xref: /dokuwiki/inc/pageutils.php (revision 50602150cdefc713863f182b761885259f3734b7) 
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// function result cache for cleanID
62global $cache_cleanid;
63$cache_cleanid = array();
64
65/**
66 * Remove unwanted chars from ID
67 *
68 * Cleans a given ID to only use allowed characters. Accented characters are
69 * converted to unaccented ones
70 *
71 * @author Andreas Gohr <andi@splitbrain.org>
72 * @param  string  $raw_id    The pageid to clean
73 * @param  boolean $ascii     Force ASCII
74 */
75function cleanID($raw_id,$ascii=false){
76  global $conf;
77  global $lang;
78  static $sepcharpat = null;
79
80  global $cache_cleanid;
81  $cache = & $cache_cleanid;
82
83  // check if it's already in the memory cache
84  if (isset($cache[$raw_id])) {
85    return $cache[$raw_id];
86	}
87
88  $sepchar = $conf['sepchar'];
89  if($sepcharpat == null) // build string only once to save clock cycles
90    $sepcharpat = '#\\'.$sepchar.'+#';
91
92  $id = trim($raw_id);
93  $id = utf8_strtolower($id);
94
95  //alternative namespace seperator
96  $id = strtr($id,';',':');
97  if($conf['useslash']){
98    $id = strtr($id,'/',':');
99  }else{
100    $id = strtr($id,'/',$sepchar);
101  }
102
103  if($conf['deaccent'] == 2 || $ascii) $id = utf8_romanize($id);
104  if($conf['deaccent'] || $ascii) $id = utf8_deaccent($id,-1);
105
106  //remove specials
107  $id = utf8_stripspecials($id,$sepchar,'\*');
108
109  if($ascii) $id = utf8_strip($id);
110
111  //clean up
112  $id = preg_replace($sepcharpat,$sepchar,$id);
113  $id = preg_replace('#:+#',':',$id);
114  $id = trim($id,':._-');
115  $id = preg_replace('#:[:\._\-]+#',':',$id);
116
117  $cache[$raw_id] = $id;
118  return($id);
119}
120
121/**
122 * Return namespacepart of a wiki ID
123 *
124 * @author Andreas Gohr <andi@splitbrain.org>
125 */
126function getNS($id){
127  $pos = strrpos($id,':');
128  if($pos!==false){
129    return substr($id,0,$pos);
130  }
131  return false;
132}
133
134/**
135 * Returns the ID without the namespace
136 *
137 * @author Andreas Gohr <andi@splitbrain.org>
138 */
139function noNS($id) {
140  $pos = strrpos($id, ':');
141  if ($pos!==false) {
142    return substr($id, $pos+1);
143  } else {
144    return $id;
145  }
146}
147
148// function result cache for wikiFN
149global $cache_wikifn;
150$cache_wikifn = array();
151
152/**
153 * returns the full path to the datafile specified by ID and
154 * optional revision
155 *
156 * The filename is URL encoded to protect Unicode chars
157 *
158 * @author Andreas Gohr <andi@splitbrain.org>
159 */
160function wikiFN($raw_id,$rev='',$clean=true){
161  global $conf;
162
163  global $cache_wikifn;
164  $cache = & $cache_wikifn;
165
166  if (isset($cache[$raw_id]) && isset($cache[$raw_id][$rev])) {
167    return $cache[$raw_id][$rev];
168  }
169
170  $id = $raw_id;
171
172  if ($clean) $id = cleanID($id);
173  $id = str_replace(':','/',$id);
174  if(empty($rev)){
175    $fn = $conf['datadir'].'/'.utf8_encodeFN($id).'.txt';
176  }else{
177    $fn = $conf['olddir'].'/'.utf8_encodeFN($id).'.'.$rev.'.txt';
178    if($conf['compression']){
179      //test for extensions here, we want to read both compressions
180       if (file_exists($fn . '.gz')){
181          $fn .= '.gz';
182       }else if(file_exists($fn . '.bz2')){
183          $fn .= '.bz2';
184       }else{
185          //file doesnt exist yet, so we take the configured extension
186          $fn .= '.' . $conf['compression'];
187       }
188    }
189  }
190
191  if (!isset($cache[$raw_id])) { $cache[$raw_id] = array(); }
192  $cache[$raw_id][$rev] = $fn;
193  return $fn;
194}
195
196/**
197 * Returns the full path to the file for locking the page while editing.
198 *
199 * @author Ben Coburn <btcoburn@silicodon.net>
200 */
201function wikiLockFN($id) {
202  global $conf;
203  return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
204}
205
206
207/**
208 * returns the full path to the meta file specified by ID and extension
209 *
210 * The filename is URL encoded to protect Unicode chars
211 *
212 * @author Steven Danz <steven-danz@kc.rr.com>
213 */
214function metaFN($id,$ext){
215  global $conf;
216  $id = cleanID($id);
217  $id = str_replace(':','/',$id);
218  $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
219  return $fn;
220}
221
222/**
223 * returns an array of full paths to all metafiles of a given ID
224 *
225 * @author Esther Brunner <esther@kaffeehaus.ch>
226 */
227function metaFiles($id){
228   $name   = noNS($id);
229   $dir    = metaFN(getNS($id),'');
230   $files  = array();
231
232   $dh = @opendir($dir);
233   if(!$dh) return $files;
234   while(($file = readdir($dh)) !== false){
235     if(strpos($file,$name.'.') === 0 && !is_dir($dir.$file))
236       $files[] = $dir.$file;
237   }
238   closedir($dh);
239
240   return $files;
241}
242
243/**
244 * returns the full path to the mediafile specified by ID
245 *
246 * The filename is URL encoded to protect Unicode chars
247 *
248 * @author Andreas Gohr <andi@splitbrain.org>
249 */
250function mediaFN($id){
251  global $conf;
252  $id = cleanID($id);
253  $id = str_replace(':','/',$id);
254    $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
255  return $fn;
256}
257
258/**
259 * Returns the full filepath to a localized textfile if local
260 * version isn't found the english one is returned
261 *
262 * @author Andreas Gohr <andi@splitbrain.org>
263 */
264function localeFN($id){
265  global $conf;
266  $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.txt';
267  if(!@file_exists($file)){
268    //fall back to english
269    $file = DOKU_INC.'inc/lang/en/'.$id.'.txt';
270  }
271  return $file;
272}
273
274/**
275 * Resolve relative paths in IDs
276 *
277 * Do not call directly use resolve_mediaid or resolve_pageid
278 * instead
279 *
280 * Partyly based on a cleanPath function found at
281 * http://www.php.net/manual/en/function.realpath.php#57016
282 *
283 * @author <bart at mediawave dot nl>
284 */
285function resolve_id($ns,$id,$clean=true){
286  // if the id starts with a dot we need to handle the
287  // relative stuff
288  if($id{0} == '.'){
289    // normalize initial dots without a colon
290    $id = preg_replace('/^(\.+)(?=[^:\.])/','\1:',$id);
291    // prepend the current namespace
292    $id = $ns.':'.$id;
293
294    // cleanup relatives
295    $result = array();
296    $pathA  = explode(':', $id);
297    if (!$pathA[0]) $result[] = '';
298    foreach ($pathA AS $key => $dir) {
299      if ($dir == '..') {
300        if (end($result) == '..') {
301          $result[] = '..';
302        } elseif (!array_pop($result)) {
303          $result[] = '..';
304        }
305      } elseif ($dir && $dir != '.') {
306        $result[] = $dir;
307      }
308    }
309    if (!end($pathA)) $result[] = '';
310    $id = implode(':', $result);
311  }elseif($ns !== false && strpos($id,':') === false){
312    //if link contains no namespace. add current namespace (if any)
313    $id = $ns.':'.$id;
314  }
315
316  if($clean) $id = cleanID($id);
317  return $id;
318}
319
320/**
321 * Returns a full media id
322 *
323 * @author Andreas Gohr <andi@splitbrain.org>
324 */
325function resolve_mediaid($ns,&$page,&$exists){
326  $page   = resolve_id($ns,$page);
327  $file   = mediaFN($page);
328  $exists = @file_exists($file);
329}
330
331/**
332 * Returns a full page id
333 *
334 * @author Andreas Gohr <andi@splitbrain.org>
335 */
336function resolve_pageid($ns,&$page,&$exists){
337  global $conf;
338  $exists = false;
339
340  //keep hashlink if exists then clean both parts
341  if (strpos($page,'#')) {
342    list($page,$hash) = split('#',$page,2);
343  } else {
344    $hash = '';
345  }
346  $hash = cleanID($hash);
347  $page = resolve_id($ns,$page,false); // resolve but don't clean, yet
348
349  // get filename (calls clean itself)
350  $file = wikiFN($page);
351
352  // if ends with colon we have a namespace link
353  if(substr($page,-1) == ':'){
354    if(@file_exists(wikiFN($page.$conf['start']))){
355      // start page inside namespace
356      $page = $page.$conf['start'];
357      $exists = true;
358    }elseif(@file_exists(wikiFN($page.noNS(cleanID($page))))){
359      // page named like the NS inside the NS
360      $page = $page.noNS(cleanID($page));
361      $exists = true;
362    }elseif(@file_exists(wikiFN($page))){
363      // page like namespace exists
364      $page = $page;
365      $exists = true;
366    }else{
367      // fall back to default
368      $page = $page.$conf['start'];
369    }
370  }else{
371    //check alternative plural/nonplural form
372    if(!@file_exists($file)){
373      if( $conf['autoplural'] ){
374        if(substr($page,-1) == 's'){
375          $try = substr($page,0,-1);
376        }else{
377          $try = $page.'s';
378        }
379        if(@file_exists(wikiFN($try))){
380          $page   = $try;
381          $exists = true;
382        }
383      }
384    }else{
385      $exists = true;
386    }
387  }
388
389  // now make sure we have a clean page
390  $page = cleanID($page);
391
392  //add hash if any
393  if(!empty($hash)) $page .= '#'.$hash;
394}
395
396/**
397 * Returns the name of a cachefile from given data
398 *
399 * The needed directory is created by this function!
400 *
401 * @author Andreas Gohr <andi@splitbrain.org>
402 *
403 * @param string $data  This data is used to create a unique md5 name
404 * @param string $ext   This is appended to the filename if given
405 * @return string       The filename of the cachefile
406 */
407function getCacheName($data,$ext=''){
408  global $conf;
409  $md5  = md5($data);
410  $file = $conf['cachedir'].'/'.$md5{0}.'/'.$md5.$ext;
411  io_makeFileDir($file);
412  return $file;
413}
414
415/**
416 * Checks a pageid against $conf['hidepages']
417 *
418 * @author Andreas Gohr <gohr@cosmocode.de>
419 */
420function isHiddenPage($id){
421  global $conf;
422  if(empty($conf['hidepages'])) return false;
423
424  if(preg_match('/'.$conf['hidepages'].'/ui',':'.$id)){
425    return true;
426  }
427  return false;
428}
429
430/**
431 * Reverse of isHiddenPage
432 *
433 * @author Andreas Gohr <gohr@cosmocode.de>
434 */
435function isVisiblePage($id){
436  return !isHiddenPage($id);
437}
438
439/**
440 * Checks and sets HTTP headers for conditional HTTP requests
441 *
442 * @author   Simon Willison <swillison@gmail.com>
443 * @link     http://simon.incutio.com/archive/2003/04/23/conditionalGet
444 * @param    timestamp $timestamp lastmodified time of the cache file
445 * @returns  void or void with previously header() commands executed
446 */
447function http_conditionalRequest($timestamp){
448  // A PHP implementation of conditional get, see
449  //   http://fishbowl.pastiche.org/archives/001132.html
450  $last_modified = substr(date('r', $timestamp), 0, -5).'GMT';
451  $etag = '"'.md5($last_modified).'"';
452  // Send the headers
453  header("Last-Modified: $last_modified");
454  header("ETag: $etag");
455  // See if the client has provided the required headers
456  if (isset($_SERVER['HTTP_IF_MODIFIED_SINCE'])){
457    $if_modified_since = stripslashes($_SERVER['HTTP_IF_MODIFIED_SINCE']);
458  }else{
459    $if_modified_since = false;
460  }
461
462  if (isset($_SERVER['HTTP_IF_NONE_MATCH'])){
463    $if_none_match = stripslashes($_SERVER['HTTP_IF_NONE_MATCH']);
464  }else{
465    $if_none_match = false;
466  }
467
468  if (!$if_modified_since && !$if_none_match){
469    return;
470  }
471
472  // At least one of the headers is there - check them
473  if ($if_none_match && $if_none_match != $etag) {
474    return; // etag is there but doesn't match
475  }
476
477  if ($if_modified_since && $if_modified_since != $last_modified) {
478    return; // if-modified-since is there but doesn't match
479  }
480
481  // Nothing has changed since their last request - serve a 304 and exit
482  header('HTTP/1.0 304 Not Modified');
483  exit;
484}
485
486//Setup VIM: ex: et ts=2 enc=utf-8 :
487