xref: /dokuwiki/inc/pageutils.php (revision fd49f8df8d347af8d355ec7b8cb2261a17e2ae5c) 
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  if($clean) $id = cleanID($id);
56  if(empty($id) && $param=='id') $id = $conf['start'];
57
58  return $id;
59}
60
61/**
62 * Remove unwanted chars from ID
63 *
64 * Cleans a given ID to only use allowed characters. Accented characters are
65 * converted to unaccented ones
66 *
67 * @author Andreas Gohr <andi@splitbrain.org>
68 * @param  string  $id    The pageid to clean
69 * @param  boolean $ascii Force ASCII
70 */
71function cleanID($id,$ascii=false){
72  global $conf;
73  global $lang;
74  static $sepcharpat = null;
75
76  $sepchar = $conf['sepchar'];
77  if($sepcharpat == null) // build string only once to save clock cycles
78    $sepcharpat = '#\\'.$sepchar.'+#';
79
80  $id = trim($id);
81  $id = utf8_strtolower($id);
82
83  //alternative namespace seperator
84  $id = strtr($id,';',':');
85  if($conf['useslash']){
86    $id = strtr($id,'/',':');
87  }else{
88    $id = strtr($id,'/',$sepchar);
89  }
90
91  if($conf['deaccent'] == 2 || $ascii) $id = utf8_romanize($id);
92  if($conf['deaccent'] || $ascii) $id = utf8_deaccent($id,-1);
93
94  //remove specials
95  $id = utf8_stripspecials($id,$sepchar,'\*');
96
97  if($ascii) $id = utf8_strip($id);
98
99  //clean up
100  $id = preg_replace($sepcharpat,$sepchar,$id);
101  $id = preg_replace('#:+#',':',$id);
102  $id = trim($id,':._-');
103  $id = preg_replace('#:[:\._\-]+#',':',$id);
104
105  return($id);
106}
107
108/**
109 * Return namespacepart of a wiki ID
110 *
111 * @author Andreas Gohr <andi@splitbrain.org>
112 */
113function getNS($id){
114  $pos = strrpos($id,':');
115  if($pos!==false){
116    return substr($id,0,$pos);
117  }
118  return false;
119}
120
121/**
122 * Returns the ID without the namespace
123 *
124 * @author Andreas Gohr <andi@splitbrain.org>
125 */
126function noNS($id) {
127  $pos = strrpos($id, ':');
128  if ($pos!==false) {
129    return substr($id, $pos+1);
130  } else {
131    return $id;
132  }
133}
134
135/**
136 * returns the full path to the datafile specified by ID and
137 * optional revision
138 *
139 * The filename is URL encoded to protect Unicode chars
140 *
141 * @author Andreas Gohr <andi@splitbrain.org>
142 */
143function wikiFN($id,$rev='',$clean=true){
144  global $conf;
145  if ($clean) $id = cleanID($id);
146  $id = str_replace(':','/',$id);
147  if(empty($rev)){
148    $fn = $conf['datadir'].'/'.utf8_encodeFN($id).'.txt';
149  }else{
150    $fn = $conf['olddir'].'/'.utf8_encodeFN($id).'.'.$rev.'.txt';
151    if($conf['compression']){
152      //test for extensions here, we want to read both compressions
153       if (file_exists($fn . '.gz')){
154          $fn .= '.gz';
155       }else if(file_exists($fn . '.bz2')){
156          $fn .= '.bz2';
157       }else{
158          //file doesnt exist yet, so we take the configured extension
159          $fn .= '.' . $conf['compression'];
160       }
161    }
162  }
163  return $fn;
164}
165
166/**
167 * Returns the full path to the file for locking the page while editing.
168 *
169 * @author Ben Coburn <btcoburn@silicodon.net>
170 */
171function wikiLockFN($id) {
172  global $conf;
173  return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
174}
175
176
177/**
178 * returns the full path to the meta file specified by ID and extension
179 *
180 * The filename is URL encoded to protect Unicode chars
181 *
182 * @author Steven Danz <steven-danz@kc.rr.com>
183 */
184function metaFN($id,$ext){
185  global $conf;
186  $id = cleanID($id);
187  $id = str_replace(':','/',$id);
188  $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
189  return $fn;
190}
191
192/**
193 * returns an array of full paths to all metafiles of a given ID
194 *
195 * @author Esther Brunner <esther@kaffeehaus.ch>
196 */
197function metaFiles($id){
198   $name   = noNS($id);
199   $dir    = metaFN(getNS($id),'');
200   $files  = array();
201
202   $dh = @opendir($dir);
203   if(!$dh) return $files;
204   while(($file = readdir($dh)) !== false){
205     if(strpos($file,$name.'.') === 0 && !is_dir($dir.$file))
206       $files[] = $dir.$file;
207   }
208   closedir($dh);
209
210   return $files;
211}
212
213/**
214 * returns the full path to the mediafile specified by ID
215 *
216 * The filename is URL encoded to protect Unicode chars
217 *
218 * @author Andreas Gohr <andi@splitbrain.org>
219 */
220function mediaFN($id){
221  global $conf;
222  $id = cleanID($id);
223  $id = str_replace(':','/',$id);
224    $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
225  return $fn;
226}
227
228/**
229 * Returns the full filepath to a localized textfile if local
230 * version isn't found the english one is returned
231 *
232 * @author Andreas Gohr <andi@splitbrain.org>
233 */
234function localeFN($id){
235  global $conf;
236  $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.txt';
237  if(!@file_exists($file)){
238    //fall back to english
239    $file = DOKU_INC.'inc/lang/en/'.$id.'.txt';
240  }
241  return $file;
242}
243
244/**
245 * Resolve relative paths in IDs
246 *
247 * Do not call directly use resolve_mediaid or resolve_pageid
248 * instead
249 *
250 * Partyly based on a cleanPath function found at
251 * http://www.php.net/manual/en/function.realpath.php#57016
252 *
253 * @author <bart at mediawave dot nl>
254 */
255function resolve_id($ns,$id,$clean=true){
256  // if the id starts with a dot we need to handle the
257  // relative stuff
258  if($id{0} == '.'){
259    // normalize initial dots without a colon
260    $id = preg_replace('/^(\.+)(?=[^:\.])/','\1:',$id);
261    // prepend the current namespace
262    $id = $ns.':'.$id;
263
264    // cleanup relatives
265    $result = array();
266    $pathA  = explode(':', $id);
267    if (!$pathA[0]) $result[] = '';
268    foreach ($pathA AS $key => $dir) {
269      if ($dir == '..') {
270        if (end($result) == '..') {
271          $result[] = '..';
272        } elseif (!array_pop($result)) {
273          $result[] = '..';
274        }
275      } elseif ($dir && $dir != '.') {
276        $result[] = $dir;
277      }
278    }
279    if (!end($pathA)) $result[] = '';
280    $id = implode(':', $result);
281  }elseif($ns !== false && strpos($id,':') === false){
282    //if link contains no namespace. add current namespace (if any)
283    $id = $ns.':'.$id;
284  }
285
286  if($clean) $id = cleanID($id);
287  return $id;
288}
289
290/**
291 * Returns a full media id
292 *
293 * @author Andreas Gohr <andi@splitbrain.org>
294 */
295function resolve_mediaid($ns,&$page,&$exists){
296  $page   = resolve_id($ns,$page);
297  $file   = mediaFN($page);
298  $exists = @file_exists($file);
299}
300
301/**
302 * Returns a full page id
303 *
304 * @author Andreas Gohr <andi@splitbrain.org>
305 */
306function resolve_pageid($ns,&$page,&$exists){
307  global $conf;
308  $exists = false;
309
310  //keep hashlink if exists then clean both parts
311  if (strpos($page,'#')) {
312    list($page,$hash) = split('#',$page,2);
313  } else {
314    $hash = '';
315  }
316  $hash = cleanID($hash);
317  $page = resolve_id($ns,$page,false); // resolve but don't clean, yet
318
319  // get filename (calls clean itself)
320  $file = wikiFN($page);
321
322  // if ends with colon we have a namespace link
323  if(substr($page,-1) == ':'){
324    if(@file_exists(wikiFN($page.$conf['start']))){
325      // start page inside namespace
326      $page = $page.$conf['start'];
327      $exists = true;
328    }elseif(@file_exists(wikiFN($page.noNS(cleanID($page))))){
329      // page named like the NS inside the NS
330      $page = $page.noNS(cleanID($page));
331      $exists = true;
332    }elseif(@file_exists(wikiFN($page))){
333      // page like namespace exists
334      $page = $page;
335      $exists = true;
336    }else{
337      // fall back to default
338      $page = $page.$conf['start'];
339    }
340  }else{
341    //check alternative plural/nonplural form
342    if(!@file_exists($file)){
343      if( $conf['autoplural'] ){
344        if(substr($page,-1) == 's'){
345          $try = substr($page,0,-1);
346        }else{
347          $try = $page.'s';
348        }
349        if(@file_exists(wikiFN($try))){
350          $page   = $try;
351          $exists = true;
352        }
353      }
354    }else{
355      $exists = true;
356    }
357  }
358
359  // now make sure we have a clean page
360  $page = cleanID($page);
361
362  //add hash if any
363  if(!empty($hash)) $page .= '#'.$hash;
364}
365
366/**
367 * Returns the name of a cachefile from given data
368 *
369 * The needed directory is created by this function!
370 *
371 * @author Andreas Gohr <andi@splitbrain.org>
372 *
373 * @param string $data  This data is used to create a unique md5 name
374 * @param string $ext   This is appended to the filename if given
375 * @return string       The filename of the cachefile
376 */
377function getCacheName($data,$ext=''){
378  global $conf;
379  $md5  = md5($data);
380  $file = $conf['cachedir'].'/'.$md5{0}.'/'.$md5.$ext;
381  io_makeFileDir($file);
382  return $file;
383}
384
385/**
386 * Checks a pageid against $conf['hidepages']
387 *
388 * @author Andreas Gohr <gohr@cosmocode.de>
389 */
390function isHiddenPage($id){
391  global $conf;
392  if(empty($conf['hidepages'])) return false;
393
394  if(preg_match('/'.$conf['hidepages'].'/ui',':'.$id)){
395    return true;
396  }
397  return false;
398}
399
400/**
401 * Reverse of isHiddenPage
402 *
403 * @author Andreas Gohr <gohr@cosmocode.de>
404 */
405function isVisiblePage($id){
406  return !isHiddenPage($id);
407}
408
409/**
410 * Checks and sets HTTP headers for conditional HTTP requests
411 *
412 * @author   Simon Willison <swillison@gmail.com>
413 * @link     http://simon.incutio.com/archive/2003/04/23/conditionalGet
414 * @param    timestamp $timestamp lastmodified time of the cache file
415 * @returns  void or void with previously header() commands executed
416 */
417function http_conditionalRequest($timestamp){
418  // A PHP implementation of conditional get, see
419  //   http://fishbowl.pastiche.org/archives/001132.html
420  $last_modified = substr(date('r', $timestamp), 0, -5).'GMT';
421  $etag = '"'.md5($last_modified).'"';
422  // Send the headers
423  header("Last-Modified: $last_modified");
424  header("ETag: $etag");
425  // See if the client has provided the required headers
426  if (isset($_SERVER['HTTP_IF_MODIFIED_SINCE'])){
427    $if_modified_since = stripslashes($_SERVER['HTTP_IF_MODIFIED_SINCE']);
428  }else{
429    $if_modified_since = false;
430  }
431
432  if (isset($_SERVER['HTTP_IF_NONE_MATCH'])){
433    $if_none_match = stripslashes($_SERVER['HTTP_IF_NONE_MATCH']);
434  }else{
435    $if_none_match = false;
436  }
437
438  if (!$if_modified_since && !$if_none_match){
439    return;
440  }
441
442  // At least one of the headers is there - check them
443  if ($if_none_match && $if_none_match != $etag) {
444    return; // etag is there but doesn't match
445  }
446
447  if ($if_modified_since && $if_modified_since != $last_modified) {
448    return; // if-modified-since is there but doesn't match
449  }
450
451  // Nothing has changed since their last request - serve a 304 and exit
452  header('HTTP/1.0 304 Not Modified');
453  exit;
454}
455
456//Setup VIM: ex: et ts=2 enc=utf-8 :
457