xref: /dokuwiki/inc/pageutils.php (revision 0da02e4da30749eeeec74b09658d8aac595a4fb7) 
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 $INPUT;
23    global $conf;
24    global $ACT;
25
26    $id = $INPUT->str($param);
27
28    //construct page id from request URI
29    if(empty($id) && $conf['userewrite'] == 2){
30        $request = $_SERVER['REQUEST_URI'];
31        $script = '';
32
33        //get the script URL
34        if($conf['basedir']){
35            $relpath = '';
36            if($param != 'id') {
37                $relpath = 'lib/exe/';
38            }
39            $script = $conf['basedir'].$relpath.utf8_basename($_SERVER['SCRIPT_FILENAME']);
40
41        }elseif($_SERVER['PATH_INFO']){
42            $request = $_SERVER['PATH_INFO'];
43        }elseif($_SERVER['SCRIPT_NAME']){
44            $script = $_SERVER['SCRIPT_NAME'];
45        }elseif($_SERVER['DOCUMENT_ROOT'] && $_SERVER['SCRIPT_FILENAME']){
46            $script = preg_replace ('/^'.preg_quote($_SERVER['DOCUMENT_ROOT'],'/').'/','',
47                    $_SERVER['SCRIPT_FILENAME']);
48            $script = '/'.$script;
49        }
50
51        //clean script and request (fixes a windows problem)
52        $script  = preg_replace('/\/\/+/','/',$script);
53        $request = preg_replace('/\/\/+/','/',$request);
54
55        //remove script URL and Querystring to gain the id
56        if(preg_match('/^'.preg_quote($script,'/').'(.*)/',$request, $match)){
57            $id = preg_replace ('/\?.*/','',$match[1]);
58        }
59        $id = urldecode($id);
60        //strip leading slashes
61        $id = preg_replace('!^/+!','',$id);
62    }
63
64    // Namespace autolinking from URL
65    if(substr($id,-1) == ':' || ($conf['useslash'] && substr($id,-1) == '/')){
66        if(page_exists($id.$conf['start'])){
67            // start page inside namespace
68            $id = $id.$conf['start'];
69        }elseif(page_exists($id.noNS(cleanID($id)))){
70            // page named like the NS inside the NS
71            $id = $id.noNS(cleanID($id));
72        }elseif(page_exists($id)){
73            // page like namespace exists
74            $id = substr($id,0,-1);
75        }else{
76            // fall back to default
77            $id = $id.$conf['start'];
78        }
79        if (isset($ACT) && $ACT === 'show') send_redirect(wl($id,'',true));
80    }
81
82    if($clean) $id = cleanID($id);
83    if(empty($id) && $param=='id') $id = $conf['start'];
84
85    return $id;
86}
87
88/**
89 * Remove unwanted chars from ID
90 *
91 * Cleans a given ID to only use allowed characters. Accented characters are
92 * converted to unaccented ones
93 *
94 * @author Andreas Gohr <andi@splitbrain.org>
95 * @param  string  $raw_id    The pageid to clean
96 * @param  boolean $ascii     Force ASCII
97 * @param  boolean $media     DEPRECATED
98 */
99function cleanID($raw_id,$ascii=false,$media=false){
100    global $conf;
101    static $sepcharpat = null;
102
103    global $cache_cleanid;
104    $cache = & $cache_cleanid;
105
106    // check if it's already in the memory cache
107    if (isset($cache[(string)$raw_id])) {
108        return $cache[(string)$raw_id];
109    }
110
111    $sepchar = $conf['sepchar'];
112    if($sepcharpat == null) // build string only once to save clock cycles
113        $sepcharpat = '#\\'.$sepchar.'+#';
114
115    $id = trim((string)$raw_id);
116    $id = utf8_strtolower($id);
117
118    //alternative namespace seperator
119    if($conf['useslash']){
120        $id = strtr($id,';/','::');
121    }else{
122        $id = strtr($id,';/',':'.$sepchar);
123    }
124
125    if($conf['deaccent'] == 2 || $ascii) $id = utf8_romanize($id);
126    if($conf['deaccent'] || $ascii) $id = utf8_deaccent($id,-1);
127
128    //remove specials
129    $id = utf8_stripspecials($id,$sepchar,'\*');
130
131    if($ascii) $id = utf8_strip($id);
132
133    //clean up
134    $id = preg_replace($sepcharpat,$sepchar,$id);
135    $id = preg_replace('#:+#',':',$id);
136    $id = trim($id,':._-');
137    $id = preg_replace('#:[:\._\-]+#',':',$id);
138    $id = preg_replace('#[:\._\-]+:#',':',$id);
139
140    $cache[(string)$raw_id] = $id;
141    return($id);
142}
143
144/**
145 * Return namespacepart of a wiki ID
146 *
147 * @author Andreas Gohr <andi@splitbrain.org>
148 */
149function getNS($id){
150    $pos = strrpos((string)$id,':');
151    if($pos!==false){
152        return substr((string)$id,0,$pos);
153    }
154    return false;
155}
156
157/**
158 * Returns the ID without the namespace
159 *
160 * @author Andreas Gohr <andi@splitbrain.org>
161 */
162function noNS($id) {
163    $pos = strrpos($id, ':');
164    if ($pos!==false) {
165        return substr($id, $pos+1);
166    } else {
167        return $id;
168    }
169}
170
171/**
172 * Returns the current namespace
173 *
174 * @author Nathan Fritz <fritzn@crown.edu>
175 */
176function curNS($id) {
177    return noNS(getNS($id));
178}
179
180/**
181 * Returns the ID without the namespace or current namespace for 'start' pages
182 *
183 * @author Nathan Fritz <fritzn@crown.edu>
184 */
185function noNSorNS($id) {
186    global $conf;
187
188    $p = noNS($id);
189    if ($p == $conf['start'] || $p == false) {
190        $p = curNS($id);
191        if ($p == false) {
192            return $conf['start'];
193        }
194    }
195    return $p;
196}
197
198/**
199 * Creates a XHTML valid linkid from a given headline title
200 *
201 * @param string  $title   The headline title
202 * @param array|bool   $check   Existing IDs (title => number)
203 * @return string the title
204 * @author Andreas Gohr <andi@splitbrain.org>
205 */
206function sectionID($title,&$check) {
207    $title = str_replace(array(':','.'),'',cleanID($title));
208    $new = ltrim($title,'0123456789_-');
209    if(empty($new)){
210        $title = 'section'.preg_replace('/[^0-9]+/','',$title); //keep numbers from headline
211    }else{
212        $title = $new;
213    }
214
215    if(is_array($check)){
216        // make sure tiles are unique
217        if (!array_key_exists ($title,$check)) {
218            $check[$title] = 0;
219        } else {
220            $title .= ++ $check[$title];
221        }
222    }
223
224    return $title;
225}
226
227
228/**
229 * Wiki page existence check
230 *
231 * parameters as for wikiFN
232 *
233 * @author Chris Smith <chris@jalakai.co.uk>
234 */
235function page_exists($id,$rev='',$clean=true) {
236    return @file_exists(wikiFN($id,$rev,$clean));
237}
238
239/**
240 * returns the full path to the datafile specified by ID and optional revision
241 *
242 * The filename is URL encoded to protect Unicode chars
243 *
244 * @param  $raw_id  string   id of wikipage
245 * @param  $rev     string   page revision, empty string for current
246 * @param  $clean   bool     flag indicating that $raw_id should be cleaned.  Only set to false
247 *                           when $id is guaranteed to have been cleaned already.
248 *
249 * @author Andreas Gohr <andi@splitbrain.org>
250 */
251function wikiFN($raw_id,$rev='',$clean=true){
252    global $conf;
253
254    global $cache_wikifn;
255    $cache = & $cache_wikifn;
256
257    if (isset($cache[$raw_id]) && isset($cache[$raw_id][$rev])) {
258        return $cache[$raw_id][$rev];
259    }
260
261    $id = $raw_id;
262
263    if ($clean) $id = cleanID($id);
264    $id = str_replace(':','/',$id);
265    if(empty($rev)){
266        $fn = $conf['datadir'].'/'.utf8_encodeFN($id).'.txt';
267    }else{
268        $fn = $conf['olddir'].'/'.utf8_encodeFN($id).'.'.$rev.'.txt';
269        if($conf['compression']){
270            //test for extensions here, we want to read both compressions
271            if (@file_exists($fn . '.gz')){
272                $fn .= '.gz';
273            }else if(@file_exists($fn . '.bz2')){
274                $fn .= '.bz2';
275            }else{
276                //file doesnt exist yet, so we take the configured extension
277                $fn .= '.' . $conf['compression'];
278            }
279        }
280    }
281
282    if (!isset($cache[$raw_id])) { $cache[$raw_id] = array(); }
283    $cache[$raw_id][$rev] = $fn;
284    return $fn;
285}
286
287/**
288 * Returns the full path to the file for locking the page while editing.
289 *
290 * @author Ben Coburn <btcoburn@silicodon.net>
291 */
292function wikiLockFN($id) {
293    global $conf;
294    return $conf['lockdir'].'/'.md5(cleanID($id)).'.lock';
295}
296
297
298/**
299 * returns the full path to the meta file specified by ID and extension
300 *
301 * @author Steven Danz <steven-danz@kc.rr.com>
302 */
303function metaFN($id,$ext){
304    global $conf;
305    $id = cleanID($id);
306    $id = str_replace(':','/',$id);
307    $fn = $conf['metadir'].'/'.utf8_encodeFN($id).$ext;
308    return $fn;
309}
310
311/**
312 * returns the full path to the media's meta file specified by ID and extension
313 *
314 * @author Kate Arzamastseva <pshns@ukr.net>
315 */
316function mediaMetaFN($id,$ext){
317    global $conf;
318    $id = cleanID($id);
319    $id = str_replace(':','/',$id);
320    $fn = $conf['mediametadir'].'/'.utf8_encodeFN($id).$ext;
321    return $fn;
322}
323
324/**
325 * returns an array of full paths to all metafiles of a given ID
326 *
327 * @author Esther Brunner <esther@kaffeehaus.ch>
328 * @author Michael Hamann <michael@content-space.de>
329 */
330function metaFiles($id){
331    $basename = metaFN($id, '');
332    $files    = glob($basename.'.*', GLOB_MARK);
333    // filter files like foo.bar.meta when $id == 'foo'
334    return    $files ? preg_grep('/^'.preg_quote($basename, '/').'\.[^.\/]*$/u', $files) : array();
335}
336
337/**
338 * returns the full path to the mediafile specified by ID
339 *
340 * The filename is URL encoded to protect Unicode chars
341 *
342 * @author Andreas Gohr <andi@splitbrain.org>
343 * @author Kate Arzamastseva <pshns@ukr.net>
344 */
345function mediaFN($id, $rev=''){
346    global $conf;
347    $id = cleanID($id);
348    $id = str_replace(':','/',$id);
349    if(empty($rev)){
350        $fn = $conf['mediadir'].'/'.utf8_encodeFN($id);
351    }else{
352        $ext = mimetype($id);
353        $name = substr($id,0, -1*strlen($ext[0])-1);
354        $fn = $conf['mediaolddir'].'/'.utf8_encodeFN($name .'.'.( (int) $rev ).'.'.$ext[0]);
355    }
356    return $fn;
357}
358
359/**
360 * Returns the full filepath to a localized file if local
361 * version isn't found the english one is returned
362 *
363 * @param  string $id  The id of the local file
364 * @param  string $ext The file extension (usually txt)
365 * @author Andreas Gohr <andi@splitbrain.org>
366 */
367function localeFN($id,$ext='txt'){
368    global $conf;
369    $file = DOKU_CONF.'/lang/'.$conf['lang'].'/'.$id.'.'.$ext;
370    if(!@file_exists($file)){
371        $file = DOKU_INC.'inc/lang/'.$conf['lang'].'/'.$id.'.'.$ext;
372        if(!@file_exists($file)){
373            //fall back to english
374            $file = DOKU_INC.'inc/lang/en/'.$id.'.'.$ext;
375        }
376    }
377    return $file;
378}
379
380/**
381 * Resolve relative paths in IDs
382 *
383 * Do not call directly use resolve_mediaid or resolve_pageid
384 * instead
385 *
386 * Partyly based on a cleanPath function found at
387 * http://www.php.net/manual/en/function.realpath.php#57016
388 *
389 * @author <bart at mediawave dot nl>
390 */
391function resolve_id($ns,$id,$clean=true){
392    global $conf;
393
394    // some pre cleaning for useslash:
395    if($conf['useslash']) $id = str_replace('/',':',$id);
396
397    // if the id starts with a dot we need to handle the
398    // relative stuff
399    if($id{0} == '.'){
400        // normalize initial dots without a colon
401        $id = preg_replace('/^(\.+)(?=[^:\.])/','\1:',$id);
402        // prepend the current namespace
403        $id = $ns.':'.$id;
404
405        // cleanup relatives
406        $result = array();
407        $pathA  = explode(':', $id);
408        if (!$pathA[0]) $result[] = '';
409        foreach ($pathA AS $key => $dir) {
410            if ($dir == '..') {
411                if (end($result) == '..') {
412                    $result[] = '..';
413                } elseif (!array_pop($result)) {
414                    $result[] = '..';
415                }
416            } elseif ($dir && $dir != '.') {
417                $result[] = $dir;
418            }
419        }
420        if (!end($pathA)) $result[] = '';
421        $id = implode(':', $result);
422    }elseif($ns !== false && strpos($id,':') === false){
423        //if link contains no namespace. add current namespace (if any)
424        $id = $ns.':'.$id;
425    }
426
427    if($clean) $id = cleanID($id);
428    return $id;
429}
430
431/**
432 * Returns a full media id
433 *
434 * @author Andreas Gohr <andi@splitbrain.org>
435 */
436function resolve_mediaid($ns,&$page,&$exists){
437    $page   = resolve_id($ns,$page);
438    $file   = mediaFN($page);
439    $exists = @file_exists($file);
440}
441
442/**
443 * Returns a full page id
444 *
445 * @author Andreas Gohr <andi@splitbrain.org>
446 */
447function resolve_pageid($ns,&$page,&$exists){
448    global $conf;
449    global $ID;
450    $exists = false;
451
452    //empty address should point to current page
453    if ($page === "") {
454        $page = $ID;
455    }
456
457    //keep hashlink if exists then clean both parts
458    if (strpos($page,'#')) {
459        list($page,$hash) = explode('#',$page,2);
460    } else {
461        $hash = '';
462    }
463    $hash = cleanID($hash);
464    $page = resolve_id($ns,$page,false); // resolve but don't clean, yet
465
466    // get filename (calls clean itself)
467    $file = wikiFN($page);
468
469    // if ends with colon or slash we have a namespace link
470    if(in_array(substr($page,-1), array(':', ';')) ||
471       ($conf['useslash'] && substr($page,-1) == '/')){
472        if(page_exists($page.$conf['start'])){
473            // start page inside namespace
474            $page = $page.$conf['start'];
475            $exists = true;
476        }elseif(page_exists($page.noNS(cleanID($page)))){
477            // page named like the NS inside the NS
478            $page = $page.noNS(cleanID($page));
479            $exists = true;
480        }elseif(page_exists($page)){
481            // page like namespace exists
482            $page = $page;
483            $exists = true;
484        }else{
485            // fall back to default
486            $page = $page.$conf['start'];
487        }
488    }else{
489        //check alternative plural/nonplural form
490        if(!@file_exists($file)){
491            if( $conf['autoplural'] ){
492                if(substr($page,-1) == 's'){
493                    $try = substr($page,0,-1);
494                }else{
495                    $try = $page.'s';
496                }
497                if(page_exists($try)){
498                    $page   = $try;
499                    $exists = true;
500                }
501            }
502        }else{
503            $exists = true;
504        }
505    }
506
507    // now make sure we have a clean page
508    $page = cleanID($page);
509
510    //add hash if any
511    if(!empty($hash)) $page .= '#'.$hash;
512}
513
514/**
515 * Returns the name of a cachefile from given data
516 *
517 * The needed directory is created by this function!
518 *
519 * @author Andreas Gohr <andi@splitbrain.org>
520 *
521 * @param string $data  This data is used to create a unique md5 name
522 * @param string $ext   This is appended to the filename if given
523 * @return string       The filename of the cachefile
524 */
525function getCacheName($data,$ext=''){
526    global $conf;
527    $md5  = md5($data);
528    $file = $conf['cachedir'].'/'.$md5{0}.'/'.$md5.$ext;
529    io_makeFileDir($file);
530    return $file;
531}
532
533/**
534 * Checks a pageid against $conf['hidepages']
535 *
536 * @author Andreas Gohr <gohr@cosmocode.de>
537 */
538function isHiddenPage($id){
539    $data = array(
540        'id' => $id,
541        'hidden' => false
542    );
543    trigger_event('PAGEUTILS_ID_HIDEPAGE', $data, '_isHiddenPage');
544    return $data['hidden'];
545}
546
547function _isHiddenPage(&$data) {
548    global $conf;
549    global $ACT;
550
551    if ($data['hidden']) return;
552    if(empty($conf['hidepages'])) return;
553    if($ACT == 'admin') return;
554
555    if(preg_match('/'.$conf['hidepages'].'/ui',':'.$data['id'])){
556        $data['hidden'] = true;
557    }
558}
559
560/**
561 * Reverse of isHiddenPage
562 *
563 * @author Andreas Gohr <gohr@cosmocode.de>
564 */
565function isVisiblePage($id){
566    return !isHiddenPage($id);
567}
568
569/**
570 * Format an id for output to a user
571 *
572 * Namespaces are denoted by a trailing “:*”. The root namespace is
573 * “*”. Output is escaped.
574 *
575 * @author Adrian Lang <lang@cosmocode.de>
576 */
577
578function prettyprint_id($id) {
579    if (!$id || $id === ':') {
580        return '*';
581    }
582    if ((substr($id, -1, 1) === ':')) {
583        $id .= '*';
584    }
585    return hsc($id);
586}
587
588/**
589 * Encode a UTF-8 filename to use on any filesystem
590 *
591 * Uses the 'fnencode' option to determine encoding
592 *
593 * When the second parameter is true the string will
594 * be encoded only if non ASCII characters are detected -
595 * This makes it safe to run it multiple times on the
596 * same string (default is true)
597 *
598 * @author Andreas Gohr <andi@splitbrain.org>
599 * @see    urlencode
600 */
601function utf8_encodeFN($file,$safe=true){
602    global $conf;
603    if($conf['fnencode'] == 'utf-8') return $file;
604
605    if($safe && preg_match('#^[a-zA-Z0-9/_\-\.%]+$#',$file)){
606        return $file;
607    }
608
609    if($conf['fnencode'] == 'safe'){
610        return SafeFN::encode($file);
611    }
612
613    $file = urlencode($file);
614    $file = str_replace('%2F','/',$file);
615    return $file;
616}
617
618/**
619 * Decode a filename back to UTF-8
620 *
621 * Uses the 'fnencode' option to determine encoding
622 *
623 * @author Andreas Gohr <andi@splitbrain.org>
624 * @see    urldecode
625 */
626function utf8_decodeFN($file){
627    global $conf;
628    if($conf['fnencode'] == 'utf-8') return $file;
629
630    if($conf['fnencode'] == 'safe'){
631        return SafeFN::decode($file);
632    }
633
634    return urldecode($file);
635}
636
637/**
638 * Find a page in the current namespace (determined from $ID) or any
639 * higher namespace
640 *
641 * Used for sidebars, but can be used other stuff as well
642 *
643 * @todo   add event hook
644 * @param  string $page the pagename you're looking for
645 * @return string|false the full page id of the found page, false if any
646 */
647function page_findnearest($page){
648    if (!$page) return false;
649    global $ID;
650
651    $ns = $ID;
652    do {
653        $ns = getNS($ns);
654        $pageid = ltrim("$ns:$page",':');
655        if(page_exists($pageid)){
656            return $pageid;
657        }
658    } while($ns);
659
660    return false;
661}
662