1 | package org.greenstone.gsdl3.service;
|
---|
2 |
|
---|
3 |
|
---|
4 | // Greenstone classes
|
---|
5 | import org.greenstone.gsdl3.util.*;
|
---|
6 |
|
---|
7 | // XML classes
|
---|
8 | import org.w3c.dom.Document;
|
---|
9 | import org.w3c.dom.Element;
|
---|
10 | import org.w3c.dom.Node;
|
---|
11 | import org.w3c.dom.Attr;
|
---|
12 | import org.w3c.dom.Text;
|
---|
13 | import org.w3c.dom.NodeList;
|
---|
14 | import org.w3c.dom.NamedNodeMap;
|
---|
15 |
|
---|
16 | // General Java classes
|
---|
17 | import java.io.File;
|
---|
18 | import java.util.Vector;
|
---|
19 | import java.util.HashMap;
|
---|
20 |
|
---|
21 |
|
---|
22 | public class XMLRetrieve extends ServiceRack {
|
---|
23 |
|
---|
24 | protected static final String CONTENT_SERVICE = "DocumentContentRetrieve";
|
---|
25 | protected static final String METADATA_SERVICE = "DocumentMetadataRetrieve";
|
---|
26 | protected static final String STRUCTURE_SERVICE = "DocumentStructureRetrieve";
|
---|
27 |
|
---|
28 | protected String toc_xsl_name = "";
|
---|
29 | protected String document_encoding = "";
|
---|
30 | protected Element collection_doc_list = null;
|
---|
31 |
|
---|
32 | protected boolean provide_content = true;
|
---|
33 | protected boolean provide_structure = true;
|
---|
34 | protected boolean provide_metadata = true;
|
---|
35 |
|
---|
36 |
|
---|
37 | public boolean configure(Element info, Element extra_info) {
|
---|
38 |
|
---|
39 | // look for the parameters
|
---|
40 | Element param_list = (Element)GSXML.getChildByTagName(info, GSXML.PARAM_ELEM+GSXML.LIST_MODIFIER);
|
---|
41 | HashMap params;
|
---|
42 | String services_to_provide = "";
|
---|
43 | if (param_list != null) {
|
---|
44 | params = GSXML.extractParams(param_list, false);
|
---|
45 | this.toc_xsl_name = (String)params.get("tocXSLT");
|
---|
46 | this.document_encoding = (String)params.get("documentEncoding");
|
---|
47 | services_to_provide = (String)params.get("provideServices");
|
---|
48 | }
|
---|
49 | if (this.toc_xsl_name == null || this.toc_xsl_name.equals("")) {
|
---|
50 | this.toc_xsl_name = "default_toc";
|
---|
51 | }
|
---|
52 | this.toc_xsl_name = this.toc_xsl_name+".xsl";
|
---|
53 |
|
---|
54 | if (this.document_encoding == null || this.document_encoding.equals("")) {
|
---|
55 | this.document_encoding = "UTF-8";
|
---|
56 | }
|
---|
57 |
|
---|
58 | if (services_to_provide != null && !services_to_provide.equals("")) {
|
---|
59 | if (services_to_provide.indexOf("content")==-1) {
|
---|
60 | provide_content = false;
|
---|
61 | }
|
---|
62 | if (services_to_provide.indexOf("metadata")==-1) {
|
---|
63 | provide_metadata = false;
|
---|
64 | }
|
---|
65 | if (services_to_provide.indexOf("structure")==-1) {
|
---|
66 | provide_structure = false;
|
---|
67 | }
|
---|
68 |
|
---|
69 | }
|
---|
70 |
|
---|
71 | // set up short_service_info_ - for now just has name and type
|
---|
72 | Element retrieve_service;
|
---|
73 | if (provide_content) {
|
---|
74 | retrieve_service = this.doc.createElement(GSXML.SERVICE_ELEM);
|
---|
75 | retrieve_service.setAttribute(GSXML.TYPE_ATT, GSXML.SERVICE_TYPE_RETRIEVE);
|
---|
76 | retrieve_service.setAttribute(GSXML.NAME_ATT, CONTENT_SERVICE);
|
---|
77 | this.short_service_info.appendChild(retrieve_service);
|
---|
78 | }
|
---|
79 | if (provide_metadata) {
|
---|
80 | retrieve_service = this.doc.createElement(GSXML.SERVICE_ELEM);
|
---|
81 | retrieve_service.setAttribute(GSXML.TYPE_ATT, GSXML.SERVICE_TYPE_RETRIEVE);
|
---|
82 | retrieve_service.setAttribute(GSXML.NAME_ATT, METADATA_SERVICE);
|
---|
83 | this.short_service_info.appendChild(retrieve_service);
|
---|
84 | }
|
---|
85 | if (provide_structure) {
|
---|
86 | retrieve_service = this.doc.createElement(GSXML.SERVICE_ELEM);
|
---|
87 | retrieve_service.setAttribute(GSXML.TYPE_ATT, GSXML.SERVICE_TYPE_RETRIEVE);
|
---|
88 | retrieve_service.setAttribute(GSXML.NAME_ATT, STRUCTURE_SERVICE);
|
---|
89 | this.short_service_info.appendChild(retrieve_service);
|
---|
90 | }
|
---|
91 | // find the doc list from the extra_info and keep it - should this be in collect.cfg or build.cfg??
|
---|
92 | collection_doc_list = (Element)GSXML.getChildByTagName(extra_info, GSXML.DOCUMENT_ELEM+GSXML.LIST_MODIFIER);
|
---|
93 |
|
---|
94 | return true;
|
---|
95 | }
|
---|
96 |
|
---|
97 | // this may get called but is not useful in the case of retrieve services
|
---|
98 | protected Element getServiceDescription(String service_id, String lang, String subset) {
|
---|
99 |
|
---|
100 | Element retrieve_service = this.doc.createElement(GSXML.SERVICE_ELEM);
|
---|
101 | retrieve_service.setAttribute(GSXML.TYPE_ATT, GSXML.SERVICE_TYPE_RETRIEVE);
|
---|
102 | retrieve_service.setAttribute(GSXML.NAME_ATT, service_id);
|
---|
103 | return retrieve_service;
|
---|
104 | }
|
---|
105 |
|
---|
106 | protected Element processDocumentContentRetrieve(Element request) {
|
---|
107 | Element result = this.doc.createElement(GSXML.RESPONSE_ELEM);
|
---|
108 | result.setAttribute(GSXML.FROM_ATT, CONTENT_SERVICE);
|
---|
109 | result.setAttribute(GSXML.TYPE_ATT, GSXML.REQUEST_TYPE_PROCESS);
|
---|
110 |
|
---|
111 | Element doc_list = (Element)GSXML.getChildByTagName(request, GSXML.DOC_NODE_ELEM+GSXML.LIST_MODIFIER);
|
---|
112 | if (doc_list == null) {
|
---|
113 | return result;
|
---|
114 | }
|
---|
115 | Element result_doc_list = (Element)this.doc.importNode(doc_list, true);
|
---|
116 | result.appendChild(result_doc_list);
|
---|
117 |
|
---|
118 | NodeList docs = result_doc_list.getElementsByTagName(GSXML.DOC_NODE_ELEM);
|
---|
119 | for (int i=0; i<docs.getLength(); i++) {
|
---|
120 |
|
---|
121 | Element doc = (Element)docs.item(i);
|
---|
122 | Element content = this.doc.createElement(GSXML.NODE_CONTENT_ELEM);
|
---|
123 | doc.appendChild(content);
|
---|
124 |
|
---|
125 | String node_id = doc.getAttribute(GSXML.NODE_ID_ATT);
|
---|
126 | String doc_name = getWorkName(node_id);
|
---|
127 |
|
---|
128 | Element doc_elem = loadDocument(doc_name); // should perhaps cache the read in docs??
|
---|
129 | if (doc_elem == null) {
|
---|
130 | continue;
|
---|
131 | }
|
---|
132 |
|
---|
133 | // if we have asked for the whole doc, just append it
|
---|
134 | if (doc_name.equals(node_id)) {
|
---|
135 | content.appendChild(this.doc.importNode(doc_elem, true));
|
---|
136 | continue;
|
---|
137 | }
|
---|
138 |
|
---|
139 | // else we only want a sub section
|
---|
140 |
|
---|
141 | Element section = getSection(doc_elem, node_id);
|
---|
142 | if (section != null) {
|
---|
143 | content.appendChild(this.doc.importNode(section, true));
|
---|
144 | }
|
---|
145 |
|
---|
146 | } // for each doc
|
---|
147 |
|
---|
148 | return result;
|
---|
149 |
|
---|
150 | }
|
---|
151 |
|
---|
152 | protected Element processDocumentStructureRetrieve(Element request) {
|
---|
153 | Element result = this.doc.createElement(GSXML.RESPONSE_ELEM);
|
---|
154 | result.setAttribute(GSXML.FROM_ATT, STRUCTURE_SERVICE);
|
---|
155 | result.setAttribute(GSXML.TYPE_ATT, GSXML.REQUEST_TYPE_PROCESS);
|
---|
156 |
|
---|
157 | Element doc_list = (Element)GSXML.getChildByTagName(request, GSXML.DOC_NODE_ELEM+GSXML.LIST_MODIFIER);
|
---|
158 | if (doc_list == null) {
|
---|
159 | System.err.println("XMLRetrieve.DocumentStructureRetrieve: no documents specified in the request. ");
|
---|
160 | return result;
|
---|
161 | }
|
---|
162 |
|
---|
163 | Element result_doc_list = (Element)this.doc.importNode(doc_list, true);
|
---|
164 | result.appendChild(result_doc_list);
|
---|
165 | // first look for the stylesheet in the collection
|
---|
166 | File stylesheet = new File(GSFile.collStylesheetFile(this.site_home, this.cluster_name, this.toc_xsl_name));
|
---|
167 | if (!stylesheet.exists()) {
|
---|
168 | // now try in the site
|
---|
169 | stylesheet = new File(GSFile.siteStylesheetFile(this.site_home, this.toc_xsl_name));
|
---|
170 | }
|
---|
171 | if (!stylesheet.exists()) {
|
---|
172 | System.err.println("XMLRetrieve.DocumentStructureRetrieve: couldn't find the stylesheet file to produce the table of contents:"+stylesheet.getPath());
|
---|
173 | return result;
|
---|
174 | }
|
---|
175 |
|
---|
176 | // for now, we dont have any params, and we always return the structure of the whole document
|
---|
177 |
|
---|
178 | XMLTransformer transformer = new XMLTransformer();
|
---|
179 | NodeList docs = result_doc_list.getElementsByTagName(GSXML.DOC_NODE_ELEM);
|
---|
180 |
|
---|
181 | for (int i=0; i<docs.getLength(); i++) {
|
---|
182 |
|
---|
183 | Element doc = (Element)docs.item(i);
|
---|
184 |
|
---|
185 | Element structure = this.doc.createElement(GSXML.NODE_STRUCTURE_ELEM);
|
---|
186 | doc.appendChild(structure);
|
---|
187 | String doc_name = doc.getAttribute(GSXML.NODE_ID_ATT);
|
---|
188 | // make sure we are at the top level
|
---|
189 | doc_name = getWorkName(doc_name);
|
---|
190 |
|
---|
191 | File doc_file = new File(GSFile.collectionIndexDir(this.site_home, this.cluster_name)+File.separator+"text"+File.separatorChar+doc_name+".xml");
|
---|
192 |
|
---|
193 | if (!doc_file.exists()) {
|
---|
194 | System.err.println("XMLRetrieve.DocumentStructureRetrieve: couldn't find file in coll "+this.cluster_name +", file "+doc_name+".xml");
|
---|
195 | } else {
|
---|
196 | try {
|
---|
197 | Node toc = transformer.transform(stylesheet, doc_file);
|
---|
198 | structure.appendChild(this.doc.importNode(toc, true));
|
---|
199 | } catch (Exception e) {
|
---|
200 | System.err.println("XMLRetrieve.DocumentStructureRetrieve: couldn't transform the document to get the toc");
|
---|
201 | }
|
---|
202 | }
|
---|
203 |
|
---|
204 | }
|
---|
205 |
|
---|
206 | return result;
|
---|
207 |
|
---|
208 | }
|
---|
209 |
|
---|
210 | // this just extracts a bit of text from the section to use as the Title
|
---|
211 | // this should be overwritten for any format that has something more suitable
|
---|
212 | protected Element processDocumentMetadataRetrieve(Element request) {
|
---|
213 | Element result = this.doc.createElement(GSXML.RESPONSE_ELEM);
|
---|
214 | result.setAttribute(GSXML.FROM_ATT, METADATA_SERVICE);
|
---|
215 | result.setAttribute(GSXML.TYPE_ATT, GSXML.REQUEST_TYPE_PROCESS);
|
---|
216 |
|
---|
217 | Element doc_list = (Element)GSXML.getChildByTagName(request, GSXML.DOC_NODE_ELEM+GSXML.LIST_MODIFIER);
|
---|
218 | if (doc_list == null) {
|
---|
219 | System.err.println("XMLRetrieve.DocumentMetadataRetrieve: no documents in the request");
|
---|
220 | return result;
|
---|
221 | }
|
---|
222 |
|
---|
223 | Element result_doc_list = (Element)this.doc.importNode(doc_list, true);
|
---|
224 | result.appendChild(result_doc_list);
|
---|
225 |
|
---|
226 | Element param_list = (Element) GSXML.getChildByTagName(request, GSXML.PARAM_ELEM+GSXML.LIST_MODIFIER);
|
---|
227 | if (param_list == null) {
|
---|
228 | System.err.println("XMLRetrieve.DocumentMetadataRetrieve: no metadata in the request");
|
---|
229 | return result;
|
---|
230 | }
|
---|
231 |
|
---|
232 | Vector meta_name_list = new Vector();
|
---|
233 | boolean all_metadata = false;
|
---|
234 | // Process the request parameters
|
---|
235 | Element param = (Element) param_list.getFirstChild();
|
---|
236 | while (param != null) {
|
---|
237 | // Identify the metadata information desired
|
---|
238 | if (param.getAttribute(GSXML.NAME_ATT).equals("metadata")) {
|
---|
239 | String metadata = GSXML.getValue(param);
|
---|
240 | if (metadata.equals("all")) {
|
---|
241 | all_metadata = true;
|
---|
242 | break;
|
---|
243 | }
|
---|
244 | meta_name_list.add(metadata);
|
---|
245 | }
|
---|
246 | param = (Element) param.getNextSibling();
|
---|
247 | }
|
---|
248 |
|
---|
249 | NodeList docs = result_doc_list.getElementsByTagName(GSXML.DOC_NODE_ELEM);
|
---|
250 | for (int i=0; i<docs.getLength(); i++) {
|
---|
251 | Element doc = (Element)docs.item(i);
|
---|
252 | String node_id = doc.getAttribute(GSXML.NODE_ID_ATT);
|
---|
253 | String doc_name = getWorkName(node_id);
|
---|
254 |
|
---|
255 | Element metadata_list = getMetadata(node_id, all_metadata, meta_name_list);
|
---|
256 | doc.appendChild(metadata_list);
|
---|
257 | }
|
---|
258 |
|
---|
259 | return result;
|
---|
260 | }
|
---|
261 |
|
---|
262 | protected Element loadDocument(String doc_name) {
|
---|
263 | // try to find the document
|
---|
264 | File doc_file = new File(GSFile.collectionIndexDir(this.site_home, this.cluster_name)+File.separator+"text"+File.separatorChar+doc_name+".xml");
|
---|
265 |
|
---|
266 | if (!doc_file.exists()) {
|
---|
267 | System.out.println("XMLRetrieve.loadDocument: couldn't find file in coll "+this.cluster_name +", file "+doc_name+".xml");
|
---|
268 | return null;
|
---|
269 | }
|
---|
270 |
|
---|
271 | Document the_doc = null;
|
---|
272 | try {
|
---|
273 | the_doc = this.converter.getDOM(doc_file, this.document_encoding);
|
---|
274 | } catch (Exception e) {
|
---|
275 | System.err.println("XMLRetrieve.loadDocument: couldn't create a DOM from file "+doc_file.getPath());
|
---|
276 | return null;
|
---|
277 | }
|
---|
278 |
|
---|
279 | return the_doc.getDocumentElement();
|
---|
280 |
|
---|
281 | }
|
---|
282 |
|
---|
283 |
|
---|
284 | protected Element getSection(Element doc_elem, String node_id) {
|
---|
285 | String [] bits = node_id.split("\\.");
|
---|
286 | if (bits.length > 4) {
|
---|
287 | System.err.println("XMLRetrieve.getSection: badly formatted node id ("+node_id +"), cant retrieve the section");
|
---|
288 | return null;
|
---|
289 | }
|
---|
290 |
|
---|
291 | String id="";
|
---|
292 | String tagname = "";
|
---|
293 | String scope = "";
|
---|
294 | if (bits.length==2) {
|
---|
295 | tagname = bits[1];
|
---|
296 | } else {
|
---|
297 | scope = bits[1];
|
---|
298 | tagname = bits[2];
|
---|
299 |
|
---|
300 | if (bits.length == 4) {
|
---|
301 | id = bits[3];
|
---|
302 | }
|
---|
303 | }
|
---|
304 | scope = translateScope(scope);
|
---|
305 | Element top=null;
|
---|
306 | if (!scope.equals("")) {
|
---|
307 | top = (Element)GSXML.getNodeByPath(doc_elem, scope);
|
---|
308 | if (top == null) {
|
---|
309 | // something gone wrong
|
---|
310 | return null;
|
---|
311 | }
|
---|
312 | } else {
|
---|
313 | top = doc_elem;
|
---|
314 | }
|
---|
315 |
|
---|
316 | NodeList elements = top.getElementsByTagName(tagname);
|
---|
317 | if (elements.getLength() == 0) {
|
---|
318 | return null;
|
---|
319 | }
|
---|
320 | // no id, just return the first one
|
---|
321 | if (id.equals("")) {
|
---|
322 | return (Element)elements.item(0);
|
---|
323 | }
|
---|
324 | // have an id, need to check and find the right one.
|
---|
325 | for (int i=0; i<elements.getLength();i++) {
|
---|
326 | Element e = (Element)elements.item(i);
|
---|
327 | if (e.getAttribute("gs3:id").equals(id)) {
|
---|
328 | return e;
|
---|
329 | }
|
---|
330 | }
|
---|
331 | return null;
|
---|
332 |
|
---|
333 | }
|
---|
334 |
|
---|
335 | protected Element getMetadata(String node_id, boolean all, Vector meta_name_list) {
|
---|
336 |
|
---|
337 | // our default strategy here is to only return Title and root:Title
|
---|
338 | // ignore all others
|
---|
339 | // the title of a section is just a little bit of the text inside it.
|
---|
340 | // the root:Title is the title from the doc info in the config file
|
---|
341 | Element metadata_list = this.doc.createElement(GSXML.METADATA_ELEM+ GSXML.LIST_MODIFIER);
|
---|
342 | String doc_name = getWorkName(node_id);
|
---|
343 | boolean node_is_root = false;
|
---|
344 | if (doc_name.equals(node_id)) {
|
---|
345 | node_is_root = true;
|
---|
346 | }
|
---|
347 |
|
---|
348 | Element this_doc = GSXML.getNamedElement(this.collection_doc_list, GSXML.DOCUMENT_ELEM, GSXML.NAME_ATT, doc_name);
|
---|
349 | Element doc_meta_list = (Element) GSXML.getChildByTagName(this_doc, GSXML.METADATA_ELEM+GSXML.LIST_MODIFIER);
|
---|
350 |
|
---|
351 | boolean get_section_title = false;
|
---|
352 |
|
---|
353 | if (all) {
|
---|
354 | if (node_is_root) {
|
---|
355 | return (Element)this.doc.importNode(doc_meta_list, true);
|
---|
356 | } else {
|
---|
357 | get_section_title = true;
|
---|
358 | }
|
---|
359 |
|
---|
360 | } else {
|
---|
361 | // have to process metadata one by one
|
---|
362 | for (int i=0; i<meta_name_list.size(); i++) {
|
---|
363 | String meta_name = (String) meta_name_list.elementAt(i);
|
---|
364 | String actual_meta_name = meta_name;
|
---|
365 | if (meta_name.startsWith("root:")) {
|
---|
366 | actual_meta_name = meta_name.substring(5);
|
---|
367 | } else {
|
---|
368 | // its a section level one - check to see if doc is root
|
---|
369 | if (!node_is_root) {
|
---|
370 | if (meta_name.equals("Title")) {
|
---|
371 | get_section_title = true;
|
---|
372 | }
|
---|
373 | continue; // move on to teh next metadata
|
---|
374 | }
|
---|
375 | }
|
---|
376 |
|
---|
377 | // here, we look for the specific meta elem in doc_meta_list
|
---|
378 | Element meta_item = GSXML.getNamedElement(doc_meta_list, GSXML.METADATA_ELEM, GSXML.NAME_ATT, actual_meta_name);
|
---|
379 | if (meta_item != null) {
|
---|
380 | meta_item = (Element)this.doc.importNode(meta_item, true);
|
---|
381 | meta_item.setAttribute(GSXML.NAME_ATT, meta_name);
|
---|
382 | metadata_list.appendChild(meta_item);
|
---|
383 | }
|
---|
384 | } // for each metadata
|
---|
385 | }
|
---|
386 |
|
---|
387 | // now we have processed all teh doc metadata, just have section one to go, if needed
|
---|
388 | if (get_section_title) {
|
---|
389 |
|
---|
390 | Element doc_elem = loadDocument(doc_name);
|
---|
391 | if (doc_elem != null) {
|
---|
392 | Element section = getSection(doc_elem, node_id);
|
---|
393 | if (section != null) {
|
---|
394 | Element title_meta = extractTitleMeta(section);
|
---|
395 | metadata_list.appendChild(title_meta);
|
---|
396 |
|
---|
397 | }
|
---|
398 | }
|
---|
399 |
|
---|
400 | }
|
---|
401 | return metadata_list;
|
---|
402 | }
|
---|
403 |
|
---|
404 | protected Element extractTitleMeta(Element section) {
|
---|
405 | Element meta_elem = this.doc.createElement(GSXML.METADATA_ELEM);
|
---|
406 | meta_elem.setAttribute(GSXML.NAME_ATT, "Title");
|
---|
407 |
|
---|
408 | String title = "dummy title";
|
---|
409 | Text t = this.doc.createTextNode(title);
|
---|
410 | meta_elem.appendChild(t);
|
---|
411 | return meta_elem;
|
---|
412 |
|
---|
413 | }
|
---|
414 | // some methods for handling nodeIDs - they may be different for different colls, so they can be overwritten
|
---|
415 |
|
---|
416 | // the full default nodeID looks like work.scope.tag.id
|
---|
417 | // the shorter versions are work, work.tag, work.scope.tag
|
---|
418 | protected String getWorkName(String node_id) {
|
---|
419 | int pos = node_id.indexOf('.');
|
---|
420 | if (pos == -1) {
|
---|
421 | return node_id;
|
---|
422 | }
|
---|
423 | return node_id.substring(0, pos);
|
---|
424 | }
|
---|
425 |
|
---|
426 | // this assumes that the scope refers to a top level node - this may be overwritten if the scope bit in the id is a shorthand of some sort
|
---|
427 | protected String translateScope(String scope) {
|
---|
428 | return scope;
|
---|
429 | }
|
---|
430 |
|
---|
431 | }
|
---|