xref: /dokuwiki/inc/pageutils.php (revision dc2c0e04646a8bbd31f08140918d61d1dda25c2e) 
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  $cache[$raw_id][$rev] = $fn;
192  return $fn;
193}
194
195/**
196 * Returns the full path to the file for locking the page while editing.
197 *
198 * @author Ben Coburn <btcoburn@silicodon.net>
199 */
200function wikiLockFN($id) {
201  global $conf;
202  return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
203}
204
205
206/**
207 * returns the full path to the meta file specified by ID and extension
208 *
209 * The filename is URL encoded to protect Unicode chars
210 *
211 * @author Steven Danz <steven-danz@kc.rr.com>
212 */
213function metaFN($id,$ext){
214  global $conf;
215  $id = cleanID($id);
216  $id = str_replace(':','/',$id);
217  $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
218  return $fn;
219}
220
221/**
222 * returns an array of full paths to all metafiles of a given ID
223 *
224 * @author Esther Brunner <esther@kaffeehaus.ch>
225 */
226function metaFiles($id){
227   $name   = noNS($id);
228   $dir    = metaFN(getNS($id),'');
229   $files  = array();
230
231   $dh = @opendir($dir);
232   if(!$dh) return $files;
233   while(($file = readdir($dh)) !== false){
234     if(strpos($file,$name.'.') === 0 && !is_dir($dir.$file))
235       $files[] = $dir.$file;
236   }
237   closedir($dh);
238
239   return $files;
240}
241
242/**
243 * returns the full path to the mediafile specified by ID
244 *
245 * The filename is URL encoded to protect Unicode chars
246 *
247 * @author Andreas Gohr <andi@splitbrain.org>
248 */
249function mediaFN($id){
250  global $conf;
251  $id = cleanID($id);
252  $id = str_replace(':','/',$id);
253    $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
254  return $fn;
255}
256
257/**
258 * Returns the full filepath to a localized textfile if local
259 * version isn't found the english one is returned
260 *
261 * @author Andreas Gohr <andi@splitbrain.org>
262 */
263function localeFN($id){
264  global $conf;
265  $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.txt';
266  if(!@file_exists($file)){
267    //fall back to english
268    $file = DOKU_INC.'inc/lang/en/'.$id.'.txt';
269  }
270  return $file;
271}
272
273/**
274 * Resolve relative paths in IDs
275 *
276 * Do not call directly use resolve_mediaid or resolve_pageid
277 * instead
278 *
279 * Partyly based on a cleanPath function found at
280 * http://www.php.net/manual/en/function.realpath.php#57016
281 *
282 * @author <bart at mediawave dot nl>
283 */
284function resolve_id($ns,$id,$clean=true){
285  // if the id starts with a dot we need to handle the
286  // relative stuff
287  if($id{0} == '.'){
288    // normalize initial dots without a colon
289    $id = preg_replace('/^(\.+)(?=[^:\.])/','\1:',$id);
290    // prepend the current namespace
291    $id = $ns.':'.$id;
292
293    // cleanup relatives
294    $result = array();
295    $pathA  = explode(':', $id);
296    if (!$pathA[0]) $result[] = '';
297    foreach ($pathA AS $key => $dir) {
298      if ($dir == '..') {
299        if (end($result) == '..') {
300          $result[] = '..';
301        } elseif (!array_pop($result)) {
302          $result[] = '..';
303        }
304      } elseif ($dir && $dir != '.') {
305        $result[] = $dir;
306      }
307    }
308    if (!end($pathA)) $result[] = '';
309    $id = implode(':', $result);
310  }elseif($ns !== false && strpos($id,':') === false){
311    //if link contains no namespace. add current namespace (if any)
312    $id = $ns.':'.$id;
313  }
314
315  if($clean) $id = cleanID($id);
316  return $id;
317}
318
319/**
320 * Returns a full media id
321 *
322 * @author Andreas Gohr <andi@splitbrain.org>
323 */
324function resolve_mediaid($ns,&$page,&$exists){
325  $page   = resolve_id($ns,$page);
326  $file   = mediaFN($page);
327  $exists = @file_exists($file);
328}
329
330/**
331 * Returns a full page id
332 *
333 * @author Andreas Gohr <andi@splitbrain.org>
334 */
335function resolve_pageid($ns,&$page,&$exists){
336  global $conf;
337  $exists = false;
338
339  //keep hashlink if exists then clean both parts
340  if (strpos($page,'#')) {
341    list($page,$hash) = split('#',$page,2);
342  } else {
343    $hash = '';
344  }
345  $hash = cleanID($hash);
346  $page = resolve_id($ns,$page,false); // resolve but don't clean, yet
347
348  // get filename (calls clean itself)
349  $file = wikiFN($page);
350
351  // if ends with colon we have a namespace link
352  if(substr($page,-1) == ':'){
353    if(@file_exists(wikiFN($page.$conf['start']))){
354      // start page inside namespace
355      $page = $page.$conf['start'];
356      $exists = true;
357    }elseif(@file_exists(wikiFN($page.noNS(cleanID($page))))){
358      // page named like the NS inside the NS
359      $page = $page.noNS(cleanID($page));
360      $exists = true;
361    }elseif(@file_exists(wikiFN($page))){
362      // page like namespace exists
363      $page = $page;
364      $exists = true;
365    }else{
366      // fall back to default
367      $page = $page.$conf['start'];
368    }
369  }else{
370    //check alternative plural/nonplural form
371    if(!@file_exists($file)){
372      if( $conf['autoplural'] ){
373        if(substr($page,-1) == 's'){
374          $try = substr($page,0,-1);
375        }else{
376          $try = $page.'s';
377        }
378        if(@file_exists(wikiFN($try))){
379          $page   = $try;
380          $exists = true;
381        }
382      }
383    }else{
384      $exists = true;
385    }
386  }
387
388  // now make sure we have a clean page
389  $page = cleanID($page);
390
391  //add hash if any
392  if(!empty($hash)) $page .= '#'.$hash;
393}
394
395/**
396 * Returns the name of a cachefile from given data
397 *
398 * The needed directory is created by this function!
399 *
400 * @author Andreas Gohr <andi@splitbrain.org>
401 *
402 * @param string $data  This data is used to create a unique md5 name
403 * @param string $ext   This is appended to the filename if given
404 * @return string       The filename of the cachefile
405 */
406function getCacheName($data,$ext=''){
407  global $conf;
408  $md5  = md5($data);
409  $file = $conf['cachedir'].'/'.$md5{0}.'/'.$md5.$ext;
410  io_makeFileDir($file);
411  return $file;
412}
413
414/**
415 * Checks a pageid against $conf['hidepages']
416 *
417 * @author Andreas Gohr <gohr@cosmocode.de>
418 */
419function isHiddenPage($id){
420  global $conf;
421  if(empty($conf['hidepages'])) return false;
422
423  if(preg_match('/'.$conf['hidepages'].'/ui',':'.$id)){
424    return true;
425  }
426  return false;
427}
428
429/**
430 * Reverse of isHiddenPage
431 *
432 * @author Andreas Gohr <gohr@cosmocode.de>
433 */
434function isVisiblePage($id){
435  return !isHiddenPage($id);
436}
437
438/**
439 * Checks and sets HTTP headers for conditional HTTP requests
440 *
441 * @author   Simon Willison <swillison@gmail.com>
442 * @link     http://simon.incutio.com/archive/2003/04/23/conditionalGet
443 * @param    timestamp $timestamp lastmodified time of the cache file
444 * @returns  void or void with previously header() commands executed
445 */
446function http_conditionalRequest($timestamp){
447  // A PHP implementation of conditional get, see
448  //   http://fishbowl.pastiche.org/archives/001132.html
449  $last_modified = substr(date('r', $timestamp), 0, -5).'GMT';
450  $etag = '"'.md5($last_modified).'"';
451  // Send the headers
452  header("Last-Modified: $last_modified");
453  header("ETag: $etag");
454  // See if the client has provided the required headers
455  if (isset($_SERVER['HTTP_IF_MODIFIED_SINCE'])){
456    $if_modified_since = stripslashes($_SERVER['HTTP_IF_MODIFIED_SINCE']);
457  }else{
458    $if_modified_since = false;
459  }
460
461  if (isset($_SERVER['HTTP_IF_NONE_MATCH'])){
462    $if_none_match = stripslashes($_SERVER['HTTP_IF_NONE_MATCH']);
463  }else{
464    $if_none_match = false;
465  }
466
467  if (!$if_modified_since && !$if_none_match){
468    return;
469  }
470
471  // At least one of the headers is there - check them
472  if ($if_none_match && $if_none_match != $etag) {
473    return; // etag is there but doesn't match
474  }
475
476  if ($if_modified_since && $if_modified_since != $last_modified) {
477    return; // if-modified-since is there but doesn't match
478  }
479
480  // Nothing has changed since their last request - serve a 304 and exit
481  header('HTTP/1.0 304 Not Modified');
482  exit;
483}
484
485//Setup VIM: ex: et ts=2 enc=utf-8 :
486