xref: /dokuwiki/inc/pageutils.php (revision ed8e6fd51df4e35299f8e53287bffc025eb1f27a) 
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 *
21 * @param string $param  the $_REQUEST variable name, default 'id'
22 * @param bool   $clean  if true, ID is cleaned
23 * @return string
24 */
25function getID($param='id',$clean=true){
26    /** @var Input $INPUT */
27    global $INPUT;
28    global $conf;
29    global $ACT;
30
31    $id = $INPUT->str($param);
32
33    //construct page id from request URI
34    if(empty($id) && $conf['userewrite'] == 2){
35        $request = $INPUT->server->str('REQUEST_URI');
36        $script = '';
37
38        //get the script URL
39        if($conf['basedir']){
40            $relpath = '';
41            if($param != 'id') {
42                $relpath = 'lib/exe/';
43            }
44            $script = $conf['basedir'].$relpath.utf8_basename($INPUT->server->str('SCRIPT_FILENAME'));
45
46        }elseif($INPUT->server->str('PATH_INFO')){
47            $request = $INPUT->server->str('PATH_INFO');
48        }elseif($INPUT->server->str('SCRIPT_NAME')){
49            $script = $INPUT->server->str('SCRIPT_NAME');
50        }elseif($INPUT->server->str('DOCUMENT_ROOT') && $INPUT->server->str('SCRIPT_FILENAME')){
51            $script = preg_replace ('/^'.preg_quote($INPUT->server->str('DOCUMENT_ROOT'),'/').'/','',
52                    $INPUT->server->str('SCRIPT_FILENAME'));
53            $script = '/'.$script;
54        }
55
56        //clean script and request (fixes a windows problem)
57        $script  = preg_replace('/\/\/+/','/',$script);
58        $request = preg_replace('/\/\/+/','/',$request);
59
60        //remove script URL and Querystring to gain the id
61        if(preg_match('/^'.preg_quote($script,'/').'(.*)/',$request, $match)){
62            $id = preg_replace ('/\?.*/','',$match[1]);
63        }
64        $id = urldecode($id);
65        //strip leading slashes
66        $id = preg_replace('!^/+!','',$id);
67    }
68
69    // Namespace autolinking from URL
70    if(substr($id,-1) == ':' || ($conf['useslash'] && substr($id,-1) == '/')){
71        if(page_exists($id.$conf['start'])){
72            // start page inside namespace
73            $id = $id.$conf['start'];
74        }elseif(page_exists($id.noNS(cleanID($id)))){
75            // page named like the NS inside the NS
76            $id = $id.noNS(cleanID($id));
77        }elseif(page_exists($id)){
78            // page like namespace exists
79            $id = substr($id,0,-1);
80        }else{
81            // fall back to default
82            $id = $id.$conf['start'];
83        }
84        if (isset($ACT) && $ACT === 'show') send_redirect(wl($id,'',true));
85    }
86
87    if($clean) $id = cleanID($id);
88    if(empty($id) && $param=='id') $id = $conf['start'];
89
90    return $id;
91}
92
93/**
94 * Remove unwanted chars from ID
95 *
96 * Cleans a given ID to only use allowed characters. Accented characters are
97 * converted to unaccented ones
98 *
99 * @author Andreas Gohr <andi@splitbrain.org>
100 *
101 * @param  string  $raw_id    The pageid to clean
102 * @param  boolean $ascii     Force ASCII
103 * @return string cleaned id
104 */
105function cleanID($raw_id,$ascii=false){
106    global $conf;
107    static $sepcharpat = null;
108
109    global $cache_cleanid;
110    $cache = & $cache_cleanid;
111
112    // check if it's already in the memory cache
113    if (!$ascii && isset($cache[(string)$raw_id])) {
114        return $cache[(string)$raw_id];
115    }
116
117    $sepchar = $conf['sepchar'];
118    if($sepcharpat == null) // build string only once to save clock cycles
119        $sepcharpat = '#\\'.$sepchar.'+#';
120
121    $id = trim((string)$raw_id);
122    $id = utf8_strtolower($id);
123
124    //alternative namespace seperator
125    if($conf['useslash']){
126        $id = strtr($id,';/','::');
127    }else{
128        $id = strtr($id,';/',':'.$sepchar);
129    }
130
131    if($conf['deaccent'] == 2 || $ascii) $id = utf8_romanize($id);
132    if($conf['deaccent'] || $ascii) $id = utf8_deaccent($id,-1);
133
134    //remove specials
135    $id = utf8_stripspecials($id,$sepchar,'\*');
136
137    if($ascii) $id = utf8_strip($id);
138
139    //clean up
140    $id = preg_replace($sepcharpat,$sepchar,$id);
141    $id = preg_replace('#:+#',':',$id);
142    $id = trim($id,':._-');
143    $id = preg_replace('#:[:\._\-]+#',':',$id);
144    $id = preg_replace('#[:\._\-]+:#',':',$id);
145
146    if (!$ascii) $cache[(string)$raw_id] = $id;
147    return($id);
148}
149
150/**
151 * Return namespacepart of a wiki ID
152 *
153 * @author Andreas Gohr <andi@splitbrain.org>
154 *
155 * @param string $id
156 * @return string|false the namespace part or false if the given ID has no namespace (root)
157 */
158function getNS($id){
159    $pos = strrpos((string)$id,':');
160    if($pos!==false){
161        return substr((string)$id,0,$pos);
162    }
163    return false;
164}
165
166/**
167 * Returns the ID without the namespace
168 *
169 * @author Andreas Gohr <andi@splitbrain.org>
170 *
171 * @param string $id
172 * @return string
173 */
174function noNS($id) {
175    $pos = strrpos($id, ':');
176    if ($pos!==false) {
177        return substr($id, $pos+1);
178    } else {
179        return $id;
180    }
181}
182
183/**
184 * Returns the current namespace
185 *
186 * @author Nathan Fritz <fritzn@crown.edu>
187 *
188 * @param string $id
189 * @return string
190 */
191function curNS($id) {
192    return noNS(getNS($id));
193}
194
195/**
196 * Returns the ID without the namespace or current namespace for 'start' pages
197 *
198 * @author Nathan Fritz <fritzn@crown.edu>
199 *
200 * @param string $id
201 * @return string
202 */
203function noNSorNS($id) {
204    global $conf;
205
206    $p = noNS($id);
207    if ($p == $conf['start'] || $p == false) {
208        $p = curNS($id);
209        if ($p == false) {
210            return $conf['start'];
211        }
212    }
213    return $p;
214}
215
216/**
217 * Creates a XHTML valid linkid from a given headline title
218 *
219 * @param string  $title   The headline title
220 * @param array|bool   $check   Existing IDs (title => number)
221 * @return string the title
222 *
223 * @author Andreas Gohr <andi@splitbrain.org>
224 */
225function sectionID($title,&$check) {
226    $title = str_replace(array(':','.'),'',cleanID($title));
227    $new = ltrim($title,'0123456789_-');
228    if(empty($new)){
229        $title = 'section'.preg_replace('/[^0-9]+/','',$title); //keep numbers from headline
230    }else{
231        $title = $new;
232    }
233
234    if(is_array($check)){
235        // make sure tiles are unique
236        if (!array_key_exists ($title,$check)) {
237            $check[$title] = 0;
238        } else {
239            $title .= ++ $check[$title];
240        }
241    }
242
243    return $title;
244}
245
246/**
247 * Wiki page existence check
248 *
249 * parameters as for wikiFN
250 *
251 * @author Chris Smith <chris@jalakai.co.uk>
252 *
253 * @param string $id page id
254 * @param string|int $rev empty or revision timestamp
255 * @param bool $clean flag indicating that $id should be cleaned (see wikiFN as well)
256 * @param bool $date_at
257 * @return bool exists?
258 */
259function page_exists($id,$rev='',$clean=true, $date_at=false) {
260    if($rev !== '' && $date_at) {
261        $pagelog = new PageChangeLog($id);
262        $pagelog_rev = $pagelog->getLastRevisionAt($rev);
263        if($pagelog_rev !== false)
264            $rev = $pagelog_rev;
265    }
266    return file_exists(wikiFN($id,$rev,$clean));
267}
268
269/**
270 * returns the full path to the datafile specified by ID and optional revision
271 *
272 * The filename is URL encoded to protect Unicode chars
273 *
274 * @param  $raw_id  string   id of wikipage
275 * @param  $rev     int|string   page revision, empty string for current
276 * @param  $clean   bool     flag indicating that $raw_id should be cleaned.  Only set to false
277 *                           when $id is guaranteed to have been cleaned already.
278 * @return string full path
279 *
280 * @author Andreas Gohr <andi@splitbrain.org>
281 */
282function wikiFN($raw_id,$rev='',$clean=true){
283    global $conf;
284
285    global $cache_wikifn;
286    $cache = & $cache_wikifn;
287
288    $id = $raw_id;
289
290    if ($clean) $id = cleanID($id);
291    $id = str_replace(':','/',$id);
292
293    if (isset($cache[$id]) && isset($cache[$id][$rev])) {
294        return $cache[$id][$rev];
295    }
296
297    if(empty($rev)){
298        $fn = $conf['datadir'].'/'.utf8_encodeFN($id).'.txt';
299    }else{
300        $fn = $conf['olddir'].'/'.utf8_encodeFN($id).'.'.$rev.'.txt';
301        if($conf['compression']){
302            //test for extensions here, we want to read both compressions
303            if (file_exists($fn . '.gz')){
304                $fn .= '.gz';
305            }else if(file_exists($fn . '.bz2')){
306                $fn .= '.bz2';
307            }else{
308                //file doesnt exist yet, so we take the configured extension
309                $fn .= '.' . $conf['compression'];
310            }
311        }
312    }
313
314    if (!isset($cache[$id])) { $cache[$id] = array(); }
315    $cache[$id][$rev] = $fn;
316    return $fn;
317}
318
319/**
320 * Returns the full path to the file for locking the page while editing.
321 *
322 * @author Ben Coburn <btcoburn@silicodon.net>
323 *
324 * @param string $id page id
325 * @return string full path
326 */
327function wikiLockFN($id) {
328    global $conf;
329    return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
330}
331
332
333/**
334 * returns the full path to the meta file specified by ID and extension
335 *
336 * @author Steven Danz <steven-danz@kc.rr.com>
337 *
338 * @param string $id   page id
339 * @param string $ext  file extension
340 * @return string full path
341 */
342function metaFN($id,$ext){
343    global $conf;
344    $id = cleanID($id);
345    $id = str_replace(':','/',$id);
346    $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
347    return $fn;
348}
349
350/**
351 * returns the full path to the media's meta file specified by ID and extension
352 *
353 * @author Kate Arzamastseva <pshns@ukr.net>
354 *
355 * @param string $id   media id
356 * @param string $ext  extension of media
357 * @return string
358 */
359function mediaMetaFN($id,$ext){
360    global $conf;
361    $id = cleanID($id);
362    $id = str_replace(':','/',$id);
363    $fn = $conf['mediametadir'].'/'.utf8_encodeFN($id).$ext;
364    return $fn;
365}
366
367/**
368 * returns an array of full paths to all metafiles of a given ID
369 *
370 * @author Esther Brunner <esther@kaffeehaus.ch>
371 * @author Michael Hamann <michael@content-space.de>
372 *
373 * @param string $id page id
374 * @return array
375 */
376function metaFiles($id){
377    $basename = metaFN($id, '');
378    $files    = glob($basename.'.*', GLOB_MARK);
379    // filter files like foo.bar.meta when $id == 'foo'
380    return    $files ? preg_grep('/^'.preg_quote($basename, '/').'\.[^.\/]*$/u', $files) : array();
381}
382
383/**
384 * returns the full path to the mediafile specified by ID
385 *
386 * The filename is URL encoded to protect Unicode chars
387 *
388 * @author Andreas Gohr <andi@splitbrain.org>
389 * @author Kate Arzamastseva <pshns@ukr.net>
390 *
391 * @param string     $id  media id
392 * @param string|int $rev empty string or revision timestamp
393 * @return string full path
394 */
395function mediaFN($id, $rev='', $clean=true){
396    global $conf;
397    if ($clean) $id = cleanID($id);
398    $id = str_replace(':','/',$id);
399    if(empty($rev)){
400        $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
401    }else{
402        $ext = mimetype($id);
403        $name = substr($id,0, -1*strlen($ext[0])-1);
404        $fn = $conf['mediaolddir'].'/'.utf8_encodeFN($name .'.'.( (int) $rev ).'.'.$ext[0]);
405    }
406    return $fn;
407}
408
409/**
410 * Returns the full filepath to a localized file if local
411 * version isn't found the english one is returned
412 *
413 * @param  string $id  The id of the local file
414 * @param  string $ext The file extension (usually txt)
415 * @return string full filepath to localized file
416 *
417 * @author Andreas Gohr <andi@splitbrain.org>
418 */
419function localeFN($id,$ext='txt'){
420    global $conf;
421    $file = DOKU_CONF.'lang/'.$conf['lang'].'/'.$id.'.'.$ext;
422    if(!file_exists($file)){
423        $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.'.$ext;
424        if(!file_exists($file)){
425            //fall back to english
426            $file = DOKU_INC.'inc/lang/en/'.$id.'.'.$ext;
427        }
428    }
429    return $file;
430}
431
432/**
433 * Resolve relative paths in IDs
434 *
435 * Do not call directly use resolve_mediaid or resolve_pageid
436 * instead
437 *
438 * Partyly based on a cleanPath function found at
439 * http://www.php.net/manual/en/function.realpath.php#57016
440 *
441 * @author <bart at mediawave dot nl>
442 *
443 * @param string $ns     namespace which is context of id
444 * @param string $id     relative id
445 * @param bool   $clean  flag indicating that id should be cleaned
446 * @return string
447 */
448function resolve_id($ns,$id,$clean=true){
449    global $conf;
450
451    // some pre cleaning for useslash:
452    if($conf['useslash']) $id = str_replace('/',':',$id);
453
454    // if the id starts with a dot we need to handle the
455    // relative stuff
456    if($id && $id{0} == '.'){
457        // normalize initial dots without a colon
458        $id = preg_replace('/^(\.+)(?=[^:\.])/','\1:',$id);
459        // prepend the current namespace
460        $id = $ns.':'.$id;
461
462        // cleanup relatives
463        $result = array();
464        $pathA  = explode(':', $id);
465        if (!$pathA[0]) $result[] = '';
466        foreach ($pathA AS $key => $dir) {
467            if ($dir == '..') {
468                if (end($result) == '..') {
469                    $result[] = '..';
470                } elseif (!array_pop($result)) {
471                    $result[] = '..';
472                }
473            } elseif ($dir && $dir != '.') {
474                $result[] = $dir;
475            }
476        }
477        if (!end($pathA)) $result[] = '';
478        $id = implode(':', $result);
479    }elseif($ns !== false && strpos($id,':') === false){
480        //if link contains no namespace. add current namespace (if any)
481        $id = $ns.':'.$id;
482    }
483
484    if($clean) $id = cleanID($id);
485    return $id;
486}
487
488/**
489 * Returns a full media id
490 *
491 * @author Andreas Gohr <andi@splitbrain.org>
492 *
493 * @param string $ns namespace which is context of id
494 * @param string &$page (reference) relative media id, updated to resolved id
495 * @param bool &$exists (reference) updated with existance of media
496 * @param int|string $rev
497 * @param bool $date_at
498 */
499function resolve_mediaid($ns,&$page,&$exists,$rev='',$date_at=false){
500    $page   = resolve_id($ns,$page);
501    if($rev !== '' &&  $date_at){
502        $medialog = new MediaChangeLog($page);
503        $medialog_rev = $medialog->getLastRevisionAt($rev);
504        if($medialog_rev !== false) {
505            $rev = $medialog_rev;
506        }
507    }
508
509    $file   = mediaFN($page,$rev);
510    $exists = file_exists($file);
511}
512
513/**
514 * Returns a full page id
515 *
516 * @author Andreas Gohr <andi@splitbrain.org>
517 *
518 * @param string $ns namespace which is context of id
519 * @param string &$page (reference) relative page id, updated to resolved id
520 * @param bool &$exists (reference) updated with existance of media
521 * @param string $rev
522 * @param bool $date_at
523 */
524function resolve_pageid($ns,&$page,&$exists,$rev='',$date_at=false ){
525    global $conf;
526    global $ID;
527    $exists = false;
528
529    //empty address should point to current page
530    if ($page === "") {
531        $page = $ID;
532    }
533
534    //keep hashlink if exists then clean both parts
535    if (strpos($page,'#')) {
536        list($page,$hash) = explode('#',$page,2);
537    } else {
538        $hash = '';
539    }
540    $hash = cleanID($hash);
541    $page = resolve_id($ns,$page,false); // resolve but don't clean, yet
542
543    // get filename (calls clean itself)
544    if($rev !== '' && $date_at) {
545        $pagelog = new PageChangeLog($page);
546        $pagelog_rev = $pagelog->getLastRevisionAt($rev);
547        if($pagelog_rev !== false)//something found
548           $rev  = $pagelog_rev;
549    }
550    $file = wikiFN($page,$rev);
551
552    // if ends with colon or slash we have a namespace link
553    if(in_array(substr($page,-1), array(':', ';')) ||
554       ($conf['useslash'] && substr($page,-1) == '/')){
555        if(page_exists($page.$conf['start'],$rev,true,$date_at)){
556            // start page inside namespace
557            $page = $page.$conf['start'];
558            $exists = true;
559        }elseif(page_exists($page.noNS(cleanID($page)),$rev,true,$date_at)){
560            // page named like the NS inside the NS
561            $page = $page.noNS(cleanID($page));
562            $exists = true;
563        }elseif(page_exists($page,$rev,true,$date_at)){
564            // page like namespace exists
565            $page = $page;
566            $exists = true;
567        }else{
568            // fall back to default
569            $page = $page.$conf['start'];
570        }
571    }else{
572        //check alternative plural/nonplural form
573        if(!file_exists($file)){
574            if( $conf['autoplural'] ){
575                if(substr($page,-1) == 's'){
576                    $try = substr($page,0,-1);
577                }else{
578                    $try = $page.'s';
579                }
580                if(page_exists($try,$rev,true,$date_at)){
581                    $page   = $try;
582                    $exists = true;
583                }
584            }
585        }else{
586            $exists = true;
587        }
588    }
589
590    // now make sure we have a clean page
591    $page = cleanID($page);
592
593    //add hash if any
594    if(!empty($hash)) $page .= '#'.$hash;
595}
596
597/**
598 * Returns the name of a cachefile from given data
599 *
600 * The needed directory is created by this function!
601 *
602 * @author Andreas Gohr <andi@splitbrain.org>
603 *
604 * @param string $data  This data is used to create a unique md5 name
605 * @param string $ext   This is appended to the filename if given
606 * @return string       The filename of the cachefile
607 */
608function getCacheName($data,$ext=''){
609    global $conf;
610    $md5  = md5($data);
611    $file = $conf['cachedir'].'/'.$md5{0}.'/'.$md5.$ext;
612    io_makeFileDir($file);
613    return $file;
614}
615
616/**
617 * Checks a pageid against $conf['hidepages']
618 *
619 * @author Andreas Gohr <gohr@cosmocode.de>
620 *
621 * @param string $id page id
622 * @return bool
623 */
624function isHiddenPage($id){
625    $data = array(
626        'id' => $id,
627        'hidden' => false
628    );
629    trigger_event('PAGEUTILS_ID_HIDEPAGE', $data, '_isHiddenPage');
630    return $data['hidden'];
631}
632
633/**
634 * callback checks if page is hidden
635 *
636 * @param array $data event data    - see isHiddenPage()
637 */
638function _isHiddenPage(&$data) {
639    global $conf;
640    global $ACT;
641
642    if ($data['hidden']) return;
643    if(empty($conf['hidepages'])) return;
644    if($ACT == 'admin') return;
645
646    if(preg_match('/'.$conf['hidepages'].'/ui',':'.$data['id'])){
647        $data['hidden'] = true;
648    }
649}
650
651/**
652 * Reverse of isHiddenPage
653 *
654 * @author Andreas Gohr <gohr@cosmocode.de>
655 *
656 * @param string $id page id
657 * @return bool
658 */
659function isVisiblePage($id){
660    return !isHiddenPage($id);
661}
662
663/**
664 * Format an id for output to a user
665 *
666 * Namespaces are denoted by a trailing “:*”. The root namespace is
667 * “*”. Output is escaped.
668 *
669 * @author Adrian Lang <lang@cosmocode.de>
670 *
671 * @param string $id page id
672 * @return string
673 */
674function prettyprint_id($id) {
675    if (!$id || $id === ':') {
676        return '*';
677    }
678    if ((substr($id, -1, 1) === ':')) {
679        $id .= '*';
680    }
681    return hsc($id);
682}
683
684/**
685 * Encode a UTF-8 filename to use on any filesystem
686 *
687 * Uses the 'fnencode' option to determine encoding
688 *
689 * When the second parameter is true the string will
690 * be encoded only if non ASCII characters are detected -
691 * This makes it safe to run it multiple times on the
692 * same string (default is true)
693 *
694 * @author Andreas Gohr <andi@splitbrain.org>
695 * @see    urlencode
696 *
697 * @param string $file file name
698 * @param bool   $safe if true, only encoded when non ASCII characters detected
699 * @return string
700 */
701function utf8_encodeFN($file,$safe=true){
702    global $conf;
703    if($conf['fnencode'] == 'utf-8') return $file;
704
705    if($safe && preg_match('#^[a-zA-Z0-9/_\-\.%]+$#',$file)){
706        return $file;
707    }
708
709    if($conf['fnencode'] == 'safe'){
710        return SafeFN::encode($file);
711    }
712
713    $file = urlencode($file);
714    $file = str_replace('%2F','/',$file);
715    return $file;
716}
717
718/**
719 * Decode a filename back to UTF-8
720 *
721 * Uses the 'fnencode' option to determine encoding
722 *
723 * @author Andreas Gohr <andi@splitbrain.org>
724 * @see    urldecode
725 *
726 * @param string $file file name
727 * @return string
728 */
729function utf8_decodeFN($file){
730    global $conf;
731    if($conf['fnencode'] == 'utf-8') return $file;
732
733    if($conf['fnencode'] == 'safe'){
734        return SafeFN::decode($file);
735    }
736
737    return urldecode($file);
738}
739
740/**
741 * Find a page in the current namespace (determined from $ID) or any
742 * higher namespace that can be accessed by the current user,
743 * this condition can be overriden by an optional parameter.
744 *
745 * Used for sidebars, but can be used other stuff as well
746 *
747 * @todo   add event hook
748 *
749 * @param  string $page the pagename you're looking for
750 * @param bool $useacl only return pages readable by the current user, false to ignore ACLs
751 * @return false|string the full page id of the found page, false if any
752 */
753function page_findnearest($page, $useacl = true){
754    if (!$page) return false;
755    global $ID;
756
757    $ns = $ID;
758    do {
759        $ns = getNS($ns);
760        $pageid = cleanID("$ns:$page");
761        if(page_exists($pageid) && (!$useacl || auth_quickaclcheck($pageid) >= AUTH_READ)){
762            return $pageid;
763        }
764    } while($ns);
765
766    return false;
767}
768