xref: /dokuwiki/inc/pageutils.php (revision 8d39e80d363eda2ef31dac8473bfdab4b9cd7ce5) 
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   $clean  if true, ID is cleaned
23 * @return mixed|string
24 */
25function getID($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 * @param  string  $raw_id    The pageid to clean
101 * @param  boolean $ascii     Force ASCII
102 * @return string cleaned id
103 */
104function cleanID($raw_id,$ascii=false){
105    global $conf;
106    static $sepcharpat = null;
107
108    global $cache_cleanid;
109    $cache = & $cache_cleanid;
110
111    // check if it's already in the memory cache
112    if (isset($cache[(string)$raw_id])) {
113        return $cache[(string)$raw_id];
114    }
115
116    $sepchar = $conf['sepchar'];
117    if($sepcharpat == null) // build string only once to save clock cycles
118        $sepcharpat = '#\\'.$sepchar.'+#';
119
120    $id = trim((string)$raw_id);
121    $id = utf8_strtolower($id);
122
123    //alternative namespace seperator
124    if($conf['useslash']){
125        $id = strtr($id,';/','::');
126    }else{
127        $id = strtr($id,';/',':'.$sepchar);
128    }
129
130    if($conf['deaccent'] == 2 || $ascii) $id = utf8_romanize($id);
131    if($conf['deaccent'] || $ascii) $id = utf8_deaccent($id,-1);
132
133    //remove specials
134    $id = utf8_stripspecials($id,$sepchar,'\*');
135
136    if($ascii) $id = utf8_strip($id);
137
138    //clean up
139    $id = preg_replace($sepcharpat,$sepchar,$id);
140    $id = preg_replace('#:+#',':',$id);
141    $id = trim($id,':._-');
142    $id = preg_replace('#:[:\._\-]+#',':',$id);
143    $id = preg_replace('#[:\._\-]+:#',':',$id);
144
145    $cache[(string)$raw_id] = $id;
146    return($id);
147}
148
149/**
150 * Return namespacepart of a wiki ID
151 *
152 * @author Andreas Gohr <andi@splitbrain.org>
153 *
154 * @param string $id
155 * @return string|bool the namespace part or false if the given ID has no namespace (root)
156 */
157function getNS($id){
158    $pos = strrpos((string)$id,':');
159    if($pos!==false){
160        return substr((string)$id,0,$pos);
161    }
162    return false;
163}
164
165/**
166 * Returns the ID without the namespace
167 *
168 * @author Andreas Gohr <andi@splitbrain.org>
169 *
170 * @param string $id
171 * @return string
172 */
173function noNS($id) {
174    $pos = strrpos($id, ':');
175    if ($pos!==false) {
176        return substr($id, $pos+1);
177    } else {
178        return $id;
179    }
180}
181
182/**
183 * Returns the current namespace
184 *
185 * @author Nathan Fritz <fritzn@crown.edu>
186 *
187 * @param string $id
188 * @return string
189 */
190function curNS($id) {
191    return noNS(getNS($id));
192}
193
194/**
195 * Returns the ID without the namespace or current namespace for 'start' pages
196 *
197 * @author Nathan Fritz <fritzn@crown.edu>
198 *
199 * @param string $id
200 * @return string
201 */
202function noNSorNS($id) {
203    global $conf;
204
205    $p = noNS($id);
206    if ($p == $conf['start'] || $p == false) {
207        $p = curNS($id);
208        if ($p == false) {
209            return $conf['start'];
210        }
211    }
212    return $p;
213}
214
215/**
216 * Creates a XHTML valid linkid from a given headline title
217 *
218 * @param string  $title   The headline title
219 * @param array|bool   $check   Existing IDs (title => number)
220 * @return string the title
221 *
222 * @author Andreas Gohr <andi@splitbrain.org>
223 */
224function sectionID($title,&$check) {
225    $title = str_replace(array(':','.'),'',cleanID($title));
226    $new = ltrim($title,'0123456789_-');
227    if(empty($new)){
228        $title = 'section'.preg_replace('/[^0-9]+/','',$title); //keep numbers from headline
229    }else{
230        $title = $new;
231    }
232
233    if(is_array($check)){
234        // make sure tiles are unique
235        if (!array_key_exists ($title,$check)) {
236            $check[$title] = 0;
237        } else {
238            $title .= ++ $check[$title];
239        }
240    }
241
242    return $title;
243}
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 $id should be cleaned (see wikiFN as well)
256 * @return bool exists?
257 */
258function page_exists($id,$rev='',$clean=true, $data_at=false) {
259    if($rev !== '' && $date_at) {
260        $pagelog = new PageChangeLog($page);
261        $pagelog_rev = $pagelog->getLastRevisionAt($rev);
262        if($pagelog_rev !== false)
263            $rev = $pagelog_rev;
264    }
265    return @file_exists(wikiFN($id,$rev,$clean));
266}
267
268/**
269 * returns the full path to the datafile specified by ID and optional revision
270 *
271 * The filename is URL encoded to protect Unicode chars
272 *
273 * @param  $raw_id  string   id of wikipage
274 * @param  $rev     string   page revision, empty string for current
275 * @param  $clean   bool     flag indicating that $raw_id should be cleaned.  Only set to false
276 *                           when $id is guaranteed to have been cleaned already.
277 * @return string full path
278 *
279 * @author Andreas Gohr <andi@splitbrain.org>
280 */
281function wikiFN($raw_id,$rev='',$clean=true){
282    global $conf;
283
284    global $cache_wikifn;
285    $cache = & $cache_wikifn;
286
287    if (isset($cache[$raw_id]) && isset($cache[$raw_id][$rev])) {
288        return $cache[$raw_id][$rev];
289    }
290
291    $id = $raw_id;
292
293    if ($clean) $id = cleanID($id);
294    $id = str_replace(':','/',$id);
295    if(empty($rev)){
296        $fn = $conf['datadir'].'/'.utf8_encodeFN($id).'.txt';
297    }else{
298        $fn = $conf['olddir'].'/'.utf8_encodeFN($id).'.'.$rev.'.txt';
299        if($conf['compression']){
300            //test for extensions here, we want to read both compressions
301            if (@file_exists($fn . '.gz')){
302                $fn .= '.gz';
303            }else if(@file_exists($fn . '.bz2')){
304                $fn .= '.bz2';
305            }else{
306                //file doesnt exist yet, so we take the configured extension
307                $fn .= '.' . $conf['compression'];
308            }
309        }
310    }
311
312    if (!isset($cache[$raw_id])) { $cache[$raw_id] = array(); }
313    $cache[$raw_id][$rev] = $fn;
314    return $fn;
315}
316
317/**
318 * Returns the full path to the file for locking the page while editing.
319 *
320 * @author Ben Coburn <btcoburn@silicodon.net>
321 *
322 * @param string $id page id
323 * @return string full path
324 */
325function wikiLockFN($id) {
326    global $conf;
327    return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
328}
329
330
331/**
332 * returns the full path to the meta file specified by ID and extension
333 *
334 * @author Steven Danz <steven-danz@kc.rr.com>
335 *
336 * @param string $id   page id
337 * @param string $ext  file extension
338 * @return string full path
339 */
340function metaFN($id,$ext){
341    global $conf;
342    $id = cleanID($id);
343    $id = str_replace(':','/',$id);
344    $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
345    return $fn;
346}
347
348/**
349 * returns the full path to the media's meta file specified by ID and extension
350 *
351 * @author Kate Arzamastseva <pshns@ukr.net>
352 *
353 * @param string $id   media id
354 * @param string $ext  extension of media
355 * @return string
356 */
357function mediaMetaFN($id,$ext){
358    global $conf;
359    $id = cleanID($id);
360    $id = str_replace(':','/',$id);
361    $fn = $conf['mediametadir'].'/'.utf8_encodeFN($id).$ext;
362    return $fn;
363}
364
365/**
366 * returns an array of full paths to all metafiles of a given ID
367 *
368 * @author Esther Brunner <esther@kaffeehaus.ch>
369 * @author Michael Hamann <michael@content-space.de>
370 *
371 * @param string $id page id
372 * @return array
373 */
374function metaFiles($id){
375    $basename = metaFN($id, '');
376    $files    = glob($basename.'.*', GLOB_MARK);
377    // filter files like foo.bar.meta when $id == 'foo'
378    return    $files ? preg_grep('/^'.preg_quote($basename, '/').'\.[^.\/]*$/u', $files) : array();
379}
380
381/**
382 * returns the full path to the mediafile specified by ID
383 *
384 * The filename is URL encoded to protect Unicode chars
385 *
386 * @author Andreas Gohr <andi@splitbrain.org>
387 * @author Kate Arzamastseva <pshns@ukr.net>
388 *
389 * @param string     $id  media id
390 * @param string|int $rev empty string or revision timestamp
391 * @return string full path
392 */
393function mediaFN($id, $rev=''){
394    global $conf;
395    $id = cleanID($id);
396    $id = str_replace(':','/',$id);
397    if(empty($rev)){
398        $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
399    }else{
400        $ext = mimetype($id);
401        $name = substr($id,0, -1*strlen($ext[0])-1);
402        $fn = $conf['mediaolddir'].'/'.utf8_encodeFN($name .'.'.( (int) $rev ).'.'.$ext[0]);
403    }
404    return $fn;
405}
406
407/**
408 * Returns the full filepath to a localized file if local
409 * version isn't found the english one is returned
410 *
411 * @param  string $id  The id of the local file
412 * @param  string $ext The file extension (usually txt)
413 * @return string full filepath to localized file
414 *
415 * @author Andreas Gohr <andi@splitbrain.org>
416 */
417function localeFN($id,$ext='txt'){
418    global $conf;
419    $file = DOKU_CONF.'lang/'.$conf['lang'].'/'.$id.'.'.$ext;
420    if(!@file_exists($file)){
421        $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.'.$ext;
422        if(!@file_exists($file)){
423            //fall back to english
424            $file = DOKU_INC.'inc/lang/en/'.$id.'.'.$ext;
425        }
426    }
427    return $file;
428}
429
430/**
431 * Resolve relative paths in IDs
432 *
433 * Do not call directly use resolve_mediaid or resolve_pageid
434 * instead
435 *
436 * Partyly based on a cleanPath function found at
437 * http://www.php.net/manual/en/function.realpath.php#57016
438 *
439 * @author <bart at mediawave dot nl>
440 *
441 * @param string $ns     namespace which is context of id
442 * @param string $id     relative id
443 * @param bool   $clean  flag indicating that id should be cleaned
444 * @return mixed|string
445 */
446function resolve_id($ns,$id,$clean=true){
447    global $conf;
448
449    // some pre cleaning for useslash:
450    if($conf['useslash']) $id = str_replace('/',':',$id);
451
452    // if the id starts with a dot we need to handle the
453    // relative stuff
454    if($id && $id{0} == '.'){
455        // normalize initial dots without a colon
456        $id = preg_replace('/^(\.+)(?=[^:\.])/','\1:',$id);
457        // prepend the current namespace
458        $id = $ns.':'.$id;
459
460        // cleanup relatives
461        $result = array();
462        $pathA  = explode(':', $id);
463        if (!$pathA[0]) $result[] = '';
464        foreach ($pathA AS $key => $dir) {
465            if ($dir == '..') {
466                if (end($result) == '..') {
467                    $result[] = '..';
468                } elseif (!array_pop($result)) {
469                    $result[] = '..';
470                }
471            } elseif ($dir && $dir != '.') {
472                $result[] = $dir;
473            }
474        }
475        if (!end($pathA)) $result[] = '';
476        $id = implode(':', $result);
477    }elseif($ns !== false && strpos($id,':') === false){
478        //if link contains no namespace. add current namespace (if any)
479        $id = $ns.':'.$id;
480    }
481
482    if($clean) $id = cleanID($id);
483    return $id;
484}
485
486/**
487 * Returns a full media id
488 *
489 * @author Andreas Gohr <andi@splitbrain.org>
490 *
491 * @param string  $ns     namespace which is context of id
492 * @param string &$page   (reference) relative media id, updated to resolved id
493 * @param bool   &$exists (reference) updated with existance of media
494 */
495function resolve_mediaid($ns,&$page,&$exists,$rev='',$date_at=false){
496    if($rev !== '' &&  $date_at){
497        $medialog = new MediaChangeLog($media_id);
498        $medialog_rev = $medialog->getLastRevisionAt($rev);
499        if($medialog_rev !== false) {
500            $rev = $medialog_rev;
501        }
502    }
503    $page   = resolve_id($ns,$page);
504    $file   = mediaFN($page,$rev);
505    $exists = @file_exists($file);
506}
507
508/**
509 * Returns a full page id
510 *
511 * @author Andreas Gohr <andi@splitbrain.org>
512 *
513 * @param string  $ns     namespace which is context of id
514 * @param string &$page   (reference) relative page id, updated to resolved id
515 * @param bool   &$exists (reference) updated with existance of media
516 */
517function resolve_pageid($ns,&$page,&$exists,$rev='',$date_at=false ){
518    global $conf;
519    global $ID;
520    $exists = false;
521
522    //empty address should point to current page
523    if ($page === "") {
524        $page = $ID;
525    }
526
527    //keep hashlink if exists then clean both parts
528    if (strpos($page,'#')) {
529        list($page,$hash) = explode('#',$page,2);
530    } else {
531        $hash = '';
532    }
533    $hash = cleanID($hash);
534    $page = resolve_id($ns,$page,false); // resolve but don't clean, yet
535
536    // get filename (calls clean itself)
537    if($rev !== '' && $date_at) {
538        $pagelog = new PageChangeLog($page);
539        $pagelog_rev = $pagelog->getLastRevisionAt($rev);
540        if($pagelog_rev !== false)//something found
541           $rev  = $pagelog_rev;
542    }
543    $file = wikiFN($page,$rev);
544
545    // if ends with colon or slash we have a namespace link
546    if(in_array(substr($page,-1), array(':', ';')) ||
547       ($conf['useslash'] && substr($page,-1) == '/')){
548        if(page_exists($page.$conf['start'],$rev,true,$date_at)){
549            // start page inside namespace
550            $page = $page.$conf['start'];
551            $exists = true;
552        }elseif(page_exists($page.noNS(cleanID($page)),$rev,true,$date_at)){
553            // page named like the NS inside the NS
554            $page = $page.noNS(cleanID($page));
555            $exists = true;
556        }elseif(page_exists($page,$rev,true,$date_at)){
557            // page like namespace exists
558            $page = $page;
559            $exists = true;
560        }else{
561            // fall back to default
562            $page = $page.$conf['start'];
563        }
564    }else{
565        //check alternative plural/nonplural form
566        if(!@file_exists($file)){
567            if( $conf['autoplural'] ){
568                if(substr($page,-1) == 's'){
569                    $try = substr($page,0,-1);
570                }else{
571                    $try = $page.'s';
572                }
573                if(page_exists($try,$rev,true,$date_at)){
574                    $page   = $try;
575                    $exists = true;
576                }
577            }
578        }else{
579            $exists = true;
580        }
581    }
582
583    // now make sure we have a clean page
584    $page = cleanID($page);
585
586    //add hash if any
587    if(!empty($hash)) $page .= '#'.$hash;
588}
589
590/**
591 * Returns the name of a cachefile from given data
592 *
593 * The needed directory is created by this function!
594 *
595 * @author Andreas Gohr <andi@splitbrain.org>
596 *
597 * @param string $data  This data is used to create a unique md5 name
598 * @param string $ext   This is appended to the filename if given
599 * @return string       The filename of the cachefile
600 */
601function getCacheName($data,$ext=''){
602    global $conf;
603    $md5  = md5($data);
604    $file = $conf['cachedir'].'/'.$md5{0}.'/'.$md5.$ext;
605    io_makeFileDir($file);
606    return $file;
607}
608
609/**
610 * Checks a pageid against $conf['hidepages']
611 *
612 * @author Andreas Gohr <gohr@cosmocode.de>
613 *
614 * @param string $id page id
615 * @return bool
616 */
617function isHiddenPage($id){
618    $data = array(
619        'id' => $id,
620        'hidden' => false
621    );
622    trigger_event('PAGEUTILS_ID_HIDEPAGE', $data, '_isHiddenPage');
623    return $data['hidden'];
624}
625
626/**
627 * callback checks if page is hidden
628 *
629 * @param array $data event data    - see isHiddenPage()
630 */
631function _isHiddenPage(&$data) {
632    global $conf;
633    global $ACT;
634
635    if ($data['hidden']) return;
636    if(empty($conf['hidepages'])) return;
637    if($ACT == 'admin') return;
638
639    if(preg_match('/'.$conf['hidepages'].'/ui',':'.$data['id'])){
640        $data['hidden'] = true;
641    }
642}
643
644/**
645 * Reverse of isHiddenPage
646 *
647 * @author Andreas Gohr <gohr@cosmocode.de>
648 *
649 * @param string $id page id
650 * @return bool
651 */
652function isVisiblePage($id){
653    return !isHiddenPage($id);
654}
655
656/**
657 * Format an id for output to a user
658 *
659 * Namespaces are denoted by a trailing “:*”. The root namespace is
660 * “*”. Output is escaped.
661 *
662 * @author Adrian Lang <lang@cosmocode.de>
663 *
664 * @param string $id page id
665 * @return string
666 */
667function prettyprint_id($id) {
668    if (!$id || $id === ':') {
669        return '*';
670    }
671    if ((substr($id, -1, 1) === ':')) {
672        $id .= '*';
673    }
674    return hsc($id);
675}
676
677/**
678 * Encode a UTF-8 filename to use on any filesystem
679 *
680 * Uses the 'fnencode' option to determine encoding
681 *
682 * When the second parameter is true the string will
683 * be encoded only if non ASCII characters are detected -
684 * This makes it safe to run it multiple times on the
685 * same string (default is true)
686 *
687 * @author Andreas Gohr <andi@splitbrain.org>
688 * @see    urlencode
689 *
690 * @param string $file file name
691 * @param bool   $safe if true, only encoded when non ASCII characters detected
692 * @return string
693 */
694function utf8_encodeFN($file,$safe=true){
695    global $conf;
696    if($conf['fnencode'] == 'utf-8') return $file;
697
698    if($safe && preg_match('#^[a-zA-Z0-9/_\-\.%]+$#',$file)){
699        return $file;
700    }
701
702    if($conf['fnencode'] == 'safe'){
703        return SafeFN::encode($file);
704    }
705
706    $file = urlencode($file);
707    $file = str_replace('%2F','/',$file);
708    return $file;
709}
710
711/**
712 * Decode a filename back to UTF-8
713 *
714 * Uses the 'fnencode' option to determine encoding
715 *
716 * @author Andreas Gohr <andi@splitbrain.org>
717 * @see    urldecode
718 *
719 * @param string $file file name
720 * @return string
721 */
722function utf8_decodeFN($file){
723    global $conf;
724    if($conf['fnencode'] == 'utf-8') return $file;
725
726    if($conf['fnencode'] == 'safe'){
727        return SafeFN::decode($file);
728    }
729
730    return urldecode($file);
731}
732
733/**
734 * Find a page in the current namespace (determined from $ID) or any
735 * higher namespace
736 *
737 * Used for sidebars, but can be used other stuff as well
738 *
739 * @todo   add event hook
740 * @param  string $page the pagename you're looking for
741 * @return string|false the full page id of the found page, false if any
742 */
743function page_findnearest($page){
744    if (!$page) return false;
745    global $ID;
746
747    $ns = $ID;
748    do {
749        $ns = getNS($ns);
750        $pageid = ltrim("$ns:$page",':');
751        if(page_exists($pageid)){
752            return $pageid;
753        }
754    } while($ns);
755
756    return false;
757}
758