xref: /dokuwiki/inc/pageutils.php (revision 671a58a68aaa7c847262d0c5c34541e6d214b5fc) 
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 = $id;
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 $id;
160  }
161}
162
163/**
164 * returns the full path to the datafile specified by ID and
165 * optional revision
166 *
167 * The filename is URL encoded to protect Unicode chars
168 *
169 * @author Andreas Gohr <andi@splitbrain.org>
170 */
171function wikiFN($raw_id,$rev='',$clean=true){
172  global $conf;
173
174  global $cache_wikifn;
175  $cache = & $cache_wikifn;
176
177  if (isset($cache[$raw_id]) && isset($cache[$raw_id][$rev])) {
178    return $cache[$raw_id][$rev];
179  }
180
181  $id = $raw_id;
182
183  if ($clean) $id = cleanID($id);
184  $id = str_replace(':','/',$id);
185  if(empty($rev)){
186    $fn = $conf['datadir'].'/'.utf8_encodeFN($id).'.txt';
187  }else{
188    $fn = $conf['olddir'].'/'.utf8_encodeFN($id).'.'.$rev.'.txt';
189    if($conf['compression']){
190      //test for extensions here, we want to read both compressions
191       if (@file_exists($fn . '.gz')){
192          $fn .= '.gz';
193       }else if(@file_exists($fn . '.bz2')){
194          $fn .= '.bz2';
195       }else{
196          //file doesnt exist yet, so we take the configured extension
197          $fn .= '.' . $conf['compression'];
198       }
199    }
200  }
201
202  if (!isset($cache[$raw_id])) { $cache[$raw_id] = array(); }
203  $cache[$raw_id][$rev] = $fn;
204  return $fn;
205}
206
207/**
208 * Returns the full path to the file for locking the page while editing.
209 *
210 * @author Ben Coburn <btcoburn@silicodon.net>
211 */
212function wikiLockFN($id) {
213  global $conf;
214  return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
215}
216
217
218/**
219 * returns the full path to the meta file specified by ID and extension
220 *
221 * The filename is URL encoded to protect Unicode chars
222 *
223 * @author Steven Danz <steven-danz@kc.rr.com>
224 */
225function metaFN($id,$ext){
226  global $conf;
227  $id = cleanID($id);
228  $id = str_replace(':','/',$id);
229  $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
230  return $fn;
231}
232
233/**
234 * returns an array of full paths to all metafiles of a given ID
235 *
236 * @author Esther Brunner <esther@kaffeehaus.ch>
237 */
238function metaFiles($id){
239   $name   = noNS($id);
240   $dir    = metaFN(getNS($id),'');
241   $files  = array();
242
243   $dh = @opendir($dir);
244   if(!$dh) return $files;
245   while(($file = readdir($dh)) !== false){
246     if(strpos($file,$name.'.') === 0 && !is_dir($dir.$file))
247       $files[] = $dir.$file;
248   }
249   closedir($dh);
250
251   return $files;
252}
253
254/**
255 * returns the full path to the mediafile specified by ID
256 *
257 * The filename is URL encoded to protect Unicode chars
258 *
259 * @author Andreas Gohr <andi@splitbrain.org>
260 */
261function mediaFN($id){
262  global $conf;
263  $id = cleanID($id);
264  $id = str_replace(':','/',$id);
265    $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
266  return $fn;
267}
268
269/**
270 * Returns the full filepath to a localized textfile if local
271 * version isn't found the english one is returned
272 *
273 * @author Andreas Gohr <andi@splitbrain.org>
274 */
275function localeFN($id){
276  global $conf;
277  $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.txt';
278  if(!@file_exists($file)){
279    //fall back to english
280    $file = DOKU_INC.'inc/lang/en/'.$id.'.txt';
281  }
282  return $file;
283}
284
285/**
286 * Resolve relative paths in IDs
287 *
288 * Do not call directly use resolve_mediaid or resolve_pageid
289 * instead
290 *
291 * Partyly based on a cleanPath function found at
292 * http://www.php.net/manual/en/function.realpath.php#57016
293 *
294 * @author <bart at mediawave dot nl>
295 */
296function resolve_id($ns,$id,$clean=true){
297  // if the id starts with a dot we need to handle the
298  // relative stuff
299  if($id{0} == '.'){
300    // normalize initial dots without a colon
301    $id = preg_replace('/^(\.+)(?=[^:\.])/','\1:',$id);
302    // prepend the current namespace
303    $id = $ns.':'.$id;
304
305    // cleanup relatives
306    $result = array();
307    $pathA  = explode(':', $id);
308    if (!$pathA[0]) $result[] = '';
309    foreach ($pathA AS $key => $dir) {
310      if ($dir == '..') {
311        if (end($result) == '..') {
312          $result[] = '..';
313        } elseif (!array_pop($result)) {
314          $result[] = '..';
315        }
316      } elseif ($dir && $dir != '.') {
317        $result[] = $dir;
318      }
319    }
320    if (!end($pathA)) $result[] = '';
321    $id = implode(':', $result);
322  }elseif($ns !== false && strpos($id,':') === false){
323    //if link contains no namespace. add current namespace (if any)
324    $id = $ns.':'.$id;
325  }
326
327  if($clean) $id = cleanID($id);
328  return $id;
329}
330
331/**
332 * Returns a full media id
333 *
334 * @author Andreas Gohr <andi@splitbrain.org>
335 */
336function resolve_mediaid($ns,&$page,&$exists){
337  $page   = resolve_id($ns,$page);
338  $file   = mediaFN($page);
339  $exists = @file_exists($file);
340}
341
342/**
343 * Returns a full page id
344 *
345 * @author Andreas Gohr <andi@splitbrain.org>
346 */
347function resolve_pageid($ns,&$page,&$exists){
348  global $conf;
349  $exists = false;
350
351  //keep hashlink if exists then clean both parts
352  if (strpos($page,'#')) {
353    list($page,$hash) = split('#',$page,2);
354  } else {
355    $hash = '';
356  }
357  $hash = cleanID($hash);
358  $page = resolve_id($ns,$page,false); // resolve but don't clean, yet
359
360  // get filename (calls clean itself)
361  $file = wikiFN($page);
362
363  // if ends with colon or slash we have a namespace link
364  if(substr($page,-1) == ':' || ($conf['useslash'] && substr($page,-1) == '/')){
365    if(@file_exists(wikiFN($page.$conf['start']))){
366      // start page inside namespace
367      $page = $page.$conf['start'];
368      $exists = true;
369    }elseif(@file_exists(wikiFN($page.noNS(cleanID($page))))){
370      // page named like the NS inside the NS
371      $page = $page.noNS(cleanID($page));
372      $exists = true;
373    }elseif(@file_exists(wikiFN($page))){
374      // page like namespace exists
375      $page = $page;
376      $exists = true;
377    }else{
378      // fall back to default
379      $page = $page.$conf['start'];
380    }
381  }else{
382    //check alternative plural/nonplural form
383    if(!@file_exists($file)){
384      if( $conf['autoplural'] ){
385        if(substr($page,-1) == 's'){
386          $try = substr($page,0,-1);
387        }else{
388          $try = $page.'s';
389        }
390        if(@file_exists(wikiFN($try))){
391          $page   = $try;
392          $exists = true;
393        }
394      }
395    }else{
396      $exists = true;
397    }
398  }
399
400  // now make sure we have a clean page
401  $page = cleanID($page);
402
403  //add hash if any
404  if(!empty($hash)) $page .= '#'.$hash;
405}
406
407/**
408 * Returns the name of a cachefile from given data
409 *
410 * The needed directory is created by this function!
411 *
412 * @author Andreas Gohr <andi@splitbrain.org>
413 *
414 * @param string $data  This data is used to create a unique md5 name
415 * @param string $ext   This is appended to the filename if given
416 * @return string       The filename of the cachefile
417 */
418function getCacheName($data,$ext=''){
419  global $conf;
420  $md5  = md5($data);
421  $file = $conf['cachedir'].'/'.$md5{0}.'/'.$md5.$ext;
422  io_makeFileDir($file);
423  return $file;
424}
425
426/**
427 * Checks a pageid against $conf['hidepages']
428 *
429 * @author Andreas Gohr <gohr@cosmocode.de>
430 */
431function isHiddenPage($id){
432  global $conf;
433  if(empty($conf['hidepages'])) return false;
434
435  if(preg_match('/'.$conf['hidepages'].'/ui',':'.$id)){
436    return true;
437  }
438  return false;
439}
440
441/**
442 * Reverse of isHiddenPage
443 *
444 * @author Andreas Gohr <gohr@cosmocode.de>
445 */
446function isVisiblePage($id){
447  return !isHiddenPage($id);
448}
449
450/**
451 * Checks and sets HTTP headers for conditional HTTP requests
452 *
453 * @author   Simon Willison <swillison@gmail.com>
454 * @link     http://simon.incutio.com/archive/2003/04/23/conditionalGet
455 * @param    timestamp $timestamp lastmodified time of the cache file
456 * @returns  void or void with previously header() commands executed
457 */
458function http_conditionalRequest($timestamp){
459  // A PHP implementation of conditional get, see
460  //   http://fishbowl.pastiche.org/archives/001132.html
461  $last_modified = substr(gmdate('r', $timestamp), 0, -5).'GMT';
462  $etag = '"'.md5($last_modified).'"';
463  // Send the headers
464  header("Last-Modified: $last_modified");
465  header("ETag: $etag");
466  // See if the client has provided the required headers
467  if (isset($_SERVER['HTTP_IF_MODIFIED_SINCE'])){
468    $if_modified_since = stripslashes($_SERVER['HTTP_IF_MODIFIED_SINCE']);
469  }else{
470    $if_modified_since = false;
471  }
472
473  if (isset($_SERVER['HTTP_IF_NONE_MATCH'])){
474    $if_none_match = stripslashes($_SERVER['HTTP_IF_NONE_MATCH']);
475  }else{
476    $if_none_match = false;
477  }
478
479  if (!$if_modified_since && !$if_none_match){
480    return;
481  }
482
483  // At least one of the headers is there - check them
484  if ($if_none_match && $if_none_match != $etag) {
485    return; // etag is there but doesn't match
486  }
487
488  if ($if_modified_since && $if_modified_since != $last_modified) {
489    return; // if-modified-since is there but doesn't match
490  }
491
492  // Nothing has changed since their last request - serve a 304 and exit
493  header('HTTP/1.0 304 Not Modified');
494  exit;
495}
496
497//Setup VIM: ex: et ts=2 enc=utf-8 :
498