|
1
|
// CatalogEntityResolver.java - SAX entityResolver using OASIS Catalogs
|
|
2
|
// Written by Norman Walsh, nwalsh@arbortext.com
|
|
3
|
// NO WARRANTY! This class is in the public domain.
|
|
4
|
|
|
5
|
package com.arbortext.catalog;
|
|
6
|
|
|
7
|
import java.io.IOException;
|
|
8
|
import java.io.InputStream;
|
|
9
|
import java.lang.Integer;
|
|
10
|
import java.net.URL;
|
|
11
|
import java.net.MalformedURLException;
|
|
12
|
|
|
13
|
import com.arbortext.catalog.Catalog;
|
|
14
|
import org.xml.sax.EntityResolver;
|
|
15
|
import org.xml.sax.InputSource;
|
|
16
|
|
|
17
|
/**
|
|
18
|
* <p>Implements SAX entityResolver using OASIS Open Catalogs.</p>
|
|
19
|
*
|
|
20
|
* <blockquote>
|
|
21
|
* <em>This module, both source code and documentation, is in the
|
|
22
|
* Public Domain, and comes with <strong>NO WARRANTY</strong>.</em>
|
|
23
|
* </blockquote>
|
|
24
|
*
|
|
25
|
* <p>This class implements the SAX entityResolver interface. It uses
|
|
26
|
* OASIS Open catalog files to provide a facility for mapping public
|
|
27
|
* or system identifiers in source documents to local system identifiers.
|
|
28
|
*
|
|
29
|
* <p>This code interrogates the following non-standard system properties:</p>
|
|
30
|
*
|
|
31
|
* <dl>
|
|
32
|
* <dt><b>xml.catalog.debug</b></dt>
|
|
33
|
* <dd><p>Sets the debug level. A value of 0 is assumed if the
|
|
34
|
* property is not set or is not a number.</p></dd>
|
|
35
|
* </dl>
|
|
36
|
*
|
|
37
|
* @see Catalog
|
|
38
|
*
|
|
39
|
* @author Arbortext, Inc.
|
|
40
|
* @author Norman Walsh
|
|
41
|
* <a href="mailto:nwalsh@arbortext.com">nwalsh@arbortext.com</a>
|
|
42
|
* @version 1.0
|
|
43
|
*/
|
|
44
|
public class CatalogEntityResolver implements EntityResolver {
|
|
45
|
/**
|
|
46
|
* <p>The debug level</p>
|
|
47
|
*
|
|
48
|
* <p>In general, higher numbers produce more information:</p>
|
|
49
|
* <ul>
|
|
50
|
* <li>0, no messages
|
|
51
|
* <li>1, minimal messages (high-level status)
|
|
52
|
* <li>2, detailed messages
|
|
53
|
* </ul>
|
|
54
|
*/
|
|
55
|
public int debug = 0;
|
|
56
|
|
|
57
|
/**
|
|
58
|
* <p>Indicates that unusable system identifiers should be ignored.</p>
|
|
59
|
*/
|
|
60
|
private boolean retryBadSystemIds = false;
|
|
61
|
|
|
62
|
/**
|
|
63
|
* <p>The OASIS Open Catalog used for entity resolution.</p>
|
|
64
|
*
|
|
65
|
* <p>This field is exposed so that the catalog can be updated
|
|
66
|
* after creating the instance of CatalogEntityResolver that will
|
|
67
|
* be used by the parser.</p>
|
|
68
|
*/
|
|
69
|
public Catalog catalog = null;
|
|
70
|
|
|
71
|
/**
|
|
72
|
* <p>Constructs a CatalogEntityResolver with an empty catalog.</p>
|
|
73
|
*/
|
|
74
|
public CatalogEntityResolver() {
|
|
75
|
String property = System.getProperty("xml.catalog.debug");
|
|
76
|
|
|
77
|
if (property != null) {
|
|
78
|
try {
|
|
79
|
debug = Integer.parseInt(property);
|
|
80
|
} catch (NumberFormatException e) {
|
|
81
|
debug = 0;
|
|
82
|
}
|
|
83
|
}
|
|
84
|
}
|
|
85
|
|
|
86
|
/**
|
|
87
|
* <p>Set the Catalog that will be used to resolve entities.</p>
|
|
88
|
*
|
|
89
|
* <p>This is a convenience method for setting the
|
|
90
|
* <a href="#catalog"><code>catalog</code></a> field.</p>
|
|
91
|
*/
|
|
92
|
public void setCatalog(Catalog cat) {
|
|
93
|
catalog = cat;
|
|
94
|
}
|
|
95
|
|
|
96
|
/**
|
|
97
|
* <p>Parse a Catalog file.</p>
|
|
98
|
*
|
|
99
|
* <p>This is really just a convenience method which calls
|
|
100
|
* <a href="#catalog"><code>catalog.parseCatalog()</code></a>.</p>
|
|
101
|
*
|
|
102
|
* @param fileName The filename of the catalog file to process
|
|
103
|
*
|
|
104
|
* @throws MalformedURLException The fileName cannot be turned into
|
|
105
|
* a valid URL.
|
|
106
|
* @throws IOException Error reading catalog file.
|
|
107
|
*/
|
|
108
|
public synchronized void parseCatalog(String fileName)
|
|
109
|
throws MalformedURLException, IOException {
|
|
110
|
catalog.parseCatalog(fileName);
|
|
111
|
}
|
|
112
|
|
|
113
|
/**
|
|
114
|
* <p>Establish whether or not bad system identifiers should be
|
|
115
|
* ignored.</p>
|
|
116
|
*
|
|
117
|
* <p>The semantics of catalog file lookup are such that if a system
|
|
118
|
* identifier is supplied in the instance document, it is possible
|
|
119
|
* that it will be used in preference to alternative system identifiers
|
|
120
|
* in the catalog.</p>
|
|
121
|
*
|
|
122
|
* <p>If this variable is <code>true</code> and the system identifier
|
|
123
|
* passed to the entity resolver would be returned, the entity resolver
|
|
124
|
* attempts to open it. If it cannot be opened, the resolver
|
|
125
|
* does another catalog search,
|
|
126
|
* ignoring the fact that a system identifier was specified. If this
|
|
127
|
* second search locates a system identifer, it will be returned.</p>
|
|
128
|
*
|
|
129
|
* <p>This setting is initially <code>false</code> meaning that
|
|
130
|
* system identifiers
|
|
131
|
* in the document will be used in preference to some entries in
|
|
132
|
* the catalog.</p>
|
|
133
|
*
|
|
134
|
* @param retry If true, the resolver will retry Catalog lookups when
|
|
135
|
* the supplied system identifer cannot be opened.
|
|
136
|
*/
|
|
137
|
public void setRetry(boolean retry) {
|
|
138
|
retryBadSystemIds = retry;
|
|
139
|
}
|
|
140
|
|
|
141
|
/**
|
|
142
|
* <p>Implements the <code>resolveEntity</code> method
|
|
143
|
* for the SAX interface.</p>
|
|
144
|
*
|
|
145
|
* <p>Presented with an optional public identifier and a system
|
|
146
|
* identifier, this function attempts to locate a mapping in the
|
|
147
|
* catalogs.</p>
|
|
148
|
*
|
|
149
|
* <p>If such a mapping is found, the resolver attempts to open
|
|
150
|
* the mapped value as an InputSource and return it. Exceptions are
|
|
151
|
* ignored and null is returned if the mapped value cannot be opened
|
|
152
|
* as an input source.</p>
|
|
153
|
*
|
|
154
|
* If no mapping is found (or an error occurs attempting to open
|
|
155
|
* the mapped value as an input source), null is returned and the system
|
|
156
|
* will use the specified system identifier as if no entityResolver
|
|
157
|
* was specified.</p>
|
|
158
|
*
|
|
159
|
* @param publicId The public identifier for the entity in question.
|
|
160
|
* This may be null.
|
|
161
|
*
|
|
162
|
* @param systemId The system identifier for the entity in question.
|
|
163
|
* XML requires a system identifier on all external entities, so this
|
|
164
|
* value is always specified.
|
|
165
|
*
|
|
166
|
* @return An InputSource for the mapped identifier, or null.
|
|
167
|
*/
|
|
168
|
public InputSource resolveEntity (String publicId, String systemId) {
|
|
169
|
String resolved = null;
|
|
170
|
|
|
171
|
if (systemId != null) {
|
|
172
|
try {
|
|
173
|
resolved = catalog.resolveSystem(systemId);
|
|
174
|
} catch (MalformedURLException me) {
|
|
175
|
debug(1, "Malformed URL exception trying to resolve",
|
|
176
|
publicId);
|
|
177
|
resolved = null;
|
|
178
|
} catch (IOException ie) {
|
|
179
|
debug(1, "I/O exception trying to resolve", publicId);
|
|
180
|
resolved = null;
|
|
181
|
}
|
|
182
|
}
|
|
183
|
|
|
184
|
if (resolved == null) {
|
|
185
|
if (publicId != null) {
|
|
186
|
try {
|
|
187
|
resolved = catalog.resolvePublic(publicId, systemId);
|
|
188
|
} catch (MalformedURLException me) {
|
|
189
|
debug(1, "Malformed URL exception trying to resolve",
|
|
190
|
publicId);
|
|
191
|
} catch (IOException ie) {
|
|
192
|
debug(1, "I/O exception trying to resolve", publicId);
|
|
193
|
}
|
|
194
|
}
|
|
195
|
|
|
196
|
if (resolved != null) {
|
|
197
|
debug(2, "Resolved", publicId, resolved);
|
|
198
|
}
|
|
199
|
} else {
|
|
200
|
debug(2, "Resolved", systemId, resolved);
|
|
201
|
}
|
|
202
|
|
|
203
|
if (resolved == null && retryBadSystemIds
|
|
204
|
&& publicId != null && systemId != null) {
|
|
205
|
URL systemURL = null;
|
|
206
|
try {
|
|
207
|
systemURL = new URL(systemId);
|
|
208
|
} catch (MalformedURLException e) {
|
|
209
|
try {
|
|
210
|
systemURL = new URL("file:///" + systemId);
|
|
211
|
} catch (MalformedURLException e2) {
|
|
212
|
systemURL = null;
|
|
213
|
}
|
|
214
|
}
|
|
215
|
|
|
216
|
if (systemURL != null) {
|
|
217
|
try {
|
|
218
|
InputStream iStream = systemURL.openStream();
|
|
219
|
|
|
220
|
// The systemId can be opened, so that's the one that
|
|
221
|
// we'll use. There's no point making the caller open
|
|
222
|
// it again though...
|
|
223
|
|
|
224
|
InputSource iSource = new InputSource(systemId);
|
|
225
|
iSource.setPublicId(publicId);
|
|
226
|
iSource.setByteStream(iStream);
|
|
227
|
return iSource;
|
|
228
|
} catch (Exception e) {
|
|
229
|
// nop
|
|
230
|
}
|
|
231
|
}
|
|
232
|
|
|
233
|
// we got here, so it must be that the systemId cannot be
|
|
234
|
// opened and the caller wants us to retry...
|
|
235
|
debug(2, "Failed to open", systemId);
|
|
236
|
debug(2, "\tAttempting catalog lookup without system identifier.");
|
|
237
|
return resolveEntity(publicId, null);
|
|
238
|
}
|
|
239
|
|
|
240
|
if (resolved != null) {
|
|
241
|
try {
|
|
242
|
InputSource iSource = new InputSource(resolved);
|
|
243
|
iSource.setPublicId(publicId);
|
|
244
|
|
|
245
|
// Ideally this method would not attempt to open the
|
|
246
|
// InputStream, but there is a bug (in Xerces, at least)
|
|
247
|
// that causes the parser to mistakenly open the wrong
|
|
248
|
// system identifier if the returned InputSource does
|
|
249
|
// not have a byteStream.
|
|
250
|
//
|
|
251
|
// It could be argued that we still shouldn't do this here,
|
|
252
|
// but since the purpose of calling the entityResolver is
|
|
253
|
// almost certainly to open the input stream, it seems to
|
|
254
|
// do little harm.
|
|
255
|
//
|
|
256
|
URL url = new URL(resolved);
|
|
257
|
InputStream iStream = url.openStream();
|
|
258
|
iSource.setByteStream(iStream);
|
|
259
|
|
|
260
|
return iSource;
|
|
261
|
} catch (Exception e) {
|
|
262
|
debug(1, "Failed to create InputSource", resolved);
|
|
263
|
return null;
|
|
264
|
}
|
|
265
|
}
|
|
266
|
return null;
|
|
267
|
}
|
|
268
|
|
|
269
|
/**
|
|
270
|
* <p>Print debug message (if the debug level is high enough).</p>
|
|
271
|
*
|
|
272
|
* @param level The debug level of this message. This message
|
|
273
|
* will only be
|
|
274
|
* displayed if the current debug level is at least equal to this
|
|
275
|
* value.
|
|
276
|
* @param message The text of the message.
|
|
277
|
*/
|
|
278
|
private void debug(int level, String message) {
|
|
279
|
if (debug >= level) {
|
|
280
|
System.out.println(message);
|
|
281
|
}
|
|
282
|
}
|
|
283
|
|
|
284
|
/**
|
|
285
|
* <p>Print debug message (if the debug level is high enough).</p>
|
|
286
|
*
|
|
287
|
* @param level The debug level of this message. This message
|
|
288
|
* will only be
|
|
289
|
* displayed if the current debug level is at least equal to this
|
|
290
|
* value.
|
|
291
|
* @param message The text of the message.
|
|
292
|
* @param spec An argument to the message.
|
|
293
|
*/
|
|
294
|
private void debug(int level, String message, String spec) {
|
|
295
|
if (debug >= level) {
|
|
296
|
System.out.println(message + ": " + spec);
|
|
297
|
}
|
|
298
|
}
|
|
299
|
|
|
300
|
/**
|
|
301
|
* <p>Print debug message (if the debug level is high enough).</p>
|
|
302
|
*
|
|
303
|
* @param level The debug level of this message. This message
|
|
304
|
* will only be
|
|
305
|
* displayed if the current debug level is at least equal to this
|
|
306
|
* value.
|
|
307
|
* @param message The text of the message.
|
|
308
|
* @param spec1 An argument to the message.
|
|
309
|
* @param spec1 Another argument to the message.
|
|
310
|
*/
|
|
311
|
private void debug(int level, String message, String spec1, String spec2) {
|
|
312
|
if (debug >= level) {
|
|
313
|
System.out.println(message + ": " + spec1);
|
|
314
|
System.out.println("\t" + spec2);
|
|
315
|
}
|
|
316
|
}
|
|
317
|
}
|