xref: /dokuwiki/inc/pageutils.php (revision 63b0c1a70fbc1d51a775e2fa30b6a520bac2a6e4) 
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 * @param  boolean $media     Allow leading or trailing _ for media files
90 */
91function cleanID($raw_id,$ascii=false,$media=false){
92  global $conf;
93  global $lang;
94  static $sepcharpat = null;
95
96  global $cache_cleanid;
97  $cache = & $cache_cleanid;
98
99  // check if it's already in the memory cache
100  if (isset($cache[(string)$raw_id])) {
101    return $cache[(string)$raw_id];
102    }
103
104  $sepchar = $conf['sepchar'];
105  if($sepcharpat == null) // build string only once to save clock cycles
106    $sepcharpat = '#\\'.$sepchar.'+#';
107
108  $id = trim((string)$raw_id);
109  $id = utf8_strtolower($id);
110
111  //alternative namespace seperator
112  $id = strtr($id,';',':');
113  if($conf['useslash']){
114    $id = strtr($id,'/',':');
115  }else{
116    $id = strtr($id,'/',$sepchar);
117  }
118
119  if($conf['deaccent'] == 2 || $ascii) $id = utf8_romanize($id);
120  if($conf['deaccent'] || $ascii) $id = utf8_deaccent($id,-1);
121
122  //remove specials
123  $id = utf8_stripspecials($id,$sepchar,'\*');
124
125  if($ascii) $id = utf8_strip($id);
126
127  //clean up
128  $id = preg_replace($sepcharpat,$sepchar,$id);
129  $id = preg_replace('#:+#',':',$id);
130  $id = ($media ? trim($id,':.-') : trim($id,':._-'));
131  $id = preg_replace('#:[:\._\-]+#',':',$id);
132
133  $cache[(string)$raw_id] = $id;
134  return($id);
135}
136
137/**
138 * Return namespacepart of a wiki ID
139 *
140 * @author Andreas Gohr <andi@splitbrain.org>
141 */
142function getNS($id){
143  $pos = strrpos((string)$id,':');
144  if($pos!==false){
145    return substr((string)$id,0,$pos);
146  }
147  return false;
148}
149
150/**
151 * Returns the ID without the namespace
152 *
153 * @author Andreas Gohr <andi@splitbrain.org>
154 */
155function noNS($id) {
156  $pos = strrpos($id, ':');
157  if ($pos!==false) {
158    return substr($id, $pos+1);
159  } else {
160    return $id;
161  }
162}
163
164/**
165* Returns the current namespace
166*
167* @author Nathan Fritz <fritzn@crown.edu>
168*/
169function curNS($id) {
170    return noNS(getNS($id));
171}
172
173/**
174* Returns the ID without the namespace or current namespace for 'start' pages
175*
176* @author Nathan Fritz <fritzn@crown.edu>
177*/
178function noNSorNS($id) {
179    global $conf;
180
181    $p = noNS($id);
182    if ($p == $conf['start']) {
183        $p = curNS($id);
184        if ($p == false) {
185            return noNS($id);
186        }
187    }
188    return $p;
189}
190
191/**
192 *  Wiki page existence check
193 *
194 *  parameters as for wikiFN
195 *
196 *  @author Chris Smith <chris@jalakai.co.uk>
197 */
198function page_exists($id,$rev='',$clean=true) {
199  return @file_exists(wikiFN($id,$rev,$clean));
200}
201
202/**
203 * returns the full path to the datafile specified by ID and optional revision
204 *
205 * The filename is URL encoded to protect Unicode chars
206 *
207 * @param  $raw_id  string   id of wikipage
208 * @param  $rev     string   page revision, empty string for current
209 * @param  $clean   bool     flag indicating that $raw_id should be cleaned.  Only set to false
210 *                           when $id is guaranteed to have been cleaned already.
211 *
212 * @author Andreas Gohr <andi@splitbrain.org>
213 */
214function wikiFN($raw_id,$rev='',$clean=true){
215  global $conf;
216
217  global $cache_wikifn;
218  $cache = & $cache_wikifn;
219
220  if (isset($cache[$raw_id]) && isset($cache[$raw_id][$rev])) {
221    return $cache[$raw_id][$rev];
222  }
223
224  $id = $raw_id;
225
226  if ($clean) $id = cleanID($id);
227  $id = str_replace(':','/',$id);
228  if(empty($rev)){
229    $fn = $conf['datadir'].'/'.utf8_encodeFN($id).'.txt';
230  }else{
231    $fn = $conf['olddir'].'/'.utf8_encodeFN($id).'.'.$rev.'.txt';
232    if($conf['compression']){
233      //test for extensions here, we want to read both compressions
234       if (@file_exists($fn . '.gz')){
235          $fn .= '.gz';
236       }else if(@file_exists($fn . '.bz2')){
237          $fn .= '.bz2';
238       }else{
239          //file doesnt exist yet, so we take the configured extension
240          $fn .= '.' . $conf['compression'];
241       }
242    }
243  }
244
245  if (!isset($cache[$raw_id])) { $cache[$raw_id] = array(); }
246  $cache[$raw_id][$rev] = $fn;
247  return $fn;
248}
249
250/**
251 * Returns the full path to the file for locking the page while editing.
252 *
253 * @author Ben Coburn <btcoburn@silicodon.net>
254 */
255function wikiLockFN($id) {
256  global $conf;
257  return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
258}
259
260
261/**
262 * returns the full path to the meta file specified by ID and extension
263 *
264 * The filename is URL encoded to protect Unicode chars
265 *
266 * @author Steven Danz <steven-danz@kc.rr.com>
267 */
268function metaFN($id,$ext){
269  global $conf;
270  $id = cleanID($id);
271  $id = str_replace(':','/',$id);
272  $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
273  return $fn;
274}
275
276/**
277 * returns an array of full paths to all metafiles of a given ID
278 *
279 * @author Esther Brunner <esther@kaffeehaus.ch>
280 */
281function metaFiles($id){
282   $name   = noNS($id);
283   $ns     = getNS($id);
284   $dir    = ($ns) ? metaFN($ns,'').'/' : metaFN($ns,'');
285   $files  = array();
286
287   $dh = @opendir($dir);
288   if(!$dh) return $files;
289   while(($file = readdir($dh)) !== false){
290     if(strpos($file,$name.'.') === 0 && !is_dir($dir.$file))
291       $files[] = $dir.$file;
292   }
293   closedir($dh);
294
295   return $files;
296}
297
298/**
299 * returns the full path to the mediafile specified by ID
300 *
301 * The filename is URL encoded to protect Unicode chars
302 *
303 * @author Andreas Gohr <andi@splitbrain.org>
304 */
305function mediaFN($id){
306  global $conf;
307  $id = cleanID($id);
308  $id = str_replace(':','/',$id);
309    $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
310  return $fn;
311}
312
313/**
314 * Returns the full filepath to a localized textfile if local
315 * version isn't found the english one is returned
316 *
317 * @author Andreas Gohr <andi@splitbrain.org>
318 */
319function localeFN($id){
320  global $conf;
321  $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.txt';
322  if(!@file_exists($file)){
323    //fall back to english
324    $file = DOKU_INC.'inc/lang/en/'.$id.'.txt';
325  }
326  return $file;
327}
328
329/**
330 * Resolve relative paths in IDs
331 *
332 * Do not call directly use resolve_mediaid or resolve_pageid
333 * instead
334 *
335 * Partyly based on a cleanPath function found at
336 * http://www.php.net/manual/en/function.realpath.php#57016
337 *
338 * @author <bart at mediawave dot nl>
339 */
340function resolve_id($ns,$id,$clean=true){
341  global $conf;
342
343  // some pre cleaning for useslash:
344  if($conf['useslash']) $id = str_replace('/',':',$id);
345
346  // if the id starts with a dot we need to handle the
347  // relative stuff
348  if($id{0} == '.'){
349    // normalize initial dots without a colon
350    $id = preg_replace('/^(\.+)(?=[^:\.])/','\1:',$id);
351    // prepend the current namespace
352    $id = $ns.':'.$id;
353
354    // cleanup relatives
355    $result = array();
356    $pathA  = explode(':', $id);
357    if (!$pathA[0]) $result[] = '';
358    foreach ($pathA AS $key => $dir) {
359      if ($dir == '..') {
360        if (end($result) == '..') {
361          $result[] = '..';
362        } elseif (!array_pop($result)) {
363          $result[] = '..';
364        }
365      } elseif ($dir && $dir != '.') {
366        $result[] = $dir;
367      }
368    }
369    if (!end($pathA)) $result[] = '';
370    $id = implode(':', $result);
371  }elseif($ns !== false && strpos($id,':') === false){
372    //if link contains no namespace. add current namespace (if any)
373    $id = $ns.':'.$id;
374  }
375
376  if($clean) $id = cleanID($id);
377  return $id;
378}
379
380/**
381 * Returns a full media id
382 *
383 * @author Andreas Gohr <andi@splitbrain.org>
384 */
385function resolve_mediaid($ns,&$page,&$exists){
386  $page   = resolve_id($ns,$page);
387  $file   = mediaFN($page);
388  $exists = @file_exists($file);
389}
390
391/**
392 * Returns a full page id
393 *
394 * @author Andreas Gohr <andi@splitbrain.org>
395 */
396function resolve_pageid($ns,&$page,&$exists){
397  global $conf;
398  $exists = false;
399
400  //keep hashlink if exists then clean both parts
401  if (strpos($page,'#')) {
402    list($page,$hash) = split('#',$page,2);
403  } else {
404    $hash = '';
405  }
406  $hash = cleanID($hash);
407  $page = resolve_id($ns,$page,false); // resolve but don't clean, yet
408
409  // get filename (calls clean itself)
410  $file = wikiFN($page);
411
412  // if ends with colon or slash we have a namespace link
413  if(substr($page,-1) == ':' || ($conf['useslash'] && substr($page,-1) == '/')){
414    if(page_exists($page.$conf['start'])){
415      // start page inside namespace
416      $page = $page.$conf['start'];
417      $exists = true;
418    }elseif(page_exists($page.noNS(cleanID($page)))){
419      // page named like the NS inside the NS
420      $page = $page.noNS(cleanID($page));
421      $exists = true;
422    }elseif(page_exists($page)){
423      // page like namespace exists
424      $page = $page;
425      $exists = true;
426    }else{
427      // fall back to default
428      $page = $page.$conf['start'];
429    }
430  }else{
431    //check alternative plural/nonplural form
432    if(!@file_exists($file)){
433      if( $conf['autoplural'] ){
434        if(substr($page,-1) == 's'){
435          $try = substr($page,0,-1);
436        }else{
437          $try = $page.'s';
438        }
439        if(page_exists($try)){
440          $page   = $try;
441          $exists = true;
442        }
443      }
444    }else{
445      $exists = true;
446    }
447  }
448
449  // now make sure we have a clean page
450  $page = cleanID($page);
451
452  //add hash if any
453  if(!empty($hash)) $page .= '#'.$hash;
454}
455
456/**
457 * Returns the name of a cachefile from given data
458 *
459 * The needed directory is created by this function!
460 *
461 * @author Andreas Gohr <andi@splitbrain.org>
462 *
463 * @param string $data  This data is used to create a unique md5 name
464 * @param string $ext   This is appended to the filename if given
465 * @return string       The filename of the cachefile
466 */
467function getCacheName($data,$ext=''){
468  global $conf;
469  $md5  = md5($data);
470  $file = $conf['cachedir'].'/'.$md5{0}.'/'.$md5.$ext;
471  io_makeFileDir($file);
472  return $file;
473}
474
475/**
476 * Checks a pageid against $conf['hidepages']
477 *
478 * @author Andreas Gohr <gohr@cosmocode.de>
479 */
480function isHiddenPage($id){
481  global $conf;
482  if(empty($conf['hidepages'])) return false;
483
484  if(preg_match('/'.$conf['hidepages'].'/ui',':'.$id)){
485    return true;
486  }
487  return false;
488}
489
490/**
491 * Reverse of isHiddenPage
492 *
493 * @author Andreas Gohr <gohr@cosmocode.de>
494 */
495function isVisiblePage($id){
496  return !isHiddenPage($id);
497}
498
499/**
500 * Checks and sets HTTP headers for conditional HTTP requests
501 *
502 * @author   Simon Willison <swillison@gmail.com>
503 * @link     http://simon.incutio.com/archive/2003/04/23/conditionalGet
504 * @param    timestamp $timestamp lastmodified time of the cache file
505 * @returns  void or void with previously header() commands executed
506 */
507function http_conditionalRequest($timestamp){
508  // A PHP implementation of conditional get, see
509  //   http://fishbowl.pastiche.org/archives/001132.html
510  $last_modified = substr(gmdate('r', $timestamp), 0, -5).'GMT';
511  $etag = '"'.md5($last_modified).'"';
512  // Send the headers
513  header("Last-Modified: $last_modified");
514  header("ETag: $etag");
515  // See if the client has provided the required headers
516  if (isset($_SERVER['HTTP_IF_MODIFIED_SINCE'])){
517    $if_modified_since = stripslashes($_SERVER['HTTP_IF_MODIFIED_SINCE']);
518  }else{
519    $if_modified_since = false;
520  }
521
522  if (isset($_SERVER['HTTP_IF_NONE_MATCH'])){
523    $if_none_match = stripslashes($_SERVER['HTTP_IF_NONE_MATCH']);
524  }else{
525    $if_none_match = false;
526  }
527
528  if (!$if_modified_since && !$if_none_match){
529    return;
530  }
531
532  // At least one of the headers is there - check them
533  if ($if_none_match && $if_none_match != $etag) {
534    return; // etag is there but doesn't match
535  }
536
537  if ($if_modified_since && $if_modified_since != $last_modified) {
538    return; // if-modified-since is there but doesn't match
539  }
540
541  // Nothing has changed since their last request - serve a 304 and exit
542  header('HTTP/1.0 304 Not Modified');
543
544  // don't produce output, even if compression is on
545  ob_end_clean();
546  exit;
547}
548
549//Setup VIM: ex: et ts=2 enc=utf-8 :
550