xref: /dokuwiki/inc/pageutils.php (revision 6d88439ada7c841b10a8de3da846f7cc1cf5842a) 
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  $raw_id    The pageid to clean
69 * @param  boolean $ascii     Force ASCII
70 */
71function cleanID($raw_id,$ascii=false){
72  global $conf;
73  global $lang;
74  static $sepcharpat = null;
75
76  global $cache_cleanid;
77  $cache = & $cache_cleanid;
78
79  // check if it's already in the memory cache
80  if (isset($cache[$raw_id])) {
81    return $cache[$raw_id];
82	}
83
84  $sepchar = $conf['sepchar'];
85  if($sepcharpat == null) // build string only once to save clock cycles
86    $sepcharpat = '#\\'.$sepchar.'+#';
87
88  $id = trim($raw_id);
89  $id = utf8_strtolower($id);
90
91  //alternative namespace seperator
92  $id = strtr($id,';',':');
93  if($conf['useslash']){
94    $id = strtr($id,'/',':');
95  }else{
96    $id = strtr($id,'/',$sepchar);
97  }
98
99  if($conf['deaccent'] == 2 || $ascii) $id = utf8_romanize($id);
100  if($conf['deaccent'] || $ascii) $id = utf8_deaccent($id,-1);
101
102  //remove specials
103  $id = utf8_stripspecials($id,$sepchar,'\*');
104
105  if($ascii) $id = utf8_strip($id);
106
107  //clean up
108  $id = preg_replace($sepcharpat,$sepchar,$id);
109  $id = preg_replace('#:+#',':',$id);
110  $id = trim($id,':._-');
111  $id = preg_replace('#:[:\._\-]+#',':',$id);
112
113  $cache[$raw_id] = $id;
114  return($id);
115}
116
117/**
118 * Return namespacepart of a wiki ID
119 *
120 * @author Andreas Gohr <andi@splitbrain.org>
121 */
122function getNS($id){
123  $pos = strrpos($id,':');
124  if($pos!==false){
125    return substr($id,0,$pos);
126  }
127  return false;
128}
129
130/**
131 * Returns the ID without the namespace
132 *
133 * @author Andreas Gohr <andi@splitbrain.org>
134 */
135function noNS($id) {
136  $pos = strrpos($id, ':');
137  if ($pos!==false) {
138    return substr($id, $pos+1);
139  } else {
140    return $id;
141  }
142}
143
144/**
145 * returns the full path to the datafile specified by ID and
146 * optional revision
147 *
148 * The filename is URL encoded to protect Unicode chars
149 *
150 * @author Andreas Gohr <andi@splitbrain.org>
151 */
152function wikiFN($raw_id,$rev='',$clean=true){
153  global $conf;
154
155  global $cache_wikifn;
156  $cache = & $cache_wikifn;
157
158  if (isset($cache[$raw_id]) && isset($cache[$raw_id][$rev])) {
159    return $cache[$raw_id][$rev];
160  }
161
162  $id = $raw_id;
163
164  if ($clean) $id = cleanID($id);
165  $id = str_replace(':','/',$id);
166  if(empty($rev)){
167    $fn = $conf['datadir'].'/'.utf8_encodeFN($id).'.txt';
168  }else{
169    $fn = $conf['olddir'].'/'.utf8_encodeFN($id).'.'.$rev.'.txt';
170    if($conf['compression']){
171      //test for extensions here, we want to read both compressions
172       if (@file_exists($fn . '.gz')){
173          $fn .= '.gz';
174       }else if(@file_exists($fn . '.bz2')){
175          $fn .= '.bz2';
176       }else{
177          //file doesnt exist yet, so we take the configured extension
178          $fn .= '.' . $conf['compression'];
179       }
180    }
181  }
182
183  if (!isset($cache[$raw_id])) { $cache[$raw_id] = array(); }
184  $cache[$raw_id][$rev] = $fn;
185  return $fn;
186}
187
188/**
189 * Returns the full path to the file for locking the page while editing.
190 *
191 * @author Ben Coburn <btcoburn@silicodon.net>
192 */
193function wikiLockFN($id) {
194  global $conf;
195  return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
196}
197
198
199/**
200 * returns the full path to the meta file specified by ID and extension
201 *
202 * The filename is URL encoded to protect Unicode chars
203 *
204 * @author Steven Danz <steven-danz@kc.rr.com>
205 */
206function metaFN($id,$ext){
207  global $conf;
208  $id = cleanID($id);
209  $id = str_replace(':','/',$id);
210  $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
211  return $fn;
212}
213
214/**
215 * returns an array of full paths to all metafiles of a given ID
216 *
217 * @author Esther Brunner <esther@kaffeehaus.ch>
218 */
219function metaFiles($id){
220   $name   = noNS($id);
221   $dir    = metaFN(getNS($id),'');
222   $files  = array();
223
224   $dh = @opendir($dir);
225   if(!$dh) return $files;
226   while(($file = readdir($dh)) !== false){
227     if(strpos($file,$name.'.') === 0 && !is_dir($dir.$file))
228       $files[] = $dir.$file;
229   }
230   closedir($dh);
231
232   return $files;
233}
234
235/**
236 * returns the full path to the mediafile specified by ID
237 *
238 * The filename is URL encoded to protect Unicode chars
239 *
240 * @author Andreas Gohr <andi@splitbrain.org>
241 */
242function mediaFN($id){
243  global $conf;
244  $id = cleanID($id);
245  $id = str_replace(':','/',$id);
246    $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
247  return $fn;
248}
249
250/**
251 * Returns the full filepath to a localized textfile if local
252 * version isn't found the english one is returned
253 *
254 * @author Andreas Gohr <andi@splitbrain.org>
255 */
256function localeFN($id){
257  global $conf;
258  $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.txt';
259  if(!@file_exists($file)){
260    //fall back to english
261    $file = DOKU_INC.'inc/lang/en/'.$id.'.txt';
262  }
263  return $file;
264}
265
266/**
267 * Resolve relative paths in IDs
268 *
269 * Do not call directly use resolve_mediaid or resolve_pageid
270 * instead
271 *
272 * Partyly based on a cleanPath function found at
273 * http://www.php.net/manual/en/function.realpath.php#57016
274 *
275 * @author <bart at mediawave dot nl>
276 */
277function resolve_id($ns,$id,$clean=true){
278  // if the id starts with a dot we need to handle the
279  // relative stuff
280  if($id{0} == '.'){
281    // normalize initial dots without a colon
282    $id = preg_replace('/^(\.+)(?=[^:\.])/','\1:',$id);
283    // prepend the current namespace
284    $id = $ns.':'.$id;
285
286    // cleanup relatives
287    $result = array();
288    $pathA  = explode(':', $id);
289    if (!$pathA[0]) $result[] = '';
290    foreach ($pathA AS $key => $dir) {
291      if ($dir == '..') {
292        if (end($result) == '..') {
293          $result[] = '..';
294        } elseif (!array_pop($result)) {
295          $result[] = '..';
296        }
297      } elseif ($dir && $dir != '.') {
298        $result[] = $dir;
299      }
300    }
301    if (!end($pathA)) $result[] = '';
302    $id = implode(':', $result);
303  }elseif($ns !== false && strpos($id,':') === false){
304    //if link contains no namespace. add current namespace (if any)
305    $id = $ns.':'.$id;
306  }
307
308  if($clean) $id = cleanID($id);
309  return $id;
310}
311
312/**
313 * Returns a full media id
314 *
315 * @author Andreas Gohr <andi@splitbrain.org>
316 */
317function resolve_mediaid($ns,&$page,&$exists){
318  $page   = resolve_id($ns,$page);
319  $file   = mediaFN($page);
320  $exists = @file_exists($file);
321}
322
323/**
324 * Returns a full page id
325 *
326 * @author Andreas Gohr <andi@splitbrain.org>
327 */
328function resolve_pageid($ns,&$page,&$exists){
329  global $conf;
330  $exists = false;
331
332  //keep hashlink if exists then clean both parts
333  if (strpos($page,'#')) {
334    list($page,$hash) = split('#',$page,2);
335  } else {
336    $hash = '';
337  }
338  $hash = cleanID($hash);
339  $page = resolve_id($ns,$page,false); // resolve but don't clean, yet
340
341  // get filename (calls clean itself)
342  $file = wikiFN($page);
343
344  // if ends with colon or slash we have a namespace link
345  if(substr($page,-1) == ':' || ($conf['useslash'] && substr($page,-1) == '/')){
346    if(@file_exists(wikiFN($page.$conf['start']))){
347      // start page inside namespace
348      $page = $page.$conf['start'];
349      $exists = true;
350    }elseif(@file_exists(wikiFN($page.noNS(cleanID($page))))){
351      // page named like the NS inside the NS
352      $page = $page.noNS(cleanID($page));
353      $exists = true;
354    }elseif(@file_exists(wikiFN($page))){
355      // page like namespace exists
356      $page = $page;
357      $exists = true;
358    }else{
359      // fall back to default
360      $page = $page.$conf['start'];
361    }
362  }else{
363    //check alternative plural/nonplural form
364    if(!@file_exists($file)){
365      if( $conf['autoplural'] ){
366        if(substr($page,-1) == 's'){
367          $try = substr($page,0,-1);
368        }else{
369          $try = $page.'s';
370        }
371        if(@file_exists(wikiFN($try))){
372          $page   = $try;
373          $exists = true;
374        }
375      }
376    }else{
377      $exists = true;
378    }
379  }
380
381  // now make sure we have a clean page
382  $page = cleanID($page);
383
384  //add hash if any
385  if(!empty($hash)) $page .= '#'.$hash;
386}
387
388/**
389 * Returns the name of a cachefile from given data
390 *
391 * The needed directory is created by this function!
392 *
393 * @author Andreas Gohr <andi@splitbrain.org>
394 *
395 * @param string $data  This data is used to create a unique md5 name
396 * @param string $ext   This is appended to the filename if given
397 * @return string       The filename of the cachefile
398 */
399function getCacheName($data,$ext=''){
400  global $conf;
401  $md5  = md5($data);
402  $file = $conf['cachedir'].'/'.$md5{0}.'/'.$md5.$ext;
403  io_makeFileDir($file);
404  return $file;
405}
406
407/**
408 * Checks a pageid against $conf['hidepages']
409 *
410 * @author Andreas Gohr <gohr@cosmocode.de>
411 */
412function isHiddenPage($id){
413  global $conf;
414  if(empty($conf['hidepages'])) return false;
415
416  if(preg_match('/'.$conf['hidepages'].'/ui',':'.$id)){
417    return true;
418  }
419  return false;
420}
421
422/**
423 * Reverse of isHiddenPage
424 *
425 * @author Andreas Gohr <gohr@cosmocode.de>
426 */
427function isVisiblePage($id){
428  return !isHiddenPage($id);
429}
430
431/**
432 * Checks and sets HTTP headers for conditional HTTP requests
433 *
434 * @author   Simon Willison <swillison@gmail.com>
435 * @link     http://simon.incutio.com/archive/2003/04/23/conditionalGet
436 * @param    timestamp $timestamp lastmodified time of the cache file
437 * @returns  void or void with previously header() commands executed
438 */
439function http_conditionalRequest($timestamp){
440  // A PHP implementation of conditional get, see
441  //   http://fishbowl.pastiche.org/archives/001132.html
442  $last_modified = substr(gmdate('r', $timestamp), 0, -5).'GMT';
443  $etag = '"'.md5($last_modified).'"';
444  // Send the headers
445  header("Last-Modified: $last_modified");
446  header("ETag: $etag");
447  // See if the client has provided the required headers
448  if (isset($_SERVER['HTTP_IF_MODIFIED_SINCE'])){
449    $if_modified_since = stripslashes($_SERVER['HTTP_IF_MODIFIED_SINCE']);
450  }else{
451    $if_modified_since = false;
452  }
453
454  if (isset($_SERVER['HTTP_IF_NONE_MATCH'])){
455    $if_none_match = stripslashes($_SERVER['HTTP_IF_NONE_MATCH']);
456  }else{
457    $if_none_match = false;
458  }
459
460  if (!$if_modified_since && !$if_none_match){
461    return;
462  }
463
464  // At least one of the headers is there - check them
465  if ($if_none_match && $if_none_match != $etag) {
466    return; // etag is there but doesn't match
467  }
468
469  if ($if_modified_since && $if_modified_since != $last_modified) {
470    return; // if-modified-since is there but doesn't match
471  }
472
473  // Nothing has changed since their last request - serve a 304 and exit
474  header('HTTP/1.0 304 Not Modified');
475  exit;
476}
477
478//Setup VIM: ex: et ts=2 enc=utf-8 :
479