xref: /dokuwiki/inc/common.php (revision ba8f83490b45737d4cf33484c72b3aaa21e8894d) 
1<?php
2/**
3 * Common DokuWiki functions
4 *
5 * @license    GPL 2 (http://www.gnu.org/licenses/gpl.html)
6 * @author     Andreas Gohr <andi@splitbrain.org>
7 */
8use dokuwiki\PassHash;
9use dokuwiki\Draft;
10use dokuwiki\Utf8\Clean;
11use dokuwiki\Utf8\PhpString;
12use dokuwiki\Utf8\Conversion;
13use dokuwiki\Cache\CacheInstructions;
14use dokuwiki\Cache\CacheRenderer;
15use dokuwiki\ChangeLog\PageChangeLog;
16use dokuwiki\File\PageFile;
17use dokuwiki\Logger;
18use dokuwiki\Subscriptions\PageSubscriptionSender;
19use dokuwiki\Subscriptions\SubscriberManager;
20use dokuwiki\Extension\AuthPlugin;
21use dokuwiki\Extension\Event;
22
23/**
24 * Wrapper around htmlspecialchars()
25 *
26 * @author Andreas Gohr <andi@splitbrain.org>
27 * @see    htmlspecialchars()
28 *
29 * @param string $string the string being converted
30 * @return string converted string
31 */
32function hsc($string)
33{
34    return htmlspecialchars($string, ENT_QUOTES | ENT_SUBSTITUTE | ENT_HTML401, 'UTF-8');
35}
36
37/**
38 * A safer explode for fixed length lists
39 *
40 * This works just like explode(), but will always return the wanted number of elements.
41 * If the $input string does not contain enough elements, the missing elements will be
42 * filled up with the $default value. If the input string contains more elements, the last
43 * one will NOT be split up and will still contain $separator
44 *
45 * @param string $separator The boundary string
46 * @param string $string The input string
47 * @param int $limit The number of expected elements
48 * @param mixed $default The value to use when filling up missing elements
49 * @see explode
50 * @return array
51 */
52function sexplode($separator, $string, $limit, $default = null)
53{
54    return array_pad(explode($separator, $string, $limit), $limit, $default);
55}
56
57/**
58 * Checks if the given input is blank
59 *
60 * This is similar to empty() but will return false for "0".
61 *
62 * Please note: when you pass uninitialized variables, they will implicitly be created
63 * with a NULL value without warning.
64 *
65 * To avoid this it's recommended to guard the call with isset like this:
66 *
67 * (isset($foo) && !blank($foo))
68 * (!isset($foo) || blank($foo))
69 *
70 * @param $in
71 * @param bool $trim Consider a string of whitespace to be blank
72 * @return bool
73 */
74function blank(&$in, $trim = false)
75{
76    if(is_null($in)) return true;
77    if(is_array($in)) return $in === [];
78    if($in === "\0") return true;
79    if($trim && trim($in) === '') return true;
80    if(strlen($in) > 0) return false;
81    return empty($in);
82}
83
84/**
85 * print a newline terminated string
86 *
87 * You can give an indention as optional parameter
88 *
89 * @author Andreas Gohr <andi@splitbrain.org>
90 *
91 * @param string $string  line of text
92 * @param int    $indent  number of spaces indention
93 */
94function ptln($string, $indent = 0)
95{
96    echo str_repeat(' ', $indent)."$string\n";
97}
98
99/**
100 * strips control characters (<32) from the given string
101 *
102 * @author Andreas Gohr <andi@splitbrain.org>
103 *
104 * @param string $string being stripped
105 * @return string
106 */
107function stripctl($string)
108{
109    return preg_replace('/[\x00-\x1F]+/s', '', $string);
110}
111
112/**
113 * Return a secret token to be used for CSRF attack prevention
114 *
115 * @author  Andreas Gohr <andi@splitbrain.org>
116 * @link    http://en.wikipedia.org/wiki/Cross-site_request_forgery
117 * @link    http://christ1an.blogspot.com/2007/04/preventing-csrf-efficiently.html
118 *
119 * @return  string
120 */
121function getSecurityToken()
122{
123    /** @var Input $INPUT */
124    global $INPUT;
125
126    $user = $INPUT->server->str('REMOTE_USER');
127    $session = session_id();
128
129    // CSRF checks are only for logged in users - do not generate for anonymous
130    if(trim($user) == '' || trim($session) == '') return '';
131    return PassHash::hmac('md5', $session.$user, auth_cookiesalt());
132}
133
134/**
135 * Check the secret CSRF token
136 *
137 * @param null|string $token security token or null to read it from request variable
138 * @return bool success if the token matched
139 */
140function checkSecurityToken($token = null)
141{
142    /** @var Input $INPUT */
143    global $INPUT;
144    if(!$INPUT->server->str('REMOTE_USER')) return true; // no logged in user, no need for a check
145
146    if(is_null($token)) $token = $INPUT->str('sectok');
147    if(getSecurityToken() != $token) {
148        msg('Security Token did not match. Possible CSRF attack.', -1);
149        return false;
150    }
151    return true;
152}
153
154/**
155 * Print a hidden form field with a secret CSRF token
156 *
157 * @author  Andreas Gohr <andi@splitbrain.org>
158 *
159 * @param bool $print  if true print the field, otherwise html of the field is returned
160 * @return string html of hidden form field
161 */
162function formSecurityToken($print = true)
163{
164    $ret = '<div class="no"><input type="hidden" name="sectok" value="'.getSecurityToken().'" /></div>'."\n";
165    if($print) echo $ret;
166    return $ret;
167}
168
169/**
170 * Determine basic information for a request of $id
171 *
172 * @author Andreas Gohr <andi@splitbrain.org>
173 * @author Chris Smith <chris@jalakai.co.uk>
174 *
175 * @param string $id         pageid
176 * @param bool   $htmlClient add info about whether is mobile browser
177 * @return array with info for a request of $id
178 *
179 */
180function basicinfo($id, $htmlClient = true)
181{
182    global $USERINFO;
183    /* @var Input $INPUT */
184    global $INPUT;
185
186    // set info about manager/admin status.
187    $info = [];
188    $info['isadmin']   = false;
189    $info['ismanager'] = false;
190    if($INPUT->server->has('REMOTE_USER')) {
191        $info['userinfo']   = $USERINFO;
192        $info['perm']       = auth_quickaclcheck($id);
193        $info['client']     = $INPUT->server->str('REMOTE_USER');
194
195        if($info['perm'] == AUTH_ADMIN) {
196            $info['isadmin']   = true;
197            $info['ismanager'] = true;
198        } elseif(auth_ismanager()) {
199            $info['ismanager'] = true;
200        }
201
202        // if some outside auth were used only REMOTE_USER is set
203        if(empty($info['userinfo']['name'])) {
204            $info['userinfo']['name'] = $INPUT->server->str('REMOTE_USER');
205        }
206
207    } else {
208        $info['perm']       = auth_aclcheck($id, '', null);
209        $info['client']     = clientIP(true);
210    }
211
212    $info['namespace'] = getNS($id);
213
214    // mobile detection
215    if ($htmlClient) {
216        $info['ismobile'] = clientismobile();
217    }
218
219    return $info;
220 }
221
222/**
223 * Return info about the current document as associative
224 * array.
225 *
226 * @author Andreas Gohr <andi@splitbrain.org>
227 *
228 * @return array with info about current document
229 */
230function pageinfo()
231{
232    global $ID;
233    global $REV;
234    global $RANGE;
235    global $lang;
236    /* @var Input $INPUT */
237    global $INPUT;
238
239    $info = basicinfo($ID);
240
241    // include ID & REV not redundant, as some parts of DokuWiki may temporarily change $ID, e.g. p_wiki_xhtml
242    // FIXME ... perhaps it would be better to ensure the temporary changes weren't necessary
243    $info['id']  = $ID;
244    $info['rev'] = $REV;
245
246    $subManager = new SubscriberManager();
247    $info['subscribed'] = $subManager->userSubscription();
248
249    $info['locked']     = checklock($ID);
250    $info['filepath']   = wikiFN($ID);
251    $info['exists']     = file_exists($info['filepath']);
252    $info['currentrev'] = @filemtime($info['filepath']);
253
254    if ($REV) {
255        //check if current revision was meant
256        if ($info['exists'] && ($info['currentrev'] == $REV)) {
257            $REV = '';
258        } elseif ($RANGE) {
259            //section editing does not work with old revisions!
260            $REV   = '';
261            $RANGE = '';
262            msg($lang['nosecedit'], 0);
263        } else {
264            //really use old revision
265            $info['filepath'] = wikiFN($ID, $REV);
266            $info['exists']   = file_exists($info['filepath']);
267        }
268    }
269    $info['rev'] = $REV;
270    if ($info['exists']) {
271        $info['writable'] = (is_writable($info['filepath']) && $info['perm'] >= AUTH_EDIT);
272    } else {
273        $info['writable'] = ($info['perm'] >= AUTH_CREATE);
274    }
275    $info['editable'] = ($info['writable'] && empty($info['locked']));
276    $info['lastmod']  = @filemtime($info['filepath']);
277
278    //load page meta data
279    $info['meta'] = p_get_metadata($ID);
280
281    //who's the editor
282    $pagelog = new PageChangeLog($ID, 1024);
283    if ($REV) {
284        $revinfo = $pagelog->getRevisionInfo($REV);
285    } elseif (!empty($info['meta']['last_change']) && is_array($info['meta']['last_change'])) {
286        $revinfo = $info['meta']['last_change'];
287    } else {
288        $revinfo = $pagelog->getRevisionInfo($info['lastmod']);
289        // cache most recent changelog line in metadata if missing and still valid
290        if ($revinfo !== false) {
291            $info['meta']['last_change'] = $revinfo;
292            p_set_metadata($ID, ['last_change' => $revinfo]);
293        }
294    }
295    //and check for an external edit
296    if ($revinfo !== false && $revinfo['date'] != $info['lastmod']) {
297        // cached changelog line no longer valid
298        $revinfo                     = false;
299        $info['meta']['last_change'] = $revinfo;
300        p_set_metadata($ID, ['last_change' => $revinfo]);
301    }
302
303    if ($revinfo !== false) {
304        $info['ip']   = $revinfo['ip'];
305        $info['user'] = $revinfo['user'];
306        $info['sum']  = $revinfo['sum'];
307        // See also $INFO['meta']['last_change'] which is the most recent log line for page $ID.
308        // Use $INFO['meta']['last_change']['type']===DOKU_CHANGE_TYPE_MINOR_EDIT in place of $info['minor'].
309
310        $info['editor'] = $revinfo['user'] ?: $revinfo['ip'];
311    } else {
312        $info['ip']     = null;
313        $info['user']   = null;
314        $info['sum']    = null;
315        $info['editor'] = null;
316    }
317
318    // draft
319    $draft = new Draft($ID, $info['client']);
320    if ($draft->isDraftAvailable()) {
321        $info['draft'] = $draft->getDraftFilename();
322    }
323
324    return $info;
325}
326
327/**
328 * Initialize and/or fill global $JSINFO with some basic info to be given to javascript
329 */
330function jsinfo()
331{
332    global $JSINFO, $ID, $INFO, $ACT;
333
334    if (!is_array($JSINFO)) {
335        $JSINFO = [];
336    }
337    //export minimal info to JS, plugins can add more
338    $JSINFO['id']                    = $ID;
339    $JSINFO['namespace']             = isset($INFO) ? (string) $INFO['namespace'] : '';
340    $JSINFO['ACT']                   = act_clean($ACT);
341    $JSINFO['useHeadingNavigation']  = (int) useHeading('navigation');
342    $JSINFO['useHeadingContent']     = (int) useHeading('content');
343}
344
345/**
346 * Return information about the current media item as an associative array.
347 *
348 * @return array with info about current media item
349 */
350function mediainfo()
351{
352    global $NS;
353    global $IMG;
354
355    $info = basicinfo("$NS:*");
356    $info['image'] = $IMG;
357
358    return $info;
359}
360
361/**
362 * Build an string of URL parameters
363 *
364 * @author Andreas Gohr
365 *
366 * @param array  $params    array with key-value pairs
367 * @param string $sep       series of pairs are separated by this character
368 * @return string query string
369 */
370function buildURLparams($params, $sep = '&amp;')
371{
372    $url = '';
373    $amp = false;
374    foreach($params as $key => $val) {
375        if($amp) $url .= $sep;
376
377        $url .= rawurlencode($key).'=';
378        $url .= rawurlencode((string) $val);
379        $amp = true;
380    }
381    return $url;
382}
383
384/**
385 * Build an string of html tag attributes
386 *
387 * Skips keys starting with '_', values get HTML encoded
388 *
389 * @author Andreas Gohr
390 *
391 * @param array $params           array with (attribute name-attribute value) pairs
392 * @param bool  $skipEmptyStrings skip empty string values?
393 * @return string
394 */
395function buildAttributes($params, $skipEmptyStrings = false)
396{
397    $url   = '';
398    $white = false;
399    foreach($params as $key => $val) {
400        if($key[0] == '_') continue;
401        if($val === '' && $skipEmptyStrings) continue;
402        if($white) $url .= ' ';
403
404        $url .= $key.'="';
405        $url .= hsc($val);
406        $url .= '"';
407        $white = true;
408    }
409    return $url;
410}
411
412/**
413 * This builds the breadcrumb trail and returns it as array
414 *
415 * @author Andreas Gohr <andi@splitbrain.org>
416 *
417 * @return string[] with the data: array(pageid=>name, ... )
418 */
419function breadcrumbs()
420{
421    // we prepare the breadcrumbs early for quick session closing
422    static $crumbs = null;
423    if($crumbs != null) return $crumbs;
424
425    global $ID;
426    global $ACT;
427    global $conf;
428    global $INFO;
429
430    //first visit?
431    $crumbs = $_SESSION[DOKU_COOKIE]['bc'] ?? [];
432    //we only save on show and existing visible readable wiki documents
433    $file = wikiFN($ID);
434    if($ACT != 'show' || $INFO['perm'] < AUTH_READ || isHiddenPage($ID) || !file_exists($file)) {
435        $_SESSION[DOKU_COOKIE]['bc'] = $crumbs;
436        return $crumbs;
437    }
438
439    // page names
440    $name = noNSorNS($ID);
441    if(useHeading('navigation')) {
442        // get page title
443        $title = p_get_first_heading($ID, METADATA_RENDER_USING_SIMPLE_CACHE);
444        if($title) {
445            $name = $title;
446        }
447    }
448
449    //remove ID from array
450    if(isset($crumbs[$ID])) {
451        unset($crumbs[$ID]);
452    }
453
454    //add to array
455    $crumbs[$ID] = $name;
456    //reduce size
457    while(count($crumbs) > $conf['breadcrumbs']) {
458        array_shift($crumbs);
459    }
460    //save to session
461    $_SESSION[DOKU_COOKIE]['bc'] = $crumbs;
462    return $crumbs;
463}
464
465/**
466 * Filter for page IDs
467 *
468 * This is run on a ID before it is outputted somewhere
469 * currently used to replace the colon with something else
470 * on Windows (non-IIS) systems and to have proper URL encoding
471 *
472 * See discussions at https://github.com/dokuwiki/dokuwiki/pull/84 and
473 * https://github.com/dokuwiki/dokuwiki/pull/173 why we use a whitelist of
474 * unaffected servers instead of blacklisting affected servers here.
475 *
476 * Urlencoding is ommitted when the second parameter is false
477 *
478 * @author Andreas Gohr <andi@splitbrain.org>
479 *
480 * @param string $id pageid being filtered
481 * @param bool   $ue apply urlencoding?
482 * @return string
483 */
484function idfilter($id, $ue = true)
485{
486    global $conf;
487    /* @var Input $INPUT */
488    global $INPUT;
489
490    $id = (string) $id;
491
492    if($conf['useslash'] && $conf['userewrite']) {
493        $id = strtr($id, ':', '/');
494    } elseif(strtoupper(substr(PHP_OS, 0, 3)) === 'WIN' &&
495        $conf['userewrite'] &&
496        strpos($INPUT->server->str('SERVER_SOFTWARE'), 'Microsoft-IIS') === false
497    ) {
498        $id = strtr($id, ':', ';');
499    }
500    if($ue) {
501        $id = rawurlencode($id);
502        $id = str_replace('%3A', ':', $id); //keep as colon
503        $id = str_replace('%3B', ';', $id); //keep as semicolon
504        $id = str_replace('%2F', '/', $id); //keep as slash
505    }
506    return $id;
507}
508
509/**
510 * This builds a link to a wikipage
511 *
512 * It handles URL rewriting and adds additional parameters
513 *
514 * @author Andreas Gohr <andi@splitbrain.org>
515 *
516 * @param string       $id             page id, defaults to start page
517 * @param string|array $urlParameters  URL parameters, associative array recommended
518 * @param bool         $absolute       request an absolute URL instead of relative
519 * @param string       $separator      parameter separator
520 * @return string
521 */
522function wl($id = '', $urlParameters = '', $absolute = false, $separator = '&amp;')
523{
524    global $conf;
525    if(is_array($urlParameters)) {
526        if(isset($urlParameters['rev']) && !$urlParameters['rev']) unset($urlParameters['rev']);
527        if(isset($urlParameters['at']) && $conf['date_at_format']) {
528            $urlParameters['at'] = date($conf['date_at_format'], $urlParameters['at']);
529        }
530        $urlParameters = buildURLparams($urlParameters, $separator);
531    } else {
532        $urlParameters = str_replace(',', $separator, $urlParameters);
533    }
534    if($id === '') {
535        $id = $conf['start'];
536    }
537    $id = idfilter($id);
538    if($absolute) {
539        $xlink = DOKU_URL;
540    } else {
541        $xlink = DOKU_BASE;
542    }
543
544    if($conf['userewrite'] == 2) {
545        $xlink .= DOKU_SCRIPT.'/'.$id;
546        if($urlParameters) $xlink .= '?'.$urlParameters;
547    } elseif($conf['userewrite']) {
548        $xlink .= $id;
549        if($urlParameters) $xlink .= '?'.$urlParameters;
550    } elseif($id !== '') {
551        $xlink .= DOKU_SCRIPT.'?id='.$id;
552        if($urlParameters) $xlink .= $separator.$urlParameters;
553    } else {
554        $xlink .= DOKU_SCRIPT;
555        if($urlParameters) $xlink .= '?'.$urlParameters;
556    }
557
558    return $xlink;
559}
560
561/**
562 * This builds a link to an alternate page format
563 *
564 * Handles URL rewriting if enabled. Follows the style of wl().
565 *
566 * @author Ben Coburn <btcoburn@silicodon.net>
567 * @param string       $id             page id, defaults to start page
568 * @param string       $format         the export renderer to use
569 * @param string|array $urlParameters  URL parameters, associative array recommended
570 * @param bool         $abs            request an absolute URL instead of relative
571 * @param string       $sep            parameter separator
572 * @return string
573 */
574function exportlink($id = '', $format = 'raw', $urlParameters = '', $abs = false, $sep = '&amp;')
575{
576    global $conf;
577    if(is_array($urlParameters)) {
578        $urlParameters = buildURLparams($urlParameters, $sep);
579    } else {
580        $urlParameters = str_replace(',', $sep, $urlParameters);
581    }
582
583    $format = rawurlencode($format);
584    $id     = idfilter($id);
585    if($abs) {
586        $xlink = DOKU_URL;
587    } else {
588        $xlink = DOKU_BASE;
589    }
590
591    if($conf['userewrite'] == 2) {
592        $xlink .= DOKU_SCRIPT.'/'.$id.'?do=export_'.$format;
593        if($urlParameters) $xlink .= $sep.$urlParameters;
594    } elseif($conf['userewrite'] == 1) {
595        $xlink .= '_export/'.$format.'/'.$id;
596        if($urlParameters) $xlink .= '?'.$urlParameters;
597    } else {
598        $xlink .= DOKU_SCRIPT.'?do=export_'.$format.$sep.'id='.$id;
599        if($urlParameters) $xlink .= $sep.$urlParameters;
600    }
601
602    return $xlink;
603}
604
605/**
606 * Build a link to a media file
607 *
608 * Will return a link to the detail page if $direct is false
609 *
610 * The $more parameter should always be given as array, the function then
611 * will strip default parameters to produce even cleaner URLs
612 *
613 * @param string  $id     the media file id or URL
614 * @param mixed   $more   string or array with additional parameters
615 * @param bool    $direct link to detail page if false
616 * @param string  $sep    URL parameter separator
617 * @param bool    $abs    Create an absolute URL
618 * @return string
619 */
620function ml($id = '', $more = '', $direct = true, $sep = '&amp;', $abs = false)
621{
622    global $conf;
623    $isexternalimage = media_isexternal($id);
624    if(!$isexternalimage) {
625        $id = cleanID($id);
626    }
627
628    if(is_array($more)) {
629        // add token for resized images
630        $w = $more['w'] ?? null;
631        $h = $more['h'] ?? null;
632        if($w || $h || $isexternalimage){
633            $more['tok'] = media_get_token($id, $w, $h);
634        }
635        // strip defaults for shorter URLs
636        if(isset($more['cache']) && $more['cache'] == 'cache') unset($more['cache']);
637        if(empty($more['w'])) unset($more['w']);
638        if(empty($more['h'])) unset($more['h']);
639        if(isset($more['id']) && $direct) unset($more['id']);
640        if(isset($more['rev']) && !$more['rev']) unset($more['rev']);
641        $more = buildURLparams($more, $sep);
642    } else {
643        $matches = [];
644        if (preg_match_all('/\b(w|h)=(\d*)\b/', $more, $matches, PREG_SET_ORDER) || $isexternalimage){
645            $resize = ['w'=>0, 'h'=>0];
646            foreach ($matches as $match){
647                $resize[$match[1]] = $match[2];
648            }
649            $more .= $more === '' ? '' : $sep;
650            $more .= 'tok='.media_get_token($id, $resize['w'], $resize['h']);
651        }
652        $more = str_replace('cache=cache', '', $more); //skip default
653        $more = str_replace(',,', ',', $more);
654        $more = str_replace(',', $sep, $more);
655    }
656
657    if($abs) {
658        $xlink = DOKU_URL;
659    } else {
660        $xlink = DOKU_BASE;
661    }
662
663    // external URLs are always direct without rewriting
664    if($isexternalimage) {
665        $xlink .= 'lib/exe/fetch.php';
666        $xlink .= '?'.$more;
667        $xlink .= $sep.'media='.rawurlencode($id);
668        return $xlink;
669    }
670
671    $id = idfilter($id);
672
673    // decide on scriptname
674    if ($direct) {
675        if($conf['userewrite'] == 1) {
676            $script = '_media';
677        } else {
678            $script = 'lib/exe/fetch.php';
679        }
680    } elseif ($conf['userewrite'] == 1) {
681        $script = '_detail';
682    } else {
683        $script = 'lib/exe/detail.php';
684    }
685
686    // build URL based on rewrite mode
687    if ($conf['userewrite']) {
688        $xlink .= $script.'/'.$id;
689        if($more) $xlink .= '?'.$more;
690    } elseif ($more) {
691        $xlink .= $script.'?'.$more;
692        $xlink .= $sep.'media='.$id;
693    } else {
694        $xlink .= $script.'?media='.$id;
695    }
696
697    return $xlink;
698}
699
700/**
701 * Returns the URL to the DokuWiki base script
702 *
703 * Consider using wl() instead, unless you absoutely need the doku.php endpoint
704 *
705 * @author Andreas Gohr <andi@splitbrain.org>
706 *
707 * @return string
708 */
709function script()
710{
711    return DOKU_BASE.DOKU_SCRIPT;
712}
713
714/**
715 * Spamcheck against wordlist
716 *
717 * Checks the wikitext against a list of blocked expressions
718 * returns true if the text contains any bad words
719 *
720 * Triggers COMMON_WORDBLOCK_BLOCKED
721 *
722 *  Action Plugins can use this event to inspect the blocked data
723 *  and gain information about the user who was blocked.
724 *
725 *  Event data:
726 *    data['matches']  - array of matches
727 *    data['userinfo'] - information about the blocked user
728 *      [ip]           - ip address
729 *      [user]         - username (if logged in)
730 *      [mail]         - mail address (if logged in)
731 *      [name]         - real name (if logged in)
732 *
733 * @author Andreas Gohr <andi@splitbrain.org>
734 * @author Michael Klier <chi@chimeric.de>
735 *
736 * @param  string $text - optional text to check, if not given the globals are used
737 * @return bool         - true if a spam word was found
738 */
739function checkwordblock($text = '')
740{
741    global $TEXT;
742    global $PRE;
743    global $SUF;
744    global $SUM;
745    global $conf;
746    global $INFO;
747    /* @var Input $INPUT */
748    global $INPUT;
749
750    if(!$conf['usewordblock']) return false;
751
752    if(!$text) $text = "$PRE $TEXT $SUF $SUM";
753
754    // we prepare the text a tiny bit to prevent spammers circumventing URL checks
755    // phpcs:disable Generic.Files.LineLength.TooLong
756    $text = preg_replace(
757        '!(\b)(www\.[\w.:?\-;,]+?\.[\w.:?\-;,]+?[\w/\#~:.?+=&%@\!\-.:?\-;,]+?)([.:?\-;,]*[^\w/\#~:.?+=&%@\!\-.:?\-;,])!i',
758        '\1http://\2 \2\3',
759        $text
760    );
761    // phpcs:enable
762
763    $wordblocks = getWordblocks();
764    // read file in chunks of 200 - this should work around the
765    // MAX_PATTERN_SIZE in modern PCRE
766    $chunksize = 200;
767
768    while($blocks = array_splice($wordblocks, 0, $chunksize)) {
769        $re = [];
770        // build regexp from blocks
771        foreach($blocks as $block) {
772            $block = preg_replace('/#.*$/', '', $block);
773            $block = trim($block);
774            if(empty($block)) continue;
775            $re[] = $block;
776        }
777        if(count($re) && preg_match('#('.implode('|', $re).')#si', $text, $matches)) {
778            // prepare event data
779            $data = [];
780            $data['matches']        = $matches;
781            $data['userinfo']['ip'] = $INPUT->server->str('REMOTE_ADDR');
782            if($INPUT->server->str('REMOTE_USER')) {
783                $data['userinfo']['user'] = $INPUT->server->str('REMOTE_USER');
784                $data['userinfo']['name'] = $INFO['userinfo']['name'];
785                $data['userinfo']['mail'] = $INFO['userinfo']['mail'];
786            }
787            $callback = static fn() => true;
788            return Event::createAndTrigger('COMMON_WORDBLOCK_BLOCKED', $data, $callback, true);
789        }
790    }
791    return false;
792}
793
794/**
795 * Return the IP of the client
796 *
797 * Honours X-Forwarded-For and X-Real-IP Proxy Headers
798 *
799 * It returns a comma separated list of IPs if the above mentioned
800 * headers are set. If the single parameter is set, it tries to return
801 * a routable public address, prefering the ones suplied in the X
802 * headers
803 *
804 * @author Andreas Gohr <andi@splitbrain.org>
805 *
806 * @param  boolean $single If set only a single IP is returned
807 * @return string
808 */
809function clientIP($single = false)
810{
811    /* @var Input $INPUT */
812    global $INPUT, $conf;
813
814    $ip   = [];
815    $ip[] = $INPUT->server->str('REMOTE_ADDR');
816    if($INPUT->server->str('HTTP_X_FORWARDED_FOR')) {
817        $ip = array_merge($ip, explode(',', str_replace(' ', '', $INPUT->server->str('HTTP_X_FORWARDED_FOR'))));
818    }
819    if($INPUT->server->str('HTTP_X_REAL_IP')) {
820        $ip = array_merge($ip, explode(',', str_replace(' ', '', $INPUT->server->str('HTTP_X_REAL_IP'))));
821    }
822
823    // remove any non-IP stuff
824    $cnt   = count($ip);
825    for($i = 0; $i < $cnt; $i++) {
826        if(filter_var($ip[$i], FILTER_VALIDATE_IP) === false) {
827            unset($ip[$i]);
828        }
829    }
830    $ip = array_values(array_unique($ip));
831    if($ip === [] || !$ip[0]) $ip[0] = '0.0.0.0'; // for some strange reason we don't have a IP
832
833    if(!$single) return implode(',', $ip);
834
835    // skip trusted local addresses
836    foreach($ip as $i) {
837        if(!empty($conf['trustedproxy']) && preg_match('/'.$conf['trustedproxy'].'/', $i)) {
838            continue;
839        } else {
840            return $i;
841        }
842    }
843
844    // still here? just use the last address
845    // this case all ips in the list are trusted
846    return $ip[count($ip)-1];
847}
848
849/**
850 * Check if the browser is on a mobile device
851 *
852 * Adapted from the example code at url below
853 *
854 * @link http://www.brainhandles.com/2007/10/15/detecting-mobile-browsers/#code
855 *
856 * @deprecated 2018-04-27 you probably want media queries instead anyway
857 * @return bool if true, client is mobile browser; otherwise false
858 */
859function clientismobile()
860{
861    /* @var Input $INPUT */
862    global $INPUT;
863
864    if($INPUT->server->has('HTTP_X_WAP_PROFILE')) return true;
865
866    if(preg_match('/wap\.|\.wap/i', $INPUT->server->str('HTTP_ACCEPT'))) return true;
867
868    if(!$INPUT->server->has('HTTP_USER_AGENT')) return false;
869
870    $uamatches = implode(
871        '|',
872        [
873            'midp', 'j2me', 'avantg', 'docomo', 'novarra', 'palmos', 'palmsource', '240x320', 'opwv',
874            'chtml', 'pda', 'windows ce', 'mmp\/', 'blackberry', 'mib\/', 'symbian', 'wireless', 'nokia',
875            'hand', 'mobi', 'phone', 'cdm', 'up\.b', 'audio', 'SIE\-', 'SEC\-', 'samsung', 'HTC', 'mot\-',
876            'mitsu', 'sagem', 'sony', 'alcatel', 'lg', 'erics', 'vx', 'NEC', 'philips', 'mmm', 'xx',
877            'panasonic', 'sharp', 'wap', 'sch', 'rover', 'pocket', 'benq', 'java', 'pt', 'pg', 'vox',
878            'amoi', 'bird', 'compal', 'kg', 'voda', 'sany', 'kdd', 'dbt', 'sendo', 'sgh', 'gradi', 'jb',
879            '\d\d\di', 'moto'
880        ]
881    );
882
883    if(preg_match("/$uamatches/i", $INPUT->server->str('HTTP_USER_AGENT'))) return true;
884
885    return false;
886}
887
888/**
889 * check if a given link is interwiki link
890 *
891 * @param string $link the link, e.g. "wiki>page"
892 * @return bool
893 */
894function link_isinterwiki($link)
895{
896    if (preg_match('/^[a-zA-Z0-9\.]+>/u', $link)) return true;
897    return false;
898}
899
900/**
901 * Convert one or more comma separated IPs to hostnames
902 *
903 * If $conf['dnslookups'] is disabled it simply returns the input string
904 *
905 * @author Glen Harris <astfgl@iamnota.org>
906 *
907 * @param  string $ips comma separated list of IP addresses
908 * @return string a comma separated list of hostnames
909 */
910function gethostsbyaddrs($ips)
911{
912    global $conf;
913    if(!$conf['dnslookups']) return $ips;
914
915    $hosts = [];
916    $ips   = explode(',', $ips);
917
918    if(is_array($ips)) {
919        foreach($ips as $ip) {
920            $hosts[] = gethostbyaddr(trim($ip));
921        }
922        return implode(',', $hosts);
923    } else {
924        return gethostbyaddr(trim($ips));
925    }
926}
927
928/**
929 * Checks if a given page is currently locked.
930 *
931 * removes stale lockfiles
932 *
933 * @author Andreas Gohr <andi@splitbrain.org>
934 *
935 * @param string $id page id
936 * @return bool page is locked?
937 */
938function checklock($id)
939{
940    global $conf;
941    /* @var Input $INPUT */
942    global $INPUT;
943
944    $lock = wikiLockFN($id);
945
946    //no lockfile
947    if(!file_exists($lock)) return false;
948
949    //lockfile expired
950    if((time() - filemtime($lock)) > $conf['locktime']) {
951        @unlink($lock);
952        return false;
953    }
954
955    //my own lock
956    @[$ip, $session] = explode("\n", io_readFile($lock));
957    if($ip == $INPUT->server->str('REMOTE_USER') || (session_id() && $session === session_id())) {
958        return false;
959    }
960
961    return $ip;
962}
963
964/**
965 * Lock a page for editing
966 *
967 * @author Andreas Gohr <andi@splitbrain.org>
968 *
969 * @param string $id page id to lock
970 */
971function lock($id)
972{
973    global $conf;
974    /* @var Input $INPUT */
975    global $INPUT;
976
977    if($conf['locktime'] == 0) {
978        return;
979    }
980
981    $lock = wikiLockFN($id);
982    if($INPUT->server->str('REMOTE_USER')) {
983        io_saveFile($lock, $INPUT->server->str('REMOTE_USER'));
984    } else {
985        io_saveFile($lock, clientIP()."\n".session_id());
986    }
987}
988
989/**
990 * Unlock a page if it was locked by the user
991 *
992 * @author Andreas Gohr <andi@splitbrain.org>
993 *
994 * @param string $id page id to unlock
995 * @return bool true if a lock was removed
996 */
997function unlock($id)
998{
999    /* @var Input $INPUT */
1000    global $INPUT;
1001
1002    $lock = wikiLockFN($id);
1003    if(file_exists($lock)) {
1004        @[$ip, $session] = explode("\n", io_readFile($lock));
1005        if($ip == $INPUT->server->str('REMOTE_USER') || $session == session_id()) {
1006            @unlink($lock);
1007            return true;
1008        }
1009    }
1010    return false;
1011}
1012
1013/**
1014 * convert line ending to unix format
1015 *
1016 * also makes sure the given text is valid UTF-8
1017 *
1018 * @see    formText() for 2crlf conversion
1019 * @author Andreas Gohr <andi@splitbrain.org>
1020 *
1021 * @param string $text
1022 * @return string
1023 */
1024function cleanText($text)
1025{
1026    $text = preg_replace("/(\015\012)|(\015)/", "\012", $text);
1027
1028    // if the text is not valid UTF-8 we simply assume latin1
1029    // this won't break any worse than it breaks with the wrong encoding
1030    // but might actually fix the problem in many cases
1031    if(!Clean::isUtf8($text)) $text = utf8_encode($text);
1032
1033    return $text;
1034}
1035
1036/**
1037 * Prepares text for print in Webforms by encoding special chars.
1038 * It also converts line endings to Windows format which is
1039 * pseudo standard for webforms.
1040 *
1041 * @see    cleanText() for 2unix conversion
1042 * @author Andreas Gohr <andi@splitbrain.org>
1043 *
1044 * @param string $text
1045 * @return string
1046 */
1047function formText($text)
1048{
1049    $text = str_replace("\012", "\015\012", $text ?? '');
1050    return htmlspecialchars($text);
1051}
1052
1053/**
1054 * Returns the specified local text in raw format
1055 *
1056 * @author Andreas Gohr <andi@splitbrain.org>
1057 *
1058 * @param string $id   page id
1059 * @param string $ext  extension of file being read, default 'txt'
1060 * @return string
1061 */
1062function rawLocale($id, $ext = 'txt')
1063{
1064    return io_readFile(localeFN($id, $ext));
1065}
1066
1067/**
1068 * Returns the raw WikiText
1069 *
1070 * @author Andreas Gohr <andi@splitbrain.org>
1071 *
1072 * @param string $id   page id
1073 * @param string|int $rev  timestamp when a revision of wikitext is desired
1074 * @return string
1075 */
1076function rawWiki($id, $rev = '')
1077{
1078    return io_readWikiPage(wikiFN($id, $rev), $id, $rev);
1079}
1080
1081/**
1082 * Returns the pagetemplate contents for the ID's namespace
1083 *
1084 * @triggers COMMON_PAGETPL_LOAD
1085 * @author Andreas Gohr <andi@splitbrain.org>
1086 *
1087 * @param string $id the id of the page to be created
1088 * @return string parsed pagetemplate content
1089 */
1090function pageTemplate($id)
1091{
1092    global $conf;
1093
1094    if(is_array($id)) $id = $id[0];
1095
1096    // prepare initial event data
1097    $data = [
1098        'id'        => $id, // the id of the page to be created
1099        'tpl'       => '', // the text used as template
1100        'tplfile'   => '', // the file above text was/should be loaded from
1101        'doreplace' => true,
1102    ];
1103
1104    $evt = new Event('COMMON_PAGETPL_LOAD', $data);
1105    if($evt->advise_before(true)) {
1106        // the before event might have loaded the content already
1107        if(empty($data['tpl'])) {
1108            // if the before event did not set a template file, try to find one
1109            if(empty($data['tplfile'])) {
1110                $path = dirname(wikiFN($id));
1111                if(file_exists($path.'/_template.txt')) {
1112                    $data['tplfile'] = $path.'/_template.txt';
1113                } else {
1114                    // search upper namespaces for templates
1115                    $len = strlen(rtrim($conf['datadir'], '/'));
1116                    while(strlen($path) >= $len) {
1117                        if(file_exists($path.'/__template.txt')) {
1118                            $data['tplfile'] = $path.'/__template.txt';
1119                            break;
1120                        }
1121                        $path = substr($path, 0, strrpos($path, '/'));
1122                    }
1123                }
1124            }
1125            // load the content
1126            $data['tpl'] = io_readFile($data['tplfile']);
1127        }
1128        if($data['doreplace']) parsePageTemplate($data);
1129    }
1130    $evt->advise_after();
1131    unset($evt);
1132
1133    return $data['tpl'];
1134}
1135
1136/**
1137 * Performs common page template replacements
1138 * This works on data from COMMON_PAGETPL_LOAD
1139 *
1140 * @author Andreas Gohr <andi@splitbrain.org>
1141 *
1142 * @param array $data array with event data
1143 * @return string
1144 */
1145function parsePageTemplate(&$data)
1146{
1147    /**
1148     * @var string $id        the id of the page to be created
1149     * @var string $tpl       the text used as template
1150     * @var string $tplfile   the file above text was/should be loaded from
1151     * @var bool   $doreplace should wildcard replacements be done on the text?
1152     */
1153    extract($data);
1154
1155    global $USERINFO;
1156    global $conf;
1157    /* @var Input $INPUT */
1158    global $INPUT;
1159
1160    // replace placeholders
1161    $file = noNS($id);
1162    $page = strtr($file, $conf['sepchar'], ' ');
1163
1164    $tpl = str_replace(
1165        [
1166            '@ID@',
1167            '@NS@',
1168            '@CURNS@',
1169            '@!CURNS@',
1170            '@!!CURNS@',
1171            '@!CURNS!@',
1172            '@FILE@',
1173            '@!FILE@',
1174            '@!FILE!@',
1175            '@PAGE@',
1176            '@!PAGE@',
1177            '@!!PAGE@',
1178            '@!PAGE!@',
1179            '@USER@',
1180            '@NAME@',
1181            '@MAIL@',
1182            '@DATE@'
1183        ],
1184        [
1185            $id,
1186            getNS($id),
1187            curNS($id),
1188            PhpString::ucfirst(curNS($id)),
1189            PhpString::ucwords(curNS($id)),
1190            PhpString::strtoupper(curNS($id)),
1191            $file,
1192            PhpString::ucfirst($file),
1193            PhpString::strtoupper($file),
1194            $page,
1195            PhpString::ucfirst($page),
1196            PhpString::ucwords($page),
1197            PhpString::strtoupper($page),
1198            $INPUT->server->str('REMOTE_USER'),
1199            $USERINFO ? $USERINFO['name'] : '',
1200            $USERINFO ? $USERINFO['mail'] : '',
1201            $conf['dformat']
1202        ],
1203        $tpl
1204    );
1205
1206    // we need the callback to work around strftime's char limit
1207    $tpl = preg_replace_callback(
1208        '/%./',
1209        static fn($m) => dformat(null, $m[0]),
1210        $tpl
1211    );
1212    $data['tpl'] = $tpl;
1213    return $tpl;
1214}
1215
1216/**
1217 * Returns the raw Wiki Text in three slices.
1218 *
1219 * The range parameter needs to have the form "from-to"
1220 * and gives the range of the section in bytes - no
1221 * UTF-8 awareness is needed.
1222 * The returned order is prefix, section and suffix.
1223 *
1224 * @author Andreas Gohr <andi@splitbrain.org>
1225 *
1226 * @param string $range in form "from-to"
1227 * @param string $id    page id
1228 * @param string $rev   optional, the revision timestamp
1229 * @return string[] with three slices
1230 */
1231function rawWikiSlices($range, $id, $rev = '')
1232{
1233    $text = io_readWikiPage(wikiFN($id, $rev), $id, $rev);
1234
1235    // Parse range
1236    [$from, $to] = sexplode('-', $range, 2);
1237    // Make range zero-based, use defaults if marker is missing
1238    $from = $from ? $from - 1 : (0);
1239    $to   = $to ? $to - 1 : (strlen($text));
1240
1241    $slices = [];
1242    $slices[0] = substr($text, 0, $from);
1243    $slices[1] = substr($text, $from, $to - $from);
1244    $slices[2] = substr($text, $to);
1245    return $slices;
1246}
1247
1248/**
1249 * Joins wiki text slices
1250 *
1251 * function to join the text slices.
1252 * When the pretty parameter is set to true it adds additional empty
1253 * lines between sections if needed (used on saving).
1254 *
1255 * @author Andreas Gohr <andi@splitbrain.org>
1256 *
1257 * @param string $pre   prefix
1258 * @param string $text  text in the middle
1259 * @param string $suf   suffix
1260 * @param bool $pretty add additional empty lines between sections
1261 * @return string
1262 */
1263function con($pre, $text, $suf, $pretty = false)
1264{
1265    if($pretty) {
1266        if($pre !== '' && substr($pre, -1) !== "\n" &&
1267            substr($text, 0, 1) !== "\n"
1268        ) {
1269            $pre .= "\n";
1270        }
1271        if($suf !== '' && substr($text, -1) !== "\n" &&
1272            substr($suf, 0, 1) !== "\n"
1273        ) {
1274            $text .= "\n";
1275        }
1276    }
1277
1278    return $pre.$text.$suf;
1279}
1280
1281/**
1282 * Checks if the current page version is newer than the last entry in the page's
1283 * changelog. If so, we assume it has been an external edit and we create an
1284 * attic copy and add a proper changelog line.
1285 *
1286 * This check is only executed when the page is about to be saved again from the
1287 * wiki, triggered in @see saveWikiText()
1288 *
1289 * @param string $id the page ID
1290 * @deprecated 2021-11-28
1291 */
1292function detectExternalEdit($id)
1293{
1294    dbg_deprecated(PageFile::class .'::detectExternalEdit()');
1295    (new PageFile($id))->detectExternalEdit();
1296}
1297
1298/**
1299 * Saves a wikitext by calling io_writeWikiPage.
1300 * Also directs changelog and attic updates.
1301 *
1302 * @author Andreas Gohr <andi@splitbrain.org>
1303 * @author Ben Coburn <btcoburn@silicodon.net>
1304 *
1305 * @param string $id       page id
1306 * @param string $text     wikitext being saved
1307 * @param string $summary  summary of text update
1308 * @param bool   $minor    mark this saved version as minor update
1309 */
1310function saveWikiText($id, $text, $summary, $minor = false)
1311{
1312
1313    // get COMMON_WIKIPAGE_SAVE event data
1314    $data = (new PageFile($id))->saveWikiText($text, $summary, $minor);
1315    if(!$data) return; // save was cancelled (for no changes or by a plugin)
1316
1317    // send notify mails
1318    ['oldRevision' => $rev, 'newRevision' => $new_rev, 'summary' => $summary] = $data;
1319    notify($id, 'admin', $rev, $summary, $minor, $new_rev);
1320    notify($id, 'subscribers', $rev, $summary, $minor, $new_rev);
1321
1322    // if useheading is enabled, purge the cache of all linking pages
1323    if (useHeading('content')) {
1324        $pages = ft_backlinks($id, true);
1325        foreach ($pages as $page) {
1326            $cache = new CacheRenderer($page, wikiFN($page), 'xhtml');
1327            $cache->removeCache();
1328        }
1329    }
1330}
1331
1332/**
1333 * moves the current version to the attic and returns its revision date
1334 *
1335 * @author Andreas Gohr <andi@splitbrain.org>
1336 *
1337 * @param string $id page id
1338 * @return int|string revision timestamp
1339 * @deprecated 2021-11-28
1340 */
1341function saveOldRevision($id)
1342{
1343    dbg_deprecated(PageFile::class .'::saveOldRevision()');
1344    return (new PageFile($id))->saveOldRevision();
1345}
1346
1347/**
1348 * Sends a notify mail on page change or registration
1349 *
1350 * @param string     $id       The changed page
1351 * @param string     $who      Who to notify (admin|subscribers|register)
1352 * @param int|string $rev      Old page revision
1353 * @param string     $summary  What changed
1354 * @param boolean    $minor    Is this a minor edit?
1355 * @param string[]   $replace  Additional string substitutions, @KEY@ to be replaced by value
1356 * @param int|string $current_rev  New page revision
1357 * @return bool
1358 *
1359 * @author Andreas Gohr <andi@splitbrain.org>
1360 */
1361function notify($id, $who, $rev = '', $summary = '', $minor = false, $replace = [], $current_rev = false)
1362{
1363    global $conf;
1364    /* @var Input $INPUT */
1365    global $INPUT;
1366
1367    // decide if there is something to do, eg. whom to mail
1368    if ($who == 'admin') {
1369        if (empty($conf['notify'])) return false; //notify enabled?
1370        $tpl = 'mailtext';
1371        $to  = $conf['notify'];
1372    } elseif ($who == 'subscribers') {
1373        if (!actionOK('subscribe')) return false; //subscribers enabled?
1374        if ($conf['useacl'] && $INPUT->server->str('REMOTE_USER') && $minor) return false; //skip minors
1375        $data = ['id' => $id, 'addresslist' => '', 'self' => false, 'replacements' => $replace];
1376        Event::createAndTrigger(
1377            'COMMON_NOTIFY_ADDRESSLIST',
1378            $data,
1379            [new SubscriberManager(), 'notifyAddresses']
1380        );
1381        $to = $data['addresslist'];
1382        if (empty($to)) return false;
1383        $tpl = 'subscr_single';
1384    } else {
1385        return false; //just to be safe
1386    }
1387
1388    // prepare content
1389    $subscription = new PageSubscriptionSender();
1390    return $subscription->sendPageDiff($to, $tpl, $id, $rev, $summary, $current_rev);
1391}
1392
1393/**
1394 * extracts the query from a search engine referrer
1395 *
1396 * @author Andreas Gohr <andi@splitbrain.org>
1397 * @author Todd Augsburger <todd@rollerorgans.com>
1398 *
1399 * @return array|string
1400 */
1401function getGoogleQuery()
1402{
1403    /* @var Input $INPUT */
1404    global $INPUT;
1405
1406    if(!$INPUT->server->has('HTTP_REFERER')) {
1407        return '';
1408    }
1409    $url = parse_url($INPUT->server->str('HTTP_REFERER'));
1410
1411    // only handle common SEs
1412    if(!array_key_exists('host', $url)) return '';
1413    if(!preg_match('/(google|bing|yahoo|ask|duckduckgo|babylon|aol|yandex)/', $url['host'])) return '';
1414
1415    $query = [];
1416    if(!array_key_exists('query', $url)) return '';
1417    parse_str($url['query'], $query);
1418
1419    $q = '';
1420    if(isset($query['q'])){
1421        $q = $query['q'];
1422    }elseif(isset($query['p'])){
1423        $q = $query['p'];
1424    }elseif(isset($query['query'])){
1425        $q = $query['query'];
1426    }
1427    $q = trim($q);
1428
1429    if(!$q) return '';
1430    // ignore if query includes a full URL
1431    if(strpos($q, '//') !== false) return '';
1432    $q = preg_split('/[\s\'"\\\\`()\]\[?:!\.{};,#+*<>\\/]+/', $q, -1, PREG_SPLIT_NO_EMPTY);
1433    return $q;
1434}
1435
1436/**
1437 * Return the human readable size of a file
1438 *
1439 * @param int $size A file size
1440 * @param int $dec A number of decimal places
1441 * @return string human readable size
1442 *
1443 * @author      Martin Benjamin <b.martin@cybernet.ch>
1444 * @author      Aidan Lister <aidan@php.net>
1445 * @version     1.0.0
1446 */
1447function filesize_h($size, $dec = 1)
1448{
1449    $sizes = ['B', 'KB', 'MB', 'GB'];
1450    $count = count($sizes);
1451    $i     = 0;
1452
1453    while($size >= 1024 && ($i < $count - 1)) {
1454        $size /= 1024;
1455        $i++;
1456    }
1457
1458    return round($size, $dec)."\xC2\xA0".$sizes[$i]; //non-breaking space
1459}
1460
1461/**
1462 * Return the given timestamp as human readable, fuzzy age
1463 *
1464 * @author Andreas Gohr <gohr@cosmocode.de>
1465 *
1466 * @param int $dt timestamp
1467 * @return string
1468 */
1469function datetime_h($dt)
1470{
1471    global $lang;
1472
1473    $ago = time() - $dt;
1474    if($ago > 24 * 60 * 60 * 30 * 12 * 2) {
1475        return sprintf($lang['years'], round($ago / (24 * 60 * 60 * 30 * 12)));
1476    }
1477    if($ago > 24 * 60 * 60 * 30 * 2) {
1478        return sprintf($lang['months'], round($ago / (24 * 60 * 60 * 30)));
1479    }
1480    if($ago > 24 * 60 * 60 * 7 * 2) {
1481        return sprintf($lang['weeks'], round($ago / (24 * 60 * 60 * 7)));
1482    }
1483    if($ago > 24 * 60 * 60 * 2) {
1484        return sprintf($lang['days'], round($ago / (24 * 60 * 60)));
1485    }
1486    if($ago > 60 * 60 * 2) {
1487        return sprintf($lang['hours'], round($ago / (60 * 60)));
1488    }
1489    if($ago > 60 * 2) {
1490        return sprintf($lang['minutes'], round($ago / (60)));
1491    }
1492    return sprintf($lang['seconds'], $ago);
1493}
1494
1495/**
1496 * Wraps around strftime but provides support for fuzzy dates
1497 *
1498 * The format default to $conf['dformat']. It is passed to
1499 * strftime - %f can be used to get the value from datetime_h()
1500 *
1501 * @see datetime_h
1502 * @author Andreas Gohr <gohr@cosmocode.de>
1503 *
1504 * @param int|null $dt      timestamp when given, null will take current timestamp
1505 * @param string   $format  empty default to $conf['dformat'], or provide format as recognized by strftime()
1506 * @return string
1507 */
1508function dformat($dt = null, $format = '')
1509{
1510    global $conf;
1511
1512    if(is_null($dt)) $dt = time();
1513    $dt = (int) $dt;
1514    if(!$format) $format = $conf['dformat'];
1515
1516    $format = str_replace('%f', datetime_h($dt), $format);
1517    return strftime($format, $dt);
1518}
1519
1520/**
1521 * Formats a timestamp as ISO 8601 date
1522 *
1523 * @author <ungu at terong dot com>
1524 * @link http://php.net/manual/en/function.date.php#54072
1525 *
1526 * @param int $int_date current date in UNIX timestamp
1527 * @return string
1528 */
1529function date_iso8601($int_date)
1530{
1531    $date_mod     = date('Y-m-d\TH:i:s', $int_date);
1532    $pre_timezone = date('O', $int_date);
1533    $time_zone    = substr($pre_timezone, 0, 3).":".substr($pre_timezone, 3, 2);
1534    $date_mod .= $time_zone;
1535    return $date_mod;
1536}
1537
1538/**
1539 * return an obfuscated email address in line with $conf['mailguard'] setting
1540 *
1541 * @author Harry Fuecks <hfuecks@gmail.com>
1542 * @author Christopher Smith <chris@jalakai.co.uk>
1543 *
1544 * @param string $email email address
1545 * @return string
1546 */
1547function obfuscate($email)
1548{
1549    global $conf;
1550
1551    switch($conf['mailguard']) {
1552        case 'visible' :
1553            $obfuscate = ['@' => ' [at] ', '.' => ' [dot] ', '-' => ' [dash] '];
1554            return strtr($email, $obfuscate);
1555
1556        case 'hex' :
1557            return Conversion::toHtml($email, true);
1558
1559        case 'none' :
1560        default :
1561            return $email;
1562    }
1563}
1564
1565/**
1566 * Removes quoting backslashes
1567 *
1568 * @author Andreas Gohr <andi@splitbrain.org>
1569 *
1570 * @param string $string
1571 * @param string $char backslashed character
1572 * @return string
1573 */
1574function unslash($string, $char = "'")
1575{
1576    return str_replace('\\'.$char, $char, $string);
1577}
1578
1579/**
1580 * Convert php.ini shorthands to byte
1581 *
1582 * On 32 bit systems values >= 2GB will fail!
1583 *
1584 * -1 (infinite size) will be reported as -1
1585 *
1586 * @link   https://www.php.net/manual/en/faq.using.php#faq.using.shorthandbytes
1587 * @param string $value PHP size shorthand
1588 * @return int
1589 */
1590function php_to_byte($value)
1591{
1592    switch (strtoupper(substr($value, -1))) {
1593        case 'G':
1594            $ret = (int) substr($value, 0, -1) * 1024 * 1024 * 1024;
1595            break;
1596        case 'M':
1597            $ret = (int) substr($value, 0, -1) * 1024 * 1024;
1598            break;
1599        case 'K':
1600            $ret = (int) substr($value, 0, -1) * 1024;
1601            break;
1602        default:
1603            $ret = (int) $value;
1604            break;
1605    }
1606    return $ret;
1607}
1608
1609/**
1610 * Wrapper around preg_quote adding the default delimiter
1611 *
1612 * @param string $string
1613 * @return string
1614 */
1615function preg_quote_cb($string)
1616{
1617    return preg_quote($string, '/');
1618}
1619
1620/**
1621 * Shorten a given string by removing data from the middle
1622 *
1623 * You can give the string in two parts, the first part $keep
1624 * will never be shortened. The second part $short will be cut
1625 * in the middle to shorten but only if at least $min chars are
1626 * left to display it. Otherwise it will be left off.
1627 *
1628 * @param string $keep   the part to keep
1629 * @param string $short  the part to shorten
1630 * @param int    $max    maximum chars you want for the whole string
1631 * @param int    $min    minimum number of chars to have left for middle shortening
1632 * @param string $char   the shortening character to use
1633 * @return string
1634 */
1635function shorten($keep, $short, $max, $min = 9, $char = '…')
1636{
1637    $max -= PhpString::strlen($keep);
1638    if($max < $min) return $keep;
1639    $len = PhpString::strlen($short);
1640    if($len <= $max) return $keep.$short;
1641    $half = floor($max / 2);
1642    return $keep .
1643        PhpString::substr($short, 0, $half - 1) .
1644        $char .
1645        PhpString::substr($short, $len - $half);
1646}
1647
1648/**
1649 * Return the users real name or e-mail address for use
1650 * in page footer and recent changes pages
1651 *
1652 * @param string|null $username or null when currently logged-in user should be used
1653 * @param bool $textonly true returns only plain text, true allows returning html
1654 * @return string html or plain text(not escaped) of formatted user name
1655 *
1656 * @author Andy Webber <dokuwiki AT andywebber DOT com>
1657 */
1658function editorinfo($username, $textonly = false)
1659{
1660    return userlink($username, $textonly);
1661}
1662
1663/**
1664 * Returns users realname w/o link
1665 *
1666 * @param string|null $username or null when currently logged-in user should be used
1667 * @param bool $textonly true returns only plain text, true allows returning html
1668 * @return string html or plain text(not escaped) of formatted user name
1669 *
1670 * @triggers COMMON_USER_LINK
1671 */
1672function userlink($username = null, $textonly = false)
1673{
1674    global $conf, $INFO;
1675    /** @var AuthPlugin $auth */
1676    global $auth;
1677    /** @var Input $INPUT */
1678    global $INPUT;
1679
1680    // prepare initial event data
1681    $data = [
1682        'username' => $username, // the unique user name
1683        'name' => '',
1684        'link' => [
1685            //setting 'link' to false disables linking
1686            'target' => '',
1687            'pre' => '',
1688            'suf' => '',
1689            'style' => '',
1690            'more' => '',
1691            'url' => '',
1692            'title' => '',
1693            'class' => '',
1694        ],
1695        'userlink' => '', // formatted user name as will be returned
1696        'textonly' => $textonly,
1697    ];
1698    if($username === null) {
1699        $data['username'] = $username = $INPUT->server->str('REMOTE_USER');
1700        if($textonly){
1701            $data['name'] = $INFO['userinfo']['name']. ' (' . $INPUT->server->str('REMOTE_USER') . ')';
1702        }else {
1703            $data['name'] = '<bdi>' . hsc($INFO['userinfo']['name']) . '</bdi> '.
1704                '(<bdi>' . hsc($INPUT->server->str('REMOTE_USER')) . '</bdi>)';
1705        }
1706    }
1707
1708    $evt = new Event('COMMON_USER_LINK', $data);
1709    if($evt->advise_before(true)) {
1710        if(empty($data['name'])) {
1711            if($auth) $info = $auth->getUserData($username);
1712            if($conf['showuseras'] != 'loginname' && isset($info) && $info) {
1713                switch($conf['showuseras']) {
1714                    case 'username':
1715                    case 'username_link':
1716                        $data['name'] = $textonly ? $info['name'] : hsc($info['name']);
1717                        break;
1718                    case 'email':
1719                    case 'email_link':
1720                        $data['name'] = obfuscate($info['mail']);
1721                        break;
1722                }
1723            } else {
1724                $data['name'] = $textonly ? $data['username'] : hsc($data['username']);
1725            }
1726        }
1727
1728        /** @var Doku_Renderer_xhtml $xhtml_renderer */
1729        static $xhtml_renderer = null;
1730
1731        if(!$data['textonly'] && empty($data['link']['url'])) {
1732
1733            if(in_array($conf['showuseras'], ['email_link', 'username_link'])) {
1734                if(!isset($info)) {
1735                    if($auth) $info = $auth->getUserData($username);
1736                }
1737                if(isset($info) && $info) {
1738                    if($conf['showuseras'] == 'email_link') {
1739                        $data['link']['url'] = 'mailto:' . obfuscate($info['mail']);
1740                    } else {
1741                        if(is_null($xhtml_renderer)) {
1742                            $xhtml_renderer = p_get_renderer('xhtml');
1743                        }
1744                        if(empty($xhtml_renderer->interwiki)) {
1745                            $xhtml_renderer->interwiki = getInterwiki();
1746                        }
1747                        $shortcut = 'user';
1748                        $exists = null;
1749                        $data['link']['url'] = $xhtml_renderer->_resolveInterWiki($shortcut, $username, $exists);
1750                        $data['link']['class'] .= ' interwiki iw_user';
1751                        if($exists !== null) {
1752                            if($exists) {
1753                                $data['link']['class'] .= ' wikilink1';
1754                            } else {
1755                                $data['link']['class'] .= ' wikilink2';
1756                                $data['link']['rel'] = 'nofollow';
1757                            }
1758                        }
1759                    }
1760                } else {
1761                    $data['textonly'] = true;
1762                }
1763
1764            } else {
1765                $data['textonly'] = true;
1766            }
1767        }
1768
1769        if($data['textonly']) {
1770            $data['userlink'] = $data['name'];
1771        } else {
1772            $data['link']['name'] = $data['name'];
1773            if(is_null($xhtml_renderer)) {
1774                $xhtml_renderer = p_get_renderer('xhtml');
1775            }
1776            $data['userlink'] = $xhtml_renderer->_formatLink($data['link']);
1777        }
1778    }
1779    $evt->advise_after();
1780    unset($evt);
1781
1782    return $data['userlink'];
1783}
1784
1785/**
1786 * Returns the path to a image file for the currently chosen license.
1787 * When no image exists, returns an empty string
1788 *
1789 * @author Andreas Gohr <andi@splitbrain.org>
1790 *
1791 * @param  string $type - type of image 'badge' or 'button'
1792 * @return string
1793 */
1794function license_img($type)
1795{
1796    global $license;
1797    global $conf;
1798    if(!$conf['license']) return '';
1799    if(!is_array($license[$conf['license']])) return '';
1800    $try   = [];
1801    $try[] = 'lib/images/license/'.$type.'/'.$conf['license'].'.png';
1802    $try[] = 'lib/images/license/'.$type.'/'.$conf['license'].'.gif';
1803    if(substr($conf['license'], 0, 3) == 'cc-') {
1804        $try[] = 'lib/images/license/'.$type.'/cc.png';
1805    }
1806    foreach($try as $src) {
1807        if(file_exists(DOKU_INC.$src)) return $src;
1808    }
1809    return '';
1810}
1811
1812/**
1813 * Checks if the given amount of memory is available
1814 *
1815 * If the memory_get_usage() function is not available the
1816 * function just assumes $bytes of already allocated memory
1817 *
1818 * @author Filip Oscadal <webmaster@illusionsoftworks.cz>
1819 * @author Andreas Gohr <andi@splitbrain.org>
1820 *
1821 * @param int  $mem    Size of memory you want to allocate in bytes
1822 * @param int  $bytes  already allocated memory (see above)
1823 * @return bool
1824 */
1825function is_mem_available($mem, $bytes = 1_048_576)
1826{
1827    $limit = trim(ini_get('memory_limit'));
1828    if(empty($limit)) return true; // no limit set!
1829    if($limit == -1) return true; // unlimited
1830
1831    // parse limit to bytes
1832    $limit = php_to_byte($limit);
1833
1834    // get used memory if possible
1835    if(function_exists('memory_get_usage')) {
1836        $used = memory_get_usage();
1837    } else {
1838        $used = $bytes;
1839    }
1840
1841    if($used + $mem > $limit) {
1842        return false;
1843    }
1844
1845    return true;
1846}
1847
1848/**
1849 * Send a HTTP redirect to the browser
1850 *
1851 * Works arround Microsoft IIS cookie sending bug. Exits the script.
1852 *
1853 * @link   http://support.microsoft.com/kb/q176113/
1854 * @author Andreas Gohr <andi@splitbrain.org>
1855 *
1856 * @param string $url url being directed to
1857 */
1858function send_redirect($url)
1859{
1860    $url = stripctl($url); // defend against HTTP Response Splitting
1861
1862    /* @var Input $INPUT */
1863    global $INPUT;
1864
1865    //are there any undisplayed messages? keep them in session for display
1866    global $MSG;
1867    if(isset($MSG) && count($MSG) && !defined('NOSESSION')) {
1868        //reopen session, store data and close session again
1869        @session_start();
1870        $_SESSION[DOKU_COOKIE]['msg'] = $MSG;
1871    }
1872
1873    // always close the session
1874    session_write_close();
1875
1876    // check if running on IIS < 6 with CGI-PHP
1877    if($INPUT->server->has('SERVER_SOFTWARE') && $INPUT->server->has('GATEWAY_INTERFACE') &&
1878        (strpos($INPUT->server->str('GATEWAY_INTERFACE'), 'CGI') !== false) &&
1879        (preg_match('|^Microsoft-IIS/(\d)\.\d$|', trim($INPUT->server->str('SERVER_SOFTWARE')), $matches)) &&
1880        $matches[1] < 6
1881    ) {
1882        header('Refresh: 0;url='.$url);
1883    } else {
1884        header('Location: '.$url);
1885    }
1886
1887    // no exits during unit tests
1888    if(defined('DOKU_UNITTEST')) {
1889        // pass info about the redirect back to the test suite
1890        $testRequest = TestRequest::getRunning();
1891        if($testRequest !== null) {
1892            $testRequest->addData('send_redirect', $url);
1893        }
1894        return;
1895    }
1896
1897    exit;
1898}
1899
1900/**
1901 * Validate a value using a set of valid values
1902 *
1903 * This function checks whether a specified value is set and in the array
1904 * $valid_values. If not, the function returns a default value or, if no
1905 * default is specified, throws an exception.
1906 *
1907 * @param string $param        The name of the parameter
1908 * @param array  $valid_values A set of valid values; Optionally a default may
1909 *                             be marked by the key “default”.
1910 * @param array  $array        The array containing the value (typically $_POST
1911 *                             or $_GET)
1912 * @param string $exc          The text of the raised exception
1913 *
1914 * @throws Exception
1915 * @return mixed
1916 * @author Adrian Lang <lang@cosmocode.de>
1917 */
1918function valid_input_set($param, $valid_values, $array, $exc = '')
1919{
1920    if(isset($array[$param]) && in_array($array[$param], $valid_values)) {
1921        return $array[$param];
1922    } elseif(isset($valid_values['default'])) {
1923        return $valid_values['default'];
1924    } else {
1925        throw new Exception($exc);
1926    }
1927}
1928
1929/**
1930 * Read a preference from the DokuWiki cookie
1931 * (remembering both keys & values are urlencoded)
1932 *
1933 * @param string $pref     preference key
1934 * @param mixed  $default  value returned when preference not found
1935 * @return string preference value
1936 */
1937function get_doku_pref($pref, $default)
1938{
1939    $enc_pref = urlencode($pref);
1940    if(isset($_COOKIE['DOKU_PREFS']) && strpos($_COOKIE['DOKU_PREFS'], $enc_pref) !== false) {
1941        $parts = explode('#', $_COOKIE['DOKU_PREFS']);
1942        $cnt   = count($parts);
1943
1944        // due to #2721 there might be duplicate entries,
1945        // so we read from the end
1946        for($i = $cnt-2; $i >= 0; $i -= 2) {
1947            if($parts[$i] === $enc_pref) {
1948                return urldecode($parts[$i + 1]);
1949            }
1950        }
1951    }
1952    return $default;
1953}
1954
1955/**
1956 * Add a preference to the DokuWiki cookie
1957 * (remembering $_COOKIE['DOKU_PREFS'] is urlencoded)
1958 * Remove it by setting $val to false
1959 *
1960 * @param string $pref  preference key
1961 * @param string $val   preference value
1962 */
1963function set_doku_pref($pref, $val)
1964{
1965    global $conf;
1966    $orig = get_doku_pref($pref, false);
1967    $cookieVal = '';
1968
1969    if ($orig !== false && ($orig !== $val)) {
1970        $parts = explode('#', $_COOKIE['DOKU_PREFS']);
1971        $cnt   = count($parts);
1972        // urlencode $pref for the comparison
1973        $enc_pref = rawurlencode($pref);
1974        $seen = false;
1975        for ($i = 0; $i < $cnt; $i += 2) {
1976            if ($parts[$i] === $enc_pref) {
1977                if (!$seen){
1978                    if ($val !== false) {
1979                        $parts[$i + 1] = rawurlencode($val ?? '');
1980                    } else {
1981                        unset($parts[$i]);
1982                        unset($parts[$i + 1]);
1983                    }
1984                    $seen = true;
1985                } else {
1986                    // no break because we want to remove duplicate entries
1987                    unset($parts[$i]);
1988                    unset($parts[$i + 1]);
1989                }
1990            }
1991        }
1992        $cookieVal = implode('#', $parts);
1993    } elseif ($orig === false && $val !== false) {
1994        $cookieVal = (isset($_COOKIE['DOKU_PREFS']) ? $_COOKIE['DOKU_PREFS'] . '#' : '') .
1995            rawurlencode($pref) . '#' . rawurlencode($val);
1996    }
1997
1998    $cookieDir = empty($conf['cookiedir']) ? DOKU_REL : $conf['cookiedir'];
1999    if(defined('DOKU_UNITTEST')) {
2000        $_COOKIE['DOKU_PREFS'] = $cookieVal;
2001    }else{
2002        setcookie('DOKU_PREFS', $cookieVal, [
2003            'expires' => time() + 365 * 24 * 3600,
2004            'path' => $cookieDir,
2005            'secure' => ($conf['securecookie'] && is_ssl()),
2006            'samesite' => 'Lax'
2007        ]);
2008    }
2009}
2010
2011/**
2012 * Strips source mapping declarations from given text #601
2013 *
2014 * @param string &$text reference to the CSS or JavaScript code to clean
2015 */
2016function stripsourcemaps(&$text)
2017{
2018    $text = preg_replace('/^(\/\/|\/\*)[@#]\s+sourceMappingURL=.*?(\*\/)?$/im', '\\1\\2', $text);
2019}
2020
2021/**
2022 * Returns the contents of a given SVG file for embedding
2023 *
2024 * Inlining SVGs saves on HTTP requests and more importantly allows for styling them through
2025 * CSS. However it should used with small SVGs only. The $maxsize setting ensures only small
2026 * files are embedded.
2027 *
2028 * This strips unneeded headers, comments and newline. The result is not a vaild standalone SVG!
2029 *
2030 * @param string $file full path to the SVG file
2031 * @param int $maxsize maximum allowed size for the SVG to be embedded
2032 * @return string|false the SVG content, false if the file couldn't be loaded
2033 */
2034function inlineSVG($file, $maxsize = 2048)
2035{
2036    $file = trim($file);
2037    if($file === '') return false;
2038    if(!file_exists($file)) return false;
2039    if(filesize($file) > $maxsize) return false;
2040    if(!is_readable($file)) return false;
2041    $content = file_get_contents($file);
2042    $content = preg_replace('/<!--.*?(-->)/s', '', $content); // comments
2043    $content = preg_replace('/<\?xml .*?\?>/i', '', $content); // xml header
2044    $content = preg_replace('/<!DOCTYPE .*?>/i', '', $content); // doc type
2045    $content = preg_replace('/>\s+</s', '><', $content); // newlines between tags
2046    $content = trim($content);
2047    if(substr($content, 0, 5) !== '<svg ') return false;
2048    return $content;
2049}
2050
2051//Setup VIM: ex: et ts=2 :
2052