 | }
|
---|
908 |
|
---|
909 | if(root.hasAttribute("numChildren")) {
|
---|
910 | String numChildren = root.getAttribute("numChildren");
|
---|
911 | Element infoEl = doc.createElement(GSXML.INFO_ATT);
|
---|
912 | infoEl.setAttribute(GSXML.NAME_ATT, "numChildren");
|
---|
913 | infoEl.setAttribute(GSXML.VALUE_ATT, numChildren);
|
---|
914 | nodeStructureInfo.appendChild(infoEl);
|
---|
915 | }
|
---|
916 | docNode.appendChild(nodeStructureInfo);
|
---|
917 | }
|
---|
918 |
|
---|
919 | // add it to our list of documentNodes
|
---|
920 | docNodeList.appendChild(docNode);
|
---|
921 | }
|
---|
922 | }
|
---|
923 |
|
---|
924 |
|
---|
925 | /**
|
---|
926 | * Takes the portion of the XML document outlining the structure of the
|
---|
927 | * document (section)--in the format this is stored in Fedora--and returns
|
---|
928 | * Greenstone 3 DOM XML format for outlining document structure.
|
---|
929 | * @return a <documentNode> element that contains a greenstone3
|
---|
930 | * DocumentStructureRetrieve XML corresponding to the parameter Element section
|
---|
931 | * (which is in fedora XML), for the document indicated by docID.
|
---|
932 | * @param requestingDocID is the identifier of the document for which the
|
---|
933 | * structure was requested. It's this document's children or descendants that
|
---|
934 | * will be returned. Note that this is not always the same as (clear from)
|
---|
935 | * parameter docID.
|
---|
936 | * @param docID is the documentID for which the section's structure is
|
---|
937 | * returned where docID = "docPID-sectionNumber".
|
---|
938 | * @param section - the fedora section XML that is being mirrored in
|
---|
939 | * greenstone3 format.
|
---|
940 | */
|
---|
941 | protected Element getStructure(Document doc, String requestingDocID,
|
---|
942 | String docID, Element section)
|
---|
943 | {
|
---|
944 | // we want to mirror the section's DOM (given in fedora XML) in
|
---|
945 | // greenstone3's XML for a DocumentStructureRetrieve response.
|
---|
946 |
|
---|
947 | // <documentNode nodeID="docID"> - the docNode on which a structure retrieve
|
---|
948 | // is being performed
|
---|
949 | Element docNode = doc.createElement(GSXML.DOC_NODE_ELEM);
|
---|
950 | Attr attribute = doc.createAttribute(GSXML.NODE_ID_ATT);
|
---|
951 | attribute.setValue(requestingDocID); //requestingDocID.replace(HYPHEN+SECTION, "")
|
---|
952 | docNode.setAttributeNode(attribute);
|
---|
953 |
|
---|
954 | // <nodeStructure>
|
---|
955 | Element nodeStructure = doc.createElement(GSXML.NODE_STRUCTURE_ELEM);
|
---|
956 |
|
---|
957 | // <documentNode nodeID="docID" docType="hierarchy" nodeType="root">
|
---|
958 | Element rootNode = createDocNodeFromSubsection(doc, section, docID);
|
---|
959 |
|
---|
960 | // fills in the subtree of the rootNode in our nodeStructure element
|
---|
961 | createDocStructure(doc, section, rootNode, docID);
|
---|
962 | //where section represents the root section
|
---|
963 |
|
---|
964 | nodeStructure.appendChild(rootNode);
|
---|
965 | docNode.appendChild(nodeStructure);
|
---|
966 | return docNode;
|
---|
967 | }
|
---|
968 |
|
---|
969 |
|
---|
970 | /** Recursive method that creates a documentStructure mirroring parameter
|
---|
971 | * section, starting from parameter parent down to all descendants
|
---|
972 | * @param section is the XML <Section> in the fedora repository's TOC
|
---|
973 | * for the docPID whose substructure is to be mirrored
|
---|
974 | * @param parent is the XML documentNode in the greenstone repository whose
|
---|
975 | * descendants created by this method will correspond to the descendants of
|
---|
976 | * parameter section.
|
---|
977 | * @param doc is the document containing the parent;
|
---|
978 | * @param docPID is the prefix of all nodeIDs in the parent's structure
|
---|
979 | */
|
---|
980 | protected void createDocStructure(
|
---|
981 | Document doc, Element section, Element parent, String docPID)
|
---|
982 | {
|
---|
983 | // get the section's children (if any)
|
---|
984 | NodeList children = section.getChildNodes();
|
---|
985 | for(int i = 0; i < children.getLength(); i++) {
|
---|
986 | Node n = children.item(i);
|
---|
987 |
|
---|
988 | if(n.getNodeName().equals(SECTION_ELEMENT)) {
|
---|
989 | //then we know it's an element AND that its tagname is "Section"
|
---|
990 | Element subsection = (Element)n;
|
---|
991 | Element child = createDocNodeFromSubsection(doc, subsection, docPID);
|
---|
992 | parent.appendChild(child);
|
---|
993 |
|
---|
994 | // recursion call on newly found child-element and subsection
|
---|
995 | createDocStructure(doc, subsection, child, docPID);
|
---|
996 | }
|
---|
997 | }
|
---|
998 | }
|
---|
999 |
|
---|
1000 | /** Given a particular subsection element, this method creates a
|
---|
1001 | * Greenstone3 DocumentNode element that mirrors it.
|
---|
1002 | * @param doc is the document that will contain the created DocumentNode
|
---|
1003 | * @param docID is the prefix of all nodeIDs in the parent's structure
|
---|
1004 | * @param subSection is the XML <Section> in the fedora repository's
|
---|
1005 | * TOC for the docPID which will be mirrored in the greenstone XML
|
---|
1006 | * documentNode that will be returned.
|
---|
1007 | * @return a greenstone <documentNode> that represents the fedora TOC's
|
---|
1008 | * <Section> element passed as parameter subSection. */
|
---|
1009 | protected Element createDocNodeFromSubsection(
|
---|
1010 | Document doc, Element subSection, String docID)
|
---|
1011 | {
|
---|
1012 | Element docNode = doc.createElement(GSXML.DOC_NODE_ELEM);
|
---|
1013 | Attr docType = doc.createAttribute(GSXML.DOC_TYPE_ATT);
|
---|
1014 | docType.setValue(GSXML.DOC_TYPE_HIERARCHY);
|
---|
1015 | docNode.setAttributeNode(docType);
|
---|
1016 |
|
---|
1017 | Attr nodeID = doc.createAttribute(GSXML.NODE_ID_ATT);
|
---|
1018 | String sectionID = subSection.hasAttribute(ID) ?
|
---|
1019 | subSection.getAttribute(ID) : "";
|
---|
1020 | if(sectionID.equals("1")
|
---|
1021 | && subSection.getElementsByTagName(SECTION_ELEMENT).getLength() > 0) { // root, non-leaf case
|
---|
1022 | // reset the attribute without the section number (just "docID" may be important for democlient?)
|
---|
1023 | nodeID.setValue(docID + HYPHEN + sectionID); // nodeID.setValue(docID);
|
---|
1024 | } else {
|
---|
1025 | nodeID.setValue(docID + HYPHEN + sectionID);
|
---|
1026 | }
|
---|
1027 | //nodeID.setValue(docID + HYPHEN + sectionID);
|
---|
1028 | docNode.setAttributeNode(nodeID);
|
---|
1029 |
|
---|
1030 | Attr nodeType = doc.createAttribute(GSXML.NODE_TYPE_ATT);
|
---|
1031 | if(subSection.hasAttribute(GSXML.NODE_TYPE_ATT)) {
|
---|
1032 | nodeType.setValue(subSection.getAttribute(GSXML.NODE_TYPE_ATT));
|
---|
1033 | }
|
---|
1034 | docNode.setAttributeNode(nodeType);
|
---|
1035 | return docNode;
|
---|
1036 | }
|
---|
1037 |
|
---|
1038 |
|
---|
1039 | /** Given an identifier that is either a docPID or a concatenation of
|
---|
1040 | * docPID+sectionID, this method works out the fedora assigned docPID and
|
---|
1041 | * sectionID and then calls getContentBody(docPID, sectionID) with those.
|
---|
1042 | * @param docID is expected to be of the form
|
---|
1043 | * "greenstone:<collectionName>-<docPID>-<sectionNumber>" or
|
---|
1044 | * "greenstone:<collectionName>-<docPID>"
|
---|
1045 | * If it is "greenstone:<collectionName>-<docPID>", then the content for
|
---|
1046 | * "greenstone:<collectionName>-1" ("greenstone:<collectionName>-Section1")
|
---|
1047 | * is returned! */
|
---|
1048 | public String getContent(String docID) {
|
---|
1049 | return this.getContent(new String[]{docID});
|
---|
1050 | }
|
---|
1051 |
|
---|
1052 | /** Given an identifier that is a concatenation of docID+sectionID, this
|
---|
1053 | * method works out the fedora assigned docPID and sectionID and then calls
|
---|
1054 | * getContentBody(docPID, sectionID) with those.
|
---|
1055 | * @param docIDs is an array of document identifiers of the form
|
---|
1056 | * "greenstone:<collectionName>-<docPID>-<sectionNumber>"
|
---|
1057 | * If it is "greenstone:<collectionName>-<docPID>", then the content for
|
---|
1058 | * "greenstone:<collectionName>-Section1" is returned! */
|
---|
1059 | public String getContent(String[] docIDs) {
|
---|
1060 | Document doc = builder.newDocument();
|
---|
1061 | FedoraGS3RunException ex = null;
|
---|
1062 |
|
---|
1063 | //<documentNodeList>
|
---|
1064 | Element docNodeList = doc.createElement(
|
---|
1065 | GSXML.DOC_NODE_ELEM+GSXML.LIST_MODIFIER);
|
---|
1066 |
|
---|
1067 | try{
|
---|
1068 | for(int i = 0; i < docIDs.length; i++) {
|
---|
1069 | // get the sectionID and docPID from the docID
|
---|
1070 | String sectionID = this.removePrefix(
|
---|
1071 | getSectionIDFromDocID(docIDs[i]), SECTION);
|
---|
1072 | String docPID = getDocPIDFromDocID(docIDs[i]);
|
---|
1073 | if(sectionID.equals("")) // if no section is specified, get
|
---|
1074 | sectionID = "1"; // get the content for Section id="1"
|
---|
1075 |
|
---|
1076 | // Get the contents for the requested section of document docPID
|
---|
1077 | String sectionContent = this.getContentBody(docPID, sectionID);
|
---|
1078 |
|
---|
1079 | // set the nodeID attribute
|
---|
1080 | Element docNode = doc.createElement(GSXML.DOC_NODE_ELEM);
|
---|
1081 | Attr nodeId = doc.createAttribute(GSXML.NODE_ID_ATT);
|
---|
1082 |
|
---|
1083 | nodeId.setValue(docIDs[i]); // just set the docID which will contain
|
---|
1084 | // the docPID (and sectionID if already present)
|
---|
1085 |
|
---|
1086 | docNode.setAttributeNode(nodeId);
|
---|
1087 | // set the text content to what was retrieved
|
---|
1088 | Element nodeContent = doc.createElement(GSXML.NODE_CONTENT_ELEM);
|
---|
1089 | Text textNode = doc.createTextNode(sectionContent.trim());
|
---|
1090 |
|
---|
1091 | nodeContent.appendChild(textNode);
|
---|
1092 | docNode.appendChild(nodeContent);
|
---|
1093 | //add the documentNode to the docNodeList
|
---|
1094 | docNodeList.appendChild(docNode);
|
---|
1095 | }
|
---|
1096 | } catch(Exception e) {
|
---|
1097 | ex = new FedoraGS3RunException(e);
|
---|
1098 | ex.setSpecifics("requested doc Section datastream");
|
---|
1099 | }
|
---|
1100 | Element responseMsg = createResponseMessage(doc, docNodeList, ex,
|
---|
1101 | GSXML.REQUEST_TYPE_PROCESS, "DocumentContentRetrieve");
|
---|
1102 | try{
|
---|
1103 | return FedoraCommons.elementToString(responseMsg);
|
---|
1104 | } catch(TransformerException e) {
|
---|
1105 | return FedoraGS3RunException.xmlToStringConversionFailureResponseMsg
|
---|
1106 | + " " + e;
|
---|
1107 | }
|
---|
1108 | }
|
---|
1109 |
|
---|
1110 | /** Gets the contents of a textNode from a section.
|
---|
1111 | * @return the text content of a section.
|
---|
1112 | * @param docPID the pid of the document from which a section's text is to
|
---|
1113 | * be retrieved.
|
---|
1114 | * @param sectionID is the section identifier of the document denoted by
|
---|
1115 | * docPID whose text is to be returned.
|
---|
1116 | */
|
---|
1117 | protected String getContentBody(String docPID, String sectionID)
|
---|
1118 | throws RemoteException, UnsupportedEncodingException,
|
---|
1119 | SAXException, IOException
|
---|
1120 | {
|
---|
1121 | String section = this.getSection(docPID, sectionID);
|
---|
1122 |
|
---|
1123 | // the content is nested inside a <Section> element,
|
---|
1124 | // we extract it from there:
|
---|
1125 | InputSource source = new InputSource(new StringReader(section));
|
---|
1126 | Document doc = builder.parse(source);
|
---|
1127 |
|
---|
1128 | // The document Element is the <Section> we want.
|
---|
1129 | // Get its text contents:
|
---|
1130 | section = FedoraCommons.getValue(doc.getDocumentElement());
|
---|
1131 |
|
---|
1132 | // we are going to remove all occurrences of "_httpdocimg_/"
|
---|
1133 | // that precede associated filenames, because that's a GS3
|
---|
1134 | // defined macro for resolving relative urls. It won't help
|
---|
1135 | // with documents stored in fedora.
|
---|
1136 | section = section.replaceAll(GS3FilePathMacro+"/", "");
|
---|
1137 | return section;
|
---|
1138 | }
|
---|
1139 |
|
---|
1140 | /** Here we create the greenstone's response message element:
|
---|
1141 | * <message≶<response><content></response></message>
|
---|
1142 | * @return a greenstone response-message element.
|
---|
1143 | * @param doc - the Document object which should me used to create the
|
---|
1144 | * <message> and <response> elements
|
---|
1145 | * @param content - the element that is to be nested inside <response>
|
---|
1146 | * @param ex - any exception that occurred when trying to create
|
---|
1147 | * the content parameter
|
---|
1148 | * @param responseType - the value for the type attribute of <response>,
|
---|
1149 | * such as "describe", "retrieve", "browse", "query"...
|
---|
1150 | * @param originator - indiates the collectionName or service (like
|
---|
1151 | * DocumentContentRetrieve) from where this response message originates
|
---|
1152 | */
|
---|
1153 | protected Element createResponseMessage(Document doc, Element content,
|
---|
1154 | Exception ex, String responseType, String originator)
|
---|
1155 | {
|
---|
1156 | Element response = doc.createElement(GSXML.RESPONSE_ELEM);
|
---|
1157 | // from = "FedoraGS3"
|
---|
1158 | Attr attribute = doc.createAttribute(GSXML.FROM_ATT);
|
---|
1159 | attribute.setValue(originator);
|
---|
1160 | response.setAttributeNode(attribute);
|
---|
1161 |
|
---|
1162 | // type = "describe" or "process" - whatever's given in requestType:
|
---|
1163 | attribute = doc.createAttribute(GSXML.TYPE_ATT);
|
---|
1164 | attribute.setValue(responseType);
|
---|
1165 | response.setAttributeNode(attribute);
|
---|
1166 |
|
---|
1167 | if(content != null)
|
---|
1168 | response.appendChild(content);
|
---|
1169 |
|
---|
1170 | // we'll create an error element for RemoteExceptions (web service problems)
|
---|
1171 | // and UnsupportedEncodingExceptions and
|
---|
1172 | if(ex != null) {
|
---|
1173 | Element error = doc.createElement(GSXML.ERROR_ELEM);
|
---|
1174 | error.appendChild(doc.createTextNode(ex.getMessage()));
|
---|
1175 | // now append the error to the <response> element (after
|
---|
1176 | // the content element whatever that was)
|
---|
1177 | response.appendChild(error);
|
---|
1178 | }
|
---|
1179 |
|
---|
1180 | Element message = doc.createElement(GSXML.MESSAGE_ELEM);
|
---|
1181 | message.appendChild(response);
|
---|
1182 | doc.appendChild(message);
|
---|
1183 | return message;
|
---|
1184 | }
|
---|
1185 |
|
---|
1186 | /** @return a <serviceList> Element as defined by GS3: containing all the
|
---|
1187 | * services (denoted by <service> elements) that are supported by FedoraGS3.
|
---|
1188 | * At present these are: DocumentContentRetrieve, DocumentMetadataRetrieve,
|
---|
1189 | * DocumentStructureRetrieve, TextQuery, FieldQuery, ClassifierBrowse,
|
---|
1190 | * ClassifierBrowseMetadataRetrieve (as indicated by member var serviceNames).
|
---|
1191 | * @param doc - the Document object which should me used to create the
|
---|
1192 | * <serviceList> element */
|
---|
1193 | protected Element createServiceList(Document doc)
|
---|
1194 | {
|
---|
1195 | Element serviceList = doc.createElement(
|
---|
1196 | GSXML.SERVICE_ELEM+GSXML.LIST_MODIFIER);
|
---|
1197 |
|
---|
1198 | for(int i = 0; i < serviceNames.length; i++) {
|
---|
1199 | // create the <service name="serviceName[i]" type="servicetype" />
|
---|
1200 | Element service = doc.createElement(GSXML.SERVICE_ELEM);
|
---|
1201 |
|
---|
1202 | Attr attribute = doc.createAttribute(GSXML.NAME_ATT);
|
---|
1203 | attribute.setValue(serviceNames[i]);
|
---|
1204 | service.setAttributeNode(attribute);
|
---|
1205 |
|
---|
1206 | attribute = doc.createAttribute(GSXML.TYPE_ATT);
|
---|
1207 | if(serviceNames[i].equals("ClassifierBrowse")) //browseTitlesByLetter
|
---|
1208 | attribute.setValue(GSXML.SERVICE_TYPE_BROWSE);
|
---|
1209 | else if(serviceNames[i].contains("Query")) // search services
|
---|
1210 | attribute.setValue(GSXML.SERVICE_TYPE_QUERY);
|
---|
1211 | else
|
---|
1212 | attribute.setValue(GSXML.SERVICE_TYPE_RETRIEVE);
|
---|
1213 | service.setAttributeNode(attribute);
|
---|
1214 |
|
---|
1215 | // add the service element to the serviceList element
|
---|
1216 | // <serviceList><service /></serviceList>
|
---|
1217 | serviceList.appendChild(service);
|
---|
1218 | }
|
---|
1219 | return serviceList;
|
---|
1220 | }
|
---|
1221 |
|
---|
1222 | /** @return a GS3 response message for a describe services request:
|
---|
1223 | * indicating the list of services supported by the Fedora-Greenstone
|
---|
1224 | * interface. These are DocumentContentRetrieve, DocumentMetadataRetrieve,
|
---|
1225 | * DocumentStructureRetrieve, ClassifierBrowse, TextQuery, FieldQuery,
|
---|
1226 | * ClassifierBrowseMetadataRetrieve - as indicated by member variable
|
---|
1227 | * serviceNames. */
|
---|
1228 | public String getServiceList()
|
---|
1229 | {
|
---|
1230 | Document doc = builder.newDocument();
|
---|
1231 | Element serviceList = createServiceList(doc);
|
---|
1232 | // make <serviceList> the body of the responseMessage:
|
---|
1233 | // <message><response><serviceList></response></message>
|
---|
1234 | Element responseMsg = createResponseMessage(doc, serviceList, null,
|
---|
1235 | GSXML.REQUEST_TYPE_DESCRIBE, "");
|
---|
1236 | try {
|
---|
1237 | return FedoraCommons.elementToString(responseMsg);
|
---|
1238 | }catch(TransformerException e) {
|
---|
1239 | return FedoraGS3RunException.xmlToStringConversionFailureResponseMsg
|
---|
1240 | + " " + e;
|
---|
1241 | }
|
---|
1242 | }
|
---|
1243 |
|
---|
1244 | /** @return a GS3 describe response message listing the collections and
|
---|
1245 | * collection-specific metadata stored in the Fedora-Greenstone repository. */
|
---|
1246 | public String getCollectionList()
|
---|
1247 | {
|
---|
1248 | Document doc = builder.newDocument();
|
---|
1249 | FedoraGS3RunException ex = null; // any RemoteException
|
---|
1250 |
|
---|
1251 | // create the <collectionList /> element
|
---|
1252 | Element collectionList = doc.createElement(
|
---|
1253 | GSXML.COLLECTION_ELEM+GSXML.LIST_MODIFIER);
|
---|
1254 | try{
|
---|
1255 | String[] collectionNames = this.getCollectionNames(
|
---|
1256 | this.getCollections()); // this line could throw RemoteException
|
---|
1257 | for(int i = 0; i < collectionNames.length; i++) {
|
---|
1258 | // create the <collection name="somename" /> element
|
---|
1259 | Element collection = doc.createElement(GSXML.COLLECTION_ELEM);
|
---|
1260 | Attr attribute = doc.createAttribute(GSXML.NAME_ATT);
|
---|
1261 | attribute.setValue(collectionNames[i]);
|
---|
1262 | collection.setAttributeNode(attribute);
|
---|
1263 |
|
---|
1264 | // append the <collection> element as child of <collectionList>
|
---|
1265 | collectionList.appendChild(collection);
|
---|
1266 |
|
---|
1267 | //if(collection.hasAttribute(GSXML.NAME_ATT))
|
---|
1268 | //LOG.debug(collection.getAttribute(GSXML.NAME_ATT));
|
---|
1269 | }
|
---|
1270 | } catch(RemoteException e) { // if this happens, perhaps it's because it
|
---|
1271 | // can't find Greenstone collections in fedora repository?
|
---|
1272 | ex = new FedoraGS3RunException(e);
|
---|
1273 | ex.setSpecifics(
|
---|
1274 | "greenstone collections in fedora repository");
|
---|
1275 | }
|
---|
1276 |
|
---|
1277 | // make <collectionList> the body of the responseMessage:
|
---|
1278 | // <message><response><collectionList></response></message>
|
---|
1279 | Element responseMsg = createResponseMessage(doc, collectionList, ex,
|
---|
1280 | GSXML.REQUEST_TYPE_DESCRIBE, "");
|
---|
1281 | try{
|
---|
1282 | return FedoraCommons.elementToString(responseMsg);
|
---|
1283 | }catch(TransformerException e) {
|
---|
1284 | return FedoraGS3RunException.xmlToStringConversionFailureResponseMsg
|
---|
1285 | + " " + e;
|
---|
1286 | }
|
---|
1287 | }
|
---|
1288 |
|
---|
1289 | /** @return a GS3 describe response message for a collection in the
|
---|
1290 | * Fedora-Greenstone repository.
|
---|
1291 | * @param collectionName - the name of the collection that is to be described.
|
---|
1292 | * It will be converted to a fedora collection pid, which is of the form
|
---|
1293 | * "greenstone:<collectionName>-collection". */
|
---|
1294 | public String describeCollection(String collectionName)
|
---|
1295 | {
|
---|
1296 | Document doc = builder.newDocument();
|
---|
1297 | FedoraGS3RunException ex = null;
|
---|
1298 |
|
---|
1299 | Element collection = doc.createElement(GSXML.COLLECTION_ELEM);
|
---|
1300 | Attr attribute = doc.createAttribute(GSXML.NAME_ATT);
|
---|
1301 | attribute.setValue(collectionName);
|
---|
1302 | collection.setAttributeNode(attribute);
|
---|
1303 |
|
---|
1304 | //<displayItem assigned="true" lang="en" name="name">
|
---|
1305 | //"some display name"</displayItem>
|
---|
1306 | Element displayItem = doc.createElement(GSXML.DISPLAY_TEXT_ELEM);
|
---|
1307 |
|
---|
1308 | attribute = doc.createAttribute(GSXML.LANG_ATT);
|
---|
1309 | attribute.setValue(this.lang);
|
---|
1310 | displayItem.setAttributeNode(attribute);
|
---|
1311 |
|
---|
1312 | attribute = doc.createAttribute(GSXML.NAME_ATT);
|
---|
1313 | attribute.setValue(GSXML.DISPLAY_TEXT_NAME);
|
---|
1314 | displayItem.setAttributeNode(attribute);
|
---|
1315 |
|
---|
1316 | try{
|
---|
1317 | Text textNode = doc.createTextNode(
|
---|
1318 | this.getCollectionTitle(getCollectionPID(collectionName)));
|
---|
1319 | displayItem.appendChild(textNode);
|
---|
1320 | } catch(Exception e) {
|
---|
1321 | // can't find Greenstone collections in fedora repository or problem
|
---|
1322 | // getting their titles from their metadata datastream?
|
---|
1323 | ex = new FedoraGS3RunException(e);
|
---|
1324 | ex.setSpecifics("greenstone collections or their metadata"
|
---|
1325 | + "in the fedora repository");
|
---|
1326 | }
|
---|
1327 | // now append the displayItem element as child of the collection element
|
---|
1328 | collection.appendChild(displayItem);
|
---|
1329 | // get the <serviceList> and add it into the collection description.
|
---|
1330 | // Services for all collections in the FedoraGS3 repository are the
|
---|
1331 | // same, offering a ClassifierBrowse to browse titles by starting letter
|
---|
1332 | // and DocRetrieve services: Content, Metadata and Structure.
|
---|
1333 |
|
---|
1334 | Element serviceList = createServiceList(doc);
|
---|
1335 | collection.appendChild(serviceList);
|
---|
1336 |
|
---|
1337 | Element responseMsg = createResponseMessage(doc, collection, ex,
|
---|
1338 | GSXML.REQUEST_TYPE_DESCRIBE, collectionName);
|
---|
1339 | try{
|
---|
1340 | return FedoraCommons.elementToString(responseMsg);
|
---|
1341 | }catch(TransformerException e) {
|
---|
1342 | return FedoraGS3RunException.xmlToStringConversionFailureResponseMsg
|
---|
1343 | + " " + e;
|
---|
1344 | }
|
---|
1345 | }
|
---|
1346 |
|
---|
1347 | /** @return a GS3 describe response message for the services of a collection
|
---|
1348 | * in the Fedora-Greenstone repository. So far, these services are the same for
|
---|
1349 | * all fedora collections: they are the services given in member variable
|
---|
1350 | * serviceNames: DocumentContent/Metadata/StructureRetrieve, ClassifierBrowse,
|
---|
1351 | * ClassifierBrowseMetadataRetrieve.
|
---|
1352 | * All collections in this Digital Library (Fedora Repository) share the
|
---|
1353 | * same services, so this method returns the same services as getServiceList();
|
---|
1354 | * @param collectionName - the name of the collection whose services are to
|
---|
1355 | * be described. It will be converted to a fedora collection pid, which is of
|
---|
1356 | * the form "greenstone:<collectionName>-collection". */
|
---|
1357 | public String describeCollectionServices(String collectionName)
|
---|
1358 | {
|
---|
1359 | Document doc = builder.newDocument();
|
---|
1360 |
|
---|
1361 | Element collection = doc.createElement(GSXML.COLLECTION_ELEM);
|
---|
1362 | Attr attribute = doc.createAttribute(GSXML.NAME_ATT);
|
---|
1363 | attribute.setValue(collectionName);
|
---|
1364 | collection.setAttributeNode(attribute);
|
---|
1365 |
|
---|
1366 | Element serviceList = createServiceList(doc);
|
---|
1367 | collection.appendChild(serviceList);
|
---|
1368 |
|
---|
1369 | Element responseMsg = createResponseMessage(doc, collection, null,
|
---|
1370 | GSXML.REQUEST_TYPE_DESCRIBE, collectionName);
|
---|
1371 | try{
|
---|
1372 | return FedoraCommons.elementToString(responseMsg);
|
---|
1373 | }catch(TransformerException e) {
|
---|
1374 | return FedoraGS3RunException.xmlToStringConversionFailureResponseMsg
|
---|
1375 | + " " + e;
|
---|
1376 | }
|
---|
1377 | }
|
---|
1378 |
|
---|
1379 | /** All collections in this Digital Library (Fedora Repository) share
|
---|
1380 | * the same services, so this method returns the same as
|
---|
1381 | * describeCollectionService(collName, serviceName).
|
---|
1382 | * @return a GS3 describe response message for the requested service
|
---|
1383 | * of the given collection. DocumentContent/Metadata/StructureRetrieve
|
---|
1384 | * return nothing special except their names; browse (and any query)
|
---|
1385 | * return more complex XML responses.
|
---|
1386 | * @param serviceName - the name of the service in the collection which is to
|
---|
1387 | * be described.*/
|
---|
1388 | public String describeService(String serviceName)
|
---|
1389 | {
|
---|
1390 | // For all the *retrieve* services (incl ClassifierBrowseMetadataRetrieve)
|
---|
1391 | // we return:
|
---|
1392 | // <message><response from="<name>Retrieve" type="describe">
|
---|
1393 | // <service name="<name>Retrieve" type="retrieve" /></response></message>
|
---|
1394 | // But for browse (and any query) service, we return the data necessary
|
---|
1395 | // for displaying it
|
---|
1396 |
|
---|
1397 | Document doc = this.builder.newDocument();
|
---|
1398 | Element service = doc.createElement(GSXML.SERVICE_ELEM);
|
---|
1399 | Attr attribute = doc.createAttribute(GSXML.NAME_ATT);
|
---|
1400 | attribute.setValue(serviceName);
|
---|
1401 | service.setAttributeNode(attribute);
|
---|
1402 |
|
---|
1403 | attribute = doc.createAttribute(GSXML.TYPE_ATT);
|
---|
1404 |
|
---|
1405 | if(serviceName.toLowerCase().endsWith("retrieve")) {
|
---|
1406 | attribute.setValue(GSXML.SERVICE_TYPE_RETRIEVE);
|
---|
1407 | }
|
---|
1408 | else if(serviceName.toLowerCase().contains("browse")) {
|
---|
1409 | attribute.setValue(GSXML.SERVICE_TYPE_BROWSE);
|
---|
1410 |
|
---|
1411 | // we need name and description <displayItem> elements
|
---|
1412 | Element displayItem
|
---|
1413 | = createNameValuePairElement(doc,
|
---|
1414 | GSXML.DISPLAY_TEXT_ELEM, GSXML.DISPLAY_TEXT_NAME, "Browse");
|
---|
1415 | service.appendChild(displayItem);
|
---|
1416 |
|
---|
1417 | displayItem = createNameValuePairElement(doc,
|
---|
1418 | GSXML.DISPLAY_TEXT_ELEM, GSXML.DISPLAY_TEXT_DESCRIPTION,
|
---|
1419 | "Browse pre-defined classification hierarchies");
|
---|
1420 | service.appendChild(displayItem);
|
---|
1421 |
|
---|
1422 | // now need a classifierList
|
---|
1423 | Element classifierList = doc.createElement(
|
---|
1424 | GSXML.CLASSIFIER_ELEM+GSXML.LIST_MODIFIER);
|
---|
1425 |
|
---|
1426 | int classifierNum = 1;
|
---|
1427 | // append a <classifier content="some letter" name="CL#">
|
---|
1428 | // for each letter of the alphabet:
|
---|
1429 | Element classifier = createClassifierElement(doc, "TitleByLetter",
|
---|
1430 | classifierNum++, "titles by letter", "Browse titles by letter");
|
---|
1431 | // now add this <classifier> to the <classifierList>
|
---|
1432 | classifierList.appendChild(classifier);
|
---|
1433 |
|
---|
1434 | // ANY MORE CLASSIFIERS? ADD THEM HERE
|
---|
1435 |
|
---|
1436 | service.appendChild(classifierList);
|
---|
1437 | } // ELSE check for whether it is a query service
|
---|
1438 | else if(serviceName.toLowerCase().contains("query")) {
|
---|
1439 | attribute.setValue(GSXML.SERVICE_TYPE_QUERY);
|
---|
1440 | if(serviceName.equals("TextQuery")) {
|
---|
1441 | describeTextQueryService(service);
|
---|
1442 | } else if(serviceName.equals("FieldQuery")) {
|
---|
1443 | describeFieldQueryService(service);
|
---|
1444 | }
|
---|
1445 | }
|
---|
1446 |
|
---|
1447 | // don't forget to add the type attribute to the service!
|
---|
1448 | service.setAttributeNode(attribute);
|
---|
1449 |
|
---|
1450 | String from = serviceName;
|
---|
1451 |
|
---|
1452 | Element responseMsg = createResponseMessage(doc, service, null,
|
---|
1453 | GSXML.REQUEST_TYPE_DESCRIBE, from);
|
---|
1454 | try{
|
---|
1455 | return FedoraCommons.elementToString(responseMsg);
|
---|
1456 | }catch(TransformerException e) {
|
---|
1457 | return FedoraGS3RunException.xmlToStringConversionFailureResponseMsg
|
---|
1458 | + " " + e;
|
---|
1459 | }
|
---|
1460 | }
|
---|
1461 |
|
---|
1462 | /** Appends children to the parameter service Element that make the
|
---|
1463 | * final service Element into a describe response XML for FedoraGS3's
|
---|
1464 | * TextQuery service.
|
---|
1465 | * @param service is the service Element that is being filled out. */
|
---|
1466 | protected void describeTextQueryService(Element service) {
|
---|
1467 | Document doc = service.getOwnerDocument();
|
---|
1468 | // we need name, submit (button) and description <displayItem> elements
|
---|
1469 | Element displayItem = createNameValuePairElement(doc,
|
---|
1470 | GSXML.DISPLAY_TEXT_ELEM, GSXML.DISPLAY_TEXT_NAME,
|
---|
1471 | "Text Search");
|
---|
1472 | service.appendChild(displayItem);
|
---|
1473 |
|
---|
1474 | displayItem = createNameValuePairElement(doc,
|
---|
1475 | GSXML.DISPLAY_TEXT_ELEM, GSXML.DISPLAY_TEXT_SUBMIT, "Search");
|
---|
1476 | service.appendChild(displayItem);
|
---|
1477 |
|
---|
1478 | displayItem = createNameValuePairElement(doc,
|
---|
1479 | GSXML.DISPLAY_TEXT_ELEM, GSXML.DISPLAY_TEXT_DESCRIPTION,
|
---|
1480 | "Title and full-text search service");
|
---|
1481 | service.appendChild(displayItem);
|
---|
1482 |
|
---|
1483 | //create the <paramList>
|
---|
1484 | Element paramList = doc.createElement(
|
---|
1485 | GSXML.PARAM_ELEM+GSXML.LIST_MODIFIER);
|
---|
1486 |
|
---|
1487 | // we ignore granularity to search at: it will always be
|
---|
1488 | // document and section level
|
---|
1489 | // we ignore casefolding: always on (that is, case is irrelevant)
|
---|
1490 | // we ignore document display order: always ranked
|
---|
1491 |
|
---|
1492 | // Constructing the following:
|
---|
1493 | // <param default="100" name="maxDocs" type="integer">
|
---|
1494 | // <displayItem name="name">Maximum hits to return</displayItem>
|
---|
1495 | // </param>
|
---|
1496 | Element param = doc.createElement(GSXML.PARAM_ELEM);
|
---|
1497 |
|
---|
1498 | Attr attribute = doc.createAttribute(GSXML.NAME_ATT);
|
---|
1499 | attribute.setValue(MAXDOCS);
|
---|
1500 | param.setAttributeNode(attribute);
|
---|
1501 |
|
---|
1502 | attribute = doc.createAttribute(GSXML.DEFAULT_ATT);
|
---|
1503 | attribute.setValue("100");
|
---|
1504 | param.setAttributeNode(attribute);
|
---|
1505 |
|
---|
1506 | attribute = doc.createAttribute(GSXML.TYPE_ATT);
|
---|
1507 | attribute.setValue(GSXML.PARAM_TYPE_INTEGER);
|
---|
1508 | param.setAttributeNode(attribute);
|
---|
1509 |
|
---|
1510 | displayItem = createNameValuePairElement(doc,
|
---|
1511 | GSXML.DISPLAY_TEXT_ELEM, GSXML.DISPLAY_TEXT_NAME,
|
---|
1512 | "Maximum hits to return");
|
---|
1513 | param.appendChild(displayItem);
|
---|
1514 |
|
---|
1515 | paramList.appendChild(param);
|
---|
1516 |
|
---|
1517 | // Constructing the following:
|
---|
1518 | // <param name="query" type="string">
|
---|
1519 | // <displayItem name="name">Query string</displayItem>
|
---|
1520 | // </param>
|
---|
1521 | param = doc.createElement(GSXML.PARAM_ELEM);
|
---|
1522 |
|
---|
1523 | attribute = doc.createAttribute(GSXML.NAME_ATT);
|
---|
1524 | attribute.setValue(QUERY);
|
---|
1525 | param.setAttributeNode(attribute);
|
---|
1526 |
|
---|
1527 | attribute = doc.createAttribute(GSXML.TYPE_ATT);
|
---|
1528 | attribute.setValue(GSXML.PARAM_TYPE_STRING);
|
---|
1529 | param.setAttributeNode(attribute);
|
---|
1530 |
|
---|
1531 | displayItem = createNameValuePairElement(doc,
|
---|
1532 | GSXML.DISPLAY_TEXT_ELEM, GSXML.DISPLAY_TEXT_NAME,
|
---|
1533 | "Query string");
|
---|
1534 | param.appendChild(displayItem);
|
---|
1535 |
|
---|
1536 | paramList.appendChild(param);
|
---|
1537 |
|
---|
1538 | service.appendChild(paramList);
|
---|
1539 | }
|
---|
1540 |
|
---|
1541 | /** Appends children to the parameter service Element that make the
|
---|
1542 | * final service Element into a describe response XML for FedoraGS3's
|
---|
1543 | * FieldQuery service.
|
---|
1544 | * @param service is the service Element that is being filled out. */
|
---|
1545 | protected void describeFieldQueryService(Element service) {
|
---|
1546 | Document doc = service.getOwnerDocument();
|
---|
1547 | // we need name, submit (button) and description <displayItem> elements
|
---|
1548 | Element displayItem = createNameValuePairElement(doc,
|
---|
1549 | GSXML.DISPLAY_TEXT_ELEM, GSXML.DISPLAY_TEXT_NAME,
|
---|
1550 | "Form Search");
|
---|
1551 | service.appendChild(displayItem);
|
---|
1552 |
|
---|
1553 | displayItem = createNameValuePairElement(doc,
|
---|
1554 | GSXML.DISPLAY_TEXT_ELEM, GSXML.DISPLAY_TEXT_SUBMIT, "Search");
|
---|
1555 | service.appendChild(displayItem);
|
---|
1556 |
|
---|
1557 | displayItem = createNameValuePairElement(doc,
|
---|
1558 | GSXML.DISPLAY_TEXT_ELEM, GSXML.DISPLAY_TEXT_DESCRIPTION,
|
---|
1559 | "Simple fielded search");
|
---|
1560 | service.appendChild(displayItem);
|
---|
1561 |
|
---|
1562 | //create the <paramList>
|
---|
1563 | Element paramList = doc.createElement(
|
---|
1564 | GSXML.PARAM_ELEM+GSXML.LIST_MODIFIER);
|
---|
1565 |
|
---|
1566 | // we ignore granularity to search at: it will always be
|
---|
1567 | // document and section level
|
---|
1568 | // we ignore casefolding: always on (that is, case is irrelevant)
|
---|
1569 | // we ignore document display order: always ranked
|
---|
1570 |
|
---|
1571 | // Constructing the following:
|
---|
1572 | // <param default="100" name="maxDocs" type="integer">
|
---|
1573 | // <displayItem name="name">Maximum hits to return</displayItem>
|
---|
1574 | // </param>
|
---|
1575 | Element param = doc.createElement(GSXML.PARAM_ELEM);
|
---|
1576 |
|
---|
1577 | Attr attribute = doc.createAttribute(GSXML.NAME_ATT);
|
---|
1578 | attribute.setValue(MAXDOCS);
|
---|
1579 | param.setAttributeNode(attribute);
|
---|
1580 |
|
---|
1581 | attribute = doc.createAttribute(GSXML.DEFAULT_ATT);
|
---|
1582 | attribute.setValue("100");
|
---|
1583 | param.setAttributeNode(attribute);
|
---|
1584 |
|
---|
1585 | attribute = doc.createAttribute(GSXML.TYPE_ATT);
|
---|
1586 | attribute.setValue(GSXML.PARAM_TYPE_INTEGER);
|
---|
1587 | param.setAttributeNode(attribute);
|
---|
1588 |
|
---|
1589 | displayItem = createNameValuePairElement(doc,
|
---|
1590 | GSXML.DISPLAY_TEXT_ELEM, GSXML.DISPLAY_TEXT_NAME,
|
---|
1591 | "Maximum hits to return");
|
---|
1592 | param.appendChild(displayItem);
|
---|
1593 |
|
---|
1594 | paramList.appendChild(param);
|
---|
1595 |
|
---|
1596 | // Constructing the following:
|
---|
1597 | // <param name="simpleField" occurs="4" type="multi">
|
---|
1598 | // <displayItem name="name"></displayItem>
|
---|
1599 | //
|
---|
1600 | // <param name="query" type="string">
|
---|
1601 | // <displayItem name="name">Word or phrase </displayItem>
|
---|
1602 | // </param>
|
---|
1603 | //
|
---|
1604 | // <param default="allFields" name="fieldname" type="enum_single">
|
---|
1605 | // <displayItem name="name">in field</displayItem>
|
---|
1606 | //
|
---|
1607 | // <option name="docTitles">
|
---|
1608 | // <displayItem name="name">document titles</displayItem>
|
---|
1609 | // </option>
|
---|
1610 | // <option name="allTitles">
|
---|
1611 | // <displayItem name="name">document and section titles</displayItem>
|
---|
1612 | // </option>
|
---|
1613 | // <option name="fullText">
|
---|
1614 | // <displayItem name="name">full text</displayItem>
|
---|
1615 | // </option>
|
---|
1616 | // <option name="all">
|
---|
1617 | // <displayItem name="name">titles and full text</displayItem>
|
---|
1618 | // </option>
|
---|
1619 | // <option name="">
|
---|
1620 | // <displayItem name="name"></displayItem>
|
---|
1621 | // </option>
|
---|
1622 | // </param>
|
---|
1623 | // </param>
|
---|
1624 | Element rowOfParams = doc.createElement(GSXML.PARAM_ELEM);
|
---|
1625 | attribute = doc.createAttribute(GSXML.NAME_ATT);
|
---|
1626 | attribute.setValue(SIMPLEFIELD_ATT);
|
---|
1627 | rowOfParams.setAttributeNode(attribute);
|
---|
1628 |
|
---|
1629 | // we want the row of controls to occur multiple times
|
---|
1630 | attribute = doc.createAttribute(GSXML.TYPE_ATT);
|
---|
1631 | attribute.setValue(GSXML.PARAM_TYPE_MULTI);
|
---|
1632 | rowOfParams.setAttributeNode(attribute);
|
---|
1633 |
|
---|
1634 | attribute = doc.createAttribute(OCCURS_ATT);
|
---|
1635 | attribute.setValue("4"); // we want this row to occur 4 times
|
---|
1636 | rowOfParams.setAttributeNode(attribute);
|
---|
1637 |
|
---|
1638 | // <param name="query" type="string">
|
---|
1639 | // <displayItem name="name">Word or phrase </displayItem>
|
---|
1640 | // </param>
|
---|
1641 | param = doc.createElement(GSXML.PARAM_ELEM);
|
---|
1642 |
|
---|
1643 | attribute = doc.createAttribute(GSXML.NAME_ATT);
|
---|
1644 | attribute.setValue(QUERY);
|
---|
1645 | param.setAttributeNode(attribute);
|
---|
1646 |
|
---|
1647 | attribute = doc.createAttribute(GSXML.TYPE_ATT);
|
---|
1648 | attribute.setValue(GSXML.PARAM_TYPE_STRING);
|
---|
1649 | param.setAttributeNode(attribute);
|
---|
1650 |
|
---|
1651 | displayItem = createNameValuePairElement(doc,
|
---|
1652 | GSXML.DISPLAY_TEXT_ELEM, GSXML.DISPLAY_TEXT_NAME,
|
---|
1653 | "Word or phrase");
|
---|
1654 | param.appendChild(displayItem);
|
---|
1655 | rowOfParams.appendChild(param);
|
---|
1656 |
|
---|
1657 | // <param default="allFields" name="fieldName" type="enum_single">
|
---|
1658 | // <displayItem name="name">in field</displayItem>
|
---|
1659 | param = doc.createElement(GSXML.PARAM_ELEM);
|
---|
1660 | attribute = doc.createAttribute(GSXML.NAME_ATT);
|
---|
1661 | attribute.setValue(FIELDNAME_ATT);
|
---|
1662 | param.setAttributeNode(attribute);
|
---|
1663 |
|
---|
1664 | attribute = doc.createAttribute(GSXML.TYPE_ATT);
|
---|
1665 | attribute.setValue(GSXML.PARAM_TYPE_ENUM_SINGLE);
|
---|
1666 | param.setAttributeNode(attribute);
|
---|
1667 |
|
---|
1668 | attribute = doc.createAttribute(GSXML.DEFAULT_ATT);
|
---|
1669 | attribute.setValue(ALL_FIELDS);
|
---|
1670 | param.setAttributeNode(attribute);
|
---|
1671 |
|
---|
1672 | displayItem = createNameValuePairElement(doc,
|
---|
1673 | GSXML.DISPLAY_TEXT_ELEM, GSXML.DISPLAY_TEXT_NAME,
|
---|
1674 | "in field");
|
---|
1675 | param.appendChild(displayItem);
|
---|
1676 |
|
---|
1677 | String[] searchFieldNames
|
---|
1678 | = {ALL_FIELDS, DOC_TITLES, ALL_TITLES, FULLTEXT};
|
---|
1679 | String[] searchFieldDisplay = {"all titles and full-text",
|
---|
1680 | "document titles only", "document and section titles",
|
---|
1681 | "full-text only"};
|
---|
1682 |
|
---|
1683 | // for each fieldName create an option element and insert
|
---|
1684 | // the option into the enum_multi drop-down param:
|
---|
1685 | // <option name="fieldName">
|
---|
1686 | // <displayItem name="name">fieldName</displayItem>
|
---|
1687 | // </option>
|
---|
1688 | for(int i = 0; i < searchFieldNames.length; i++) {
|
---|
1689 | Element option = doc.createElement(GSXML.PARAM_OPTION_ELEM);
|
---|
1690 | attribute = doc.createAttribute(GSXML.NAME_ATT);
|
---|
1691 | attribute.setValue(searchFieldNames[i]);
|
---|
1692 | option.setAttributeNode(attribute);
|
---|
1693 |
|
---|
1694 | displayItem = createNameValuePairElement(doc,
|
---|
1695 | GSXML.DISPLAY_TEXT_ELEM, GSXML.DISPLAY_TEXT_NAME,
|
---|
1696 | searchFieldDisplay[i]);
|
---|
1697 | option.appendChild(displayItem);
|
---|
1698 | param.appendChild(option); // add option to the drop-down box
|
---|
1699 | }
|
---|
1700 |
|
---|
1701 | rowOfParams.appendChild(param);
|
---|
1702 | paramList.appendChild(rowOfParams);
|
---|
1703 | service.appendChild(paramList);
|
---|
1704 | }
|
---|
1705 |
|
---|
1706 | /**
|
---|
1707 | * @return a GS3 describe response message for the requested service
|
---|
1708 | * of the given collection. DocumentContent/Metadata/StructureRetrieve
|
---|
1709 | * return nothing special except their names; browse (and any query)
|
---|
1710 | * return more complex XML responses.
|
---|
1711 | * All collections in this Digital Library (Fedora Repository) share
|
---|
1712 | * the same services, so this method returns the same as
|
---|
1713 | * describeService(serviceName).
|
---|
1714 | * @param collectionName - the name of the collection whose service is to
|
---|
1715 | * be described. It will be converted to a fedora collection pid, which is of
|
---|
1716 | * the form "greenstone:<collectionName>-collection".
|
---|
1717 | * @param serviceName - the name of the service in the collection which is to
|
---|
1718 | * be described. */
|
---|
1719 | public String describeCollectionService(String collectionName,
|
---|
1720 | String serviceName) {
|
---|
1721 | // collectionName can be ignored, because all services are FedoraGS3
|
---|
1722 | // services and are not unique to any particular (greenstone) collection.
|
---|
1723 | return describeService(serviceName);
|
---|
1724 | }
|
---|
1725 |
|
---|
1726 | /** This method performs the implemented browse operation: allowing the
|
---|
1727 | * user to browse the titles of documents in the given collection by letter
|
---|
1728 | * and returning the results.
|
---|
1729 | * @param collectionName is the name of the collection whose documents
|
---|
1730 | * starting with the given letter will be returned.
|
---|
1731 | * @param classifierIDs are the ids of the classifiers on which to browse. In
|
---|
1732 | * this case, the classifier indicates whether we browse titles by letter, or
|
---|
1733 | * browse (documents) by collection; and it is of the form <CL(letter)>.
|
---|
1734 | * @param structures - the requested browse substructure. Can be any combination
|
---|
1735 | * of ancestors, parent, siblings, children, descendants.
|
---|
1736 | * @param infos - the requested structural info. Can be numSiblings,
|
---|
1737 | * siblingPosition, numChildren.
|
---|
1738 | * @return a GS3 ClassifierBrowse response message which lists all
|
---|
1739 | * the documents that start with the letter indicated by parameter classifier.
|
---|
1740 | */
|
---|
1741 | public String browse(String collectionName, String[] classifierIDs,
|
---|
1742 | String[] structures, String[] infos)
|
---|
1743 | {
|
---|
1744 | // Construct one string from the structures and structural info arrays
|
---|
1745 | String structure = "";
|
---|
1746 | String info = "";
|
---|
1747 | for(int i = 0; i < structures.length; i++) {
|
---|
1748 | structure = structure + structures[i] + "|";
|
---|
1749 | }
|
---|
1750 | for(int i = 0; i < infos.length; i++) {
|
---|
1751 | info = info + infos[i] + "|";
|
---|
1752 | }
|
---|
1753 |
|
---|
1754 | Document doc = builder.newDocument();
|
---|
1755 | FedoraGS3RunException ex = null; //any RemoteException or UnsupportedEncodingException
|
---|
1756 |
|
---|
1757 | // <classifierNodeList>
|
---|
1758 | Element classifierNodeList = doc.createElement(GSXML.CLASS_NODE_ELEM+GSXML.LIST_MODIFIER);
|
---|
1759 |
|
---|
1760 | for(int i = 0; i < classifierIDs.length; i++) {
|
---|
1761 | if(classifierIDs[i].startsWith("CL1")) { // browse by titles
|
---|
1762 | browseTitlesByLetterClassifier(doc, classifierNodeList,
|
---|
1763 | collectionName, classifierIDs[i],
|
---|
1764 | structure, info);
|
---|
1765 | }
|
---|
1766 | }
|
---|
1767 |
|
---|
1768 | Element responseMsg = createResponseMessage(doc, classifierNodeList, ex,
|
---|
1769 | GSXML.REQUEST_TYPE_DESCRIBE, /*collectionName+/ */"ClassifierBrowse");
|
---|
1770 | try {
|
---|
1771 | return FedoraCommons.elementToString(responseMsg);
|
---|
1772 | } catch(TransformerException e) {
|
---|
1773 | return FedoraGS3RunException.xmlToStringConversionFailureResponseMsg
|
---|
1774 | + " " + e;
|
---|
1775 | }
|
---|
1776 | }
|
---|
1777 |
|
---|
1778 | /** CL1 browsing classifier: browsing titles by starting letter.
|
---|
1779 | * The browsing structure is retrieved.
|
---|
1780 | * @param doc - the document object that will contain the CL1 browsing structure.
|
---|
1781 | * @param classifierNodeList - the classifiers will be added to this nodeList.
|
---|
1782 | * @param collectionName - name of the collection through which we are browsing CL1.
|
---|
1783 | * @param classifierID - the ID of the (sub)classifier. Can be CL1, CL1.x, where x is
|
---|
1784 | * a letter.
|
---|
1785 | * @param structure - the requested browse substructure. Can be any combination
|
---|
1786 | * of ancestors, parent, siblings, children, descendants. siblings not yet implemented.
|
---|
1787 | * @param info - the requested structural info. Can be numSiblings, siblingPosition,
|
---|
1788 | * numChildren.
|
---|
1789 | * @return the classifierNodeList with the CL1 classifier browse structure.
|
---|
1790 | */
|
---|
1791 | public Element browseTitlesByLetterClassifier(Document doc, Element classifierNodeList,
|
---|
1792 | String collectionName, String classifierID,
|
---|
1793 | String structure, String info)
|
---|
1794 | {
|
---|
1795 | FedoraGS3RunException ex = null; //any RemoteException or UnsupportedEncodingException
|
---|
1796 |
|
---|
1797 | // TODO
|
---|
1798 | if(structure.indexOf("siblings") != -1) {
|
---|
1799 | LOG.error("Structure: siblings. Not yet implemented\n");
|
---|
1800 | }
|
---|
1801 |
|
---|
1802 | if(structure.indexOf("entire") != -1) {
|
---|
1803 | structure = structure + "ancestors|descendants";
|
---|
1804 | }
|
---|
1805 |
|
---|
1806 | // Structure of ancestors and children only at this stage
|
---|
1807 | int firstLevel = classifierID.indexOf('.');
|
---|
1808 | int secondLevel = classifierID.lastIndexOf('.');
|
---|
1809 |
|
---|
1810 | // <nodeStructure>
|
---|
1811 | Element nodeStructure = doc.createElement(GSXML.NODE_STRUCTURE_ELEM);
|
---|
1812 |
|
---|
1813 | // requested classifier node
|
---|
1814 | Element classNode = doc.createElement(GSXML.CLASS_NODE_ELEM);
|
---|
1815 | Attr attribute = doc.createAttribute(GSXML.NODE_ID_ATT);
|
---|
1816 | attribute.setValue(classifierID);
|
---|
1817 | classNode.setAttributeNode(attribute);
|
---|
1818 |
|
---|
1819 | if(firstLevel == -1) { // CL1 - toplevel node
|
---|
1820 | Element root = (Element)classNode.cloneNode(true); // clone the node before appending children
|
---|
1821 |
|
---|
1822 | classifierNodeList.appendChild(classNode);
|
---|
1823 | classNode.appendChild(nodeStructure);
|
---|
1824 |
|
---|
1825 | nodeStructure.appendChild(root);
|
---|
1826 | if(structure.indexOf("descendants") != -1) {
|
---|
1827 | getTitlesByLetterStructure(collectionName, root, classifierID, true);
|
---|
1828 | } else if(structure.indexOf("children") != -1) {
|
---|
1829 | getTitlesByLetterStructure(collectionName, root, classifierID, false);
|
---|
1830 | }
|
---|
1831 | }
|
---|
1832 | else if(firstLevel == secondLevel) { // CL1.x, where x is a number
|
---|
1833 |
|
---|
1834 | if(structure.indexOf("parent") != -1 || structure.indexOf("ancestors") != -1) {
|
---|
1835 | String toplevelID = classifierID.substring(0, firstLevel);
|
---|
1836 | Element toplevelNode = doc.createElement(GSXML.CLASS_NODE_ELEM);
|
---|
1837 | attribute = doc.createAttribute(GSXML.NODE_ID_ATT);
|
---|
1838 | attribute.setValue(toplevelID);
|
---|
1839 | toplevelNode.setAttributeNode(attribute);
|
---|
1840 | Element node = (Element)toplevelNode.cloneNode(true); // clone nodes before appending children
|
---|
1841 |
|
---|
1842 | classifierNodeList.appendChild(toplevelNode);
|
---|
1843 | toplevelNode.appendChild(nodeStructure);
|
---|
1844 |
|
---|
1845 | nodeStructure.appendChild(node);
|
---|
1846 | node.appendChild(classNode);
|
---|
1847 | } else {
|
---|
1848 | Element node = (Element)classNode.cloneNode(true);
|
---|
1849 | classifierNodeList.appendChild(node);
|
---|
1850 | node.appendChild(nodeStructure);
|
---|
1851 | nodeStructure.appendChild(classNode);
|
---|
1852 | }
|
---|
1853 |
|
---|
1854 | int num = Integer.parseInt(classifierID.substring(firstLevel+1)); // get x from CL1.x
|
---|
1855 | char ch = (char)(num - 1 + 'A');
|
---|
1856 | if(structure.indexOf("descendants") != -1) {
|
---|
1857 | getTitlesForLetter(ch, collectionName, classNode, "descendants");
|
---|
1858 | } else if(structure.indexOf("children") != -1) {
|
---|
1859 | getTitlesForLetter(ch, collectionName, classNode, "children");
|
---|
1860 | }
|
---|
1861 | }
|
---|
1862 | else { // ought to be a doc structure retrieve request, not classifierbrowse structure retrieve
|
---|
1863 | LOG.error("ClassifierID: " + classifierID + ". Shouldn't be in browse method");
|
---|
1864 | }
|
---|
1865 |
|
---|
1866 | return classifierNodeList;
|
---|
1867 | }
|
---|
1868 |
|
---|
1869 | /** Creates a (CL1) subclassifier element for the docs whose titles start with
|
---|
1870 | * the given letter.
|
---|
1871 | * @param ch - the starting letter of the document titles to retrieve.
|
---|
1872 | * @param collectionName - name of the collection through which we are browsing CL1.
|
---|
1873 | * @param classifierNode - the docNodes found will be appended to this node.
|
---|
1874 | * @param depthStructure - can be descendants or children. Specifies what to retrieve:
|
---|
1875 | * gets descendants of any documents found, otherwise gets just the children.
|
---|
1876 | * @return the given classifierNode which will have the child (or descendant) documents
|
---|
1877 | * appended to it.
|
---|
1878 | */
|
---|
1879 | public Element getTitlesForLetter(char ch, String collectionName,
|
---|
1880 | Element classifierNode, String depthStructure)
|
---|
1881 | {
|
---|
1882 | Document doc = classifierNode.getOwnerDocument();
|
---|
1883 | FedoraGS3RunException ex = null; //any RemoteException or UnsupportedEncodingException
|
---|
1884 |
|
---|
1885 |
|
---|
1886 | // Retrieve the document structure for each subClassifierID:
|
---|
1887 | // all the documents that begin with its letter.
|
---|
1888 | String letter = String.valueOf(ch);
|
---|
1889 | try {
|
---|
1890 | String[] docPIDs = this.browseTitlesByLetter(collectionName, letter);
|
---|
1891 | if(docPIDs.length == 0) {
|
---|
1892 | return classifierNode; // skip letters that don't have any kids
|
---|
1893 | }
|
---|
1894 |
|
---|
1895 | for(int i = 0; i < docPIDs.length; i++) {
|
---|
1896 | // work out the document's fedora PID and section ID
|
---|
1897 | String sectionID = getSectionIDFromDocID(docPIDs[i]);
|
---|
1898 | String docPID = getDocPIDFromDocID(docPIDs[i]);
|
---|
1899 |
|
---|
1900 | // get the required section, along with children or descendants
|
---|
1901 | Element section = getSectionStructureXML(docPID, sectionID, depthStructure, "");
|
---|
1902 |
|
---|
1903 | // <documentNode nodeID="docID" docType="hierarchy" nodeType="root">
|
---|
1904 | Element docRootNode = createDocNodeFromSubsection(doc, section, docPID);
|
---|
1905 |
|
---|
1906 | // fills in the subtree of the rootNode in our nodeStructure element
|
---|
1907 | createDocStructure(doc, section, docRootNode, docPID); //where section represents the root section
|
---|
1908 | classifierNode.appendChild(docRootNode);
|
---|
1909 | }
|
---|
1910 | } catch(Exception e) {
|
---|
1911 | ex = new FedoraGS3RunException(e);
|
---|
1912 | ex.setSpecifics("requested portion of TOC file or trouble with fielded search ");
|
---|
1913 | }
|
---|
1914 |
|
---|
1915 | return classifierNode;
|
---|
1916 | }
|
---|
1917 |
|
---|
1918 |
|
---|
1919 | /** Creates all the subclassifiers (CL1.x) for CL1, the classifier to browse by the
|
---|
1920 | * starting letter of the alphabet. X is each letter of the alphabet for which there
|
---|
1921 | * are matching document titles.
|
---|
1922 | * @param collectionName - name of the collection through which we are browsing CL1.
|
---|
1923 | * @param classifierNode - the docNodes found will be appended to this node.
|
---|
1924 | * @param classifierID - the ID of parent classifier, i.e. CL1, which is used to create
|
---|
1925 | * the IDs for the subclassifiers (CL.x).
|
---|
1926 | * @param getDescendants - if true, get descendants of any documents found, otherwise
|
---|
1927 | * get just the children.
|
---|
1928 | * @return the given classifierNode, with the CL.x subclassifiers for the letters of
|
---|
1929 | * the alphabet that are represented in the document titles.
|
---|
1930 | */
|
---|
1931 | public Element getTitlesByLetterStructure(String collectionName, Element classifierNode,
|
---|
1932 | String classifierID, boolean getDescendants)
|
---|
1933 | {
|
---|
1934 | Document doc = classifierNode.getOwnerDocument();
|
---|
1935 | FedoraGS3RunException ex = null; // any RemoteException or UnsupportedEncodingException
|
---|
1936 |
|
---|
1937 | // We're going to loop to the end of the alphabet
|
---|
1938 | int count = 1;
|
---|
1939 | for(char ch = 'A'; ch <= 'Z'; ch++, count++) {
|
---|
1940 | // Retrieve the document structure for each subClassifierID:
|
---|
1941 | // all the documents that begin with its letter.
|
---|
1942 | String letter = String.valueOf(ch);
|
---|
1943 | try {
|
---|
1944 | String[] docPIDs = this.browseTitlesByLetter(collectionName, letter);
|
---|
1945 | if(docPIDs.length == 0) {
|
---|
1946 | continue; // skip letters that don't have any kids
|
---|
1947 | }
|
---|
1948 |
|
---|
1949 | // <classifierNode nodeID="CL1.x">
|
---|
1950 | Element subClassifier = doc.createElement(GSXML.CLASS_NODE_ELEM);
|
---|
1951 | Attr attribute = doc.createAttribute(GSXML.NODE_ID_ATT);
|
---|
1952 | attribute.setValue(classifierID+"."+count);
|
---|
1953 | subClassifier.setAttributeNode(attribute);
|
---|
1954 | classifierNode.appendChild(subClassifier);
|
---|
1955 |
|
---|
1956 | if(getDescendants) { // get the documents
|
---|
1957 |
|
---|
1958 | // append the <docNodes> for the docPIDs found as children
|
---|
1959 | // of subclassifier
|
---|
1960 |
|
---|
1961 | for(int i = 0; i < docPIDs.length; i++) {
|
---|
1962 | // work out the document's fedora PID and section ID
|
---|
1963 | String sectionID = getSectionIDFromDocID(docPIDs[i]);
|
---|
1964 | String docPID = getDocPIDFromDocID(docPIDs[i]);
|
---|
1965 |
|
---|
1966 | // get the required section, along with children or descendants
|
---|
1967 | Element section = getSectionStructureXML(docPID, sectionID, "descendants", "");
|
---|
1968 |
|
---|
1969 | // <documentNode nodeID="docID" docType="hierarchy" nodeType="root">
|
---|
1970 | Element rootNode = createDocNodeFromSubsection(doc, section, docPID);
|
---|
1971 |
|
---|
1972 | // fills in the subtree of the rootNode in our nodeStructure element
|
---|
1973 | createDocStructure(doc, section, rootNode, docPID); //where section represents the root section
|
---|
1974 | subClassifier.appendChild(rootNode);
|
---|
1975 | }
|
---|
1976 | }
|
---|
1977 | } catch(Exception e) {
|
---|
1978 | ex = new FedoraGS3RunException(e);
|
---|
1979 | ex.setSpecifics("requested portion of TOC file or "
|
---|
1980 | + "trouble with fielded search ");
|
---|
1981 | }
|
---|
1982 | }
|
---|
1983 | return classifierNode;
|
---|
1984 | }
|
---|
1985 |
|
---|
1986 |
|
---|
1987 | /** This method performs something equivalent to a greenstone3
|
---|
1988 | * ClassifierBrowseMetadataRetrieve on the classifierNodeIDs
|
---|
1989 | * @param classNodeIDs array of classifierNode IDs for which the metadata
|
---|
1990 | * needs to be returned.
|
---|
1991 | * @param metafields are the classifier metadata fields that are to be returned.
|
---|
1992 | * At present this method ignores them/pretends the requested metafields are
|
---|
1993 | * "all" and always returns the Title meta for the requested classifier nodes
|
---|
1994 | * (because that is all the metadata this Fedora classifier has at present).
|
---|
1995 | * @return a GS3 ClassifierBrowseMetadataRetrieve response message which
|
---|
1996 | * lists the metadata for all the classifierNodes passed as parameter.*/
|
---|
1997 | public String browseMetadataRetrieve(String[] classNodeIDs, String[] metafields)
|
---|
1998 | {
|
---|
1999 | Document doc = this.builder.newDocument();
|
---|
2000 | // <classifierNodeList>
|
---|
2001 | Element classifierNodeList = doc.createElement(
|
---|
2002 | GSXML.CLASS_NODE_ELEM+GSXML.LIST_MODIFIER);
|
---|
2003 |
|
---|
2004 | // create <classifierNode><metadataList><metadata>s
|
---|
2005 | // </metadataList></classifierNode> for all letters of the alphabet
|
---|
2006 | for(int i = 0; i < classNodeIDs.length; i++) {
|
---|
2007 | // strip ID of everything before the first '.' (i.e. remove "CL#.")
|
---|
2008 | int index = classNodeIDs[i].indexOf('.');
|
---|
2009 | String subClassifierNumber = classNodeIDs[i].substring(index+1);
|
---|
2010 | index = subClassifierNumber.indexOf('.'); // find next decimal point, if any
|
---|
2011 | if(index != -1) {
|
---|
2012 | subClassifierNumber = subClassifierNumber.substring(0, index);
|
---|
2013 | }
|
---|
2014 | int subClassifierNum = Integer.parseInt(subClassifierNumber);
|
---|
2015 | String classifierName = "";
|
---|
2016 | if(subClassifierNum == 0) { // no document titles started with a letter
|
---|
2017 | classifierName = "A-Z";
|
---|
2018 | } else {
|
---|
2019 | char letter = (char)('A' + subClassifierNum - 1); // A = 1
|
---|
2020 | classifierName = String.valueOf(letter);
|
---|
2021 | }
|
---|
2022 |
|
---|
2023 | // <classifierNode nodeID="CL#.subNum">
|
---|
2024 | Element classifierNode = doc.createElement(GSXML.CLASS_NODE_ELEM);
|
---|
2025 | Attr attribute = doc.createAttribute(GSXML.NODE_ID_ATT);
|
---|
2026 | attribute.setValue(classNodeIDs[i]);
|
---|
2027 | classifierNode.setAttributeNode(attribute);
|
---|
2028 |
|
---|
2029 | // <metadataList>
|
---|
2030 | Element metadataList = doc.createElement(
|
---|
2031 | GSXML.METADATA_ELEM+GSXML.LIST_MODIFIER);
|
---|
2032 |
|
---|
2033 | // at least one metadata element: that of the title of this
|
---|
2034 | // classifierNode:
|
---|
2035 | // <metadata name="Title">letter</metadata>
|
---|
2036 | Element metadata = this.createNameValuePairElement(doc,
|
---|
2037 | GSXML.METADATA_ELEM, "Title", classifierName);
|
---|
2038 |
|
---|
2039 | // now connect up everything
|
---|
2040 | metadataList.appendChild(metadata);
|
---|
2041 | classifierNode.appendChild(metadataList);
|
---|
2042 | classifierNodeList.appendChild(classifierNode);
|
---|
2043 | }
|
---|
2044 |
|
---|
2045 | Element responseMsg = createResponseMessage(doc, classifierNodeList, null,
|
---|
2046 | GSXML.REQUEST_TYPE_PROCESS, //collName +
|
---|
2047 | "ClassifierBrowseMetadataRetrieve");
|
---|
2048 | try{
|
---|
2049 | return FedoraCommons.elementToString(responseMsg);
|
---|
2050 | }catch(TransformerException e) {
|
---|
2051 | return FedoraGS3RunException.xmlToStringConversionFailureResponseMsg
|
---|
2052 | + " " + e;
|
---|
2053 | }
|
---|
2054 | }
|
---|
2055 |
|
---|
2056 | /** @return a newly created element of the following format:
|
---|
2057 | * <classifier content="somecontent" name="CL+num">
|
---|
2058 | * <displayItem name="name">someClassifierName</displayItem>
|
---|
2059 | * <displayItem name="description">Browse by classifier name</displayItem>
|
---|
2060 | * </classifier>
|
---|
2061 | * @param doc - the document used to create the element
|
---|
2062 | * @param content - value of the content attribute
|
---|
2063 | * @param classifierNum - the number suffixed to the CL, together forming
|
---|
2064 | * the classifier Node's ID
|
---|
2065 | * @param displayNameVal is the bodytext of a named displayItem element
|
---|
2066 | * @param displayDescrVal is the bodytext of a displayItem element with
|
---|
2067 | * description */
|
---|
2068 | protected Element createClassifierElement(Document doc, String content,
|
---|
2069 | int classifierNum, String displayNameVal, String displayDescrVal)
|
---|
2070 | {
|
---|
2071 | final String CL = "CL";
|
---|
2072 | Element classifier = doc.createElement(GSXML.CLASSIFIER_ELEM);
|
---|
2073 | // content attribute
|
---|
2074 | Attr att = doc.createAttribute(GSXML.CLASSIFIER_CONTENT_ATT);
|
---|
2075 | att.setValue(content);
|
---|
2076 | classifier.setAttributeNode(att);
|
---|
2077 | // name attribute
|
---|
2078 | att = doc.createAttribute(GSXML.NAME_ATT);
|
---|
2079 | att.setValue(CL + classifierNum);
|
---|
2080 | classifier.setAttributeNode(att);
|
---|
2081 |
|
---|
2082 | // now create the displayItem children for classifier:
|
---|
2083 | // <displayItem name="name">#letter</displayItem>
|
---|
2084 | // <displayItem name="description">Browse titles starting with #letter</displayItem>
|
---|
2085 | Element displayItem = createNameValuePairElement(doc,
|
---|
2086 | GSXML.DISPLAY_TEXT_ELEM, GSXML.DISPLAY_TEXT_NAME, displayNameVal);
|
---|
2087 | classifier.appendChild(displayItem);
|
---|
2088 | displayItem = createNameValuePairElement(doc, GSXML.DISPLAY_TEXT_ELEM,
|
---|
2089 | GSXML.DISPLAY_TEXT_DESCRIPTION, displayDescrVal);
|
---|
2090 | classifier.appendChild(displayItem);
|
---|
2091 |
|
---|
2092 | return classifier;
|
---|
2093 | }
|
---|
2094 |
|
---|
2095 |
|
---|
2096 | /** @return a newly created element of the following format:
|
---|
2097 | * <elementName name="somename">"some display value"</elementName>
|
---|
2098 | * @param doc - the document used to create the element
|
---|
2099 | * @param elementName - the tag name
|
---|
2100 | * @param name - value of attribute name
|
---|
2101 | * @param value - the body text of the element */
|
---|
2102 | protected Element createNameValuePairElement(Document doc, String elementName,
|
---|
2103 | String name, String value) {
|
---|
2104 | // <elementName name="somename">"some display value"</elementName>
|
---|
2105 | Element element = doc.createElement(elementName);
|
---|
2106 | Attr attribute = doc.createAttribute(GSXML.NAME_ATT);
|
---|
2107 | attribute.setValue(name);
|
---|
2108 | element.setAttributeNode(attribute);
|
---|
2109 |
|
---|
2110 | element.appendChild(doc.createTextNode(value));
|
---|
2111 | return element;
|
---|
2112 | }
|
---|
2113 |
|
---|
2114 | /**
|
---|
2115 | * @param collection is the collection to search in
|
---|
2116 | * @param query is the query term to search for. It won't specify the
|
---|
2117 | * indexed field to search in, which will mean that GSearch will
|
---|
2118 | * search all default indexed fields.
|
---|
2119 | * @param maxDocs is the maximum number of results to return (which
|
---|
2120 | * at present we consider equivalent to FedoraGSearch's hitpageSize).
|
---|
2121 | */
|
---|
2122 | public String[] textQuery(String collection, String query,
|
---|
2123 | int maxDocs)
|
---|
2124 | throws Exception
|
---|
2125 | {
|
---|
2126 | // no need to search there is no query or query is empty spaces
|
---|
2127 | if(query.trim().equals(""))
|
---|
2128 | return new String[]{};
|
---|
2129 |
|
---|
2130 | // QUERY value won't specify indexed field to search, Fedora
|
---|
2131 | // Gsearch will take that as meaning all default indexed fields.
|
---|
2132 | // Params to search() method below: string of fielded query terms;
|
---|
2133 | // hitpageStart, hitpageEnd, snippetsMax (leave that 0)
|
---|
2134 | query = query + " " + "PID" + COLON + GREENSTONE;
|
---|
2135 |
|
---|
2136 | String searchResult = this.fedoraGSearch.search(query, 1, maxDocs, 0);
|
---|
2137 | // now we have the XML returned by FedoraGSearch, get the pids
|
---|
2138 | // of the documents returned (if any)
|
---|
2139 | String[] pids = this.fedoraGSearch.getPIDsFromSearchResult(
|
---|
2140 | collection, searchResult);
|
---|
2141 | return pids;
|
---|
2142 | }
|
---|
2143 |
|
---|
2144 | /**
|
---|
2145 | * This method performs a fieldquery, searching for x number of phrases
|
---|
2146 | * in each of the 4 indexed fields.
|
---|
2147 | * @param collection is the collection to search in
|
---|
2148 | * @param nameValParamsMap is a Map of several(key, value) entries,
|
---|
2149 | * 4 of which we're concerned with here:
|
---|
2150 | * - the keys are ALL_FIELDS, DOC_TITLES, ALL_TITLES, FULLTEXT
|
---|
2151 | * - the values are a comma separated list of terms (phrases or single
|
---|
2152 | * words) to search that field in. There may be more than 1 or
|
---|
2153 | * there may be none (in which case there may be N empty values or
|
---|
2154 | * spaces separated by commas).
|
---|
2155 | * @param maxDocs is the maximum number of results to return (which
|
---|
2156 | * at present we consider equivalent to FedoraGSearch's hitpageSize).
|
---|
2157 | * */
|
---|
2158 | public String[] fieldQuery(String collection, Map nameValParamsMap,
|
---|
2159 | int maxDocs)
|
---|
2160 | throws Exception
|
---|
2161 | {
|
---|
2162 | // we're going to maintain a list of UNIQUE pids that were returned
|
---|
2163 | // in search results. Hence we use Set:
|
---|
2164 | java.util.Set set = new java.util.HashSet();
|
---|
2165 |
|
---|
2166 | // (1) Use Fedora's search to search document titles, if they were
|
---|
2167 | // specified:
|
---|
2168 | String[] docTitlepids = {};
|
---|
2169 |
|
---|
2170 | String docTitleTerms = (String)nameValParamsMap.get(DOC_TITLES);
|
---|
2171 | if(docTitleTerms != null) { // no doc titles may have been specified
|
---|
2172 | String[] phrases = docTitleTerms.split(COMMA);
|
---|
2173 |
|
---|
2174 | // search the individual phrases first:
|
---|
2175 | for(int i = 0; i < phrases.length; i++) {
|
---|
2176 | if(phrases.equals("") || phrases.equals(" "))
|
---|
2177 | continue; //skip when there are no terms
|
---|
2178 | docTitlepids = this.searchDocumentTitles(
|
---|
2179 | collection, phrases[i], false);
|
---|
2180 | for(int j = 0; j < docTitlepids.length; j++)
|
---|
2181 | set.add(docTitlepids[j]);
|
---|
2182 | }
|
---|
2183 | }
|
---|
2184 | // (2) use FedoraGSearch to search doc AND section titles, and
|
---|
2185 | // fulltext (in case these were specified in nameValParamsMap):
|
---|
2186 | String searchResult = this.fedoraGSearch.search(
|
---|
2187 | nameValParamsMap, 1, maxDocs);
|
---|
2188 |
|
---|
2189 | String[] pids = this.fedoraGSearch.getPIDsFromSearchResult(
|
---|
2190 | collection, searchResult);
|
---|
2191 |
|
---|
2192 | for(int i = 0; i < pids.length; i++)
|
---|
2193 | set.add(pids[i]);
|
---|
2194 |
|
---|
2195 | pids = null;
|
---|
2196 | pids = new String[set.size()];
|
---|
2197 | set.toArray(pids); // unique pids
|
---|
2198 | return pids;
|
---|
2199 | }
|
---|
2200 |
|
---|
2201 | /** @return a String representing Greenstone3 XML for a query process
|
---|
2202 | * response returning the results for the query denoted by parameter
|
---|
2203 | * nameValParamsMap.
|
---|
2204 | * @param nameValParamsMap is a Hashmap of name and value pairs for all the
|
---|
2205 | * query field data values. The names match the field names that
|
---|
2206 | * describeCollectionService() would have returned for the query service.
|
---|
2207 | * @param collection is the name of the collection
|
---|
2208 | * @param service is the name of the query service
|
---|
2209 | * This method is only ever called when any of the services in the digital
|
---|
2210 | * library described themselves as type=query. Therefore any digital
|
---|
2211 | * libraries that have no query services, can just return emtpy message
|
---|
2212 | * strings (or even "") since this method will never be called on them
|
---|
2213 | * anyway. */
|
---|
2214 | public String query(String collection, String service,
|
---|
2215 | Map nameValParamsMap)
|
---|
2216 | {
|
---|
2217 | FedoraGS3RunException ex = null;
|
---|
2218 | // (1) obtain the requested number of maximum result documents
|
---|
2219 | int maxDocs = 100;
|
---|
2220 | try{
|
---|
2221 | maxDocs = Integer.parseInt((String)nameValParamsMap.get(MAXDOCS));
|
---|
2222 | } catch(NumberFormatException e) {
|
---|
2223 | maxDocs = 100;
|
---|
2224 | }
|
---|
2225 |
|
---|
2226 | String pids[] = {};
|
---|
2227 | // (2) for Textquery, we simply search ALL_FIELDS using FedoraGSearch
|
---|
2228 | if(service.endsWith("TextQuery")) {
|
---|
2229 | try {
|
---|
2230 | // get the Query field:
|
---|
2231 | String query = (String)nameValParamsMap.get(QUERY);
|
---|
2232 | pids = textQuery(collection, query, maxDocs);
|
---|
2233 | }
|
---|
2234 | catch(Exception e) {
|
---|
2235 | LOG.error("Error in TextQuery processing: " + e);
|
---|
2236 | ex = new FedoraGS3RunException(
|
---|
2237 | "When trying to use FedoraGenericSearch for a TextQuery", e);
|
---|
2238 |
|
---|
2239 | }
|
---|
2240 | } else { // (3) FieldQuery
|
---|
2241 | // first get the comma-separated lists
|
---|
2242 | String listOfFieldNames = (String)nameValParamsMap.get(FIELDNAME_ATT);
|
---|
2243 | String listOfSearchTerms = (String)nameValParamsMap.get(QUERY);
|
---|
2244 | // both are comma separated lists, so split both on 'comma'
|
---|
2245 | String[] fieldNames = listOfFieldNames.split(COMMA);
|
---|
2246 | String[] searchTerms = listOfSearchTerms.split(COMMA);
|
---|
2247 |
|
---|
2248 | // In the fieldNames and searchTerms lists of nameValParamsMap,
|
---|
2249 | // each searchTerm element was matched with its correspondingly
|
---|
2250 | // indexed fieldName.
|
---|
2251 | // A new map is going to reorganise this, by putting all terms
|
---|
2252 | // for a particular fieldName together in a comma separated list
|
---|
2253 | // and associating that with the fieldName. I.e. (key, value) ->
|
---|
2254 | // (fieldName, comma-separated list of all terms in that field)
|
---|
2255 | Map map = new HashMap();
|
---|
2256 | for(int i = 0; i < searchTerms.length; i++) {
|
---|
2257 | // there may be fewer searchTerms than fieldNames (since some
|
---|
2258 | // fieldNames may have been left empty), so loop on searchTerms
|
---|
2259 | if(map.containsKey(fieldNames[i])) { // fieldName is already
|
---|
2260 | // in the list, so append comma with new value
|
---|
2261 | String termsList = (String)map.get(fieldNames[i]);
|
---|
2262 | termsList = termsList + COMMA + searchTerms[i];
|
---|
2263 | map.put(fieldNames[i], termsList);
|
---|
2264 | } else { // this is the first time this fieldName occurred
|
---|
2265 | // just put the fieldName with searchTerm as-is
|
---|
2266 | map.put(fieldNames[i], searchTerms[i]);
|
---|
2267 | }
|
---|
2268 | }
|
---|
2269 |
|
---|
2270 | try {
|
---|
2271 | // For fieldquery, we search on all the fieldNames specified
|
---|
2272 | // - if DOC_TITLES is specified then we use Fedora's search
|
---|
2273 | // - for all other fieldNames specified, we use FedoraGSearch
|
---|
2274 | pids = fieldQuery(collection, map, maxDocs);
|
---|
2275 | }
|
---|
2276 | catch(Exception e) {
|
---|
2277 | LOG.error("Error in FieldQuery processing: " + e);
|
---|
2278 | ex = new FedoraGS3RunException(
|
---|
2279 | "When trying to use FedoraGenericSearch for a FieldQuery", e);
|
---|
2280 | }
|
---|
2281 | }
|
---|
2282 |
|
---|
2283 | // Build Greenstone XML Query response message for from
|
---|
2284 | // the pids (which should be document identifiers)
|
---|
2285 | Document doc = builder.newDocument();
|
---|
2286 | // <metadataList><metadata name="numDocsMatched" value="n" />
|
---|
2287 | // </metadataList>
|
---|
2288 | Element metadataList = doc.createElement(
|
---|
2289 | GSXML.METADATA_ELEM+GSXML.LIST_MODIFIER);
|
---|
2290 | Element metadata = doc.createElement(GSXML.METADATA_ELEM);
|
---|
2291 |
|
---|
2292 | Attr attribute = doc.createAttribute(GSXML.NAME_ATT);
|
---|
2293 | attribute.setValue(NUM_DOCS_MATCHED);
|
---|
2294 | metadata.setAttributeNode(attribute);
|
---|
2295 |
|
---|
2296 | attribute = doc.createAttribute(GSXML.VALUE_ATT);
|
---|
2297 | attribute.setValue(Integer.toString(pids.length));
|
---|
2298 | metadata.setAttributeNode(attribute);
|
---|
2299 |
|
---|
2300 | metadataList.appendChild(metadata);
|
---|
2301 |
|
---|
2302 | // <documentNodeList>
|
---|
2303 | // <documentNode nodeID="HASHac0a04dd14571c60d7fbfd.4.2"
|
---|
2304 | // docType='hierarchy' nodeType="leaf" />
|
---|
2305 | // ...
|
---|
2306 | // ...
|
---|
2307 | // </documentNodeList>
|
---|
2308 | Element docNodeList = doc.createElement(
|
---|
2309 | GSXML.DOC_NODE_ELEM+GSXML.LIST_MODIFIER);
|
---|
2310 | // for each
|
---|
2311 | for(int i = 0; i < pids.length; i++) {
|
---|
2312 | Element docNode = doc.createElement(GSXML.DOC_NODE_ELEM);
|
---|
2313 | attribute = doc.createAttribute(GSXML.NODE_ID_ATT);
|
---|
2314 | attribute.setValue(pids[i]);
|
---|
2315 | docNode.setAttributeNode(attribute);
|
---|
2316 |
|
---|
2317 | attribute = doc.createAttribute(GSXML.DOC_TYPE_ATT);
|
---|
2318 | attribute.setValue("hierarchy");
|
---|
2319 | docNode.setAttributeNode(attribute);
|
---|
2320 |
|
---|
2321 | attribute = doc.createAttribute(GSXML.NODE_TYPE_ATT);
|
---|
2322 | attribute.setValue("root");
|
---|
2323 | docNode.setAttributeNode(attribute);
|
---|
2324 | docNodeList.appendChild(docNode);
|
---|
2325 | }
|
---|
2326 |
|
---|
2327 | Element responseMsg = createResponseMessage(doc, docNodeList, ex,
|
---|
2328 | GSXML.REQUEST_TYPE_PROCESS, service);
|
---|
2329 | try{
|
---|
2330 | return FedoraCommons.elementToString(responseMsg);
|
---|
2331 | }catch(TransformerException e) {
|
---|
2332 | return FedoraGS3RunException.xmlToStringConversionFailureResponseMsg
|
---|
2333 | + " " + e;
|
---|
2334 | }
|
---|
2335 | }
|
---|
2336 |
|
---|
2337 | public static void main(String args[]) {
|
---|
2338 | try{
|
---|
2339 | // testing default constructor
|
---|
2340 | //FedoraGS3Connection con = new FedoraGS3Connection();
|
---|
2341 |
|
---|
2342 | // testing constructor that takes properties file to show initial
|
---|
2343 | // fedora server values
|
---|
2344 | java.io.File propertyFilename
|
---|
2345 | = new java.io.File("fedoraGS3.properties");
|
---|
2346 | FedoraGS3Connection con = new FedoraGS3Connection(propertyFilename);
|
---|
2347 |
|
---|
2348 | // DESCRIBE: serviceList, collectionList
|
---|
2349 | System.out.println("serviceList:\n" + con.getServiceList());
|
---|
2350 |
|
---|
2351 | System.out.println("collectionList:\n" + con.getCollectionList());
|
---|
2352 |
|
---|
2353 | String[] colPIDs = con.getCollections();
|
---|
2354 | String[] collectionNames = con.getCollectionNames(con.getCollections());
|
---|
2355 |
|
---|
2356 |
|
---|
2357 | for(int i = 0; i < collectionNames.length; i++) {
|
---|
2358 | System.out.println("Describing collections:\n");
|
---|
2359 | System.out.println(con.describeCollection(collectionNames[i]));
|
---|
2360 | System.out.println("Describing collection services:\n"
|
---|
2361 | + con.describeCollectionServices(collectionNames[i]));
|
---|
2362 | }
|
---|
2363 |
|
---|
2364 | String[] serviceNames = con.getServiceNames();
|
---|
2365 | for(int i = 0; i < serviceNames.length; i++) {
|
---|
2366 | System.out.println("Describing " + serviceNames[i] + ":\n"
|
---|
2367 | + con.describeCollectionService("demo", serviceNames[i]));
|
---|
2368 | }
|
---|
2369 |
|
---|
2370 |
|
---|
2371 | // TRYING OUT SPECIAL top-level document metadata retrieval (DLS, DC)
|
---|
2372 | // along with EX of the top-level document:
|
---|
2373 | System.out.println("\nGET META for greenstone:gs2mgdemo-HASH01d667303fe98545f03c14ae:");
|
---|
2374 | System.out.println(con.getDocumentMetadata(new String[]{"greenstone:gs2mgdemo-HASH01d667303fe98545f03c14ae"}, new String[]{"all"}));
|
---|
2375 |
|
---|
2376 |
|
---|
2377 | String[] docIDs = con.getCollectionDocs(colPIDs[0]);
|
---|
2378 | System.out.println("\nGET CONTENT:");
|
---|
2379 | for(int i = 0; i < docIDs.length; i++) {
|
---|
2380 | System.out.println(con.getContent(docIDs[i]));
|
---|
2381 | }
|
---|
2382 |
|
---|
2383 | System.out.println("\nGET META:");
|
---|
2384 | for(int i = 0; i < docIDs.length; i++) {
|
---|
2385 | System.out.println(con.getDocumentMetadata(docIDs[i], new String[]{"all"}));
|
---|
2386 | }
|
---|
2387 |
|
---|
2388 | String[] getTitlesFor = {
|
---|
2389 | "greenstone:gs2mgdemo-HASH01d667303fe98545f03c14ae",
|
---|
2390 | "greenstone:gs2mgdemo-HASHa568bac1d8d7bd12a0938b",
|
---|
2391 | "greenstone:gs2mgdemo-HASHa568bac1d8d7bd12a0938b-1",
|
---|
2392 | "greenstone:gs2mgdemo-HASH01d667303fe98545f03c14ae-1.7",
|
---|
2393 | "greenstone:gs2mgdemo-HASHa568bac1d8d7bd12a0938b-1.5.1"
|
---|
2394 | };
|
---|
2395 |
|
---|
2396 | // first let's display the regular meta for top-level docs and
|
---|
2397 | // their sections
|
---|
2398 | for(int i = 0; i < getTitlesFor.length; i++) {
|
---|
2399 | System.out.println(con.getDocumentMetadata(getTitlesFor[i], new String[]{"all"}));
|
---|
2400 | }
|
---|
2401 |
|
---|
2402 | System.out.println("\nTitles are:");
|
---|
2403 | System.out.println(con.getTitleMetadata(getTitlesFor));
|
---|
2404 |
|
---|
2405 | System.out.println("\nGET STRUCTURE:");
|
---|
2406 | for(int i = 0; i < docIDs.length; i++) {
|
---|
2407 | System.out.println("Descendents and numChildren:\n"
|
---|
2408 | + con.getDocumentStructure(docIDs[i], new String[] {"descendants"}, new String[] {"numChildren"}));
|
---|
2409 | System.out.println("Parent and numSiblings:\n"
|
---|
2410 | + con.getDocumentStructure(docIDs[i], new String[] {"parent"}, new String[] {"numSiblings"}));
|
---|
2411 | }
|
---|
2412 |
|
---|
2413 | // TEST ERROR CASES:
|
---|
2414 | System.out.println("\nTESTING ERROR CASES");
|
---|
2415 | System.out.println(con.getContent("greenstone:demo-pinky"));
|
---|
2416 | String[] errorCases = { "greenstone:demo-HASH23d1019b589e2ef6a680e3-1.5.1.5",
|
---|
2417 | "greenstone:demo-pinky" };
|
---|
2418 | System.out.println(con.getContent(errorCases));
|
---|
2419 | System.out.println(con.getDocumentMetadata(errorCases, new String[]{"all"}));
|
---|
2420 | System.out.println(con.getDocumentStructure(errorCases, new String[] {"descendants"}, new String[] {"numChildren"}));
|
---|
2421 |
|
---|
2422 | System.out.println("\nCLASSIFIER BROWSE");
|
---|
2423 | System.out.println(con.browse("gs2mgdemo", //"ClassifierBrowse",
|
---|
2424 | new String[]{"CL1"}, new String[] {""}, new String[] {""}));
|
---|
2425 |
|
---|
2426 | System.out.println("\nCLASSIFIER BROWSE METADATA RETRIEVE");
|
---|
2427 | String[] classNodeIDs = new String[26];
|
---|
2428 | for(int i = 0; i < classNodeIDs.length; i++) {
|
---|
2429 | int subClassifierNum = i + 1;
|
---|
2430 | classNodeIDs[i] = "CL1." + subClassifierNum;
|
---|
2431 | }
|
---|
2432 | System.out.println(con.browseMetadataRetrieve(//"gs2mgdemo",
|
---|
2433 | classNodeIDs, new String[]{"all"}));
|
---|
2434 |
|
---|
2435 | System.out.println("Testing query services");
|
---|
2436 | System.out.println("TEXT QUERY:");
|
---|
2437 | Map formControlValsMap = new HashMap();
|
---|
2438 | formControlValsMap.put(MAXDOCS, "100");
|
---|
2439 | formControlValsMap.put(QUERY, "snails");
|
---|
2440 | String searchResponse
|
---|
2441 | = con.query("gs2mgdemo", "TextQuery", formControlValsMap);
|
---|
2442 | System.out.println(searchResponse);
|
---|
2443 |
|
---|
2444 | System.out.println("FIELD QUERY:");
|
---|
2445 | formControlValsMap.clear();
|
---|
2446 | formControlValsMap.put(MAXDOCS, "100");
|
---|
2447 | formControlValsMap.put(QUERY, "interview,Gender equality,cyclone");
|
---|
2448 | formControlValsMap.put(FIELDNAME_ATT,
|
---|
2449 | "allFields,docTitles,allFields,allFields");
|
---|
2450 | searchResponse
|
---|
2451 | = con.query("gs2mgdemo", "FieldQuery", formControlValsMap);
|
---|
2452 | System.out.println(searchResponse);
|
---|
2453 |
|
---|
2454 | System.exit(0);
|
---|
2455 | }catch(Exception e) {
|
---|
2456 | JOptionPane.showMessageDialog(
|
---|
2457 | null, e, "Error", JOptionPane.ERROR_MESSAGE);
|
---|
2458 | //System.err.println("ERROR: " + e);
|
---|
2459 | e.printStackTrace();
|
---|
2460 | }
|
---|
2461 | }
|
---|
2462 | }
|
---|