xref: /dokuwiki/inc/pageutils.php (revision da8154baf873777d30aa8cb74a5c1c61454fce2a) 
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 = $_REQUEST[$param];
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  return preg_replace('/.*:/','',$id);
128}
129
130/**
131 * returns the full path to the datafile specified by ID and
132 * optional revision
133 *
134 * The filename is URL encoded to protect Unicode chars
135 *
136 * @author Andreas Gohr <andi@splitbrain.org>
137 */
138function wikiFN($id,$rev=''){
139  global $conf;
140  $id = cleanID($id);
141  $id = str_replace(':','/',$id);
142  if(empty($rev)){
143    $fn = $conf['datadir'].'/'.utf8_encodeFN($id).'.txt';
144  }else{
145    $fn = $conf['olddir'].'/'.utf8_encodeFN($id).'.'.$rev.'.txt';
146    if($conf['usegzip'] && !@file_exists($fn)){
147      //return gzip if enabled and plaintext doesn't exist
148      $fn .= '.gz';
149    }
150  }
151  return $fn;
152}
153
154/**
155 * returns the full path to the meta file specified by ID and extension
156 *
157 * The filename is URL encoded to protect Unicode chars
158 *
159 * @author Steven Danz <steven-danz@kc.rr.com>
160 */
161function metaFN($id,$ext){
162  global $conf;
163  $id = cleanID($id);
164  $id = str_replace(':','/',$id);
165  $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
166  return $fn;
167}
168
169/**
170 * returns an array of full paths to all metafiles of a given ID
171 *
172 * @author Esther Brunner <esther@kaffeehaus.ch>
173 */
174function metaFiles($id){
175   $name   = noNS($id);
176   $dir    = metaFN(getNS($id),'');
177   $files  = array();
178
179   $dh = @opendir($dir);
180   if(!$dh) return $files;
181   while(($file = readdir($dh)) !== false){
182     if(strpos($file,$name.'.') === 0 && !is_dir($dir.$file))
183       $files[] = $dir.$file;
184   }
185   closedir($dh);
186
187   return $files;
188}
189
190/**
191 * returns the full path to the mediafile specified by ID
192 *
193 * The filename is URL encoded to protect Unicode chars
194 *
195 * @author Andreas Gohr <andi@splitbrain.org>
196 */
197function mediaFN($id){
198  global $conf;
199  $id = cleanID($id);
200  $id = str_replace(':','/',$id);
201    $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
202  return $fn;
203}
204
205/**
206 * Returns the full filepath to a localized textfile if local
207 * version isn't found the english one is returned
208 *
209 * @author Andreas Gohr <andi@splitbrain.org>
210 */
211function localeFN($id){
212  global $conf;
213  $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.txt';
214  if(!@file_exists($file)){
215    //fall back to english
216    $file = DOKU_INC.'inc/lang/en/'.$id.'.txt';
217  }
218  return $file;
219}
220
221/**
222 * Resolve relative paths in IDs
223 *
224 * Do not call directly use resolve_mediaid or resolve_pageid
225 * instead
226 *
227 * Partyly based on a cleanPath function found at
228 * http://www.php.net/manual/en/function.realpath.php#57016
229 *
230 * @author <bart at mediawave dot nl>
231 */
232function resolve_id($ns,$id,$clean=true){
233  // if the id starts with a dot we need to handle the
234  // relative stuff
235  if($id{0} == '.'){
236    // normalize initial dots without a colon
237    $id = preg_replace('/^(\.+)(?=[^:\.])/','\1:',$id);
238    // prepend the current namespace
239    $id = $ns.':'.$id;
240
241    // cleanup relatives
242    $result = array();
243    $pathA  = explode(':', $id);
244    if (!$pathA[0]) $result[] = '';
245    foreach ($pathA AS $key => $dir) {
246      if ($dir == '..') {
247        if (end($result) == '..') {
248          $result[] = '..';
249        } elseif (!array_pop($result)) {
250          $result[] = '..';
251        }
252      } elseif ($dir && $dir != '.') {
253        $result[] = $dir;
254      }
255    }
256    if (!end($pathA)) $result[] = '';
257    $id = implode(':', $result);
258  }elseif($ns !== false && strpos($id,':') === false){
259    //if link contains no namespace. add current namespace (if any)
260    $id = $ns.':'.$id;
261  }
262
263  if($clean) $id = cleanID($id);
264  return $id;
265}
266
267/**
268 * Returns a full media id
269 *
270 * @author Andreas Gohr <andi@splitbrain.org>
271 */
272function resolve_mediaid($ns,&$page,&$exists){
273  $page   = resolve_id($ns,$page);
274  $file   = mediaFN($page);
275  $exists = @file_exists($file);
276}
277
278/**
279 * Returns a full page id
280 *
281 * @author Andreas Gohr <andi@splitbrain.org>
282 */
283function resolve_pageid($ns,&$page,&$exists){
284  global $conf;
285  $exists = false;
286
287  //keep hashlink if exists then clean both parts
288  list($page,$hash) = split('#',$page,2);
289  $hash = cleanID($hash);
290  $page = resolve_id($ns,$page,false); // resolve but don't clean, yet
291
292  // get filename (calls clean itself)
293  $file = wikiFN($page);
294
295  // if ends with colon we have a namespace link
296  if(substr($page,-1) == ':'){
297    if(@file_exists(wikiFN($page.$conf['start']))){
298      // start page inside namespace
299      $page = $page.$conf['start'];
300      $exists = true;
301    }elseif(@file_exists(wikiFN($page.noNS(cleanID($page))))){
302      // page named like the NS inside the NS
303      $page = $page.noNS(cleanID($page));
304      $exists = true;
305    }elseif(@file_exists(wikiFN($page))){
306      // page like namespace exists
307      $page = $page;
308      $exists = true;
309    }else{
310      // fall back to default
311      $page = $page.$conf['start'];
312    }
313  }else{
314    //check alternative plural/nonplural form
315    if(!@file_exists($file)){
316      if( $conf['autoplural'] ){
317        if(substr($page,-1) == 's'){
318          $try = substr($page,0,-1);
319        }else{
320          $try = $page.'s';
321        }
322        if(@file_exists(wikiFN($try))){
323          $page   = $try;
324          $exists = true;
325        }
326      }
327    }else{
328      $exists = true;
329    }
330  }
331
332  // now make sure we have a clean page
333  $page = cleanID($page);
334
335  //add hash if any
336  if(!empty($hash)) $page .= '#'.$hash;
337}
338
339/**
340 * Returns the name of a cachefile from given data
341 *
342 * The needed directory is created by this function!
343 *
344 * @author Andreas Gohr <andi@splitbrain.org>
345 *
346 * @param string $data  This data is used to create a unique md5 name
347 * @param string $ext   This is appended to the filename if given
348 * @return string       The filename of the cachefile
349 */
350function getCacheName($data,$ext=''){
351  global $conf;
352  $md5  = md5($data);
353  $file = $conf['cachedir'].'/'.$md5{0}.'/'.$md5.$ext;
354  io_makeFileDir($file);
355  return $file;
356}
357
358/**
359 * Checks a pageid against $conf['hidepages']
360 *
361 * @author Andreas Gohr <gohr@cosmocode.de>
362 */
363function isHiddenPage($id){
364  global $conf;
365  if(empty($conf['hidepages'])) return false;
366
367  if(preg_match('/'.$conf['hidepages'].'/ui',':'.$id)){
368    return true;
369  }
370  return false;
371}
372
373/**
374 * Reverse of isHiddenPage
375 *
376 * @author Andreas Gohr <gohr@cosmocode.de>
377 */
378function isVisiblePage($id){
379  return !isHiddenPage($id);
380}
381
382/**
383 * Checks and sets HTTP headers for conditional HTTP requests
384 *
385 * @author Simon Willison <swillison@gmail.com>
386 * @link   http://simon.incutio.com/archive/2003/04/23/conditionalGet
387 */
388function http_conditionalRequest($timestamp){
389    // A PHP implementation of conditional get, see
390    //   http://fishbowl.pastiche.org/archives/001132.html
391    $last_modified = substr(date('r', $timestamp), 0, -5).'GMT';
392    $etag = '"'.md5($last_modified).'"';
393    // Send the headers
394    header("Last-Modified: $last_modified");
395    header("ETag: $etag");
396    // See if the client has provided the required headers
397    $if_modified_since = isset($_SERVER['HTTP_IF_MODIFIED_SINCE']) ?
398        stripslashes($_SERVER['HTTP_IF_MODIFIED_SINCE']) :
399        false;
400    $if_none_match = isset($_SERVER['HTTP_IF_NONE_MATCH']) ?
401        stripslashes($_SERVER['HTTP_IF_NONE_MATCH']) :
402        false;
403    if (!$if_modified_since && !$if_none_match) {
404        return;
405    }
406    // At least one of the headers is there - check them
407    if ($if_none_match && $if_none_match != $etag) {
408        return; // etag is there but doesn't match
409    }
410    if ($if_modified_since && $if_modified_since != $last_modified) {
411        return; // if-modified-since is there but doesn't match
412    }
413    // Nothing has changed since their last request - serve a 304 and exit
414    header('HTTP/1.0 304 Not Modified');
415    exit;
416}
417
418//Setup VIM: ex: et ts=2 enc=utf-8 :
419