xref: /dokuwiki/inc/pageutils.php (revision 7e089c87a86bb19bfcb72fe835f4cfd3984eb992) 
1<?php
2/**
3 * Utilities for handling pagenames
4 *
5 * @license    GPL 2 (http://www.gnu.org/licenses/gpl.html)
6 * @author     Andreas Gohr <andi@splitbrain.org>
7 * @todo       Combine similar functions like {wiki,media,meta}FN()
8 */
9
10/**
11 * Fetch the an ID from request
12 *
13 * Uses either standard $_REQUEST variable or extracts it from
14 * the full request URI when userewrite is set to 2
15 *
16 * For $param='id' $conf['start'] is returned if no id was found.
17 * If the second parameter is true (default) the ID is cleaned.
18 *
19 * @author Andreas Gohr <andi@splitbrain.org>
20 */
21function getID($param='id',$clean=true){
22  global $conf;
23
24  $id = isset($_REQUEST[$param]) ? $_REQUEST[$param] : null;
25
26  //construct page id from request URI
27  if(empty($id) && $conf['userewrite'] == 2){
28    //get the script URL
29    if($conf['basedir']){
30      $relpath = '';
31      if($param != 'id') {
32        $relpath = 'lib/exe/';
33      }
34      $script = $conf['basedir'].$relpath.basename($_SERVER['SCRIPT_FILENAME']);
35    }elseif($_SERVER['DOCUMENT_ROOT'] && $_SERVER['SCRIPT_FILENAME']){
36      $script = preg_replace ('/^'.preg_quote($_SERVER['DOCUMENT_ROOT'],'/').'/','',
37                              $_SERVER['SCRIPT_FILENAME']);
38      $script = '/'.$script;
39    }else{
40      $script = $_SERVER['SCRIPT_NAME'];
41    }
42
43    //clean script and request (fixes a windows problem)
44    $script  = preg_replace('/\/\/+/','/',$script);
45    $request = preg_replace('/\/\/+/','/',$_SERVER['REQUEST_URI']);
46
47    //remove script URL and Querystring to gain the id
48    if(preg_match('/^'.preg_quote($script,'/').'(.*)/',$request, $match)){
49      $id = preg_replace ('/\?.*/','',$match[1]);
50    }
51    $id = urldecode($id);
52    //strip leading slashes
53    $id = preg_replace('!^/+!','',$id);
54  }
55
56  // Namespace autolinking from URL
57  if(substr($id,-1) == ':' || ($conf['useslash'] && substr($id,-1) == '/')){
58    if(page_exists($id.$conf['start'])){
59      // start page inside namespace
60      $id = $id.$conf['start'];
61    }elseif(page_exists($id.noNS(cleanID($id)))){
62      // page named like the NS inside the NS
63      $id = $id.noNS(cleanID($id));
64    }elseif(page_exists($id)){
65      // page like namespace exists
66      $id = substr($id,0,-1);
67    }else{
68      // fall back to default
69      $id = $id.$conf['start'];
70    }
71    header("Location: ".wl($id,'',true));
72  }
73
74  if($clean) $id = cleanID($id);
75  if(empty($id) && $param=='id') $id = $conf['start'];
76
77  return $id;
78}
79
80/**
81 * Remove unwanted chars from ID
82 *
83 * Cleans a given ID to only use allowed characters. Accented characters are
84 * converted to unaccented ones
85 *
86 * @author Andreas Gohr <andi@splitbrain.org>
87 * @param  string  $raw_id    The pageid to clean
88 * @param  boolean $ascii     Force ASCII
89 */
90function cleanID($raw_id,$ascii=false){
91  global $conf;
92  global $lang;
93  static $sepcharpat = null;
94
95  global $cache_cleanid;
96  $cache = & $cache_cleanid;
97
98  // check if it's already in the memory cache
99  if (isset($cache[$raw_id])) {
100    return $cache[$raw_id];
101    }
102
103  $sepchar = $conf['sepchar'];
104  if($sepcharpat == null) // build string only once to save clock cycles
105    $sepcharpat = '#\\'.$sepchar.'+#';
106
107  $id = trim($raw_id);
108  $id = utf8_strtolower($id);
109
110  //alternative namespace seperator
111  $id = strtr($id,';',':');
112  if($conf['useslash']){
113    $id = strtr($id,'/',':');
114  }else{
115    $id = strtr($id,'/',$sepchar);
116  }
117
118  if($conf['deaccent'] == 2 || $ascii) $id = utf8_romanize($id);
119  if($conf['deaccent'] || $ascii) $id = utf8_deaccent($id,-1);
120
121  //remove specials
122  $id = utf8_stripspecials($id,$sepchar,'\*');
123
124  if($ascii) $id = utf8_strip($id);
125
126  //clean up
127  $id = preg_replace($sepcharpat,$sepchar,$id);
128  $id = preg_replace('#:+#',':',$id);
129  $id = trim($id,':._-');
130  $id = preg_replace('#:[:\._\-]+#',':',$id);
131
132  $cache[$raw_id] = $id;
133  return($id);
134}
135
136/**
137 * Return namespacepart of a wiki ID
138 *
139 * @author Andreas Gohr <andi@splitbrain.org>
140 */
141function getNS($id){
142  $pos = strrpos($id,':');
143  if($pos!==false){
144    return substr($id,0,$pos);
145  }
146  return false;
147}
148
149/**
150 * Returns the ID without the namespace
151 *
152 * @author Andreas Gohr <andi@splitbrain.org>
153 */
154function noNS($id) {
155  $pos = strrpos($id, ':');
156  if ($pos!==false) {
157    return substr($id, $pos+1);
158  } else {
159    return $id;
160  }
161}
162
163/**
164* Returns the current namespace
165*
166* @author Nathan Fritz <fritzn@crown.edu>
167*/
168function curNS($id) {
169    return noNS(getNS($id));
170}
171
172/**
173* Returns the ID without the namespace or current namespace for 'start' pages
174*
175* @author Nathan Fritz <fritzn@crown.edu>
176*/
177function noNSorNS($id) {
178    global $conf;
179
180    $p = noNS($id);
181    if ($p == $conf['start']) {
182        $p = curNS($id);
183        if ($p == false) {
184            return noNS($id);
185        }
186    }
187    return $p;
188}
189
190/**
191 *  Wiki page existence check
192 *
193 *  parameters as for wikiFN
194 *
195 *  @author Chris Smith <chris@jalakai.co.uk>
196 */
197function page_exists($id,$rev='',$clean=true) {
198  return @file_exists(wikiFN($id,$rev,$clean));
199}
200
201/**
202 * returns the full path to the datafile specified by ID and optional revision
203 *
204 * The filename is URL encoded to protect Unicode chars
205 *
206 * @param  $raw_id  string   id of wikipage
207 * @param  $rev     string   page revision, empty string for current
208 * @param  $clean   bool     flag indicating that $raw_id should be cleaned.  Only set to false
209 *                           when $id is guaranteed to have been cleaned already.
210 *
211 * @author Andreas Gohr <andi@splitbrain.org>
212 */
213function wikiFN($raw_id,$rev='',$clean=true){
214  global $conf;
215
216  global $cache_wikifn;
217  $cache = & $cache_wikifn;
218
219  if (isset($cache[$raw_id]) && isset($cache[$raw_id][$rev])) {
220    return $cache[$raw_id][$rev];
221  }
222
223  $id = $raw_id;
224
225  if ($clean) $id = cleanID($id);
226  $id = str_replace(':','/',$id);
227  if(empty($rev)){
228    $fn = $conf['datadir'].'/'.utf8_encodeFN($id).'.txt';
229  }else{
230    $fn = $conf['olddir'].'/'.utf8_encodeFN($id).'.'.$rev.'.txt';
231    if($conf['compression']){
232      //test for extensions here, we want to read both compressions
233       if (@file_exists($fn . '.gz')){
234          $fn .= '.gz';
235       }else if(@file_exists($fn . '.bz2')){
236          $fn .= '.bz2';
237       }else{
238          //file doesnt exist yet, so we take the configured extension
239          $fn .= '.' . $conf['compression'];
240       }
241    }
242  }
243
244  if (!isset($cache[$raw_id])) { $cache[$raw_id] = array(); }
245  $cache[$raw_id][$rev] = $fn;
246  return $fn;
247}
248
249/**
250 * Returns the full path to the file for locking the page while editing.
251 *
252 * @author Ben Coburn <btcoburn@silicodon.net>
253 */
254function wikiLockFN($id) {
255  global $conf;
256  return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
257}
258
259
260/**
261 * returns the full path to the meta file specified by ID and extension
262 *
263 * The filename is URL encoded to protect Unicode chars
264 *
265 * @author Steven Danz <steven-danz@kc.rr.com>
266 */
267function metaFN($id,$ext){
268  global $conf;
269  $id = cleanID($id);
270  $id = str_replace(':','/',$id);
271  $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
272  return $fn;
273}
274
275/**
276 * returns an array of full paths to all metafiles of a given ID
277 *
278 * @author Esther Brunner <esther@kaffeehaus.ch>
279 */
280function metaFiles($id){
281   $name   = noNS($id);
282   $ns     = getNS($id);
283   $dir    = ($ns) ? metaFN($ns,'').'/' : metaFN($ns,'');
284   $files  = array();
285
286   $dh = @opendir($dir);
287   if(!$dh) return $files;
288   while(($file = readdir($dh)) !== false){
289     if(strpos($file,$name.'.') === 0 && !is_dir($dir.$file))
290       $files[] = $dir.$file;
291   }
292   closedir($dh);
293
294   return $files;
295}
296
297/**
298 * returns the full path to the mediafile specified by ID
299 *
300 * The filename is URL encoded to protect Unicode chars
301 *
302 * @author Andreas Gohr <andi@splitbrain.org>
303 */
304function mediaFN($id){
305  global $conf;
306  $id = cleanID($id);
307  $id = str_replace(':','/',$id);
308    $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
309  return $fn;
310}
311
312/**
313 * Returns the full filepath to a localized textfile if local
314 * version isn't found the english one is returned
315 *
316 * @author Andreas Gohr <andi@splitbrain.org>
317 */
318function localeFN($id){
319  global $conf;
320  $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.txt';
321  if(!@file_exists($file)){
322    //fall back to english
323    $file = DOKU_INC.'inc/lang/en/'.$id.'.txt';
324  }
325  return $file;
326}
327
328/**
329 * Resolve relative paths in IDs
330 *
331 * Do not call directly use resolve_mediaid or resolve_pageid
332 * instead
333 *
334 * Partyly based on a cleanPath function found at
335 * http://www.php.net/manual/en/function.realpath.php#57016
336 *
337 * @author <bart at mediawave dot nl>
338 */
339function resolve_id($ns,$id,$clean=true){
340  global $conf;
341
342  // some pre cleaning for useslash:
343  if($conf['useslash']) $id = str_replace('/',':',$id);
344
345  // if the id starts with a dot we need to handle the
346  // relative stuff
347  if($id{0} == '.'){
348    // normalize initial dots without a colon
349    $id = preg_replace('/^(\.+)(?=[^:\.])/','\1:',$id);
350    // prepend the current namespace
351    $id = $ns.':'.$id;
352
353    // cleanup relatives
354    $result = array();
355    $pathA  = explode(':', $id);
356    if (!$pathA[0]) $result[] = '';
357    foreach ($pathA AS $key => $dir) {
358      if ($dir == '..') {
359        if (end($result) == '..') {
360          $result[] = '..';
361        } elseif (!array_pop($result)) {
362          $result[] = '..';
363        }
364      } elseif ($dir && $dir != '.') {
365        $result[] = $dir;
366      }
367    }
368    if (!end($pathA)) $result[] = '';
369    $id = implode(':', $result);
370  }elseif($ns !== false && strpos($id,':') === false){
371    //if link contains no namespace. add current namespace (if any)
372    $id = $ns.':'.$id;
373  }
374
375  if($clean) $id = cleanID($id);
376  return $id;
377}
378
379/**
380 * Returns a full media id
381 *
382 * @author Andreas Gohr <andi@splitbrain.org>
383 */
384function resolve_mediaid($ns,&$page,&$exists){
385  $page   = resolve_id($ns,$page);
386  $file   = mediaFN($page);
387  $exists = @file_exists($file);
388}
389
390/**
391 * Returns a full page id
392 *
393 * @author Andreas Gohr <andi@splitbrain.org>
394 */
395function resolve_pageid($ns,&$page,&$exists){
396  global $conf;
397  $exists = false;
398
399  //keep hashlink if exists then clean both parts
400  if (strpos($page,'#')) {
401    list($page,$hash) = split('#',$page,2);
402  } else {
403    $hash = '';
404  }
405  $hash = cleanID($hash);
406  $page = resolve_id($ns,$page,false); // resolve but don't clean, yet
407
408  // get filename (calls clean itself)
409  $file = wikiFN($page);
410
411  // if ends with colon or slash we have a namespace link
412  if(substr($page,-1) == ':' || ($conf['useslash'] && substr($page,-1) == '/')){
413    if(page_exists($page.$conf['start'])){
414      // start page inside namespace
415      $page = $page.$conf['start'];
416      $exists = true;
417    }elseif(page_exists($page.noNS(cleanID($page)))){
418      // page named like the NS inside the NS
419      $page = $page.noNS(cleanID($page));
420      $exists = true;
421    }elseif(page_exists($page)){
422      // page like namespace exists
423      $page = $page;
424      $exists = true;
425    }else{
426      // fall back to default
427      $page = $page.$conf['start'];
428    }
429  }else{
430    //check alternative plural/nonplural form
431    if(!@file_exists($file)){
432      if( $conf['autoplural'] ){
433        if(substr($page,-1) == 's'){
434          $try = substr($page,0,-1);
435        }else{
436          $try = $page.'s';
437        }
438        if(page_exists($try)){
439          $page   = $try;
440          $exists = true;
441        }
442      }
443    }else{
444      $exists = true;
445    }
446  }
447
448  // now make sure we have a clean page
449  $page = cleanID($page);
450
451  //add hash if any
452  if(!empty($hash)) $page .= '#'.$hash;
453}
454
455/**
456 * Returns the name of a cachefile from given data
457 *
458 * The needed directory is created by this function!
459 *
460 * @author Andreas Gohr <andi@splitbrain.org>
461 *
462 * @param string $data  This data is used to create a unique md5 name
463 * @param string $ext   This is appended to the filename if given
464 * @return string       The filename of the cachefile
465 */
466function getCacheName($data,$ext=''){
467  global $conf;
468  $md5  = md5($data);
469  $file = $conf['cachedir'].'/'.$md5{0}.'/'.$md5.$ext;
470  io_makeFileDir($file);
471  return $file;
472}
473
474/**
475 * Checks a pageid against $conf['hidepages']
476 *
477 * @author Andreas Gohr <gohr@cosmocode.de>
478 */
479function isHiddenPage($id){
480  global $conf;
481  if(empty($conf['hidepages'])) return false;
482
483  if(preg_match('/'.$conf['hidepages'].'/ui',':'.$id)){
484    return true;
485  }
486  return false;
487}
488
489/**
490 * Reverse of isHiddenPage
491 *
492 * @author Andreas Gohr <gohr@cosmocode.de>
493 */
494function isVisiblePage($id){
495  return !isHiddenPage($id);
496}
497
498/**
499 * Checks and sets HTTP headers for conditional HTTP requests
500 *
501 * @author   Simon Willison <swillison@gmail.com>
502 * @link     http://simon.incutio.com/archive/2003/04/23/conditionalGet
503 * @param    timestamp $timestamp lastmodified time of the cache file
504 * @returns  void or void with previously header() commands executed
505 */
506function http_conditionalRequest($timestamp){
507  // A PHP implementation of conditional get, see
508  //   http://fishbowl.pastiche.org/archives/001132.html
509  $last_modified = substr(gmdate('r', $timestamp), 0, -5).'GMT';
510  $etag = '"'.md5($last_modified).'"';
511  // Send the headers
512  header("Last-Modified: $last_modified");
513  header("ETag: $etag");
514  // See if the client has provided the required headers
515  if (isset($_SERVER['HTTP_IF_MODIFIED_SINCE'])){
516    $if_modified_since = stripslashes($_SERVER['HTTP_IF_MODIFIED_SINCE']);
517  }else{
518    $if_modified_since = false;
519  }
520
521  if (isset($_SERVER['HTTP_IF_NONE_MATCH'])){
522    $if_none_match = stripslashes($_SERVER['HTTP_IF_NONE_MATCH']);
523  }else{
524    $if_none_match = false;
525  }
526
527  if (!$if_modified_since && !$if_none_match){
528    return;
529  }
530
531  // At least one of the headers is there - check them
532  if ($if_none_match && $if_none_match != $etag) {
533    return; // etag is there but doesn't match
534  }
535
536  if ($if_modified_since && $if_modified_since != $last_modified) {
537    return; // if-modified-since is there but doesn't match
538  }
539
540  // Nothing has changed since their last request - serve a 304 and exit
541  header('HTTP/1.0 304 Not Modified');
542
543  // don't produce output, even if compression is on
544  ob_end_clean();
545  exit;
546}
547
548//Setup VIM: ex: et ts=2 enc=utf-8 :
549