1 /*
2 * Copyright (c) 1995, 2016, Oracle and/or its affiliates. All rights reserved.
3 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
4 *
5 * This code is free software; you can redistribute it and/or modify it
6 * under the terms of the GNU General Public License version 2 only, as
7 * published by the Free Software Foundation. Oracle designates this
8 * particular file as subject to the "Classpath" exception as provided
9 * by Oracle in the LICENSE file that accompanied this code.
10 *
11 * This code is distributed in the hope that it will be useful, but WITHOUT
12 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
13 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
14 * version 2 for more details (a copy is included in the LICENSE file that
15 * accompanied this code).
16 *
17 * You should have received a copy of the GNU General Public License version
18 * 2 along with this work; if not, write to the Free Software Foundation,
19 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
20 *
21 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
22 * or visit www.oracle.com if you need additional information or have any
23 * questions.
24 */
25
26 package java.net;
27
28 import java.io.IOException;
29 import java.io.InputStream;
30 import java.io.OutputStream;
31 import java.security.PrivilegedAction;
32 import java.util.Hashtable;
33 import java.util.Date;
34 import java.util.Iterator;
35 import java.util.Objects;
36 import java.util.ServiceConfigurationError;
37 import java.util.ServiceLoader;
38 import java.util.StringTokenizer;
39 import java.util.Collections;
40 import java.util.Map;
41 import java.util.List;
42 import java.security.Permission;
43 import java.security.AccessController;
44 import sun.security.util.SecurityConstants;
45 import sun.net.www.MessageHeader;
46 import sun.security.action.GetPropertyAction;
47
48 /**
49 * The abstract class {@code URLConnection} is the superclass
50 * of all classes that represent a communications link between the
51 * application and a URL. Instances of this class can be used both to
52 * read from and to write to the resource referenced by the URL. In
53 * general, creating a connection to a URL is a multistep process:
54 *
55 * <center><table border=2 summary="Describes the process of creating a connection to a URL: openConnection() and connect() over time.">
56 * <tr><th>{@code openConnection()}</th>
57 * <th>{@code connect()}</th></tr>
58 * <tr><td>Manipulate parameters that affect the connection to the remote
59 * resource.</td>
60 * <td>Interact with the resource; query header fields and
61 * contents.</td></tr>
62 * </table>
63 * ---------------------------->
64 * <br>time</center>
65 *
66 * <ol>
67 * <li>The connection object is created by invoking the
68 * {@code openConnection} method on a URL.
69 * <li>The setup parameters and general request properties are manipulated.
70 * <li>The actual connection to the remote object is made, using the
71 * {@code connect} method.
72 * <li>The remote object becomes available. The header fields and the contents
73 * of the remote object can be accessed.
74 * </ol>
75 * <p>
76 * The setup parameters are modified using the following methods:
77 * <ul>
78 * <li>{@code setAllowUserInteraction}
79 * <li>{@code setDoInput}
80 * <li>{@code setDoOutput}
81 * <li>{@code setIfModifiedSince}
82 * <li>{@code setUseCaches}
83 * </ul>
84 * <p>
85 * and the general request properties are modified using the method:
86 * <ul>
87 * <li>{@code setRequestProperty}
88 * </ul>
89 * <p>
90 * Default values for the {@code AllowUserInteraction} and
91 * {@code UseCaches} parameters can be set using the methods
92 * {@code setDefaultAllowUserInteraction} and
93 * {@code setDefaultUseCaches}.
94 * <p>
95 * Each of the above {@code set} methods has a corresponding
96 * {@code get} method to retrieve the value of the parameter or
97 * general request property. The specific parameters and general
98 * request properties that are applicable are protocol specific.
99 * <p>
100 * The following methods are used to access the header fields and
101 * the contents after the connection is made to the remote object:
102 * <ul>
103 * <li>{@code getContent}
104 * <li>{@code getHeaderField}
105 * <li>{@code getInputStream}
106 * <li>{@code getOutputStream}
107 * </ul>
108 * <p>
109 * Certain header fields are accessed frequently. The methods:
110 * <ul>
111 * <li>{@code getContentEncoding}
112 * <li>{@code getContentLength}
113 * <li>{@code getContentType}
114 * <li>{@code getDate}
115 * <li>{@code getExpiration}
116 * <li>{@code getLastModified}
117 * </ul>
118 * <p>
119 * provide convenient access to these fields. The
120 * {@code getContentType} method is used by the
121 * {@code getContent} method to determine the type of the remote
122 * object; subclasses may find it convenient to override the
123 * {@code getContentType} method.
124 * <p>
125 * In the common case, all of the pre-connection parameters and
126 * general request properties can be ignored: the pre-connection
127 * parameters and request properties default to sensible values. For
128 * most clients of this interface, there are only two interesting
129 * methods: {@code getInputStream} and {@code getContent},
130 * which are mirrored in the {@code URL} class by convenience methods.
131 * <p>
132 * More information on the request properties and header fields of
133 * an {@code http} connection can be found at:
134 * <blockquote><pre>
135 * <a href="http://www.ietf.org/rfc/rfc2616.txt">http://www.ietf.org/rfc/rfc2616.txt</a>
136 * </pre></blockquote>
137 *
138 * Invoking the {@code close()} methods on the {@code InputStream} or {@code OutputStream} of an
139 * {@code URLConnection} after a request may free network resources associated with this
140 * instance, unless particular protocol specifications specify different behaviours
141 * for it.
142 *
143 * @author James Gosling
144 * @see java.net.URL#openConnection()
145 * @see java.net.URLConnection#connect()
146 * @see java.net.URLConnection#getContent()
147 * @see java.net.URLConnection#getContentEncoding()
148 * @see java.net.URLConnection#getContentLength()
149 * @see java.net.URLConnection#getContentType()
150 * @see java.net.URLConnection#getDate()
151 * @see java.net.URLConnection#getExpiration()
152 * @see java.net.URLConnection#getHeaderField(int)
153 * @see java.net.URLConnection#getHeaderField(java.lang.String)
154 * @see java.net.URLConnection#getInputStream()
155 * @see java.net.URLConnection#getLastModified()
156 * @see java.net.URLConnection#getOutputStream()
157 * @see java.net.URLConnection#setAllowUserInteraction(boolean)
158 * @see java.net.URLConnection#setDefaultUseCaches(boolean)
159 * @see java.net.URLConnection#setDoInput(boolean)
160 * @see java.net.URLConnection#setDoOutput(boolean)
161 * @see java.net.URLConnection#setIfModifiedSince(long)
162 * @see java.net.URLConnection#setRequestProperty(java.lang.String, java.lang.String)
163 * @see java.net.URLConnection#setUseCaches(boolean)
164 * @since 1.0
165 */
166 public abstract class URLConnection {
167
168 /**
169 * The URL represents the remote object on the World Wide Web to
170 * which this connection is opened.
171 * <p>
172 * The value of this field can be accessed by the
173 * {@code getURL} method.
174 * <p>
175 * The default value of this variable is the value of the URL
176 * argument in the {@code URLConnection} constructor.
177 *
178 * @see java.net.URLConnection#getURL()
179 * @see java.net.URLConnection#url
180 */
181 protected URL url;
182
183 /**
184 * This variable is set by the {@code setDoInput} method. Its
185 * value is returned by the {@code getDoInput} method.
186 * <p>
187 * A URL connection can be used for input and/or output. Setting the
188 * {@code doInput} flag to {@code true} indicates that
189 * the application intends to read data from the URL connection.
190 * <p>
191 * The default value of this field is {@code true}.
192 *
193 * @see java.net.URLConnection#getDoInput()
194 * @see java.net.URLConnection#setDoInput(boolean)
195 */
196 protected boolean doInput = true;
197
198 /**
199 * This variable is set by the {@code setDoOutput} method. Its
200 * value is returned by the {@code getDoOutput} method.
201 * <p>
202 * A URL connection can be used for input and/or output. Setting the
203 * {@code doOutput} flag to {@code true} indicates
204 * that the application intends to write data to the URL connection.
205 * <p>
206 * The default value of this field is {@code false}.
207 *
208 * @see java.net.URLConnection#getDoOutput()
209 * @see java.net.URLConnection#setDoOutput(boolean)
210 */
211 protected boolean doOutput = false;
212
213 private static boolean defaultAllowUserInteraction = false;
214
215 /**
216 * If {@code true}, this {@code URL} is being examined in
217 * a context in which it makes sense to allow user interactions such
218 * as popping up an authentication dialog. If {@code false},
219 * then no user interaction is allowed.
220 * <p>
221 * The value of this field can be set by the
222 * {@code setAllowUserInteraction} method.
223 * Its value is returned by the
224 * {@code getAllowUserInteraction} method.
225 * Its default value is the value of the argument in the last invocation
226 * of the {@code setDefaultAllowUserInteraction} method.
227 *
228 * @see java.net.URLConnection#getAllowUserInteraction()
229 * @see java.net.URLConnection#setAllowUserInteraction(boolean)
230 * @see java.net.URLConnection#setDefaultAllowUserInteraction(boolean)
231 */
232 protected boolean allowUserInteraction = defaultAllowUserInteraction;
233
234 private static boolean defaultUseCaches = true;
235
236 /**
237 * If {@code true}, the protocol is allowed to use caching
238 * whenever it can. If {@code false}, the protocol must always
239 * try to get a fresh copy of the object.
240 * <p>
241 * This field is set by the {@code setUseCaches} method. Its
242 * value is returned by the {@code getUseCaches} method.
243 * <p>
244 * Its default value is the value given in the last invocation of the
245 * {@code setDefaultUseCaches} method.
246 *
247 * @see java.net.URLConnection#setUseCaches(boolean)
248 * @see java.net.URLConnection#getUseCaches()
249 * @see java.net.URLConnection#setDefaultUseCaches(boolean)
250 */
251 protected boolean useCaches = defaultUseCaches;
252
253 /**
254 * Some protocols support skipping the fetching of the object unless
255 * the object has been modified more recently than a certain time.
256 * <p>
257 * A nonzero value gives a time as the number of milliseconds since
258 * January 1, 1970, GMT. The object is fetched only if it has been
259 * modified more recently than that time.
260 * <p>
261 * This variable is set by the {@code setIfModifiedSince}
262 * method. Its value is returned by the
263 * {@code getIfModifiedSince} method.
264 * <p>
265 * The default value of this field is {@code 0}, indicating
266 * that the fetching must always occur.
267 *
268 * @see java.net.URLConnection#getIfModifiedSince()
269 * @see java.net.URLConnection#setIfModifiedSince(long)
270 */
271 protected long ifModifiedSince = 0;
272
273 /**
274 * If {@code false}, this connection object has not created a
275 * communications link to the specified URL. If {@code true},
276 * the communications link has been established.
277 */
278 protected boolean connected = false;
279
280 /**
281 * @since 1.5
282 */
283 private int connectTimeout;
284 private int readTimeout;
285
286 /**
287 * @since 1.6
288 */
289 private MessageHeader requests;
290
291 /**
292 * @since 1.1
293 */
294 private static FileNameMap fileNameMap;
295
296 /**
297 * @since 1.2.2
298 */
299 private static boolean fileNameMapLoaded = false;
300
301 /**
302 * Loads filename map (a mimetable) from a data file. It will
303 * first try to load the user-specific table, defined
304 * by "content.types.user.table" property. If that fails,
305 * it tries to load the default built-in table.
306 *
307 * @return the FileNameMap
308 * @since 1.2
309 * @see #setFileNameMap(java.net.FileNameMap)
310 */
311 public static synchronized FileNameMap getFileNameMap() {
312 if ((fileNameMap == null) && !fileNameMapLoaded) {
313 fileNameMap = sun.net.www.MimeTable.loadTable();
314 fileNameMapLoaded = true;
315 }
316
317 return new FileNameMap() {
318 private FileNameMap map = fileNameMap;
319 public String getContentTypeFor(String fileName) {
320 return map.getContentTypeFor(fileName);
321 }
322 };
323 }
324
325 /**
326 * Sets the FileNameMap.
327 * <p>
328 * If there is a security manager, this method first calls
329 * the security manager's {@code checkSetFactory} method
330 * to ensure the operation is allowed.
331 * This could result in a SecurityException.
332 *
333 * @param map the FileNameMap to be set
334 * @exception SecurityException if a security manager exists and its
335 * {@code checkSetFactory} method doesn't allow the operation.
336 * @see SecurityManager#checkSetFactory
337 * @see #getFileNameMap()
338 * @since 1.2
339 */
340 public static void setFileNameMap(FileNameMap map) {
341 SecurityManager sm = System.getSecurityManager();
342 if (sm != null) sm.checkSetFactory();
343 fileNameMap = map;
344 }
345
346 /**
347 * Opens a communications link to the resource referenced by this
348 * URL, if such a connection has not already been established.
349 * <p>
350 * If the {@code connect} method is called when the connection
351 * has already been opened (indicated by the {@code connected}
352 * field having the value {@code true}), the call is ignored.
353 * <p>
354 * URLConnection objects go through two phases: first they are
355 * created, then they are connected. After being created, and
356 * before being connected, various options can be specified
357 * (e.g., doInput and UseCaches). After connecting, it is an
358 * error to try to set them. Operations that depend on being
359 * connected, like getContentLength, will implicitly perform the
360 * connection, if necessary.
361 *
362 * @throws SocketTimeoutException if the timeout expires before
363 * the connection can be established
364 * @exception IOException if an I/O error occurs while opening the
365 * connection.
366 * @see java.net.URLConnection#connected
367 * @see #getConnectTimeout()
368 * @see #setConnectTimeout(int)
369 */
370 public abstract void connect() throws IOException;
371
372 /**
373 * Sets a specified timeout value, in milliseconds, to be used
374 * when opening a communications link to the resource referenced
375 * by this URLConnection. If the timeout expires before the
376 * connection can be established, a
377 * java.net.SocketTimeoutException is raised. A timeout of zero is
378 * interpreted as an infinite timeout.
379
380 * <p> Some non-standard implementation of this method may ignore
381 * the specified timeout. To see the connect timeout set, please
382 * call getConnectTimeout().
383 *
384 * @param timeout an {@code int} that specifies the connect
385 * timeout value in milliseconds
386 * @throws IllegalArgumentException if the timeout parameter is negative
387 *
388 * @see #getConnectTimeout()
389 * @see #connect()
390 * @since 1.5
391 */
392 public void setConnectTimeout(int timeout) {
393 if (timeout < 0) {
394 throw new IllegalArgumentException("timeout can not be negative");
395 }
396 connectTimeout = timeout;
397 }
398
399 /**
400 * Returns setting for connect timeout.
401 * <p>
402 * 0 return implies that the option is disabled
403 * (i.e., timeout of infinity).
404 *
405 * @return an {@code int} that indicates the connect timeout
406 * value in milliseconds
407 * @see #setConnectTimeout(int)
408 * @see #connect()
409 * @since 1.5
410 */
411 public int getConnectTimeout() {
412 return connectTimeout;
413 }
414
415 /**
416 * Sets the read timeout to a specified timeout, in
417 * milliseconds. A non-zero value specifies the timeout when
418 * reading from Input stream when a connection is established to a
419 * resource. If the timeout expires before there is data available
420 * for read, a java.net.SocketTimeoutException is raised. A
421 * timeout of zero is interpreted as an infinite timeout.
422 *
423 *<p> Some non-standard implementation of this method ignores the
424 * specified timeout. To see the read timeout set, please call
425 * getReadTimeout().
426 *
427 * @param timeout an {@code int} that specifies the timeout
428 * value to be used in milliseconds
429 * @throws IllegalArgumentException if the timeout parameter is negative
430 *
431 * @see #getReadTimeout()
432 * @see InputStream#read()
433 * @since 1.5
434 */
435 public void setReadTimeout(int timeout) {
436 if (timeout < 0) {
437 throw new IllegalArgumentException("timeout can not be negative");
438 }
439 readTimeout = timeout;
440 }
441
442 /**
443 * Returns setting for read timeout. 0 return implies that the
444 * option is disabled (i.e., timeout of infinity).
445 *
446 * @return an {@code int} that indicates the read timeout
447 * value in milliseconds
448 *
449 * @see #setReadTimeout(int)
450 * @see InputStream#read()
451 * @since 1.5
452 */
453 public int getReadTimeout() {
454 return readTimeout;
455 }
456
457 /**
458 * Constructs a URL connection to the specified URL. A connection to
459 * the object referenced by the URL is not created.
460 *
461 * @param url the specified URL.
462 */
463 protected URLConnection(URL url) {
464 this.url = url;
465 }
466
467 /**
468 * Returns the value of this {@code URLConnection}'s {@code URL}
469 * field.
470 *
471 * @return the value of this {@code URLConnection}'s {@code URL}
472 * field.
473 * @see java.net.URLConnection#url
474 */
475 public URL getURL() {
476 return url;
477 }
478
479 /**
480 * Returns the value of the {@code content-length} header field.
481 * <P>
482 * <B>Note</B>: {@link #getContentLengthLong() getContentLengthLong()}
483 * should be preferred over this method, since it returns a {@code long}
484 * instead and is therefore more portable.</P>
485 *
486 * @return the content length of the resource that this connection's URL
487 * references, {@code -1} if the content length is not known,
488 * or if the content length is greater than Integer.MAX_VALUE.
489 */
490 public int getContentLength() {
491 long l = getContentLengthLong();
492 if (l > Integer.MAX_VALUE)
493 return -1;
494 return (int) l;
495 }
496
497 /**
498 * Returns the value of the {@code content-length} header field as a
499 * long.
500 *
501 * @return the content length of the resource that this connection's URL
502 * references, or {@code -1} if the content length is
503 * not known.
504 * @since 1.7
505 */
506 public long getContentLengthLong() {
507 return getHeaderFieldLong("content-length", -1);
508 }
509
510 /**
511 * Returns the value of the {@code content-type} header field.
512 *
513 * @return the content type of the resource that the URL references,
514 * or {@code null} if not known.
515 * @see java.net.URLConnection#getHeaderField(java.lang.String)
516 */
517 public String getContentType() {
518 return getHeaderField("content-type");
519 }
520
521 /**
522 * Returns the value of the {@code content-encoding} header field.
523 *
524 * @return the content encoding of the resource that the URL references,
525 * or {@code null} if not known.
526 * @see java.net.URLConnection#getHeaderField(java.lang.String)
527 */
528 public String getContentEncoding() {
529 return getHeaderField("content-encoding");
530 }
531
532 /**
533 * Returns the value of the {@code expires} header field.
534 *
535 * @return the expiration date of the resource that this URL references,
536 * or 0 if not known. The value is the number of milliseconds since
537 * January 1, 1970 GMT.
538 * @see java.net.URLConnection#getHeaderField(java.lang.String)
539 */
540 public long getExpiration() {
541 return getHeaderFieldDate("expires", 0);
542 }
543
544 /**
545 * Returns the value of the {@code date} header field.
546 *
547 * @return the sending date of the resource that the URL references,
548 * or {@code 0} if not known. The value returned is the
549 * number of milliseconds since January 1, 1970 GMT.
550 * @see java.net.URLConnection#getHeaderField(java.lang.String)
551 */
552 public long getDate() {
553 return getHeaderFieldDate("date", 0);
554 }
555
556 /**
557 * Returns the value of the {@code last-modified} header field.
558 * The result is the number of milliseconds since January 1, 1970 GMT.
559 *
560 * @return the date the resource referenced by this
561 * {@code URLConnection} was last modified, or 0 if not known.
562 * @see java.net.URLConnection#getHeaderField(java.lang.String)
563 */
564 public long getLastModified() {
565 return getHeaderFieldDate("last-modified", 0);
566 }
567
568 /**
569 * Returns the value of the named header field.
570 * <p>
571 * If called on a connection that sets the same header multiple times
572 * with possibly different values, only the last value is returned.
573 *
574 *
575 * @param name the name of a header field.
576 * @return the value of the named header field, or {@code null}
577 * if there is no such field in the header.
578 */
579 public String getHeaderField(String name) {
580 return null;
581 }
582
583 /**
584 * Returns an unmodifiable Map of the header fields.
585 * The Map keys are Strings that represent the
586 * response-header field names. Each Map value is an
587 * unmodifiable List of Strings that represents
588 * the corresponding field values.
589 *
590 * @return a Map of header fields
591 * @since 1.4
592 */
593 public Map<String,List<String>> getHeaderFields() {
594 return Collections.emptyMap();
595 }
596
597 /**
598 * Returns the value of the named field parsed as a number.
599 * <p>
600 * This form of {@code getHeaderField} exists because some
601 * connection types (e.g., {@code http-ng}) have pre-parsed
602 * headers. Classes for that connection type can override this method
603 * and short-circuit the parsing.
604 *
605 * @param name the name of the header field.
606 * @param Default the default value.
607 * @return the value of the named field, parsed as an integer. The
608 * {@code Default} value is returned if the field is
609 * missing or malformed.
610 */
611 public int getHeaderFieldInt(String name, int Default) {
612 String value = getHeaderField(name);
613 try {
614 return Integer.parseInt(value);
615 } catch (Exception e) { }
616 return Default;
617 }
618
619 /**
620 * Returns the value of the named field parsed as a number.
621 * <p>
622 * This form of {@code getHeaderField} exists because some
623 * connection types (e.g., {@code http-ng}) have pre-parsed
624 * headers. Classes for that connection type can override this method
625 * and short-circuit the parsing.
626 *
627 * @param name the name of the header field.
628 * @param Default the default value.
629 * @return the value of the named field, parsed as a long. The
630 * {@code Default} value is returned if the field is
631 * missing or malformed.
632 * @since 1.7
633 */
634 public long getHeaderFieldLong(String name, long Default) {
635 String value = getHeaderField(name);
636 try {
637 return Long.parseLong(value);
638 } catch (Exception e) { }
639 return Default;
640 }
641
642 /**
643 * Returns the value of the named field parsed as date.
644 * The result is the number of milliseconds since January 1, 1970 GMT
645 * represented by the named field.
646 * <p>
647 * This form of {@code getHeaderField} exists because some
648 * connection types (e.g., {@code http-ng}) have pre-parsed
649 * headers. Classes for that connection type can override this method
650 * and short-circuit the parsing.
651 *
652 * @param name the name of the header field.
653 * @param Default a default value.
654 * @return the value of the field, parsed as a date. The value of the
655 * {@code Default} argument is returned if the field is
656 * missing or malformed.
657 */
658 @SuppressWarnings("deprecation")
659 public long getHeaderFieldDate(String name, long Default) {
660 String value = getHeaderField(name);
661 try {
662 return Date.parse(value);
663 } catch (Exception e) { }
664 return Default;
665 }
666
667 /**
668 * Returns the key for the {@code n}<sup>th</sup> header field.
669 * It returns {@code null} if there are fewer than {@code n+1} fields.
670 *
671 * @param n an index, where {@code n>=0}
672 * @return the key for the {@code n}<sup>th</sup> header field,
673 * or {@code null} if there are fewer than {@code n+1}
674 * fields.
675 */
676 public String getHeaderFieldKey(int n) {
677 return null;
678 }
679
680 /**
681 * Returns the value for the {@code n}<sup>th</sup> header field.
682 * It returns {@code null} if there are fewer than
683 * {@code n+1}fields.
684 * <p>
685 * This method can be used in conjunction with the
686 * {@link #getHeaderFieldKey(int) getHeaderFieldKey} method to iterate through all
687 * the headers in the message.
688 *
689 * @param n an index, where {@code n>=0}
690 * @return the value of the {@code n}<sup>th</sup> header field
691 * or {@code null} if there are fewer than {@code n+1} fields
692 * @see java.net.URLConnection#getHeaderFieldKey(int)
693 */
694 public String getHeaderField(int n) {
695 return null;
696 }
697
698 /**
699 * Retrieves the contents of this URL connection.
700 * <p>
701 * This method first determines the content type of the object by
702 * calling the {@code getContentType} method. If this is
703 * the first time that the application has seen that specific content
704 * type, a content handler for that content type is created.
705 * <p> This is done as follows:
706 * <ol>
707 * <li>If the application has set up a content handler factory instance
708 * using the {@code setContentHandlerFactory} method, the
709 * {@code createContentHandler} method of that instance is called
710 * with the content type as an argument; the result is a content
711 * handler for that content type.
712 * <li>If no {@code ContentHandlerFactory} has yet been set up,
713 * or if the factory's {@code createContentHandler} method
714 * returns {@code null}, then the {@linkplain java.util.ServiceLoader
715 * ServiceLoader} mechanism is used to locate {@linkplain
716 * java.net.ContentHandlerFactory ContentHandlerFactory}
717 * implementations using the system class
718 * loader. The order that factories are located is implementation
719 * specific, and an implementation is free to cache the located
720 * factories. A {@linkplain java.util.ServiceConfigurationError
721 * ServiceConfigurationError}, {@code Error} or {@code RuntimeException}
722 * thrown from the {@code createContentHandler}, if encountered, will
723 * be propagated to the calling thread. The {@code
724 * createContentHandler} method of each factory, if instantiated, is
725 * invoked, with the content type, until a factory returns non-null,
726 * or all factories have been exhausted.
727 * <li>Failing that, this method tries to load a content handler
728 * class as defined by {@link java.net.ContentHandler ContentHandler}.
729 * If the class does not exist, or is not a subclass of {@code
730 * ContentHandler}, then an {@code UnknownServiceException} is thrown.
731 * </ol>
732 *
733 * @return the object fetched. The {@code instanceof} operator
734 * should be used to determine the specific kind of object
735 * returned.
736 * @exception IOException if an I/O error occurs while
737 * getting the content.
738 * @exception UnknownServiceException if the protocol does not support
739 * the content type.
740 * @see java.net.ContentHandlerFactory#createContentHandler(java.lang.String)
741 * @see java.net.URLConnection#getContentType()
742 * @see java.net.URLConnection#setContentHandlerFactory(java.net.ContentHandlerFactory)
743 */
744 public Object getContent() throws IOException {
745 // Must call getInputStream before GetHeaderField gets called
746 // so that FileNotFoundException has a chance to be thrown up
747 // from here without being caught.
748 getInputStream();
749 return getContentHandler().getContent(this);
750 }
751
752 /**
753 * Retrieves the contents of this URL connection.
754 *
755 * @param classes the {@code Class} array
756 * indicating the requested types
757 * @return the object fetched that is the first match of the type
758 * specified in the classes array. null if none of
759 * the requested types are supported.
760 * The {@code instanceof} operator should be used to
761 * determine the specific kind of object returned.
762 * @exception IOException if an I/O error occurs while
763 * getting the content.
764 * @exception UnknownServiceException if the protocol does not support
765 * the content type.
766 * @see java.net.URLConnection#getContent()
767 * @see java.net.ContentHandlerFactory#createContentHandler(java.lang.String)
768 * @see java.net.URLConnection#getContent(java.lang.Class[])
769 * @see java.net.URLConnection#setContentHandlerFactory(java.net.ContentHandlerFactory)
770 * @since 1.3
771 */
772 public Object getContent(Class<?>[] classes) throws IOException {
773 // Must call getInputStream before GetHeaderField gets called
774 // so that FileNotFoundException has a chance to be thrown up
775 // from here without being caught.
776 getInputStream();
777 return getContentHandler().getContent(this, classes);
778 }
779
780 /**
781 * Returns a permission object representing the permission
782 * necessary to make the connection represented by this
783 * object. This method returns null if no permission is
784 * required to make the connection. By default, this method
785 * returns {@code java.security.AllPermission}. Subclasses
786 * should override this method and return the permission
787 * that best represents the permission required to make a
788 * a connection to the URL. For example, a {@code URLConnection}
789 * representing a {@code file:} URL would return a
790 * {@code java.io.FilePermission} object.
791 *
792 * <p>The permission returned may dependent upon the state of the
793 * connection. For example, the permission before connecting may be
794 * different from that after connecting. For example, an HTTP
795 * sever, say foo.com, may redirect the connection to a different
796 * host, say bar.com. Before connecting the permission returned by
797 * the connection will represent the permission needed to connect
798 * to foo.com, while the permission returned after connecting will
799 * be to bar.com.
800 *
801 * <p>Permissions are generally used for two purposes: to protect
802 * caches of objects obtained through URLConnections, and to check
803 * the right of a recipient to learn about a particular URL. In
804 * the first case, the permission should be obtained
805 * <em>after</em> the object has been obtained. For example, in an
806 * HTTP connection, this will represent the permission to connect
807 * to the host from which the data was ultimately fetched. In the
808 * second case, the permission should be obtained and tested
809 * <em>before</em> connecting.
810 *
811 * @return the permission object representing the permission
812 * necessary to make the connection represented by this
813 * URLConnection.
814 *
815 * @exception IOException if the computation of the permission
816 * requires network or file I/O and an exception occurs while
817 * computing it.
818 */
819 public Permission getPermission() throws IOException {
820 return SecurityConstants.ALL_PERMISSION;
821 }
822
823 /**
824 * Returns an input stream that reads from this open connection.
825 *
826 * A SocketTimeoutException can be thrown when reading from the
827 * returned input stream if the read timeout expires before data
828 * is available for read.
829 *
830 * @return an input stream that reads from this open connection.
831 * @exception IOException if an I/O error occurs while
832 * creating the input stream.
833 * @exception UnknownServiceException if the protocol does not support
834 * input.
835 * @see #setReadTimeout(int)
836 * @see #getReadTimeout()
837 */
838 public InputStream getInputStream() throws IOException {
839 throw new UnknownServiceException("protocol doesn't support input");
840 }
841
842 /**
843 * Returns an output stream that writes to this connection.
844 *
845 * @return an output stream that writes to this connection.
846 * @exception IOException if an I/O error occurs while
847 * creating the output stream.
848 * @exception UnknownServiceException if the protocol does not support
849 * output.
850 */
851 public OutputStream getOutputStream() throws IOException {
852 throw new UnknownServiceException("protocol doesn't support output");
853 }
854
855 /**
856 * Returns a {@code String} representation of this URL connection.
857 *
858 * @return a string representation of this {@code URLConnection}.
859 */
860 public String toString() {
861 return this.getClass().getName() + ":" + url;
862 }
863
864 /**
865 * Sets the value of the {@code doInput} field for this
866 * {@code URLConnection} to the specified value.
867 * <p>
868 * A URL connection can be used for input and/or output. Set the DoInput
869 * flag to true if you intend to use the URL connection for input,
870 * false if not. The default is true.
871 *
872 * @param doinput the new value.
873 * @throws IllegalStateException if already connected
874 * @see java.net.URLConnection#doInput
875 * @see #getDoInput()
876 */
877 public void setDoInput(boolean doinput) {
878 checkConnected();
879 doInput = doinput;
880 }
881
882 /**
883 * Returns the value of this {@code URLConnection}'s
884 * {@code doInput} flag.
885 *
886 * @return the value of this {@code URLConnection}'s
887 * {@code doInput} flag.
888 * @see #setDoInput(boolean)
889 */
890 public boolean getDoInput() {
891 return doInput;
892 }
893
894 /**
895 * Sets the value of the {@code doOutput} field for this
896 * {@code URLConnection} to the specified value.
897 * <p>
898 * A URL connection can be used for input and/or output. Set the DoOutput
899 * flag to true if you intend to use the URL connection for output,
900 * false if not. The default is false.
901 *
902 * @param dooutput the new value.
903 * @throws IllegalStateException if already connected
904 * @see #getDoOutput()
905 */
906 public void setDoOutput(boolean dooutput) {
907 checkConnected();
908 doOutput = dooutput;
909 }
910
911 /**
912 * Returns the value of this {@code URLConnection}'s
913 * {@code doOutput} flag.
914 *
915 * @return the value of this {@code URLConnection}'s
916 * {@code doOutput} flag.
917 * @see #setDoOutput(boolean)
918 */
919 public boolean getDoOutput() {
920 return doOutput;
921 }
922
923 /**
924 * Set the value of the {@code allowUserInteraction} field of
925 * this {@code URLConnection}.
926 *
927 * @param allowuserinteraction the new value.
928 * @throws IllegalStateException if already connected
929 * @see #getAllowUserInteraction()
930 */
931 public void setAllowUserInteraction(boolean allowuserinteraction) {
932 checkConnected();
933 allowUserInteraction = allowuserinteraction;
934 }
935
936 /**
937 * Returns the value of the {@code allowUserInteraction} field for
938 * this object.
939 *
940 * @return the value of the {@code allowUserInteraction} field for
941 * this object.
942 * @see #setAllowUserInteraction(boolean)
943 */
944 public boolean getAllowUserInteraction() {
945 return allowUserInteraction;
946 }
947
948 /**
949 * Sets the default value of the
950 * {@code allowUserInteraction} field for all future
951 * {@code URLConnection} objects to the specified value.
952 *
953 * @param defaultallowuserinteraction the new value.
954 * @see #getDefaultAllowUserInteraction()
955 */
956 public static void setDefaultAllowUserInteraction(boolean defaultallowuserinteraction) {
957 defaultAllowUserInteraction = defaultallowuserinteraction;
958 }
959
960 /**
961 * Returns the default value of the {@code allowUserInteraction}
962 * field.
963 * <p>
964 * Ths default is "sticky", being a part of the static state of all
965 * URLConnections. This flag applies to the next, and all following
966 * URLConnections that are created.
967 *
968 * @return the default value of the {@code allowUserInteraction}
969 * field.
970 * @see #setDefaultAllowUserInteraction(boolean)
971 */
972 public static boolean getDefaultAllowUserInteraction() {
973 return defaultAllowUserInteraction;
974 }
975
976 /**
977 * Sets the value of the {@code useCaches} field of this
978 * {@code URLConnection} to the specified value.
979 * <p>
980 * Some protocols do caching of documents. Occasionally, it is important
981 * to be able to "tunnel through" and ignore the caches (e.g., the
982 * "reload" button in a browser). If the UseCaches flag on a connection
983 * is true, the connection is allowed to use whatever caches it can.
984 * If false, caches are to be ignored.
985 * The default value comes from DefaultUseCaches, which defaults to
986 * true.
987 *
988 * @param usecaches a {@code boolean} indicating whether
989 * or not to allow caching
990 * @throws IllegalStateException if already connected
991 * @see #getUseCaches()
992 */
993 public void setUseCaches(boolean usecaches) {
994 checkConnected();
995 useCaches = usecaches;
996 }
997
998 /**
999 * Returns the value of this {@code URLConnection}'s
1000 * {@code useCaches} field.
1001 *
1002 * @return the value of this {@code URLConnection}'s
1003 * {@code useCaches} field.
1004 * @see #setUseCaches(boolean)
1005 */
1006 public boolean getUseCaches() {
1007 return useCaches;
1008 }
1009
1010 /**
1011 * Sets the value of the {@code ifModifiedSince} field of
1012 * this {@code URLConnection} to the specified value.
1013 *
1014 * @param ifmodifiedsince the new value.
1015 * @throws IllegalStateException if already connected
1016 * @see #getIfModifiedSince()
1017 */
1018 public void setIfModifiedSince(long ifmodifiedsince) {
1019 checkConnected();
1020 ifModifiedSince = ifmodifiedsince;
1021 }
1022
1023 /**
1024 * Returns the value of this object's {@code ifModifiedSince} field.
1025 *
1026 * @return the value of this object's {@code ifModifiedSince} field.
1027 * @see #setIfModifiedSince(long)
1028 */
1029 public long getIfModifiedSince() {
1030 return ifModifiedSince;
1031 }
1032
1033 /**
1034 * Returns the default value of a {@code URLConnection}'s
1035 * {@code useCaches} flag.
1036 * <p>
1037 * Ths default is "sticky", being a part of the static state of all
1038 * URLConnections. This flag applies to the next, and all following
1039 * URLConnections that are created.
1040 *
1041 * @return the default value of a {@code URLConnection}'s
1042 * {@code useCaches} flag.
1043 * @see #setDefaultUseCaches(boolean)
1044 */
1045 public boolean getDefaultUseCaches() {
1046 return defaultUseCaches;
1047 }
1048
1049 /**
1050 * Sets the default value of the {@code useCaches} field to the
1051 * specified value.
1052 *
1053 * @param defaultusecaches the new value.
1054 * @see #getDefaultUseCaches()
1055 */
1056 public void setDefaultUseCaches(boolean defaultusecaches) {
1057 defaultUseCaches = defaultusecaches;
1058 }
1059
1060 /**
1061 * Sets the general request property. If a property with the key already
1062 * exists, overwrite its value with the new value.
1063 *
1064 * <p> NOTE: HTTP requires all request properties which can
1065 * legally have multiple instances with the same key
1066 * to use a comma-separated list syntax which enables multiple
1067 * properties to be appended into a single property.
1068 *
1069 * @param key the keyword by which the request is known
1070 * (e.g., "{@code Accept}").
1071 * @param value the value associated with it.
1072 * @throws IllegalStateException if already connected
1073 * @throws NullPointerException if key is {@code null}
1074 * @see #getRequestProperty(java.lang.String)
1075 */
1076 public void setRequestProperty(String key, String value) {
1077 checkConnected();
1078 if (key == null)
1079 throw new NullPointerException ("key is null");
1080
1081 if (requests == null)
1082 requests = new MessageHeader();
1083
1084 requests.set(key, value);
1085 }
1086
1087 /**
1088 * Adds a general request property specified by a
1089 * key-value pair. This method will not overwrite
1090 * existing values associated with the same key.
1091 *
1092 * @param key the keyword by which the request is known
1093 * (e.g., "{@code Accept}").
1094 * @param value the value associated with it.
1095 * @throws IllegalStateException if already connected
1096 * @throws NullPointerException if key is null
1097 * @see #getRequestProperties()
1098 * @since 1.4
1099 */
1100 public void addRequestProperty(String key, String value) {
1101 checkConnected();
1102 if (key == null)
1103 throw new NullPointerException ("key is null");
1104
1105 if (requests == null)
1106 requests = new MessageHeader();
1107
1108 requests.add(key, value);
1109 }
1110
1111
1112 /**
1113 * Returns the value of the named general request property for this
1114 * connection.
1115 *
1116 * @param key the keyword by which the request is known (e.g., "Accept").
1117 * @return the value of the named general request property for this
1118 * connection. If key is null, then null is returned.
1119 * @throws IllegalStateException if already connected
1120 * @see #setRequestProperty(java.lang.String, java.lang.String)
1121 */
1122 public String getRequestProperty(String key) {
1123 checkConnected();
1124
1125 if (requests == null)
1126 return null;
1127
1128 return requests.findValue(key);
1129 }
1130
1131 /**
1132 * Returns an unmodifiable Map of general request
1133 * properties for this connection. The Map keys
1134 * are Strings that represent the request-header
1135 * field names. Each Map value is a unmodifiable List
1136 * of Strings that represents the corresponding
1137 * field values.
1138 *
1139 * @return a Map of the general request properties for this connection.
1140 * @throws IllegalStateException if already connected
1141 * @since 1.4
1142 */
1143 public Map<String,List<String>> getRequestProperties() {
1144 checkConnected();
1145
1146 if (requests == null)
1147 return Collections.emptyMap();
1148
1149 return requests.getHeaders(null);
1150 }
1151
1152 /**
1153 * Sets the default value of a general request property. When a
1154 * {@code URLConnection} is created, it is initialized with
1155 * these properties.
1156 *
1157 * @param key the keyword by which the request is known
1158 * (e.g., "{@code Accept}").
1159 * @param value the value associated with the key.
1160 *
1161 * @see java.net.URLConnection#setRequestProperty(java.lang.String,java.lang.String)
1162 *
1163 * @deprecated The instance specific setRequestProperty method
1164 * should be used after an appropriate instance of URLConnection
1165 * is obtained. Invoking this method will have no effect.
1166 *
1167 * @see #getDefaultRequestProperty(java.lang.String)
1168 */
1169 @Deprecated
1170 public static void setDefaultRequestProperty(String key, String value) {
1171 }
1172
1173 /**
1174 * Returns the value of the default request property. Default request
1175 * properties are set for every connection.
1176 *
1177 * @param key the keyword by which the request is known (e.g., "Accept").
1178 * @return the value of the default request property
1179 * for the specified key.
1180 *
1181 * @see java.net.URLConnection#getRequestProperty(java.lang.String)
1182 *
1183 * @deprecated The instance specific getRequestProperty method
1184 * should be used after an appropriate instance of URLConnection
1185 * is obtained.
1186 *
1187 * @see #setDefaultRequestProperty(java.lang.String, java.lang.String)
1188 */
1189 @Deprecated
1190 public static String getDefaultRequestProperty(String key) {
1191 return null;
1192 }
1193
1194 /**
1195 * The ContentHandler factory.
1196 */
1197 private static volatile ContentHandlerFactory factory;
1198
1199 /**
1200 * Sets the {@code ContentHandlerFactory} of an
1201 * application. It can be called at most once by an application.
1202 * <p>
1203 * The {@code ContentHandlerFactory} instance is used to
1204 * construct a content handler from a content type
1205 * <p>
1206 * If there is a security manager, this method first calls
1207 * the security manager's {@code checkSetFactory} method
1208 * to ensure the operation is allowed.
1209 * This could result in a SecurityException.
1210 *
1211 * @param fac the desired factory.
1212 * @exception Error if the factory has already been defined.
1213 * @exception SecurityException if a security manager exists and its
1214 * {@code checkSetFactory} method doesn't allow the operation.
1215 * @see java.net.ContentHandlerFactory
1216 * @see java.net.URLConnection#getContent()
1217 * @see SecurityManager#checkSetFactory
1218 */
1219 public static synchronized void setContentHandlerFactory(ContentHandlerFactory fac) {
1220 if (factory != null) {
1221 throw new Error("factory already defined");
1222 }
1223 SecurityManager security = System.getSecurityManager();
1224 if (security != null) {
1225 security.checkSetFactory();
1226 }
1227 factory = fac;
1228 }
1229
1230 private static final Hashtable<String, ContentHandler> handlers = new Hashtable<>();
1231
1232 /**
1233 * Gets the Content Handler appropriate for this connection.
1234 */
1235 private ContentHandler getContentHandler() throws UnknownServiceException {
1236 String contentType = stripOffParameters(getContentType());
1237 if (contentType == null) {
1238 throw new UnknownServiceException("no content-type");
1239 }
1240
1241 ContentHandler handler = handlers.get(contentType);
1242 if (handler != null)
1243 return handler;
1244
1245 if (factory != null) {
1246 handler = factory.createContentHandler(contentType);
1247 if (handler != null)
1248 return handler;
1249 }
1250
1251 handler = lookupContentHandlerViaProvider(contentType);
1252
1253 if (handler != null) {
1254 ContentHandler h = handlers.putIfAbsent(contentType, handler);
1255 return Objects.requireNonNullElse(h, handler);
1256 }
1257
1258 try {
1259 handler = lookupContentHandlerClassFor(contentType);
1260 } catch (Exception e) {
1261 e.printStackTrace();
1262 handler = UnknownContentHandler.INSTANCE;
1263 }
1264
1265 assert handler != null;
1266
1267 ContentHandler h = handlers.putIfAbsent(contentType, handler);
1268 return Objects.requireNonNullElse(h, handler);
1269 }
1270
1271 /*
1272 * Media types are in the format: type/subtype*(; parameter).
1273 * For looking up the content handler, we should ignore those
1274 * parameters.
1275 */
1276 private String stripOffParameters(String contentType)
1277 {
1278 if (contentType == null)
1279 return null;
1280 int index = contentType.indexOf(';');
1281
1282 if (index > 0)
1283 return contentType.substring(0, index);
1284 else
1285 return contentType;
1286 }
1287
1288 private static final String contentClassPrefix = "sun.net.www.content";
1289 private static final String contentPathProp = "java.content.handler.pkgs";
1290
1291 /**
1292 * Looks for a content handler in a user-definable set of places.
1293 * By default it looks in {@value #contentClassPrefix}, but users can define
1294 * a vertical-bar delimited set of class prefixes to search through in
1295 * addition by defining the {@value #contentPathProp} property.
1296 * The class name must be of the form:
1297 * <pre>
1298 * {package-prefix}.{major}.{minor}
1299 * e.g.
1300 * YoyoDyne.experimental.text.plain
1301 * </pre>
1302 */
1303 private ContentHandler lookupContentHandlerClassFor(String contentType) {
1304 String contentHandlerClassName = typeToPackageName(contentType);
1305
1306 String contentHandlerPkgPrefixes = getContentHandlerPkgPrefixes();
1307
1308 StringTokenizer packagePrefixIter =
1309 new StringTokenizer(contentHandlerPkgPrefixes, "|");
1310
1311 while (packagePrefixIter.hasMoreTokens()) {
1312 String packagePrefix = packagePrefixIter.nextToken().trim();
1313
1314 try {
1315 String clsName = packagePrefix + "." + contentHandlerClassName;
1316 Class<?> cls = null;
1317 try {
1318 cls = Class.forName(clsName);
1319 } catch (ClassNotFoundException e) {
1320 ClassLoader cl = ClassLoader.getSystemClassLoader();
1321 if (cl != null) {
1322 cls = cl.loadClass(clsName);
1323 }
1324 }
1325 if (cls != null) {
1326 return (ContentHandler) cls.newInstance();
1327 }
1328 } catch(Exception ignored) { }
1329 }
1330
1331 return UnknownContentHandler.INSTANCE;
1332 }
1333
1334 private ContentHandler lookupContentHandlerViaProvider(String contentType) {
1335 return AccessController.doPrivileged(
1336 new PrivilegedAction<>() {
1337 @Override
1338 public ContentHandler run() {
1339 ClassLoader cl = ClassLoader.getSystemClassLoader();
1340 ServiceLoader<ContentHandlerFactory> sl =
1341 ServiceLoader.load(ContentHandlerFactory.class, cl);
1342
1343 Iterator<ContentHandlerFactory> iterator = sl.iterator();
1344
1345 ContentHandler handler = null;
1346 while (iterator.hasNext()) {
1347 ContentHandlerFactory f;
1348 try {
1349 f = iterator.next();
1350 } catch (ServiceConfigurationError e) {
1351 if (e.getCause() instanceof SecurityException) {
1352 continue;
1353 }
1354 throw e;
1355 }
1356 handler = f.createContentHandler(contentType);
1357 if (handler != null) {
1358 break;
1359 }
1360 }
1361 return handler;
1362 }
1363 });
1364 }
1365
1366 /**
1367 * Utility function to map a MIME content type into an equivalent
1368 * pair of class name components. For example: "text/html" would
1369 * be returned as "text.html"
1370 */
1371 private String typeToPackageName(String contentType) {
1372 // make sure we canonicalize the class name: all lower case
1373 contentType = contentType.toLowerCase();
1374 int len = contentType.length();
1375 char nm[] = new char[len];
1376 contentType.getChars(0, len, nm, 0);
1377 for (int i = 0; i < len; i++) {
1378 char c = nm[i];
1379 if (c == '/') {
1380 nm[i] = '.';
1381 } else if (!('A' <= c && c <= 'Z' ||
1382 'a' <= c && c <= 'z' ||
1383 '0' <= c && c <= '9')) {
1384 nm[i] = '_';
1385 }
1386 }
1387 return new String(nm);
1388 }
1389
1390
1391 /**
1392 * Returns a vertical bar separated list of package prefixes for potential
1393 * content handlers. Tries to get the java.content.handler.pkgs property
1394 * to use as a set of package prefixes to search. Whether or not
1395 * that property has been defined, the {@value #contentClassPrefix}
1396 * is always the last one on the returned package list.
1397 */
1398 private String getContentHandlerPkgPrefixes() {
1399 String packagePrefixList =
1400 GetPropertyAction.getProperty(contentPathProp, "");
1401
1402 if (packagePrefixList != "") {
1403 packagePrefixList += "|";
1404 }
1405
1406 return packagePrefixList + contentClassPrefix;
1407 }
1408
1409 /**
1410 * Tries to determine the content type of an object, based
1411 * on the specified "file" component of a URL.
1412 * This is a convenience method that can be used by
1413 * subclasses that override the {@code getContentType} method.
1414 *
1415 * @param fname a filename.
1416 * @return a guess as to what the content type of the object is,
1417 * based upon its file name.
1418 * @see java.net.URLConnection#getContentType()
1419 */
1420 public static String guessContentTypeFromName(String fname) {
1421 return getFileNameMap().getContentTypeFor(fname);
1422 }
1423
1424 /**
1425 * Tries to determine the type of an input stream based on the
1426 * characters at the beginning of the input stream. This method can
1427 * be used by subclasses that override the
1428 * {@code getContentType} method.
1429 * <p>
1430 * Ideally, this routine would not be needed. But many
1431 * {@code http} servers return the incorrect content type; in
1432 * addition, there are many nonstandard extensions. Direct inspection
1433 * of the bytes to determine the content type is often more accurate
1434 * than believing the content type claimed by the {@code http} server.
1435 *
1436 * @param is an input stream that supports marks.
1437 * @return a guess at the content type, or {@code null} if none
1438 * can be determined.
1439 * @exception IOException if an I/O error occurs while reading the
1440 * input stream.
1441 * @see java.io.InputStream#mark(int)
1442 * @see java.io.InputStream#markSupported()
1443 * @see java.net.URLConnection#getContentType()
1444 */
1445 public static String guessContentTypeFromStream(InputStream is)
1446 throws IOException {
1447 // If we can't read ahead safely, just give up on guessing
1448 if (!is.markSupported())
1449 return null;
1450
1451 is.mark(16);
1452 int c1 = is.read();
1453 int c2 = is.read();
1454 int c3 = is.read();
1455 int c4 = is.read();
1456 int c5 = is.read();
1457 int c6 = is.read();
1458 int c7 = is.read();
1459 int c8 = is.read();
1460 int c9 = is.read();
1461 int c10 = is.read();
1462 int c11 = is.read();
1463 int c12 = is.read();
1464 int c13 = is.read();
1465 int c14 = is.read();
1466 int c15 = is.read();
1467 int c16 = is.read();
1468 is.reset();
1469
1470 if (c1 == 0xCA && c2 == 0xFE && c3 == 0xBA && c4 == 0xBE) {
1471 return "application/java-vm";
1472 }
1473
1474 if (c1 == 0xAC && c2 == 0xED) {
1475 // next two bytes are version number, currently 0x00 0x05
1476 return "application/x-java-serialized-object";
1477 }
1478
1479 if (c1 == '<') {
1480 if (c2 == '!'
1481 || ((c2 == 'h' && (c3 == 't' && c4 == 'm' && c5 == 'l' ||
1482 c3 == 'e' && c4 == 'a' && c5 == 'd') ||
1483 (c2 == 'b' && c3 == 'o' && c4 == 'd' && c5 == 'y'))) ||
1484 ((c2 == 'H' && (c3 == 'T' && c4 == 'M' && c5 == 'L' ||
1485 c3 == 'E' && c4 == 'A' && c5 == 'D') ||
1486 (c2 == 'B' && c3 == 'O' && c4 == 'D' && c5 == 'Y')))) {
1487 return "text/html";
1488 }
1489
1490 if (c2 == '?' && c3 == 'x' && c4 == 'm' && c5 == 'l' && c6 == ' ') {
1491 return "application/xml";
1492 }
1493 }
1494
1495 // big and little (identical) endian UTF-8 encodings, with BOM
1496 if (c1 == 0xef && c2 == 0xbb && c3 == 0xbf) {
1497 if (c4 == '<' && c5 == '?' && c6 == 'x') {
1498 return "application/xml";
1499 }
1500 }
1501
1502 // big and little endian UTF-16 encodings, with byte order mark
1503 if (c1 == 0xfe && c2 == 0xff) {
1504 if (c3 == 0 && c4 == '<' && c5 == 0 && c6 == '?' &&
1505 c7 == 0 && c8 == 'x') {
1506 return "application/xml";
1507 }
1508 }
1509
1510 if (c1 == 0xff && c2 == 0xfe) {
1511 if (c3 == '<' && c4 == 0 && c5 == '?' && c6 == 0 &&
1512 c7 == 'x' && c8 == 0) {
1513 return "application/xml";
1514 }
1515 }
1516
1517 // big and little endian UTF-32 encodings, with BOM
1518 if (c1 == 0x00 && c2 == 0x00 && c3 == 0xfe && c4 == 0xff) {
1519 if (c5 == 0 && c6 == 0 && c7 == 0 && c8 == '<' &&
1520 c9 == 0 && c10 == 0 && c11 == 0 && c12 == '?' &&
1521 c13 == 0 && c14 == 0 && c15 == 0 && c16 == 'x') {
1522 return "application/xml";
1523 }
1524 }
1525
1526 if (c1 == 0xff && c2 == 0xfe && c3 == 0x00 && c4 == 0x00) {
1527 if (c5 == '<' && c6 == 0 && c7 == 0 && c8 == 0 &&
1528 c9 == '?' && c10 == 0 && c11 == 0 && c12 == 0 &&
1529 c13 == 'x' && c14 == 0 && c15 == 0 && c16 == 0) {
1530 return "application/xml";
1531 }
1532 }
1533
1534 if (c1 == 'G' && c2 == 'I' && c3 == 'F' && c4 == '8') {
1535 return "image/gif";
1536 }
1537
1538 if (c1 == '#' && c2 == 'd' && c3 == 'e' && c4 == 'f') {
1539 return "image/x-bitmap";
1540 }
1541
1542 if (c1 == '!' && c2 == ' ' && c3 == 'X' && c4 == 'P' &&
1543 c5 == 'M' && c6 == '2') {
1544 return "image/x-pixmap";
1545 }
1546
1547 if (c1 == 137 && c2 == 80 && c3 == 78 &&
1548 c4 == 71 && c5 == 13 && c6 == 10 &&
1549 c7 == 26 && c8 == 10) {
1550 return "image/png";
1551 }
1552
1553 if (c1 == 0xFF && c2 == 0xD8 && c3 == 0xFF) {
1554 if (c4 == 0xE0 || c4 == 0xEE) {
1555 return "image/jpeg";
1556 }
1557
1558 /**
1559 * File format used by digital cameras to store images.
1560 * Exif Format can be read by any application supporting
1561 * JPEG. Exif Spec can be found at:
1562 * http://www.pima.net/standards/it10/PIMA15740/Exif_2-1.PDF
1563 */
1564 if ((c4 == 0xE1) &&
1565 (c7 == 'E' && c8 == 'x' && c9 == 'i' && c10 =='f' &&
1566 c11 == 0)) {
1567 return "image/jpeg";
1568 }
1569 }
1570
1571 if ((c1 == 0x49 && c2 == 0x49 && c3 == 0x2a && c4 == 0x00)
1572 || (c1 == 0x4d && c2 == 0x4d && c3 == 0x00 && c4 == 0x2a)) {
1573 return "image/tiff";
1574 }
1575
1576 if (c1 == 0xD0 && c2 == 0xCF && c3 == 0x11 && c4 == 0xE0 &&
1577 c5 == 0xA1 && c6 == 0xB1 && c7 == 0x1A && c8 == 0xE1) {
1578
1579 /* Above is signature of Microsoft Structured Storage.
1580 * Below this, could have tests for various SS entities.
1581 * For now, just test for FlashPix.
1582 */
1583 if (checkfpx(is)) {
1584 return "image/vnd.fpx";
1585 }
1586 }
1587
1588 if (c1 == 0x2E && c2 == 0x73 && c3 == 0x6E && c4 == 0x64) {
1589 return "audio/basic"; // .au format, big endian
1590 }
1591
1592 if (c1 == 0x64 && c2 == 0x6E && c3 == 0x73 && c4 == 0x2E) {
1593 return "audio/basic"; // .au format, little endian
1594 }
1595
1596 if (c1 == 'R' && c2 == 'I' && c3 == 'F' && c4 == 'F') {
1597 /* I don't know if this is official but evidence
1598 * suggests that .wav files start with "RIFF" - brown
1599 */
1600 return "audio/x-wav";
1601 }
1602 return null;
1603 }
1604
1605 /**
1606 * Check for FlashPix image data in InputStream is. Return true if
1607 * the stream has FlashPix data, false otherwise. Before calling this
1608 * method, the stream should have already been checked to be sure it
1609 * contains Microsoft Structured Storage data.
1610 */
1611 private static boolean checkfpx(InputStream is) throws IOException {
1612
1613 /* Test for FlashPix image data in Microsoft Structured Storage format.
1614 * In general, should do this with calls to an SS implementation.
1615 * Lacking that, need to dig via offsets to get to the FlashPix
1616 * ClassID. Details:
1617 *
1618 * Offset to Fpx ClsID from beginning of stream should be:
1619 *
1620 * FpxClsidOffset = rootEntryOffset + clsidOffset
1621 *
1622 * where: clsidOffset = 0x50.
1623 * rootEntryOffset = headerSize + sectorSize*sectDirStart
1624 * + 128*rootEntryDirectory
1625 *
1626 * where: headerSize = 0x200 (always)
1627 * sectorSize = 2 raised to power of uSectorShift,
1628 * which is found in the header at
1629 * offset 0x1E.
1630 * sectDirStart = found in the header at offset 0x30.
1631 * rootEntryDirectory = in general, should search for
1632 * directory labelled as root.
1633 * We will assume value of 0 (i.e.,
1634 * rootEntry is in first directory)
1635 */
1636
1637 // Mark the stream so we can reset it. 0x100 is enough for the first
1638 // few reads, but the mark will have to be reset and set again once
1639 // the offset to the root directory entry is computed. That offset
1640 // can be very large and isn't know until the stream has been read from
1641 is.mark(0x100);
1642
1643 // Get the byte ordering located at 0x1E. 0xFE is Intel,
1644 // 0xFF is other
1645 long toSkip = (long)0x1C;
1646 long posn;
1647
1648 if ((posn = skipForward(is, toSkip)) < toSkip) {
1649 is.reset();
1650 return false;
1651 }
1652
1653 int c[] = new int[16];
1654 if (readBytes(c, 2, is) < 0) {
1655 is.reset();
1656 return false;
1657 }
1658
1659 int byteOrder = c[0];
1660
1661 posn+=2;
1662 int uSectorShift;
1663 if (readBytes(c, 2, is) < 0) {
1664 is.reset();
1665 return false;
1666 }
1667
1668 if(byteOrder == 0xFE) {
1669 uSectorShift = c[0];
1670 uSectorShift += c[1] << 8;
1671 }
1672 else {
1673 uSectorShift = c[0] << 8;
1674 uSectorShift += c[1];
1675 }
1676
1677 posn += 2;
1678 toSkip = (long)0x30 - posn;
1679 long skipped = 0;
1680 if ((skipped = skipForward(is, toSkip)) < toSkip) {
1681 is.reset();
1682 return false;
1683 }
1684 posn += skipped;
1685
1686 if (readBytes(c, 4, is) < 0) {
1687 is.reset();
1688 return false;
1689 }
1690
1691 int sectDirStart;
1692 if(byteOrder == 0xFE) {
1693 sectDirStart = c[0];
1694 sectDirStart += c[1] << 8;
1695 sectDirStart += c[2] << 16;
1696 sectDirStart += c[3] << 24;
1697 } else {
1698 sectDirStart = c[0] << 24;
1699 sectDirStart += c[1] << 16;
1700 sectDirStart += c[2] << 8;
1701 sectDirStart += c[3];
1702 }
1703 posn += 4;
1704 is.reset(); // Reset back to the beginning
1705
1706 toSkip = 0x200L + (long)(1<<uSectorShift)*sectDirStart + 0x50L;
1707
1708 // Sanity check!
1709 if (toSkip < 0) {
1710 return false;
1711 }
1712
1713 /*
1714 * How far can we skip? Is there any performance problem here?
1715 * This skip can be fairly long, at least 0x4c650 in at least
1716 * one case. Have to assume that the skip will fit in an int.
1717 * Leave room to read whole root dir
1718 */
1719 is.mark((int)toSkip+0x30);
1720
1721 if ((skipForward(is, toSkip)) < toSkip) {
1722 is.reset();
1723 return false;
1724 }
1725
1726 /* should be at beginning of ClassID, which is as follows
1727 * (in Intel byte order):
1728 * 00 67 61 56 54 C1 CE 11 85 53 00 AA 00 A1 F9 5B
1729 *
1730 * This is stored from Windows as long,short,short,char[8]
1731 * so for byte order changes, the order only changes for
1732 * the first 8 bytes in the ClassID.
1733 *
1734 * Test against this, ignoring second byte (Intel) since
1735 * this could change depending on part of Fpx file we have.
1736 */
1737
1738 if (readBytes(c, 16, is) < 0) {
1739 is.reset();
1740 return false;
1741 }
1742
1743 // intel byte order
1744 if (byteOrder == 0xFE &&
1745 c[0] == 0x00 && c[2] == 0x61 && c[3] == 0x56 &&
1746 c[4] == 0x54 && c[5] == 0xC1 && c[6] == 0xCE &&
1747 c[7] == 0x11 && c[8] == 0x85 && c[9] == 0x53 &&
1748 c[10]== 0x00 && c[11]== 0xAA && c[12]== 0x00 &&
1749 c[13]== 0xA1 && c[14]== 0xF9 && c[15]== 0x5B) {
1750 is.reset();
1751 return true;
1752 }
1753
1754 // non-intel byte order
1755 else if (c[3] == 0x00 && c[1] == 0x61 && c[0] == 0x56 &&
1756 c[5] == 0x54 && c[4] == 0xC1 && c[7] == 0xCE &&
1757 c[6] == 0x11 && c[8] == 0x85 && c[9] == 0x53 &&
1758 c[10]== 0x00 && c[11]== 0xAA && c[12]== 0x00 &&
1759 c[13]== 0xA1 && c[14]== 0xF9 && c[15]== 0x5B) {
1760 is.reset();
1761 return true;
1762 }
1763 is.reset();
1764 return false;
1765 }
1766
1767 /**
1768 * Tries to read the specified number of bytes from the stream
1769 * Returns -1, If EOF is reached before len bytes are read, returns 0
1770 * otherwise
1771 */
1772 private static int readBytes(int c[], int len, InputStream is)
1773 throws IOException {
1774
1775 byte buf[] = new byte[len];
1776 if (is.read(buf, 0, len) < len) {
1777 return -1;
1778 }
1779
1780 // fill the passed in int array
1781 for (int i = 0; i < len; i++) {
1782 c[i] = buf[i] & 0xff;
1783 }
1784 return 0;
1785 }
1786
1787
1788 /**
1789 * Skips through the specified number of bytes from the stream
1790 * until either EOF is reached, or the specified
1791 * number of bytes have been skipped
1792 */
1793 private static long skipForward(InputStream is, long toSkip)
1794 throws IOException {
1795
1796 long eachSkip = 0;
1797 long skipped = 0;
1798
1799 while (skipped != toSkip) {
1800 eachSkip = is.skip(toSkip - skipped);
1801
1802 // check if EOF is reached
1803 if (eachSkip <= 0) {
1804 if (is.read() == -1) {
1805 return skipped ;
1806 } else {
1807 skipped++;
1808 }
1809 }
1810 skipped += eachSkip;
1811 }
1812 return skipped;
1813 }
1814
1815 private void checkConnected() {
1816 if (connected)
1817 throw new IllegalStateException("Already connected");
1818 }
1819 }
1820
1821 class UnknownContentHandler extends ContentHandler {
1822 static final ContentHandler INSTANCE = new UnknownContentHandler();
1823
1824 public Object getContent(URLConnection uc) throws IOException {
1825 return uc.getInputStream();
1826 }
1827 }