xref: /dokuwiki/inc/pageutils.php (revision 4ab889ea63838db0bbb33dd0d316eac03ab69cf2) 
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($id,-1) == '/')){
58    if(@file_exists(wikiFN($id.$conf['start']))){
59      // start page inside namespace
60      $id = $id.$conf['start'];
61    }elseif(@file_exists(wikiFN($id.noNS(cleanID($id))))){
62      // page named like the NS inside the NS
63      $id = $id.noNS(cleanID($id));
64    }elseif(@file_exists(wikiFN($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 $p;
188}
189
190/**
191 * returns the full path to the datafile specified by ID and
192 * optional revision
193 *
194 * The filename is URL encoded to protect Unicode chars
195 *
196 * @author Andreas Gohr <andi@splitbrain.org>
197 */
198function wikiFN($raw_id,$rev='',$clean=true){
199  global $conf;
200
201  global $cache_wikifn;
202  $cache = & $cache_wikifn;
203
204  if (isset($cache[$raw_id]) && isset($cache[$raw_id][$rev])) {
205    return $cache[$raw_id][$rev];
206  }
207
208  $id = $raw_id;
209
210  if ($clean) $id = cleanID($id);
211  $id = str_replace(':','/',$id);
212  if(empty($rev)){
213    $fn = $conf['datadir'].'/'.utf8_encodeFN($id).'.txt';
214  }else{
215    $fn = $conf['olddir'].'/'.utf8_encodeFN($id).'.'.$rev.'.txt';
216    if($conf['compression']){
217      //test for extensions here, we want to read both compressions
218       if (@file_exists($fn . '.gz')){
219          $fn .= '.gz';
220       }else if(@file_exists($fn . '.bz2')){
221          $fn .= '.bz2';
222       }else{
223          //file doesnt exist yet, so we take the configured extension
224          $fn .= '.' . $conf['compression'];
225       }
226    }
227  }
228
229  if (!isset($cache[$raw_id])) { $cache[$raw_id] = array(); }
230  $cache[$raw_id][$rev] = $fn;
231  return $fn;
232}
233
234/**
235 * Returns the full path to the file for locking the page while editing.
236 *
237 * @author Ben Coburn <btcoburn@silicodon.net>
238 */
239function wikiLockFN($id) {
240  global $conf;
241  return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
242}
243
244
245/**
246 * returns the full path to the meta file specified by ID and extension
247 *
248 * The filename is URL encoded to protect Unicode chars
249 *
250 * @author Steven Danz <steven-danz@kc.rr.com>
251 */
252function metaFN($id,$ext){
253  global $conf;
254  $id = cleanID($id);
255  $id = str_replace(':','/',$id);
256  $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
257  return $fn;
258}
259
260/**
261 * returns an array of full paths to all metafiles of a given ID
262 *
263 * @author Esther Brunner <esther@kaffeehaus.ch>
264 */
265function metaFiles($id){
266   $name   = noNS($id);
267   $dir    = metaFN(getNS($id),'');
268   $files  = array();
269
270   $dh = @opendir($dir);
271   if(!$dh) return $files;
272   while(($file = readdir($dh)) !== false){
273     if(strpos($file,$name.'.') === 0 && !is_dir($dir.$file))
274       $files[] = $dir.$file;
275   }
276   closedir($dh);
277
278   return $files;
279}
280
281/**
282 * returns the full path to the mediafile specified by ID
283 *
284 * The filename is URL encoded to protect Unicode chars
285 *
286 * @author Andreas Gohr <andi@splitbrain.org>
287 */
288function mediaFN($id){
289  global $conf;
290  $id = cleanID($id);
291  $id = str_replace(':','/',$id);
292    $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
293  return $fn;
294}
295
296/**
297 * Returns the full filepath to a localized textfile if local
298 * version isn't found the english one is returned
299 *
300 * @author Andreas Gohr <andi@splitbrain.org>
301 */
302function localeFN($id){
303  global $conf;
304  $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.txt';
305  if(!@file_exists($file)){
306    //fall back to english
307    $file = DOKU_INC.'inc/lang/en/'.$id.'.txt';
308  }
309  return $file;
310}
311
312/**
313 * Resolve relative paths in IDs
314 *
315 * Do not call directly use resolve_mediaid or resolve_pageid
316 * instead
317 *
318 * Partyly based on a cleanPath function found at
319 * http://www.php.net/manual/en/function.realpath.php#57016
320 *
321 * @author <bart at mediawave dot nl>
322 */
323function resolve_id($ns,$id,$clean=true){
324  global $conf;
325
326  // some pre cleaning for useslash:
327  if($conf['useslash']) $id = str_replace('/',':',$id);
328
329  // if the id starts with a dot we need to handle the
330  // relative stuff
331  if($id{0} == '.'){
332    // normalize initial dots without a colon
333    $id = preg_replace('/^(\.+)(?=[^:\.])/','\1:',$id);
334    // prepend the current namespace
335    $id = $ns.':'.$id;
336
337    // cleanup relatives
338    $result = array();
339    $pathA  = explode(':', $id);
340    if (!$pathA[0]) $result[] = '';
341    foreach ($pathA AS $key => $dir) {
342      if ($dir == '..') {
343        if (end($result) == '..') {
344          $result[] = '..';
345        } elseif (!array_pop($result)) {
346          $result[] = '..';
347        }
348      } elseif ($dir && $dir != '.') {
349        $result[] = $dir;
350      }
351    }
352    if (!end($pathA)) $result[] = '';
353    $id = implode(':', $result);
354  }elseif($ns !== false && strpos($id,':') === false){
355    //if link contains no namespace. add current namespace (if any)
356    $id = $ns.':'.$id;
357  }
358
359  if($clean) $id = cleanID($id);
360  return $id;
361}
362
363/**
364 * Returns a full media id
365 *
366 * @author Andreas Gohr <andi@splitbrain.org>
367 */
368function resolve_mediaid($ns,&$page,&$exists){
369  $page   = resolve_id($ns,$page);
370  $file   = mediaFN($page);
371  $exists = @file_exists($file);
372}
373
374/**
375 * Returns a full page id
376 *
377 * @author Andreas Gohr <andi@splitbrain.org>
378 */
379function resolve_pageid($ns,&$page,&$exists){
380  global $conf;
381  $exists = false;
382
383  //keep hashlink if exists then clean both parts
384  if (strpos($page,'#')) {
385    list($page,$hash) = split('#',$page,2);
386  } else {
387    $hash = '';
388  }
389  $hash = cleanID($hash);
390  $page = resolve_id($ns,$page,false); // resolve but don't clean, yet
391
392  // get filename (calls clean itself)
393  $file = wikiFN($page);
394
395  // if ends with colon or slash we have a namespace link
396  if(substr($page,-1) == ':' || ($conf['useslash'] && substr($page,-1) == '/')){
397    if(@file_exists(wikiFN($page.$conf['start']))){
398      // start page inside namespace
399      $page = $page.$conf['start'];
400      $exists = true;
401    }elseif(@file_exists(wikiFN($page.noNS(cleanID($page))))){
402      // page named like the NS inside the NS
403      $page = $page.noNS(cleanID($page));
404      $exists = true;
405    }elseif(@file_exists(wikiFN($page))){
406      // page like namespace exists
407      $page = $page;
408      $exists = true;
409    }else{
410      // fall back to default
411      $page = $page.$conf['start'];
412    }
413  }else{
414    //check alternative plural/nonplural form
415    if(!@file_exists($file)){
416      if( $conf['autoplural'] ){
417        if(substr($page,-1) == 's'){
418          $try = substr($page,0,-1);
419        }else{
420          $try = $page.'s';
421        }
422        if(@file_exists(wikiFN($try))){
423          $page   = $try;
424          $exists = true;
425        }
426      }
427    }else{
428      $exists = true;
429    }
430  }
431
432  // now make sure we have a clean page
433  $page = cleanID($page);
434
435  //add hash if any
436  if(!empty($hash)) $page .= '#'.$hash;
437}
438
439/**
440 * Returns the name of a cachefile from given data
441 *
442 * The needed directory is created by this function!
443 *
444 * @author Andreas Gohr <andi@splitbrain.org>
445 *
446 * @param string $data  This data is used to create a unique md5 name
447 * @param string $ext   This is appended to the filename if given
448 * @return string       The filename of the cachefile
449 */
450function getCacheName($data,$ext=''){
451  global $conf;
452  $md5  = md5($data);
453  $file = $conf['cachedir'].'/'.$md5{0}.'/'.$md5.$ext;
454  io_makeFileDir($file);
455  return $file;
456}
457
458/**
459 * Checks a pageid against $conf['hidepages']
460 *
461 * @author Andreas Gohr <gohr@cosmocode.de>
462 */
463function isHiddenPage($id){
464  global $conf;
465  if(empty($conf['hidepages'])) return false;
466
467  if(preg_match('/'.$conf['hidepages'].'/ui',':'.$id)){
468    return true;
469  }
470  return false;
471}
472
473/**
474 * Reverse of isHiddenPage
475 *
476 * @author Andreas Gohr <gohr@cosmocode.de>
477 */
478function isVisiblePage($id){
479  return !isHiddenPage($id);
480}
481
482/**
483 * Checks and sets HTTP headers for conditional HTTP requests
484 *
485 * @author   Simon Willison <swillison@gmail.com>
486 * @link     http://simon.incutio.com/archive/2003/04/23/conditionalGet
487 * @param    timestamp $timestamp lastmodified time of the cache file
488 * @returns  void or void with previously header() commands executed
489 */
490function http_conditionalRequest($timestamp){
491  // A PHP implementation of conditional get, see
492  //   http://fishbowl.pastiche.org/archives/001132.html
493  $last_modified = substr(gmdate('r', $timestamp), 0, -5).'GMT';
494  $etag = '"'.md5($last_modified).'"';
495  // Send the headers
496  header("Last-Modified: $last_modified");
497  header("ETag: $etag");
498  // See if the client has provided the required headers
499  if (isset($_SERVER['HTTP_IF_MODIFIED_SINCE'])){
500    $if_modified_since = stripslashes($_SERVER['HTTP_IF_MODIFIED_SINCE']);
501  }else{
502    $if_modified_since = false;
503  }
504
505  if (isset($_SERVER['HTTP_IF_NONE_MATCH'])){
506    $if_none_match = stripslashes($_SERVER['HTTP_IF_NONE_MATCH']);
507  }else{
508    $if_none_match = false;
509  }
510
511  if (!$if_modified_since && !$if_none_match){
512    return;
513  }
514
515  // At least one of the headers is there - check them
516  if ($if_none_match && $if_none_match != $etag) {
517    return; // etag is there but doesn't match
518  }
519
520  if ($if_modified_since && $if_modified_since != $last_modified) {
521    return; // if-modified-since is there but doesn't match
522  }
523
524  // Nothing has changed since their last request - serve a 304 and exit
525  header('HTTP/1.0 304 Not Modified');
526  exit;
527}
528
529//Setup VIM: ex: et ts=2 enc=utf-8 :
530