xref: /dokuwiki/inc/pageutils.php (revision b14a97aef26bc5ae134ea206ab27467c3a30828e) 
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
56  // Namespace autolinking from URL
57  if(substr($id,-1) == ':' || ($conf['useslash'] && substr($id,-1) == '/')){
58    if(page_exists($id.$conf['start'])){
59      // start page inside namespace
60      $id = $id.$conf['start'];
61    }elseif(page_exists($id.noNS(cleanID($id)))){
62      // page named like the NS inside the NS
63      $id = $id.noNS(cleanID($id));
64    }elseif(page_exists($id)){
65      // page like namespace exists
66      $id = substr($id,0,-1);
67    }else{
68      // fall back to default
69      $id = $id.$conf['start'];
70    }
71    send_redirect(wl($id,'',true));
72  }
73
74  if($clean) $id = cleanID($id);
75  if(empty($id) && $param=='id') $id = $conf['start'];
76
77  return $id;
78}
79
80/**
81 * Remove unwanted chars from ID
82 *
83 * Cleans a given ID to only use allowed characters. Accented characters are
84 * converted to unaccented ones
85 *
86 * @author Andreas Gohr <andi@splitbrain.org>
87 * @param  string  $raw_id    The pageid to clean
88 * @param  boolean $ascii     Force ASCII
89 * @param  boolean $media     Allow leading or trailing _ for media files
90 */
91function cleanID($raw_id,$ascii=false,$media=false){
92  global $conf;
93  global $lang;
94  static $sepcharpat = null;
95
96  global $cache_cleanid;
97  $cache = & $cache_cleanid;
98
99  // check if it's already in the memory cache
100  if (isset($cache[(string)$raw_id])) {
101    return $cache[(string)$raw_id];
102    }
103
104  $sepchar = $conf['sepchar'];
105  if($sepcharpat == null) // build string only once to save clock cycles
106    $sepcharpat = '#\\'.$sepchar.'+#';
107
108  $id = trim((string)$raw_id);
109  $id = utf8_strtolower($id);
110
111  //alternative namespace seperator
112  $id = strtr($id,';',':');
113  if($conf['useslash']){
114    $id = strtr($id,'/',':');
115  }else{
116    $id = strtr($id,'/',$sepchar);
117  }
118
119  if($conf['deaccent'] == 2 || $ascii) $id = utf8_romanize($id);
120  if($conf['deaccent'] || $ascii) $id = utf8_deaccent($id,-1);
121
122  //remove specials
123  $id = utf8_stripspecials($id,$sepchar,'\*');
124
125  if($ascii) $id = utf8_strip($id);
126
127  //clean up
128  $id = preg_replace($sepcharpat,$sepchar,$id);
129  $id = preg_replace('#:+#',':',$id);
130  $id = ($media ? trim($id,':.-') : trim($id,':._-'));
131  $id = preg_replace('#:[:\._\-]+#',':',$id);
132
133  $cache[(string)$raw_id] = $id;
134  return($id);
135}
136
137/**
138 * Return namespacepart of a wiki ID
139 *
140 * @author Andreas Gohr <andi@splitbrain.org>
141 */
142function getNS($id){
143  $pos = strrpos((string)$id,':');
144  if($pos!==false){
145    return substr((string)$id,0,$pos);
146  }
147  return false;
148}
149
150/**
151 * Returns the ID without the namespace
152 *
153 * @author Andreas Gohr <andi@splitbrain.org>
154 */
155function noNS($id) {
156  $pos = strrpos($id, ':');
157  if ($pos!==false) {
158    return substr($id, $pos+1);
159  } else {
160    return $id;
161  }
162}
163
164/**
165 * Returns the current namespace
166 *
167 * @author Nathan Fritz <fritzn@crown.edu>
168 */
169function curNS($id) {
170    return noNS(getNS($id));
171}
172
173/**
174 * Returns the ID without the namespace or current namespace for 'start' pages
175 *
176 * @author Nathan Fritz <fritzn@crown.edu>
177 */
178function noNSorNS($id) {
179    global $conf;
180
181    $p = noNS($id);
182    if ($p == $conf['start']) {
183        $p = curNS($id);
184        if ($p == false) {
185            return noNS($id);
186        }
187    }
188    return $p;
189}
190
191/**
192 * Creates a XHTML valid linkid from a given headline title
193 *
194 * @param string  $title   The headline title
195 * @param array   $check   List of existing IDs
196 * @author Andreas Gohr <andi@splitbrain.org>
197 */
198function sectionID($title,&$check) {
199    $title = str_replace(':','',cleanID($title));
200    $new = ltrim($title,'0123456789._-');
201    if(empty($new)){
202        $title = 'section'.preg_replace('/[^0-9]+/','',$title); //keep numbers from headline
203    }else{
204        $title = $new;
205    }
206
207    if(is_array($check)){
208        // make sure tiles are unique
209        $num = '';
210        while(in_array($title.$num,$check)){
211            ($num) ? $num++ : $num = 1;
212        }
213        $title = $title.$num;
214        $check[] = $title;
215    }
216
217    return $title;
218}
219
220
221/**
222 *  Wiki page existence check
223 *
224 *  parameters as for wikiFN
225 *
226 *  @author Chris Smith <chris@jalakai.co.uk>
227 */
228function page_exists($id,$rev='',$clean=true) {
229  return @file_exists(wikiFN($id,$rev,$clean));
230}
231
232/**
233 * returns the full path to the datafile specified by ID and optional revision
234 *
235 * The filename is URL encoded to protect Unicode chars
236 *
237 * @param  $raw_id  string   id of wikipage
238 * @param  $rev     string   page revision, empty string for current
239 * @param  $clean   bool     flag indicating that $raw_id should be cleaned.  Only set to false
240 *                           when $id is guaranteed to have been cleaned already.
241 *
242 * @author Andreas Gohr <andi@splitbrain.org>
243 */
244function wikiFN($raw_id,$rev='',$clean=true){
245  global $conf;
246
247  global $cache_wikifn;
248  $cache = & $cache_wikifn;
249
250  if (isset($cache[$raw_id]) && isset($cache[$raw_id][$rev])) {
251    return $cache[$raw_id][$rev];
252  }
253
254  $id = $raw_id;
255
256  if ($clean) $id = cleanID($id);
257  $id = str_replace(':','/',$id);
258  if(empty($rev)){
259    $fn = $conf['datadir'].'/'.utf8_encodeFN($id).'.txt';
260  }else{
261    $fn = $conf['olddir'].'/'.utf8_encodeFN($id).'.'.$rev.'.txt';
262    if($conf['compression']){
263      //test for extensions here, we want to read both compressions
264       if (@file_exists($fn . '.gz')){
265          $fn .= '.gz';
266       }else if(@file_exists($fn . '.bz2')){
267          $fn .= '.bz2';
268       }else{
269          //file doesnt exist yet, so we take the configured extension
270          $fn .= '.' . $conf['compression'];
271       }
272    }
273  }
274
275  if (!isset($cache[$raw_id])) { $cache[$raw_id] = array(); }
276  $cache[$raw_id][$rev] = $fn;
277  return $fn;
278}
279
280/**
281 * Returns the full path to the file for locking the page while editing.
282 *
283 * @author Ben Coburn <btcoburn@silicodon.net>
284 */
285function wikiLockFN($id) {
286  global $conf;
287  return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
288}
289
290
291/**
292 * returns the full path to the meta file specified by ID and extension
293 *
294 * The filename is URL encoded to protect Unicode chars
295 *
296 * @author Steven Danz <steven-danz@kc.rr.com>
297 */
298function metaFN($id,$ext){
299  global $conf;
300  $id = cleanID($id);
301  $id = str_replace(':','/',$id);
302  $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
303  return $fn;
304}
305
306/**
307 * returns an array of full paths to all metafiles of a given ID
308 *
309 * @author Esther Brunner <esther@kaffeehaus.ch>
310 */
311function metaFiles($id){
312   $name   = noNS($id);
313   $ns     = getNS($id);
314   $dir    = ($ns) ? metaFN($ns,'').'/' : metaFN($ns,'');
315   $files  = array();
316
317   $dh = @opendir($dir);
318   if(!$dh) return $files;
319   while(($file = readdir($dh)) !== false){
320     if(strpos($file,$name.'.') === 0 && !is_dir($dir.$file))
321       $files[] = $dir.$file;
322   }
323   closedir($dh);
324
325   return $files;
326}
327
328/**
329 * returns the full path to the mediafile specified by ID
330 *
331 * The filename is URL encoded to protect Unicode chars
332 *
333 * @author Andreas Gohr <andi@splitbrain.org>
334 */
335function mediaFN($id){
336  global $conf;
337  $id = cleanID($id);
338  $id = str_replace(':','/',$id);
339    $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
340  return $fn;
341}
342
343/**
344 * Returns the full filepath to a localized textfile if local
345 * version isn't found the english one is returned
346 *
347 * @author Andreas Gohr <andi@splitbrain.org>
348 */
349function localeFN($id){
350  global $conf;
351  $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.txt';
352  if(!@file_exists($file)){
353    //fall back to english
354    $file = DOKU_INC.'inc/lang/en/'.$id.'.txt';
355  }
356  return $file;
357}
358
359/**
360 * Resolve relative paths in IDs
361 *
362 * Do not call directly use resolve_mediaid or resolve_pageid
363 * instead
364 *
365 * Partyly based on a cleanPath function found at
366 * http://www.php.net/manual/en/function.realpath.php#57016
367 *
368 * @author <bart at mediawave dot nl>
369 */
370function resolve_id($ns,$id,$clean=true){
371  global $conf;
372
373  // some pre cleaning for useslash:
374  if($conf['useslash']) $id = str_replace('/',':',$id);
375
376  // if the id starts with a dot we need to handle the
377  // relative stuff
378  if($id{0} == '.'){
379    // normalize initial dots without a colon
380    $id = preg_replace('/^(\.+)(?=[^:\.])/','\1:',$id);
381    // prepend the current namespace
382    $id = $ns.':'.$id;
383
384    // cleanup relatives
385    $result = array();
386    $pathA  = explode(':', $id);
387    if (!$pathA[0]) $result[] = '';
388    foreach ($pathA AS $key => $dir) {
389      if ($dir == '..') {
390        if (end($result) == '..') {
391          $result[] = '..';
392        } elseif (!array_pop($result)) {
393          $result[] = '..';
394        }
395      } elseif ($dir && $dir != '.') {
396        $result[] = $dir;
397      }
398    }
399    if (!end($pathA)) $result[] = '';
400    $id = implode(':', $result);
401  }elseif($ns !== false && strpos($id,':') === false){
402    //if link contains no namespace. add current namespace (if any)
403    $id = $ns.':'.$id;
404  }
405
406  if($clean) $id = cleanID($id);
407  return $id;
408}
409
410/**
411 * Returns a full media id
412 *
413 * @author Andreas Gohr <andi@splitbrain.org>
414 */
415function resolve_mediaid($ns,&$page,&$exists){
416  $page   = resolve_id($ns,$page);
417  $file   = mediaFN($page);
418  $exists = @file_exists($file);
419}
420
421/**
422 * Returns a full page id
423 *
424 * @author Andreas Gohr <andi@splitbrain.org>
425 */
426function resolve_pageid($ns,&$page,&$exists){
427  global $conf;
428  $exists = false;
429
430  //keep hashlink if exists then clean both parts
431  if (strpos($page,'#')) {
432    list($page,$hash) = split('#',$page,2);
433  } else {
434    $hash = '';
435  }
436  $hash = cleanID($hash);
437  $page = resolve_id($ns,$page,false); // resolve but don't clean, yet
438
439  // get filename (calls clean itself)
440  $file = wikiFN($page);
441
442  // if ends with colon or slash we have a namespace link
443  if(substr($page,-1) == ':' || ($conf['useslash'] && substr($page,-1) == '/')){
444    if(page_exists($page.$conf['start'])){
445      // start page inside namespace
446      $page = $page.$conf['start'];
447      $exists = true;
448    }elseif(page_exists($page.noNS(cleanID($page)))){
449      // page named like the NS inside the NS
450      $page = $page.noNS(cleanID($page));
451      $exists = true;
452    }elseif(page_exists($page)){
453      // page like namespace exists
454      $page = $page;
455      $exists = true;
456    }else{
457      // fall back to default
458      $page = $page.$conf['start'];
459    }
460  }else{
461    //check alternative plural/nonplural form
462    if(!@file_exists($file)){
463      if( $conf['autoplural'] ){
464        if(substr($page,-1) == 's'){
465          $try = substr($page,0,-1);
466        }else{
467          $try = $page.'s';
468        }
469        if(page_exists($try)){
470          $page   = $try;
471          $exists = true;
472        }
473      }
474    }else{
475      $exists = true;
476    }
477  }
478
479  // now make sure we have a clean page
480  $page = cleanID($page);
481
482  //add hash if any
483  if(!empty($hash)) $page .= '#'.$hash;
484}
485
486/**
487 * Returns the name of a cachefile from given data
488 *
489 * The needed directory is created by this function!
490 *
491 * @author Andreas Gohr <andi@splitbrain.org>
492 *
493 * @param string $data  This data is used to create a unique md5 name
494 * @param string $ext   This is appended to the filename if given
495 * @return string       The filename of the cachefile
496 */
497function getCacheName($data,$ext=''){
498  global $conf;
499  $md5  = md5($data);
500  $file = $conf['cachedir'].'/'.$md5{0}.'/'.$md5.$ext;
501  io_makeFileDir($file);
502  return $file;
503}
504
505/**
506 * Checks a pageid against $conf['hidepages']
507 *
508 * @author Andreas Gohr <gohr@cosmocode.de>
509 */
510function isHiddenPage($id){
511  global $conf;
512  global $ACT;
513  if(empty($conf['hidepages'])) return false;
514  if($ACT == 'admin') return false;
515
516  if(preg_match('/'.$conf['hidepages'].'/ui',':'.$id)){
517    return true;
518  }
519  return false;
520}
521
522/**
523 * Reverse of isHiddenPage
524 *
525 * @author Andreas Gohr <gohr@cosmocode.de>
526 */
527function isVisiblePage($id){
528  return !isHiddenPage($id);
529}
530
531/**
532 * Checks and sets HTTP headers for conditional HTTP requests
533 *
534 * @author   Simon Willison <swillison@gmail.com>
535 * @link     http://simon.incutio.com/archive/2003/04/23/conditionalGet
536 * @param    timestamp $timestamp lastmodified time of the cache file
537 * @returns  void or exits with previously header() commands executed
538 */
539function http_conditionalRequest($timestamp){
540  // A PHP implementation of conditional get, see
541  //   http://fishbowl.pastiche.org/archives/001132.html
542  $last_modified = substr(gmdate('r', $timestamp), 0, -5).'GMT';
543  $etag = '"'.md5($last_modified).'"';
544  // Send the headers
545  header("Last-Modified: $last_modified");
546  header("ETag: $etag");
547  // See if the client has provided the required headers
548  if (isset($_SERVER['HTTP_IF_MODIFIED_SINCE'])){
549    $if_modified_since = stripslashes($_SERVER['HTTP_IF_MODIFIED_SINCE']);
550  }else{
551    $if_modified_since = false;
552  }
553
554  if (isset($_SERVER['HTTP_IF_NONE_MATCH'])){
555    $if_none_match = stripslashes($_SERVER['HTTP_IF_NONE_MATCH']);
556  }else{
557    $if_none_match = false;
558  }
559
560  if (!$if_modified_since && !$if_none_match){
561    return;
562  }
563
564  // At least one of the headers is there - check them
565  if ($if_none_match && $if_none_match != $etag) {
566    return; // etag is there but doesn't match
567  }
568
569  if ($if_modified_since && $if_modified_since != $last_modified) {
570    return; // if-modified-since is there but doesn't match
571  }
572
573  // Nothing has changed since their last request - serve a 304 and exit
574  header('HTTP/1.0 304 Not Modified');
575
576  // don't produce output, even if compression is on
577  ob_end_clean();
578  exit;
579}
580
581/**
582 * Let the webserver send the given file vi x-sendfile method
583 *
584 * @author Chris Smith <chris.eureka@jalakai.co.uk>
585 * @returns  void or exits with previously header() commands executed
586 */
587function http_sendfile($file) {
588  global $conf;
589
590  //use x-sendfile header to pass the delivery to compatible webservers
591  if($conf['xsendfile'] == 1){
592    header("X-LIGHTTPD-send-file: $file");
593    ob_end_clean();
594    exit;
595  }elseif($conf['xsendfile'] == 2){
596    header("X-Sendfile: $file");
597    ob_end_clean();
598    exit;
599  }elseif($conf['xsendfile'] == 3){
600    header("X-Accel-Redirect: $file");
601    ob_end_clean();
602    exit;
603  }
604
605  return false;
606}
607
608/**
609 * Check for a gzipped version and create if necessary
610 *
611 * return true if there exists a gzip version of the uncompressed file
612 * (samepath/samefilename.sameext.gz) created after the uncompressed file
613 *
614 * @author Chris Smith <chris.eureka@jalakai.co.uk>
615 */
616function http_gzip_valid($uncompressed_file) {
617  $gzip = $uncompressed_file.'.gz';
618  if (filemtime($gzip) < filemtime($uncompressed_file)) {    // filemtime returns false (0) if file doesn't exist
619    return copy($uncompressed_file, 'compress.zlib://'.$gzip);
620  }
621
622  return true;
623}
624//Setup VIM: ex: et ts=2 enc=utf-8 :
625