1 | /**********************************************************************
|
---|
2 | *
|
---|
3 | * ApplyXSLT.java
|
---|
4 | *
|
---|
5 | * Copyright 2006-2010 The New Zealand Digital Library Project
|
---|
6 | *
|
---|
7 | * A component of the Greenstone digital library software
|
---|
8 | * from the New Zealand Digital Library Project at the
|
---|
9 | * University of Waikato, New Zealand.
|
---|
10 | *
|
---|
11 | * This program is free software; you can redistribute it and/or modify
|
---|
12 | * it under the terms of the GNU General Public License as published by
|
---|
13 | * the Free Software Foundation; either version 2 of the License, or
|
---|
14 | * (at your option) any later version.
|
---|
15 | *
|
---|
16 | * This program is distributed in the hope that it will be useful,
|
---|
17 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
18 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
19 | * GNU General Public License for more details.
|
---|
20 | *
|
---|
21 | * You should have received a copy of the GNU General Public License
|
---|
22 | * along with this program; if not, write to the Free Software
|
---|
23 | * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
24 | *
|
---|
25 | *********************************************************************/
|
---|
26 |
|
---|
27 |
|
---|
28 | package org.nzdl.gsdl;
|
---|
29 |
|
---|
30 | import java.io.*;
|
---|
31 |
|
---|
32 | import javax.xml.transform.Transformer;
|
---|
33 | import javax.xml.transform.TransformerConfigurationException;
|
---|
34 | import javax.xml.transform.TransformerException;
|
---|
35 | import javax.xml.transform.TransformerFactory;
|
---|
36 | import javax.xml.transform.stream.StreamResult;
|
---|
37 | import javax.xml.transform.stream.StreamSource;
|
---|
38 |
|
---|
39 | import javax.xml.parsers.*;
|
---|
40 | import javax.xml.transform.dom.*;
|
---|
41 | import org.w3c.dom.*;
|
---|
42 |
|
---|
43 |
|
---|
44 |
|
---|
45 | /**
|
---|
46 | * Use the TraX interface to perform a transformation in the simplest manner possible
|
---|
47 | * (3 statements).
|
---|
48 | */
|
---|
49 | public class ApplyXSLT
|
---|
50 | {
|
---|
51 |
|
---|
52 | public static final String DOC_START = new String ("<?DocStart?>");
|
---|
53 | public static final String DOC_END = new String ("<?DocEnd?>");
|
---|
54 | public static final String INPUT_END = new String ("<?Done?>");
|
---|
55 |
|
---|
56 | private static final String RECORD_ELEMENT = "record";
|
---|
57 | private static final String CONTROLFIELD_ELEMENT = "controlfield";
|
---|
58 | private static final String SUBFIELD_ELEMENT = "subfield";
|
---|
59 | private static final String LEADER_ELEMENT = "leader";
|
---|
60 |
|
---|
61 | private final int BEFORE_READING = 0;
|
---|
62 | private final int IS_READING = 1;
|
---|
63 | private String xsl_file;
|
---|
64 | private String mapping_file;
|
---|
65 |
|
---|
66 | public ApplyXSLT(){}
|
---|
67 |
|
---|
68 | public ApplyXSLT(String xsl_file)
|
---|
69 | {
|
---|
70 | this.xsl_file = xsl_file;
|
---|
71 | }
|
---|
72 |
|
---|
73 | public ApplyXSLT(String xsl_file,String mapping_file){
|
---|
74 | this.xsl_file = xsl_file;
|
---|
75 | this.mapping_file = mapping_file;
|
---|
76 | }
|
---|
77 |
|
---|
78 | private boolean process()
|
---|
79 | {
|
---|
80 | try{
|
---|
81 |
|
---|
82 | // Use System InputStream to receive piped data from the perl program
|
---|
83 | InputStreamReader ir = new InputStreamReader(System.in, "UTF8");
|
---|
84 | BufferedReader br = new BufferedReader(ir);
|
---|
85 |
|
---|
86 | int system_status = BEFORE_READING;
|
---|
87 | StringBuffer a_doc = new StringBuffer();
|
---|
88 | String output_file = new String();
|
---|
89 |
|
---|
90 |
|
---|
91 | while (br.ready()) {
|
---|
92 |
|
---|
93 | String this_line = br.readLine();
|
---|
94 | if(system_status == BEFORE_READING){
|
---|
95 | if(this_line.compareTo(DOC_START) == 0){
|
---|
96 | output_file = br.readLine(); // read the next line as the output file name
|
---|
97 | system_status = IS_READING;
|
---|
98 | a_doc = new StringBuffer();
|
---|
99 | }
|
---|
100 | else if(this_line.compareTo(INPUT_END) == 0){
|
---|
101 | return true;
|
---|
102 | }
|
---|
103 | else{
|
---|
104 | System.err.println("Undefined process status:" + this_line);
|
---|
105 | system_status = BEFORE_READING;
|
---|
106 | }
|
---|
107 |
|
---|
108 | }
|
---|
109 | else if(system_status == IS_READING){
|
---|
110 | if(this_line.compareTo(DOC_END) == 0){
|
---|
111 | boolean result = false;
|
---|
112 | if (mapping_file !=null && !mapping_file.equals("")){
|
---|
113 | result = translateXMLWithMapping(a_doc.toString(), output_file);
|
---|
114 | }
|
---|
115 | else{
|
---|
116 | result = translateXML(a_doc.toString(), output_file);
|
---|
117 | }
|
---|
118 |
|
---|
119 | if (!result){
|
---|
120 | System.err.println("Translation Failed!!");
|
---|
121 | return false;
|
---|
122 | }
|
---|
123 |
|
---|
124 | system_status = BEFORE_READING;
|
---|
125 |
|
---|
126 | }
|
---|
127 | else{
|
---|
128 | a_doc.append(this_line + "\n");
|
---|
129 | }
|
---|
130 | }
|
---|
131 | else{
|
---|
132 | System.err.println ("Undefined system status in ApplyXSLT.java main().");
|
---|
133 | System.exit(-1);
|
---|
134 | }
|
---|
135 |
|
---|
136 | }
|
---|
137 | }catch (Exception e)
|
---|
138 | {
|
---|
139 | System.err.println("Receiving piped data error!" + e.toString());
|
---|
140 | }
|
---|
141 |
|
---|
142 | return false;
|
---|
143 | }
|
---|
144 |
|
---|
145 |
|
---|
146 | private boolean translateXML(String full_doc, String output_file)
|
---|
147 | throws IOException,TransformerException, TransformerConfigurationException, FileNotFoundException
|
---|
148 | {
|
---|
149 |
|
---|
150 | StringReader str = new StringReader(full_doc) ;
|
---|
151 |
|
---|
152 | TransformerFactory tFactory = TransformerFactory.newInstance();
|
---|
153 | Transformer transformer = tFactory.newTransformer(new StreamSource(xsl_file));
|
---|
154 | transformer.transform(new StreamSource(str), new StreamResult(new FileOutputStream(output_file)));
|
---|
155 | return true;
|
---|
156 | }
|
---|
157 |
|
---|
158 | private boolean translateXMLWithMapping(String full_doc, String output_file)
|
---|
159 | throws IOException,TransformerException, TransformerConfigurationException, FileNotFoundException
|
---|
160 | {
|
---|
161 | StringReader str = new StringReader(full_doc) ;
|
---|
162 |
|
---|
163 | try{
|
---|
164 | TransformerFactory tFactory = TransformerFactory.newInstance();
|
---|
165 | Transformer transformer = tFactory.newTransformer(new StreamSource(xsl_file));
|
---|
166 |
|
---|
167 | Document mapping_doc = DocumentBuilderFactory.newInstance().newDocumentBuilder().parse(mapping_file);
|
---|
168 | Element mapping =mapping_doc.getDocumentElement();
|
---|
169 |
|
---|
170 | transformer.setParameter("mapping",mapping);
|
---|
171 |
|
---|
172 | Document output_doc = DocumentBuilderFactory.newInstance().newDocumentBuilder().newDocument();
|
---|
173 |
|
---|
174 | transformer.transform(new StreamSource(str), new DOMResult(output_doc));
|
---|
175 |
|
---|
176 | calculateRecordsLength(output_doc);
|
---|
177 |
|
---|
178 | transformer = tFactory.newTransformer();
|
---|
179 |
|
---|
180 | transformer.transform(new DOMSource(output_doc), new StreamResult(new FileOutputStream(output_file)));
|
---|
181 |
|
---|
182 | }
|
---|
183 | catch(Exception e){
|
---|
184 | e.printStackTrace();
|
---|
185 | return false;
|
---|
186 | }
|
---|
187 |
|
---|
188 | return true;
|
---|
189 | }
|
---|
190 |
|
---|
191 | private void calculateRecordsLength(Document output_doc){
|
---|
192 | NodeList records = output_doc.getDocumentElement().getElementsByTagName(RECORD_ELEMENT);
|
---|
193 |
|
---|
194 | for(int i=0;i<records.getLength();i++){
|
---|
195 | Element record = (Element)records.item(i);
|
---|
196 | calculateRecordLength(record);
|
---|
197 | }
|
---|
198 | }
|
---|
199 |
|
---|
200 | private void calculateRecordLength(Element record){
|
---|
201 | int total_length =0;
|
---|
202 | NodeList controlfileds = record.getElementsByTagName(CONTROLFIELD_ELEMENT);
|
---|
203 | for(int i=0;i<controlfileds.getLength();i++){
|
---|
204 | Element controlfiled = (Element)controlfileds.item(i);
|
---|
205 | total_length +=getElementTextValue(controlfiled).length();
|
---|
206 | }
|
---|
207 |
|
---|
208 | NodeList subfileds = record.getElementsByTagName(SUBFIELD_ELEMENT);
|
---|
209 | for(int i=0;i<subfileds.getLength();i++){
|
---|
210 | Element subfiled = (Element)subfileds.item(i);
|
---|
211 | total_length +=getElementTextValue(subfiled).length();
|
---|
212 | }
|
---|
213 |
|
---|
214 | String record_length = total_length+"";
|
---|
215 | //fill in a extra digit as record length needs to be five characters long
|
---|
216 | if (total_length < 10000){
|
---|
217 | record_length = "0"+record_length;
|
---|
218 | if (total_length < 1000){
|
---|
219 | record_length = "0"+record_length;
|
---|
220 | }
|
---|
221 | if (total_length < 100){
|
---|
222 | record_length = "0"+record_length;
|
---|
223 | }
|
---|
224 | if (total_length < 10){
|
---|
225 | record_length = "0"+record_length;
|
---|
226 | }
|
---|
227 |
|
---|
228 | }
|
---|
229 |
|
---|
230 | NodeList leaders = record.getElementsByTagName(LEADER_ELEMENT);
|
---|
231 |
|
---|
232 | //only one leader element
|
---|
233 | if (leaders.getLength() >0){
|
---|
234 | Element leader_element = (Element)leaders.item(0);
|
---|
235 | removeFirstTextNode(leader_element);
|
---|
236 | leader_element.insertBefore(leader_element.getOwnerDocument().createTextNode(record_length),leader_element.getFirstChild());
|
---|
237 | }
|
---|
238 |
|
---|
239 | }
|
---|
240 |
|
---|
241 | private void removeFirstTextNode(Element element){
|
---|
242 | //remove the first text node
|
---|
243 | NodeList children_nodelist = element.getChildNodes();
|
---|
244 | for (int i = 0; i < children_nodelist.getLength(); i++) {
|
---|
245 | Node child_node = children_nodelist.item(i);
|
---|
246 | if (child_node.getNodeType() == Node.TEXT_NODE) {
|
---|
247 | element.removeChild(child_node);
|
---|
248 | return;
|
---|
249 | }
|
---|
250 | }
|
---|
251 |
|
---|
252 | }
|
---|
253 |
|
---|
254 | private String getElementTextValue(Element element)
|
---|
255 | {
|
---|
256 | String text ="";
|
---|
257 |
|
---|
258 | // Find the node child
|
---|
259 | NodeList children_nodelist = element.getChildNodes();
|
---|
260 | for (int i = 0; i < children_nodelist.getLength(); i++) {
|
---|
261 | Node child_node = children_nodelist.item(i);
|
---|
262 | if (child_node.getNodeType() == Node.TEXT_NODE) {
|
---|
263 | text +=child_node.getNodeValue();
|
---|
264 | }
|
---|
265 | }
|
---|
266 |
|
---|
267 | return text;
|
---|
268 | }
|
---|
269 |
|
---|
270 |
|
---|
271 | private void setMappingVariable(Document style_doc){
|
---|
272 | Node child = style_doc.getDocumentElement().getFirstChild();
|
---|
273 | while(child != null) {
|
---|
274 | String name = child.getNodeName();
|
---|
275 | if (name.equals("xsl:variable")) {
|
---|
276 | Element variable_element = (Element)child;
|
---|
277 | if ( variable_element.getAttribute("name").trim().equals("mapping")){
|
---|
278 | variable_element.setAttribute("select","document('"+mapping_file+"')/Mapping");
|
---|
279 | }
|
---|
280 | }
|
---|
281 | child = child.getNextSibling();
|
---|
282 | }
|
---|
283 |
|
---|
284 | }
|
---|
285 |
|
---|
286 |
|
---|
287 | private void translate(String xml_file, String xsl_file, String output_file)throws IOException,TransformerException, TransformerConfigurationException, FileNotFoundException, IOException{
|
---|
288 |
|
---|
289 | TransformerFactory tFactory = TransformerFactory.newInstance();
|
---|
290 | Transformer transformer = tFactory.newTransformer(new StreamSource(xsl_file));
|
---|
291 |
|
---|
292 | OutputStreamWriter output = null;
|
---|
293 | if (output_file.equals("")) {
|
---|
294 | output = new OutputStreamWriter(System.out, "UTF-8");
|
---|
295 | }
|
---|
296 | else{
|
---|
297 | output = new OutputStreamWriter(new FileOutputStream(output_file), "UTF-8");
|
---|
298 | }
|
---|
299 |
|
---|
300 | transformer.transform(new StreamSource(new File(xml_file)),new StreamResult(output));
|
---|
301 |
|
---|
302 | }
|
---|
303 |
|
---|
304 | static public String replaceAll(String source_string, String match_regexp, String replace_string)
|
---|
305 | {
|
---|
306 | return source_string.replaceAll(match_regexp, replace_string);
|
---|
307 | }
|
---|
308 |
|
---|
309 |
|
---|
310 | public static void main(String[] args)
|
---|
311 | {
|
---|
312 |
|
---|
313 | String xml_file="";
|
---|
314 | String xsl_file="";
|
---|
315 | String mapping_file="";
|
---|
316 | String output_file="";
|
---|
317 |
|
---|
318 | // Checking Arguments
|
---|
319 | if(args.length < 1)
|
---|
320 | {
|
---|
321 | printUsage();
|
---|
322 | }
|
---|
323 |
|
---|
324 | for (int i=0;i<args.length;i++){
|
---|
325 | if (args[i].equals("-m") && i+1 < args.length && !args[i+1].startsWith("-")){
|
---|
326 | mapping_file = args[++i];
|
---|
327 | checkFile(mapping_file.replaceAll("file:///",""));
|
---|
328 | }
|
---|
329 | else if (args[i].equals("-x") && i+1 < args.length && !args[i+1].startsWith("-")){
|
---|
330 | xml_file = args[++i];
|
---|
331 | checkFile(xml_file.replaceAll("file:///",""));
|
---|
332 | }
|
---|
333 | else if(args[i].equals("-t") && i+1 < args.length && !args[i+1].startsWith("-")){
|
---|
334 | xsl_file = args[++i];
|
---|
335 | checkFile( xsl_file.replaceAll("file:///",""));
|
---|
336 | }
|
---|
337 | else if(args[i].equals("-o") && i+1 < args.length && !args[i+1].startsWith("-")){
|
---|
338 | output_file = args[++i];
|
---|
339 |
|
---|
340 | }
|
---|
341 | else if(args[i].equals("-h")){
|
---|
342 | printUsage();
|
---|
343 | }
|
---|
344 | else{
|
---|
345 | printUsage();
|
---|
346 | }
|
---|
347 |
|
---|
348 | }
|
---|
349 |
|
---|
350 |
|
---|
351 | ApplyXSLT core = null;
|
---|
352 |
|
---|
353 | if (xml_file.equals("") && !xsl_file.equals("")){//read from pipe line
|
---|
354 | if (mapping_file.equals("")){
|
---|
355 | core = new ApplyXSLT(xsl_file);
|
---|
356 | }
|
---|
357 | else{
|
---|
358 | core = new ApplyXSLT(xsl_file,mapping_file);
|
---|
359 | }
|
---|
360 |
|
---|
361 | if (core != null){
|
---|
362 | core.process();
|
---|
363 | }
|
---|
364 | else{
|
---|
365 | printUsage();
|
---|
366 | }
|
---|
367 | }
|
---|
368 | else if(!xml_file.equals("") && !xsl_file.equals("")){
|
---|
369 | core = new ApplyXSLT();
|
---|
370 | try {
|
---|
371 | core.translate(xml_file,xsl_file,output_file);
|
---|
372 | }
|
---|
373 | catch(Exception e){e.printStackTrace();}
|
---|
374 | }
|
---|
375 | else{
|
---|
376 | printUsage();
|
---|
377 | }
|
---|
378 |
|
---|
379 | }
|
---|
380 |
|
---|
381 | private static void checkFile(String filename){
|
---|
382 | File file = new File(filename);
|
---|
383 | if (!file.exists()){
|
---|
384 | System.out.println("Error: "+filename+" doesn't exist!");
|
---|
385 | System.exit(-1);
|
---|
386 | }
|
---|
387 | }
|
---|
388 |
|
---|
389 | private static void printUsage(){
|
---|
390 | System.out.println("Usage:ApplyXSLT -x File -t File [-m File] [-o File]");
|
---|
391 | System.out.println("\t-x specifies the xml file (Note: optional for piped xml data)");
|
---|
392 | System.out.println("\t-t specifies the xsl file");
|
---|
393 | System.out.println("\t-m specifies the mapping file (for MARCXMLPlugout.pm only)");
|
---|
394 | System.out.println("\t-o specifies the output file name (output to screen if this option is absent)");
|
---|
395 | System.exit(-1);
|
---|
396 | }
|
---|
397 | }
|
---|
398 |
|
---|
399 |
|
---|