Move the check for legacy/UTF character conversion on incoming URLs from
[lhc/web/wiklou.git] / includes / WebRequest.php
1 <?php
2 /**
3 * Deal with importing all those nasssty globals and things
4 * @package MediaWiki
5 */
6
7 # Copyright (C) 2003 Brion Vibber <brion@pobox.com>
8 # http://www.mediawiki.org/
9 #
10 # This program is free software; you can redistribute it and/or modify
11 # it under the terms of the GNU General Public License as published by
12 # the Free Software Foundation; either version 2 of the License, or
13 # (at your option) any later version.
14 #
15 # This program is distributed in the hope that it will be useful,
16 # but WITHOUT ANY WARRANTY; without even the implied warranty of
17 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
18 # GNU General Public License for more details.
19 #
20 # You should have received a copy of the GNU General Public License along
21 # with this program; if not, write to the Free Software Foundation, Inc.,
22 # 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
23 # http://www.gnu.org/copyleft/gpl.html
24
25 /**
26 * The WebRequest class encapsulates getting at data passed in the
27 * URL or via a POSTed form, handling remove of "magic quotes" slashes,
28 * stripping illegal input characters and normalizing Unicode sequences.
29 *
30 * Usually this is used via a global singleton, $wgRequest. You should
31 * not create a second WebRequest object; make a FauxRequest object if
32 * you want to pass arbitrary data to some function in place of the web
33 * input.
34 *
35 * @package MediaWiki
36 */
37 class WebRequest {
38 function WebRequest() {
39 $this->checkMagicQuotes();
40 global $wgUsePathInfo;
41 if( isset( $_SERVER['PATH_INFO'] ) && $wgUsePathInfo ) {
42 # Stuff it!
43 $_GET['title'] = $_REQUEST['title'] =
44 substr( $_SERVER['PATH_INFO'], 1 );
45 }
46 }
47
48 /**
49 * Recursively strips slashes from the given array;
50 * used for undoing the evil that is magic_quotes_gpc.
51 * @param array &$arr will be modified
52 * @return array the original array
53 * @private
54 */
55 function &fix_magic_quotes( &$arr ) {
56 foreach( $arr as $key => $val ) {
57 if( is_array( $val ) ) {
58 $this->fix_magic_quotes( $arr[$key] );
59 } else {
60 $arr[$key] = stripslashes( $val );
61 }
62 }
63 return $arr;
64 }
65
66 /**
67 * If magic_quotes_gpc option is on, run the global arrays
68 * through fix_magic_quotes to strip out the stupid dlashes.
69 * WARNING: This should only be done once! Running a second
70 * time could damage the values.
71 * @private
72 */
73 function checkMagicQuotes() {
74 if ( get_magic_quotes_gpc() ) {
75 $this->fix_magic_quotes( $_COOKIE );
76 $this->fix_magic_quotes( $_ENV );
77 $this->fix_magic_quotes( $_GET );
78 $this->fix_magic_quotes( $_POST );
79 $this->fix_magic_quotes( $_REQUEST );
80 $this->fix_magic_quotes( $_SERVER );
81 }
82 }
83
84 /**
85 * Recursively normalizes UTF-8 strings in the given array.
86 * @param array $data string or array
87 * @return cleaned-up version of the given
88 * @private
89 */
90 function normalizeUnicode( $data ) {
91 if( is_array( $data ) ) {
92 foreach( $data as $key => $val ) {
93 $data[$key] = $this->normalizeUnicode( $val );
94 }
95 } else {
96 $data = UtfNormal::cleanUp( $data );
97 }
98 return $data;
99 }
100
101 /**
102 * Fetch a value from the given array or return $default if it's not set.
103 * @param array &$arr
104 * @param string $name
105 * @param mixed $default
106 * @return mixed
107 * @private
108 */
109 function getGPCVal( &$arr, $name, $default ) {
110 if( isset( $arr[$name] ) ) {
111 global $wgUseLatin1, $wgServer, $wgLang;
112 $data = $arr[$name];
113 if( isset( $_GET[$name] ) &&
114 ( empty( $_SERVER['HTTP_REFERER'] ) ||
115 strncmp($wgServer, $_SERVER['HTTP_REFERER'], strlen( $wgServer ) ) ) ) {
116 # For links that came from outside, check for alternate/legacy
117 # character encoding.
118 $data = $wgLang->checkTitleEncoding( $data );
119 }
120 if( !$wgUseLatin1 ) {
121 require_once( 'normal/UtfNormal.php' );
122 $data = $this->normalizeUnicode( $data );
123 }
124 return $data;
125 } else {
126 return $default;
127 }
128 }
129
130 /**
131 * Fetch a value from the given array or return $default if it's not set.
132 * \r is stripped from the text, and with some language modules there is
133 * an input transliteration applied.
134 * @param array &$arr
135 * @param string $name
136 * @param string $default
137 * @return string
138 * @private
139 */
140 function getGPCText( &$arr, $name, $default ) {
141 # Text fields may be in an alternate encoding which we should check.
142 # Also, strip CRLF line endings down to LF to achieve consistency.
143 global $wgLang;
144 if( isset( $arr[$name] ) ) {
145 return str_replace( "\r\n", "\n", $wgLang->recodeInput( $arr[$name] ) );
146 } else {
147 return $default;
148 }
149 }
150
151 /**
152 * Fetch a value from the input or return $default if it's not set.
153 * Value may be of a string or array, and is not altered.
154 * @param string $name
155 * @param mixed $default optional default (or NULL)
156 * @return mixed
157 */
158 function getVal( $name, $default = NULL ) {
159 return $this->getGPCVal( $_REQUEST, $name, $default );
160 }
161
162 /**
163 * Fetch an integer value from the input or return $default if not set.
164 * Guaranteed to return an integer; non-numeric input will typically
165 * return 0.
166 * @param string $name
167 * @param int $default
168 * @return int
169 */
170 function getInt( $name, $default = 0 ) {
171 return IntVal( $this->getVal( $name, $default ) );
172 }
173
174 /**
175 * Fetch a boolean value from the input or return $default if not set.
176 * Guaranteed to return true or false, with normal PHP semantics for
177 * boolean interpretation of strings.
178 * @param string $name
179 * @param bool $default
180 * @return bool
181 */
182 function getBool( $name, $default = false ) {
183 return $this->getVal( $name, $default ) ? true : false;
184 }
185
186 /**
187 * Return true if the named value is set in the input, whatever that
188 * value is (even "0"). Return false if the named value is not set.
189 * Example use is checking for the presence of check boxes in forms.
190 * @param string $name
191 * @return bool
192 */
193 function getCheck( $name ) {
194 # Checkboxes and buttons are only present when clicked
195 # Presence connotes truth, abscense false
196 $val = $this->getVal( $name, NULL );
197 return isset( $val );
198 }
199
200 /**
201 * Fetch a text string from the given array or return $default if it's not
202 * set. \r is stripped from the text, and with some language modules there
203 * is an input transliteration applied. This should generally be used for
204 * form <textarea> and <input> fields.
205 *
206 * @param string $name
207 * @param string $default optional
208 * @return string
209 */
210 function getText( $name, $default = '' ) {
211 return $this->getGPCText( $_REQUEST, $name, $default );
212 }
213
214 /**
215 * Extracts the given named values into an array.
216 * If no arguments are given, returns all input values.
217 * No transformation is performed on the values.
218 */
219 function getValues() {
220 $names = func_get_args();
221 if ( count( $names ) == 0 ) {
222 $names = array_keys( $_REQUEST );
223 }
224
225 $retVal = array();
226 foreach ( $names as $name ) {
227 $value = $this->getVal( $name );
228 if ( !is_null( $value ) ) {
229 $retVal[$name] = $value;
230 }
231 }
232 return $retVal;
233 }
234
235 /**
236 * Returns true if the present request was reached by a POST operation,
237 * false otherwise (GET, HEAD, or command-line).
238 *
239 * Note that values retrieved by the object may come from the
240 * GET URL etc even on a POST request.
241 *
242 * @return bool
243 */
244 function wasPosted() {
245 return $_SERVER['REQUEST_METHOD'] == 'POST';
246 }
247
248 /**
249 * Returns true if there is a session cookie set.
250 * This does not necessarily mean that the user is logged in!
251 *
252 * @return bool
253 */
254 function checkSessionCookie() {
255 return isset( $_COOKIE[ini_get('session.name')] );
256 }
257
258 /**
259 * Return the path portion of the request URI.
260 * @return string
261 */
262 function getRequestURL() {
263 return $_SERVER['REQUEST_URI'];
264 }
265
266 /**
267 * Return the request URI with the canonical service and hostname.
268 * @return string
269 */
270 function getFullRequestURL() {
271 global $wgServer;
272 return $wgServer . $this->getRequestURL();
273 }
274
275 /**
276 * Take an arbitrary query and rewrite the present URL to include it
277 * @param string $query Query string fragment; do not include initial '?'
278 * @return string
279 */
280 function appendQuery( $query ) {
281 global $wgTitle;
282 $basequery = '';
283 foreach( $_GET as $var => $val ) {
284 if( $var == 'title' ) continue;
285 $basequery .= '&' . urlencode( $var ) . '=' . urlencode( $val );
286 }
287 $basequery .= '&' . $query;
288
289 # Trim the extra &
290 $basequery = substr( $basequery, 1 );
291 return $wgTitle->getLocalURL( $basequery );
292 }
293
294 /**
295 * HTML-safe version of appendQuery().
296 * @param string $query Query string fragment; do not include initial '?'
297 * @return string
298 */
299 function escapeAppendQuery( $query ) {
300 return htmlspecialchars( $this->appendQuery( $query ) );
301 }
302
303 /**
304 * Check for limit and offset parameters on the input, and return sensible
305 * defaults if not given. The limit must be positive and is capped at 5000.
306 * Offset must be positive but is not capped.
307 *
308 * @param int $deflimit Limit to use if no input and the user hasn't set the option.
309 * @param string $optionname To specify an option other than rclimit to pull from.
310 * @return array first element is limit, second is offset
311 */
312 function getLimitOffset( $deflimit = 50, $optionname = 'rclimit' ) {
313 global $wgUser;
314
315 $limit = $this->getInt( 'limit', 0 );
316 if( $limit < 0 ) $limit = 0;
317 if( ( $limit == 0 ) && ( $optionname != '' ) ) {
318 $limit = (int)$wgUser->getOption( $optionname );
319 }
320 if( $limit <= 0 ) $limit = $deflimit;
321 if( $limit > 5000 ) $limit = 5000; # We have *some* limits...
322
323 $offset = $this->getInt( 'offset', 0 );
324 if( $offset < 0 ) $offset = 0;
325
326 return array( $limit, $offset );
327 }
328
329 /**
330 * Return the path to the temporary file where PHP has stored the upload.
331 * @param string $key
332 * @return string or NULL if no such file.
333 */
334 function getFileTempname( $key ) {
335 if( !isset( $_FILES[$key] ) ) {
336 return NULL;
337 }
338 return $_FILES[$key]['tmp_name'];
339 }
340
341 /**
342 * Return the size of the upload, or 0.
343 * @param string $key
344 * @return integer
345 */
346 function getFileSize( $key ) {
347 if( !isset( $_FILES[$key] ) ) {
348 return 0;
349 }
350 return $_FILES[$key]['size'];
351 }
352
353 /**
354 * Return the original filename of the uploaded file, as reported by
355 * the submitting user agent. HTML-style character entities are
356 * interpreted and normalized to Unicode normalization form C, in part
357 * to deal with weird input from Safari with non-ASCII filenames.
358 *
359 * Other than this the name is not verified for being a safe filename.
360 *
361 * @param string $key
362 * @return string or NULL if no such file.
363 */
364 function getFileName( $key ) {
365 if( !isset( $_FILES[$key] ) ) {
366 return NULL;
367 }
368 $name = $_FILES[$key]['name'];
369
370 # Safari sends filenames in HTML-encoded Unicode form D...
371 # Horrid and evil! Let's try to make some kind of sense of it.
372 global $wgUseLatin1;
373 if( $wgUseLatin1 ) {
374 $name = utf8_encode( $name );
375 }
376 $name = wfMungeToUtf8( $name );
377 $name = UtfNormal::cleanUp( $name );
378 if( $wgUseLatin1 ) {
379 $name = utf8_decode( $name );
380 }
381 wfDebug( "WebRequest::getFileName() '" . $_FILES[$key]['name'] . "' normalized to '$name'\n" );
382 return $name;
383 }
384 }
385
386 /**
387 * WebRequest clone which takes values from a provided array.
388 *
389 * @package MediaWiki
390 */
391 class FauxRequest extends WebRequest {
392 var $data = null;
393 var $wasPosted = false;
394
395 function WebRequest( $data, $wasPosted = false ) {
396 if( is_array( $data ) ) {
397 $this->data = $data;
398 } else {
399 wfDebugDieBacktrace( "FauxReqeust() got bogus data" );
400 }
401 $this->wasPosted = $wasPosted;
402 }
403
404 function getVal( $name, $default = NULL ) {
405 return $this->getGPCVal( $this->data, $name, $default );
406 }
407
408 function getText( $name, $default = '' ) {
409 # Override; don't recode since we're using internal data
410 return $this->getVal( $name, $default );
411 }
412
413 function getValues() {
414 return $this->data;
415 }
416
417 function wasPosted() {
418 return $this->wasPosted;
419 }
420
421 function checkSessionCookie() {
422 return false;
423 }
424
425 function getRequestURL() {
426 wfDebugDieBacktrace( 'FauxRequest::getRequestURL() not implemented' );
427 }
428
429 function appendQuery( $query ) {
430 wfDebugDieBacktrace( 'FauxRequest::appendQuery() not implemented' );
431 }
432
433 }
434
435 ?>