1 /*
2 * reserved comment block
3 * DO NOT REMOVE OR ALTER!
4 */
5 /*
6 * Copyright 2003-2005 The Apache Software Foundation.
7 *
8 * Licensed under the Apache License, Version 2.0 (the "License");
9 * you may not use this file except in compliance with the License.
10 * You may obtain a copy of the License at
11 *
12 * http://www.apache.org/licenses/LICENSE-2.0
13 *
14 * Unless required by applicable law or agreed to in writing, software
15 * distributed under the License is distributed on an "AS IS" BASIS,
16 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
17 * See the License for the specific language governing permissions and
18 * limitations under the License.
19 */
20 package com.sun.org.apache.xerces.internal.xinclude;
21
22 import java.io.BufferedInputStream;
23 import java.io.IOException;
24 import java.io.InputStream;
25 import java.io.InputStreamReader;
26 import java.io.Reader;
27 import java.net.HttpURLConnection;
28 import java.net.URL;
29 import java.net.URLConnection;
30 import java.util.Iterator;
31 import java.util.Locale;
32 import java.util.Map;
33
34 import com.sun.org.apache.xerces.internal.impl.XMLEntityManager;
35 import com.sun.org.apache.xerces.internal.impl.XMLErrorReporter;
36 import com.sun.org.apache.xerces.internal.impl.io.ASCIIReader;
37 import com.sun.org.apache.xerces.internal.impl.io.UTF8Reader;
38 import com.sun.org.apache.xerces.internal.impl.msg.XMLMessageFormatter;
39 import com.sun.org.apache.xerces.internal.util.EncodingMap;
40 import com.sun.org.apache.xerces.internal.util.HTTPInputSource;
41 import com.sun.org.apache.xerces.internal.util.MessageFormatter;
42 import com.sun.org.apache.xerces.internal.util.XMLChar;
43 import com.sun.org.apache.xerces.internal.xni.XMLString;
44 import com.sun.org.apache.xerces.internal.xni.parser.XMLInputSource;
45
46 /**
47 * This class is used for reading resources requested in <include> elements,
48 * when the parse attribute of the <include> element is "text". Using this
49 * class will open the location, detect the encoding, and discard the byte order
50 * mark, if applicable.
51 *
52 * REVISIT:
53 * Much of the code in this class is taken from XMLEntityManager. It would be nice
54 * if this code could be shared in some way. However, since XMLEntityManager is used
55 * for reading files as XML, and this needs to read files as text, there would need
56 * to be some refactoring done.
57 *
58 * @author Michael Glavassevich, IBM
59 * @author Peter McCracken, IBM
60 * @author Ankit Pasricha, IBM
61 * @author Arun Yadav, Sun Microsystems Inc.
62 *
63 *
64 * @see XIncludeHandler
65 */
66 public class XIncludeTextReader {
67
68 private Reader fReader;
69 private XIncludeHandler fHandler;
70 private XMLInputSource fSource;
71 private XMLErrorReporter fErrorReporter;
72 private XMLString fTempString = new XMLString();
73
74 /**
75 * Construct the XIncludeReader using the XMLInputSource and XIncludeHandler.
76 *
77 * @param source The XMLInputSource to use.
78 * @param handler The XIncludeHandler to use.
79 * @param bufferSize The size of this text reader's buffer.
80 */
81 public XIncludeTextReader(XMLInputSource source, XIncludeHandler handler, int bufferSize)
82 throws IOException {
83 fHandler = handler;
84 fSource = source;
85 fTempString = new XMLString(new char[bufferSize + 1], 0, 0);
86 }
87
88 /**
89 * Sets the XMLErrorReporter used for reporting errors while
90 * reading the text include.
91 *
92 * @param errorReporter the XMLErrorReporter to be used for
93 * reporting errors.
94 */
95 public void setErrorReporter(XMLErrorReporter errorReporter) {
96 fErrorReporter = errorReporter;
97 }
98
99 /**
100 * Return the Reader for given XMLInputSource.
101 *
102 * @param source The XMLInputSource to use.
103 */
104 protected Reader getReader(XMLInputSource source) throws IOException {
105 if (source.getCharacterStream() != null) {
106 return source.getCharacterStream();
107 }
108 else {
109 InputStream stream = null;
110
111 String encoding = source.getEncoding();
112 if (encoding == null) {
113 encoding = "UTF-8";
114 }
115 if (source.getByteStream() != null) {
116 stream = source.getByteStream();
117 // Wrap the InputStream so that it is possible to rewind it.
118 if (!(stream instanceof BufferedInputStream)) {
119 stream = new BufferedInputStream(stream, fTempString.ch.length);
120 }
121 }
122 else {
123 String expandedSystemId = XMLEntityManager.expandSystemId(source.getSystemId(), source.getBaseSystemId(), false);
124
125 URL url = new URL(expandedSystemId);
126 URLConnection urlCon = url.openConnection();
127
128 // If this is an HTTP connection attach any request properties to the request.
129 if (urlCon instanceof HttpURLConnection && source instanceof HTTPInputSource) {
130 final HttpURLConnection urlConnection = (HttpURLConnection) urlCon;
131 final HTTPInputSource httpInputSource = (HTTPInputSource) source;
132
133 // set request properties
134 Iterator propIter = httpInputSource.getHTTPRequestProperties();
135 while (propIter.hasNext()) {
136 Map.Entry entry = (Map.Entry) propIter.next();
137 urlConnection.setRequestProperty((String) entry.getKey(), (String) entry.getValue());
138 }
139
140 // set preference for redirection
141 boolean followRedirects = httpInputSource.getFollowHTTPRedirects();
142 if (!followRedirects) {
143 XMLEntityManager.setInstanceFollowRedirects(urlConnection, followRedirects);
144 }
145 }
146
147 // Wrap the InputStream so that it is possible to rewind it.
148 stream = new BufferedInputStream(urlCon.getInputStream());
149
150 // content type will be string like "text/xml; charset=UTF-8" or "text/xml"
151 String rawContentType = urlCon.getContentType();
152
153 // text/xml and application/xml offer only one optional parameter
154 int index = (rawContentType != null) ? rawContentType.indexOf(';') : -1;
155
156 String contentType = null;
157 String charset = null;
158 if (index != -1) {
159 // this should be something like "text/xml"
160 contentType = rawContentType.substring(0, index).trim();
161
162 // this should be something like "charset=UTF-8", but we want to
163 // strip it down to just "UTF-8"
164 charset = rawContentType.substring(index + 1).trim();
165 if (charset.startsWith("charset=")) {
166 // 8 is the length of "charset="
167 charset = charset.substring(8).trim();
168 // strip quotes, if present
169 if ((charset.charAt(0) == '"'
170 && charset.charAt(charset.length() - 1) == '"')
171 || (charset.charAt(0) == '\''
172 && charset.charAt(charset.length() - 1)
173 == '\'')) {
174 charset =
175 charset.substring(1, charset.length() - 1);
176 }
177 }
178 else {
179 charset = null;
180 }
181 }
182 else {
183 contentType = rawContentType.trim();
184 }
185
186 String detectedEncoding = null;
187 /** The encoding of such a resource is determined by:
188 1 external encoding information, if available, otherwise
189 -- the most common type of external information is the "charset" parameter of a MIME package
190 2 if the media type of the resource is text/xml, application/xml, or matches the conventions text/*+xml or application/*+xml as described in XML Media Types [IETF RFC 3023], the encoding is recognized as specified in XML 1.0, otherwise
191 3 the value of the encoding attribute if one exists, otherwise
192 4 UTF-8.
193 **/
194 if (contentType.equals("text/xml")) {
195 if (charset != null) {
196 detectedEncoding = charset;
197 }
198 else {
199 // see RFC2376 or 3023, section 3.1
200 detectedEncoding = "US-ASCII";
201 }
202 }
203 else if (contentType.equals("application/xml")) {
204 if (charset != null) {
205 detectedEncoding = charset;
206 }
207 else {
208 // see RFC2376 or 3023, section 3.2
209 detectedEncoding = getEncodingName(stream);
210 }
211 }
212 else if (contentType.endsWith("+xml")) {
213 detectedEncoding = getEncodingName(stream);
214 }
215
216 if (detectedEncoding != null) {
217 encoding = detectedEncoding;
218 }
219 // else 3 or 4.
220 }
221
222 encoding = encoding.toUpperCase(Locale.ENGLISH);
223
224 // eat the Byte Order Mark
225 encoding = consumeBOM(stream, encoding);
226
227 // If the document is UTF-8 or US-ASCII use
228 // the Xerces readers for these encodings. For
229 // US-ASCII consult the encoding map since
230 // this encoding has many aliases.
231 if (encoding.equals("UTF-8")) {
232 return new UTF8Reader(stream,
233 fTempString.ch.length,
234 fErrorReporter.getMessageFormatter(XMLMessageFormatter.XML_DOMAIN),
235 fErrorReporter.getLocale() );
236 }
237
238 // Try to use a Java reader.
239 String javaEncoding = EncodingMap.getIANA2JavaMapping(encoding);
240
241 // If the specified encoding wasn't a recognized IANA encoding throw an IOException.
242 // The XIncludeHandler will report this as a ResourceError and then will
243 // attempt to include a fallback if there is one.
244 if (javaEncoding == null) {
245 MessageFormatter aFormatter =
246 fErrorReporter.getMessageFormatter(XMLMessageFormatter.XML_DOMAIN);
247 Locale aLocale = fErrorReporter.getLocale();
248 throw new IOException( aFormatter.formatMessage( aLocale,
249 "EncodingDeclInvalid",
250 new Object[] {encoding} ) );
251 }
252 else if (javaEncoding.equals("ASCII")) {
253 return new ASCIIReader(stream,
254 fTempString.ch.length,
255 fErrorReporter.getMessageFormatter(XMLMessageFormatter.XML_DOMAIN),
256 fErrorReporter.getLocale() );
257 }
258
259 return new InputStreamReader(stream, javaEncoding);
260 }
261 }
262
263 /**
264 * XMLEntityManager cares about endian-ness, since it creates its own optimized
265 * readers. Since we're just using generic Java readers for now, we're not caring
266 * about endian-ness. If this changes, even more code needs to be copied from
267 * XMLEntity manager. -- PJM
268 */
269 protected String getEncodingName(InputStream stream) throws IOException {
270 final byte[] b4 = new byte[4];
271 String encoding = null;
272
273 // this has the potential to throw an exception
274 // it will be fixed when we ensure the stream is rewindable (see note above)
275 stream.mark(4);
276 int count = stream.read(b4, 0, 4);
277 stream.reset();
278 if (count == 4) {
279 encoding = getEncodingName(b4);
280 }
281
282 return encoding;
283 }
284
285 /**
286 * Removes the byte order mark from the stream, if
287 * it exists and returns the encoding name.
288 *
289 * @param stream
290 * @param encoding
291 * @throws IOException
292 */
293 protected String consumeBOM(InputStream stream, String encoding)
294 throws IOException {
295
296 byte[] b = new byte[3];
297 int count = 0;
298 stream.mark(3);
299 if (encoding.equals("UTF-8")) {
300 count = stream.read(b, 0, 3);
301 if (count == 3) {
302 final int b0 = b[0] & 0xFF;
303 final int b1 = b[1] & 0xFF;
304 final int b2 = b[2] & 0xFF;
305 if (b0 != 0xEF || b1 != 0xBB || b2 != 0xBF) {
306 // First three bytes are not BOM, so reset.
307 stream.reset();
308 }
309 }
310 else {
311 stream.reset();
312 }
313 }
314 else if (encoding.startsWith("UTF-16")) {
315 count = stream.read(b, 0, 2);
316 if (count == 2) {
317 final int b0 = b[0] & 0xFF;
318 final int b1 = b[1] & 0xFF;
319 if (b0 == 0xFE && b1 == 0xFF) {
320 return "UTF-16BE";
321 }
322 else if (b0 == 0xFF && b1 == 0xFE) {
323 return "UTF-16LE";
324 }
325 }
326 // First two bytes are not BOM, so reset.
327 stream.reset();
328 }
329 // We could do UTF-32, but since the getEncodingName() doesn't support that
330 // we won't support it here.
331 // To implement UTF-32, look for: 00 00 FE FF for big-endian
332 // or FF FE 00 00 for little-endian
333 return encoding;
334 }
335
336 /**
337 * REVISIT: This code is taken from com.sun.org.apache.xerces.internal.impl.XMLEntityManager.
338 * Is there any way we can share the code, without having it implemented twice?
339 * I think we should make it public and static in XMLEntityManager. --PJM
340 *
341 * Returns the IANA encoding name that is auto-detected from
342 * the bytes specified, with the endian-ness of that encoding where appropriate.
343 *
344 * @param b4 The first four bytes of the input.
345 * @return the encoding name, or null if no encoding could be detected
346 */
347 protected String getEncodingName(byte[] b4) {
348
349 // UTF-16, with BOM
350 int b0 = b4[0] & 0xFF;
351 int b1 = b4[1] & 0xFF;
352 if (b0 == 0xFE && b1 == 0xFF) {
353 // UTF-16, big-endian
354 return "UTF-16BE";
355 }
356 if (b0 == 0xFF && b1 == 0xFE) {
357 // UTF-16, little-endian
358 return "UTF-16LE";
359 }
360
361 // UTF-8 with a BOM
362 int b2 = b4[2] & 0xFF;
363 if (b0 == 0xEF && b1 == 0xBB && b2 == 0xBF) {
364 return "UTF-8";
365 }
366
367 // other encodings
368 int b3 = b4[3] & 0xFF;
369 if (b0 == 0x00 && b1 == 0x00 && b2 == 0x00 && b3 == 0x3C) {
370 // UCS-4, big endian (1234)
371 return "ISO-10646-UCS-4";
372 }
373 if (b0 == 0x3C && b1 == 0x00 && b2 == 0x00 && b3 == 0x00) {
374 // UCS-4, little endian (4321)
375 return "ISO-10646-UCS-4";
376 }
377 if (b0 == 0x00 && b1 == 0x00 && b2 == 0x3C && b3 == 0x00) {
378 // UCS-4, unusual octet order (2143)
379 return "ISO-10646-UCS-4";
380 }
381 if (b0 == 0x00 && b1 == 0x3C && b2 == 0x00 && b3 == 0x00) {
382 // UCS-4, unusual octect order (3412)
383 return "ISO-10646-UCS-4";
384 }
385 if (b0 == 0x00 && b1 == 0x3C && b2 == 0x00 && b3 == 0x3F) {
386 // UTF-16, big-endian, no BOM
387 // (or could turn out to be UCS-2...
388 return "UTF-16BE";
389 }
390 if (b0 == 0x3C && b1 == 0x00 && b2 == 0x3F && b3 == 0x00) {
391 // UTF-16, little-endian, no BOM
392 // (or could turn out to be UCS-2...
393 return "UTF-16LE";
394 }
395 if (b0 == 0x4C && b1 == 0x6F && b2 == 0xA7 && b3 == 0x94) {
396 // EBCDIC
397 // a la xerces1, return CP037 instead of EBCDIC here
398 return "CP037";
399 }
400
401 // this signals us to use the value from the encoding attribute
402 return null;
403
404 } // getEncodingName(byte[]):Object[]
405
406 /**
407 * Read the input stream as text, and pass the text on to the XIncludeHandler
408 * using calls to characters(). This will read all of the text it can from the
409 * resource.
410 *
411 * @throws IOException
412 */
413 public void parse() throws IOException {
414
415 fReader = getReader(fSource);
416 fSource = null;
417 int readSize = fReader.read(fTempString.ch, 0, fTempString.ch.length - 1);
418 while (readSize != -1) {
419 for (int i = 0; i < readSize; ++i) {
420 char ch = fTempString.ch[i];
421 if (!isValid(ch)) {
422 if (XMLChar.isHighSurrogate(ch)) {
423 int ch2;
424 // retrieve next character
425 if (++i < readSize) {
426 ch2 = fTempString.ch[i];
427 }
428 // handle rare boundary case
429 else {
430 ch2 = fReader.read();
431 if (ch2 != -1) {
432 fTempString.ch[readSize++] = (char) ch2;
433 }
434 }
435 if (XMLChar.isLowSurrogate(ch2)) {
436 // convert surrogates to a supplemental character
437 int sup = XMLChar.supplemental(ch, (char)ch2);
438 if (!isValid(sup)) {
439 fErrorReporter.reportError(XMLMessageFormatter.XML_DOMAIN,
440 "InvalidCharInContent",
441 new Object[] { Integer.toString(sup, 16) },
442 XMLErrorReporter.SEVERITY_FATAL_ERROR);
443 }
444 }
445 else {
446 fErrorReporter.reportError(XMLMessageFormatter.XML_DOMAIN,
447 "InvalidCharInContent",
448 new Object[] { Integer.toString(ch2, 16) },
449 XMLErrorReporter.SEVERITY_FATAL_ERROR);
450 }
451 }
452 else {
453 fErrorReporter.reportError(XMLMessageFormatter.XML_DOMAIN,
454 "InvalidCharInContent",
455 new Object[] { Integer.toString(ch, 16) },
456 XMLErrorReporter.SEVERITY_FATAL_ERROR);
457 }
458 }
459 }
460 if (fHandler != null && readSize > 0) {
461 fTempString.offset = 0;
462 fTempString.length = readSize;
463 fHandler.characters(
464 fTempString,
465 fHandler.modifyAugmentations(null, true));
466 }
467 readSize = fReader.read(fTempString.ch, 0, fTempString.ch.length - 1);
468 }
469
470 }
471
472 /**
473 * Sets the input source on this text reader.
474 *
475 * @param source The XMLInputSource to use.
476 */
477 public void setInputSource(XMLInputSource source) {
478 fSource = source;
479 }
480
481 /**
482 * Closes the stream. Call this after parse(), or when there is no longer any need
483 * for this object.
484 *
485 * @throws IOException
486 */
487 public void close() throws IOException {
488 if (fReader != null) {
489 fReader.close();
490 fReader = null;
491 }
492 }
493
494 /**
495 * Returns true if the specified character is a valid XML character
496 * as per the rules of XML 1.0.
497 *
498 * @param ch The character to check.
499 */
500 protected boolean isValid(int ch) {
501 return XMLChar.isValid(ch);
502 }
503
504 /**
505 * Sets the buffer size property for the reader which decides the chunk sizes that are parsed
506 * by the reader at a time and passed to the handler
507 *
508 * @param bufferSize The size of the buffer desired
509 */
510 protected void setBufferSize(int bufferSize) {
511 if (fTempString.ch.length != ++bufferSize) {
512 fTempString.ch = new char[bufferSize];
513 }
514 }
515
516 }