]> git.mxchange.org Git - friendica.git/blob - src/Content/Text/BBCode.php
Improve quote share ("message_id" added)
[friendica.git] / src / Content / Text / BBCode.php
1 <?php
2 /**
3  * @copyright Copyright (C) 2010-2022, the Friendica project
4  *
5  * @license GNU AGPL version 3 or any later version
6  *
7  * This program is free software: you can redistribute it and/or modify
8  * it under the terms of the GNU Affero General Public License as
9  * published by the Free Software Foundation, either version 3 of the
10  * License, or (at your option) any later version.
11  *
12  * This program is distributed in the hope that it will be useful,
13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
15  * GNU Affero General Public License for more details.
16  *
17  * You should have received a copy of the GNU Affero General Public License
18  * along with this program.  If not, see <https://www.gnu.org/licenses/>.
19  *
20  */
21
22 namespace Friendica\Content\Text;
23
24 use DOMDocument;
25 use DOMXPath;
26 use Exception;
27 use Friendica\Content\ContactSelector;
28 use Friendica\Content\Item;
29 use Friendica\Content\OEmbed;
30 use Friendica\Content\PageInfo;
31 use Friendica\Content\Smilies;
32 use Friendica\Core\Hook;
33 use Friendica\Core\Logger;
34 use Friendica\Core\Protocol;
35 use Friendica\Core\Renderer;
36 use Friendica\DI;
37 use Friendica\Model\Contact;
38 use Friendica\Model\Event;
39 use Friendica\Model\Photo;
40 use Friendica\Model\Post;
41 use Friendica\Model\Tag;
42 use Friendica\Network\HTTPClient\Client\HttpClientAccept;
43 use Friendica\Network\HTTPClient\Client\HttpClientOptions;
44 use Friendica\Object\Image;
45 use Friendica\Protocol\Activity;
46 use Friendica\Util\Images;
47 use Friendica\Util\Map;
48 use Friendica\Util\ParseUrl;
49 use Friendica\Util\Proxy;
50 use Friendica\Util\Strings;
51 use Friendica\Util\XML;
52
53 class BBCode
54 {
55         // Update this value to the current date whenever changes are made to BBCode::convert
56         const VERSION = '2021-07-28';
57
58         const INTERNAL = 0;
59         const EXTERNAL = 1;
60         const API = 2;
61         const DIASPORA = 3;
62         const CONNECTORS = 4;
63         const OSTATUS = 7;
64         const TWITTER = 8;
65         const BACKLINK = 8;
66         const ACTIVITYPUB = 9;
67
68         const TOP_ANCHOR = '<br class="top-anchor">';
69         const BOTTOM_ANCHOR = '<br class="button-anchor">';
70         /**
71          * Fetches attachment data that were generated the old way
72          *
73          * @param string $body Message body
74          * @return array
75          *                     'type' -> Message type ('link', 'video', 'photo')
76          *                     'text' -> Text before the shared message
77          *                     'after' -> Text after the shared message
78          *                     'image' -> Preview image of the message
79          *                     'url' -> Url to the attached message
80          *                     'title' -> Title of the attachment
81          *                     'description' -> Description of the attachment
82          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
83          */
84         private static function getOldAttachmentData(string $body): array
85         {
86                 $post = [];
87
88                 // Simplify image codes
89                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
90
91                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
92                         foreach ($attached as $data) {
93                                 if (!in_array($data[1], ['type-link', 'type-video', 'type-photo'])) {
94                                         continue;
95                                 }
96
97                                 $post['type'] = substr($data[1], 5);
98
99                                 $pos = strpos($body, $data[0]);
100                                 if ($pos > 0) {
101                                         $post['text'] = trim(substr($body, 0, $pos));
102                                         $post['after'] = trim(substr($body, $pos + strlen($data[0])));
103                                 } else {
104                                         $post['text'] = trim(str_replace($data[0], '', $body));
105                                         $post['after'] = '';
106                                 }
107
108                                 $attacheddata = $data[2];
109
110                                 if (preg_match("/\[img\](.*?)\[\/img\]/ism", $attacheddata, $matches)) {
111
112                                         $picturedata = Images::getInfoFromURLCached($matches[1]);
113
114                                         if ($picturedata) {
115                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
116                                                         $post['image'] = $matches[1];
117                                                 } else {
118                                                         $post['preview'] = $matches[1];
119                                                 }
120                                         }
121                                 }
122
123                                 if (preg_match("/\[bookmark\=(.*?)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
124                                         $post['url'] = $matches[1];
125                                         $post['title'] = $matches[2];
126                                 }
127                                 if (!empty($post['url']) && (in_array($post['type'], ['link', 'video']))
128                                         && preg_match("/\[url\=(.*?)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
129                                         $post['url'] = $matches[1];
130                                 }
131
132                                 // Search for description
133                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
134                                         $post['description'] = $matches[1];
135                                 }
136                         }
137                 }
138                 return $post;
139         }
140
141         /**
142          * Fetches attachment data that were generated with the "attachment" element
143          *
144          * @param string $body Message body
145          * @return array
146          *                     'type' -> Message type ('link', 'video', 'photo')
147          *                     'text' -> Text before the shared message
148          *                     'after' -> Text after the shared message
149          *                     'image' -> Preview image of the message
150          *                     'url' -> Url to the attached message
151          *                     'title' -> Title of the attachment
152          *                     'description' -> Description of the attachment
153          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
154          */
155         public static function getAttachmentData(string $body): array
156         {
157                 DI::profiler()->startRecording('rendering');
158                 $data = [
159                         'type'          => '',
160                         'text'          => '',
161                         'after'         => '',
162                         'image'         => null,
163                         'url'           => '',
164                         'author_name'   => '',
165                         'author_url'    => '',
166                         'provider_name' => '',
167                         'provider_url'  => '',
168                         'title'         => '',
169                         'description'   => '',
170                 ];
171
172                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
173                         DI::profiler()->stopRecording();
174                         return self::getOldAttachmentData($body);
175                 }
176
177                 $attributes = $match[2];
178
179                 $data['text'] = trim($match[1]);
180
181                 foreach (['type', 'url', 'title', 'image', 'preview', 'publisher_name', 'publisher_url', 'author_name', 'author_url'] as $field) {
182                         preg_match('/' . preg_quote($field, '/') . '=("|\')(.*?)\1/ism', $attributes, $matches);
183                         $value = $matches[2] ?? '';
184
185                         if ($value != '') {
186                                 switch ($field) {
187                                         case 'publisher_name':
188                                                 $data['provider_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
189                                                 break;
190
191                                         case 'publisher_url':
192                                                 $data['provider_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
193                                                 break;
194
195                                         case 'author_name':
196                                                 $data['author_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
197                                                 if ($data['provider_name'] == $data['author_name']) {
198                                                         $data['author_name'] = '';
199                                                 }
200                                                 break;
201
202                                         case 'author_url':
203                                                 $data['author_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
204                                                 if ($data['provider_url'] == $data['author_url']) {
205                                                         $data['author_url'] = '';
206                                                 }
207                                                 break;
208
209                                         case 'title':
210                                                 $value = self::convert(html_entity_decode($value, ENT_QUOTES, 'UTF-8'), false, true);
211                                                 $value = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
212                                                 $value = str_replace(['[', ']'], ['&#91;', '&#93;'], $value);
213                                                 $data['title'] = $value;
214
215                                         default:
216                                                 $data[$field] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
217                                                 break;
218                                 }
219                         }
220                 }
221
222                 if (!in_array($data['type'], ['link', 'audio', 'photo', 'video'])) {
223                         DI::profiler()->stopRecording();
224                         return [];
225                 }
226
227                 $data['description'] = trim($match[3]);
228
229                 $data['after'] = trim($match[4]);
230
231                 $parts = parse_url($data['url']);
232                 if (!empty($parts['scheme']) && !empty($parts['host'])) {
233                         if (empty($data['provider_name'])) {
234                                 $data['provider_name'] = $parts['host'];
235                         }
236                         if (empty($data['provider_url']) || empty(parse_url($data['provider_url'], PHP_URL_SCHEME))) {
237                                 $data['provider_url'] = $parts['scheme'] . '://' . $parts['host'];
238
239                                 if (!empty($parts['port'])) {
240                                         $data['provider_url'] .= ':' . $parts['port'];
241                                 }
242                         }
243                 }
244
245                 DI::profiler()->stopRecording();
246                 return $data;
247         }
248
249         public static function getAttachedData(string $body, array $item = []): array
250         {
251                 /*
252                 - text:
253                 - type: link, video, photo
254                 - title:
255                 - url:
256                 - image:
257                 - description:
258                 - (thumbnail)
259                 */
260
261                 DI::profiler()->startRecording('rendering');
262                 $has_title = !empty($item['title']);
263                 $plink = $item['plink'] ?? '';
264                 $post = self::getAttachmentData($body);
265
266                 // Get all linked images with alternative image description
267                 if (preg_match_all("/\[img=(http[^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
268                         foreach ($pictures as $picture) {
269                                 if (Photo::isLocal($picture[1])) {
270                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]];
271                                 } else {
272                                         $post['remote_images'][] = ['url' => $picture[1], 'description' => $picture[2]];
273                                 }
274                         }
275                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
276                                 $post['image_description'] = $post['images'][0]['description'];
277                         }
278                 }
279
280                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
281                         foreach ($pictures as $picture) {
282                                 if (Photo::isLocal($picture[1])) {
283                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
284                                 } else {
285                                         $post['remote_images'][] = ['url' => $picture[1], 'description' => ''];
286                                 }
287                         }
288                 }
289
290                 if (!isset($post['type'])) {
291                         $post['text'] = $body;
292                 }
293
294                 // Simplify image codes
295                 $post['text'] = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $post['text']);
296                 $post['text'] = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $post['text']);
297                 
298                 // if nothing is found, it maybe having an image.
299                 if (!isset($post['type'])) {
300                         if (preg_match_all("#\[url=([^\]]+?)\]\s*\[img\]([^\[]+?)\[/img\]\s*\[/url\]#ism", $post['text'], $pictures, PREG_SET_ORDER)) {
301                                 if ((count($pictures) == 1) && !$has_title) {
302                                         if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) {
303                                                 // Replace the preview picture with the real picture
304                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
305                                                 $data = ['url' => $url, 'type' => 'photo'];
306                                         } else {
307                                                 // Checking, if the link goes to a picture
308                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1]);
309                                         }
310
311                                         // Workaround:
312                                         // Sometimes photo posts to the own album are not detected at the start.
313                                         // So we seem to cannot use the cache for these cases. That's strange.
314                                         if (($data['type'] != 'photo') && strstr($pictures[0][1], '/photos/')) {
315                                                 $data = ParseUrl::getSiteinfo($pictures[0][1]);
316                                         }
317
318                                         if ($data['type'] == 'photo') {
319                                                 $post['type'] = 'photo';
320                                                 if (isset($data['images'][0])) {
321                                                         $post['image'] = $data['images'][0]['src'];
322                                                         $post['url'] = $data['url'];
323                                                 } else {
324                                                         $post['image'] = $data['url'];
325                                                 }
326
327                                                 $post['preview'] = $pictures[0][2];
328                                                 $post['text'] = trim(str_replace($pictures[0][0], '', $post['text']));
329                                         } else {
330                                                 $imgdata = Images::getInfoFromURLCached($pictures[0][1]);
331                                                 if (($imgdata) && substr($imgdata['mime'], 0, 6) == 'image/') {
332                                                         $post['type'] = 'photo';
333                                                         $post['image'] = $pictures[0][1];
334                                                         $post['preview'] = $pictures[0][2];
335                                                         $post['text'] = trim(str_replace($pictures[0][0], '', $post['text']));
336                                                 }
337                                         }
338                                 } elseif (count($pictures) > 0) {
339                                         if (count($pictures) > 4) {
340                                                 $post['type'] = 'link';
341                                                 $post['url'] = $plink;
342                                         } else {
343                                                 $post['type'] = 'photo';
344                                         }
345
346                                         $post['image'] = $pictures[0][2];
347
348                                         foreach ($pictures as $picture) {
349                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
350                                         }
351                                 }
352                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $post['text'], $pictures, PREG_SET_ORDER)) {
353                                 if ($has_title) {
354                                         $post['type'] = 'link';
355                                         $post['url'] = $plink;
356                                 } else {
357                                         $post['type'] = 'photo';
358                                 }
359
360                                 $post['image'] = $pictures[0][1];
361                                 foreach ($pictures as $picture) {
362                                         $post['text'] = trim(str_replace($picture[0], '', $post['text']));
363                                 }
364                         }
365
366                         // Test for the external links
367                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER);
368                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER);
369
370                         $links = array_merge($links1, $links2);
371
372                         // If there is only a single one, then use it.
373                         // This should cover link posts via API.
374                         if ((count($links) == 1) && !isset($post['preview']) && !$has_title) {
375                                 $post['type'] = 'link';
376                                 $post['url'] = $links[0][1];
377                         }
378
379                         // Simplify "video" element
380                         $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']);
381
382                         // Now count the number of external media links
383                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
384                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
385                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
386                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
387
388                         // Add them to the other external links
389                         $links = array_merge($links, $links1, $links2, $links3, $links4);
390
391                         // Are there more than one?
392                         if (count($links) > 1) {
393                                 // The post will be the type "text", which means a blog post
394                                 unset($post['type']);
395                                 $post['url'] = $plink;
396                         }
397
398                         if (!isset($post['type'])) {
399                                 $post['type'] = 'text';
400                         }
401
402                         if (($post['type'] == 'photo') && empty($post['images']) && !empty($post['remote_images'])) {
403                                 $post['images'] = $post['remote_images'];
404                                 $post['image'] = $post['images'][0]['url'];
405                                 if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
406                                         $post['image_description'] = $post['images'][0]['description'];
407                                 }
408                         }
409                         unset($post['remote_images']);
410                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
411                         $data = ParseUrl::getSiteinfoCached($post['url']);
412
413                         if (isset($data['images'][0])) {
414                                 $post['image'] = $data['images'][0]['src'];
415                         }
416                 } elseif (preg_match_all("#\[url=([^\]]+?)\]\s*\[img\]([^\[]+?)\[/img\]\s*\[/url\]#ism", $post['text'], $pictures, PREG_SET_ORDER)) {
417                         foreach ($pictures as $picture) {
418                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
419                         }
420                 }
421
422                 DI::profiler()->stopRecording();
423                 return $post;
424         }
425
426         /**
427          * Remove [attachment] BBCode and replaces it with a regular [url]
428          *
429          * @param string  $body
430          * @param boolean $no_link_desc No link description
431          * @return string with replaced body
432          */
433         public static function removeAttachment(string $body, bool $no_link_desc = false): string
434         {
435                 return preg_replace_callback("/\s*\[attachment (.*?)\](.*?)\[\/attachment\]\s*/ism",
436                         function ($match) use ($body, $no_link_desc) {
437                                 $attach_data = self::getAttachmentData($match[0]);
438                                 if (empty($attach_data['url'])) {
439                                         return $match[0];
440                                 } elseif (strpos(str_replace($match[0], '', $body), $attach_data['url']) !== false) {
441                                         return '';
442                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
443                                         return " \n[url]" . $attach_data['url'] . "[/url]\n";
444                                 } else {
445                                         return " \n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
446                                 }
447                 }, $body);
448         }
449
450         /**
451          * Converts a BBCode text into plaintext
452          *
453          * @param string $text
454          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
455          * @return string
456          */
457         public static function toPlaintext(string $text, bool $keep_urls = true): string
458         {
459                 DI::profiler()->startRecording('rendering');
460                 // Remove pictures in advance to avoid unneeded proxy calls
461                 $text = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $2 ', $text);
462                 $text = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $text);
463
464                 // Remove attachment
465                 $text = self::removeAttachment($text);
466
467                 $naked_text = HTML::toPlaintext(self::convert($text, false, 0, true), 0, !$keep_urls);
468
469                 DI::profiler()->stopRecording();
470                 return $naked_text;
471         }
472
473         private static function proxyUrl(string $image, int $simplehtml = self::INTERNAL, int $uriid = 0, string $size = ''): string
474         {
475                 // Only send proxied pictures to API and for internal display
476                 if (!in_array($simplehtml, [self::INTERNAL, self::API])) {
477                         return $image;
478                 } elseif ($uriid > 0) {
479                         return Post\Link::getByLink($uriid, $image, $size);
480                 } else {
481                         return Proxy::proxifyUrl($image, $size);
482                 }
483         }
484
485         /**
486          * This function changing the visual size (not the real size) of images.
487          * The function does not work for pictures with an alternate text description.
488          * This could only be changed by using some new "img" BBCode format.
489          *
490          * @param string $srctext The body with images
491          * @return string The body with possibly scaled images
492          */
493         public static function scaleExternalImages(string $srctext): string
494         {
495                 DI::profiler()->startRecording('rendering');
496                 $s = $srctext;
497
498                 // Simplify image links
499                 $s = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $s);
500
501                 $matches = null;
502                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
503                 if ($c) {
504                         foreach ($matches as $mtch) {
505                                 Logger::debug('scale_external_image', ['image' => $mtch[1]]);
506
507                                 $hostname = str_replace('www.', '', substr(DI::baseUrl(), strpos(DI::baseUrl(), '://') + 3));
508                                 if (stristr($mtch[1], $hostname)) {
509                                         continue;
510                                 }
511
512                                 $curlResult = DI::httpClient()->get($mtch[1], HttpClientAccept::IMAGE);
513                                 if (!$curlResult->isSuccess()) {
514                                         continue;
515                                 }
516
517                                 Logger::debug('Got picture', ['Content-Type' => $curlResult->getHeader('Content-Type'), 'url' => $mtch[1]]);
518
519                                 $i = $curlResult->getBody();
520                                 $type = $curlResult->getContentType();
521                                 $type = Images::getMimeTypeByData($i, $mtch[1], $type);
522
523                                 if ($i) {
524                                         $Image = new Image($i, $type);
525                                         if ($Image->isValid()) {
526                                                 $orig_width = $Image->getWidth();
527                                                 $orig_height = $Image->getHeight();
528
529                                                 if ($orig_width > 640 || $orig_height > 640) {
530                                                         $Image->scaleDown(640);
531                                                         $new_width = $Image->getWidth();
532                                                         $new_height = $Image->getHeight();
533                                                         Logger::debug('External images scaled', ['orig_width' => $orig_width, 'new_width' => $new_width, 'orig_height' => $orig_height, 'new_height' => $new_height, 'match' => $mtch[0]]);
534                                                         $s = str_replace(
535                                                                 $mtch[0],
536                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]'
537                                                                 . "\n",
538                                                                 $s
539                                                         );
540                                                         Logger::debug('New string', ['image' => $s]);
541                                                 }
542                                         }
543                                 }
544                         }
545                 }
546
547                 DI::profiler()->stopRecording();
548                 return $s;
549         }
550
551         /**
552          * Truncates imported message body string length to max_import_size
553          *
554          * The purpose of this function is to apply system message length limits to
555          * imported messages without including any embedded photos in the length
556          *
557          * @param string $body
558          * @return string
559          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
560          */
561         public static function limitBodySize(string $body): string
562         {
563                 DI::profiler()->startRecording('rendering');
564                 $maxlen = DI::config()->get('config', 'max_import_size', 0);
565
566                 // If the length of the body, including the embedded images, is smaller
567                 // than the maximum, then don't waste time looking for the images
568                 if ($maxlen && (strlen($body) > $maxlen)) {
569
570                         Logger::info('the total body length exceeds the limit', ['maxlen' => $maxlen, 'body_len' => strlen($body)]);
571
572                         $orig_body = $body;
573                         $new_body = '';
574                         $textlen = 0;
575
576                         $img_start = strpos($orig_body, '[img');
577                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
578                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
579                         while (($img_st_close !== false) && ($img_end !== false)) {
580
581                                 $img_st_close++; // make it point to AFTER the closing bracket
582                                 $img_end += $img_start;
583                                 $img_end += strlen('[/img]');
584
585                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
586                                         // This is an embedded image
587
588                                         if (($textlen + $img_start) > $maxlen) {
589                                                 if ($textlen < $maxlen) {
590                                                         Logger::debug('the limit happens before an embedded image');
591                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
592                                                         $textlen = $maxlen;
593                                                 }
594                                         } else {
595                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
596                                                 $textlen += $img_start;
597                                         }
598
599                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
600                                 } else {
601
602                                         if (($textlen + $img_end) > $maxlen) {
603                                                 if ($textlen < $maxlen) {
604                                                         Logger::debug('the limit happens before the end of a non-embedded image');
605                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
606                                                         $textlen = $maxlen;
607                                                 }
608                                         } else {
609                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
610                                                 $textlen += $img_end;
611                                         }
612                                 }
613                                 $orig_body = substr($orig_body, $img_end);
614
615                                 if ($orig_body === false) {
616                                         // in case the body ends on a closing image tag
617                                         $orig_body = '';
618                                 }
619
620                                 $img_start = strpos($orig_body, '[img');
621                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
622                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
623                         }
624
625                         if (($textlen + strlen($orig_body)) > $maxlen) {
626                                 if ($textlen < $maxlen) {
627                                         Logger::debug('the limit happens after the end of the last image');
628                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
629                                 }
630                         } else {
631                                 Logger::debug('the text size with embedded images extracted did not violate the limit');
632                                 $new_body = $new_body . $orig_body;
633                         }
634
635                         DI::profiler()->stopRecording();
636                         return $new_body;
637                 } else {
638                         DI::profiler()->stopRecording();
639                         return $body;
640                 }
641         }
642
643         /**
644          * Processes [attachment] tags
645          *
646          * Note: Can produce a [bookmark] tag in the returned string
647          *
648          * @param string  $text
649          * @param integer $simplehtml
650          * @param bool    $tryoembed
651          * @param array   $data
652          * @param int     $uriid
653          * @return string
654          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
655          */
656         public static function convertAttachment(string $text, int $simplehtml = self::INTERNAL, bool $tryoembed = true, array $data = [], int $uriid = 0): string
657         {
658                 DI::profiler()->startRecording('rendering');
659                 $data = $data ?: self::getAttachmentData($text);
660                 if (empty($data) || empty($data['url'])) {
661                         DI::profiler()->stopRecording();
662                         return $text;
663                 }
664
665                 if (isset($data['title'])) {
666                         $data['title'] = strip_tags($data['title']);
667                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
668                 } else {
669                         $data['title'] = '';
670                 }
671
672                 if (((strpos($data['text'], '[img=') !== false) || (strpos($data['text'], '[img]') !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) {
673                         $data['preview'] = $data['image'];
674                         $data['image'] = '';
675                 }
676
677                 $return = '';
678                 try {
679                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
680                                 $return = OEmbed::getHTML($data['url'], $data['title']);
681                         } else {
682                                 throw new Exception('OEmbed is disabled for this attachment.');
683                         }
684                 } catch (Exception $e) {
685                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
686
687                         if ($simplehtml != self::CONNECTORS) {
688                                 $return = sprintf('<div class="type-%s">', $data['type']);
689                         }
690
691                         if (!empty($data['title']) && !empty($data['url'])) {
692                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
693                                         $return .= sprintf('<a href="%s" target="_blank" rel="noopener noreferrer"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data['url'], self::proxyUrl($data['image'], $simplehtml, $uriid), $data['title']);
694                                 } else {
695                                         if (!empty($data['image'])) {
696                                                 $return .= sprintf('<a href="%s" target="_blank" rel="noopener noreferrer"><img src="%s" alt="" title="%s" class="attachment-image" /></a><br>', $data['url'], self::proxyUrl($data['image'], $simplehtml, $uriid), $data['title']);
697                                         } elseif (!empty($data['preview'])) {
698                                                 $return .= sprintf('<a href="%s" target="_blank" rel="noopener noreferrer"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br>', $data['url'], self::proxyUrl($data['preview'], $simplehtml, $uriid), $data['title']);
699                                         }
700                                         $return .= sprintf('<h4><a href="%s" target="_blank" rel="noopener noreferrer">%s</a></h4>', $data['url'], $data['title']);
701                                 }
702                         }
703
704                         if (!empty($data['description']) && $data['description'] != $data['title']) {
705                                 // Sanitize the HTML
706                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(HTML::purify($data['description'])));
707                         }
708
709                         if (!empty($data['provider_url']) && !empty($data['provider_name'])) {
710                                 if (!empty($data['author_name'])) {
711                                         $return .= sprintf('<sup><a href="%s" target="_blank" rel="noopener noreferrer">%s (%s)</a></sup>', $data['provider_url'], $data['author_name'], $data['provider_name']);
712                                 } else {
713                                         $return .= sprintf('<sup><a href="%s" target="_blank" rel="noopener noreferrer">%s</a></sup>', $data['provider_url'], $data['provider_name']);
714                                 }
715                         }
716
717                         if ($simplehtml != self::CONNECTORS) {
718                                 $return .= '</div>';
719                         }
720                 }
721
722                 DI::profiler()->stopRecording();
723                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
724         }
725
726         public static function removeShareInformation(string $text, bool $plaintext = false, bool $nolink = false): string
727         {
728                 DI::profiler()->startRecording('rendering');
729                 $data = self::getAttachmentData($text);
730
731                 if (!$data) {
732                         DI::profiler()->stopRecording();
733                         return $text;
734                 } elseif ($nolink) {
735                         DI::profiler()->stopRecording();
736                         return $data['text'] . ($data['after'] ?? '');
737                 }
738
739                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
740                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
741                 if ($plaintext || (($title != '') && strstr($text, $title))) {
742                         $data['title'] = $data['url'];
743                 } elseif (($text != '') && strstr($title, $text)) {
744                         $data['text'] = $data['title'];
745                         $data['title'] = $data['url'];
746                 }
747
748                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
749                         DI::profiler()->stopRecording();
750                         return $data['title'] . $data['after'];
751                 }
752
753                 // If the link already is included in the post, don't add it again
754                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
755                         DI::profiler()->stopRecording();
756                         return $data['text'] . $data['after'];
757                 }
758
759                 $text = $data['text'];
760
761                 if (!empty($data['url']) && !empty($data['title'])) {
762                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
763                 } elseif (!empty($data['url'])) {
764                         $text .= "\n[url]" . $data['url'] . '[/url]';
765                 }
766
767                 DI::profiler()->stopRecording();
768                 return $text . "\n" . $data['after'];
769         }
770
771         /**
772          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
773          *
774          * @param array $match Array with the matching values
775          * @return string reformatted link including HTML codes
776          */
777         private static function convertUrlForActivityPubCallback(array $match): string
778         {
779                 $url = $match[1];
780
781                 if (isset($match[2]) && ($match[1] != $match[2])) {
782                         return $match[0];
783                 }
784
785                 $parts = parse_url($url);
786                 if (!isset($parts['scheme'])) {
787                         return $match[0];
788                 }
789
790                 return self::convertUrlForActivityPub($url);
791         }
792
793         /**
794          * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
795          *
796          * @param string $url URL that is about to be reformatted
797          * @return string reformatted link including HTML codes
798          */
799         private static function convertUrlForActivityPub(string $url): string
800         {
801                 return sprintf('<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>', $url, self::getStyledURL($url));
802         }
803
804         /**
805          * Converts an URL in a nicer format (without the scheme and possibly shortened)
806          *
807          * @param string $url URL that is about to be reformatted
808          * @return string reformatted link
809          */
810         private static function getStyledURL(string $url): string
811         {
812                 $parts = parse_url($url);
813                 $scheme = $parts['scheme'] . '://';
814                 $styled_url = str_replace($scheme, '', $url);
815
816                 if (strlen($styled_url) > 30) {
817                         $styled_url = substr($styled_url, 0, 30) . "…";
818                 }
819
820                 return $styled_url;
821         }
822
823         /*
824          * [noparse][i]italic[/i][/noparse] turns into
825          * [noparse][ i ]italic[ /i ][/noparse],
826          * to hide them from parser.
827          *
828          * @param array $match
829          * @return string
830          */
831         private static function escapeNoparseCallback(array $match): string
832         {
833                 $whole_match = $match[0];
834                 $captured = $match[1];
835                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
836                 $new_str = str_replace($captured, $spacefied, $whole_match);
837                 return $new_str;
838         }
839
840         /*
841          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
842          * now turns back and the [noparse] tags are trimed
843          * returning [i]italic[/i]
844          *
845          * @param array $match
846          * @return string
847          */
848         private static function unescapeNoparseCallback(array $match): string
849         {
850                 $captured = $match[1];
851                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
852                 return $unspacefied;
853         }
854
855         /**
856          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
857          * occurrences
858          *
859          * @param string $text        Text to search
860          * @param string $name        Tag name
861          * @param int    $occurrences Number of first occurrences to skip
862          * @return boolean|array
863          */
864         public static function getTagPosition(string $text, string $name, int $occurrences = 0)
865         {
866                 DI::profiler()->startRecording('rendering');
867                 if ($occurrences < 0) {
868                         $occurrences = 0;
869                 }
870
871                 $start_open = -1;
872                 for ($i = 0; $i <= $occurrences; $i++) {
873                         if ($start_open !== false) {
874                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
875                         }
876                 }
877
878                 if ($start_open === false) {
879                         DI::profiler()->stopRecording();
880                         return false;
881                 }
882
883                 $start_equal = strpos($text, '=', $start_open);
884                 $start_close = strpos($text, ']', $start_open);
885
886                 if ($start_close === false) {
887                         DI::profiler()->stopRecording();
888                         return false;
889                 }
890
891                 $start_close++;
892
893                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
894
895                 if ($end_open === false) {
896                         DI::profiler()->stopRecording();
897                         return false;
898                 }
899
900                 $res = [
901                         'start' => [
902                                 'open' => $start_open,
903                                 'close' => $start_close
904                         ],
905                         'end' => [
906                                 'open' => $end_open,
907                                 'close' => $end_open + strlen('[/' . $name . ']')
908                         ],
909                 ];
910
911                 if ($start_equal !== false) {
912                         $res['start']['equal'] = $start_equal + 1;
913                 }
914
915                 DI::profiler()->stopRecording();
916                 return $res;
917         }
918
919         /**
920          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
921          *
922          * @param string $pattern Preg pattern string
923          * @param string $replace Preg replace string
924          * @param string $name    BBCode tag name
925          * @param string $text    Text to search
926          * @return string
927          */
928         public static function pregReplaceInTag(string $pattern, string $replace, string $name, string $text): string
929         {
930                 DI::profiler()->startRecording('rendering');
931                 $occurrences = 0;
932                 $pos = self::getTagPosition($text, $name, $occurrences);
933                 while ($pos !== false && $occurrences++ < 1000) {
934                         $start = substr($text, 0, $pos['start']['open']);
935                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
936                         $end = substr($text, $pos['end']['close']);
937                         if ($end === false) {
938                                 $end = '';
939                         }
940
941                         $subject = preg_replace($pattern, $replace, $subject);
942                         $text = $start . $subject . $end;
943
944                         $pos = self::getTagPosition($text, $name, $occurrences);
945                 }
946
947                 DI::profiler()->stopRecording();
948                 return $text;
949         }
950
951         private static function extractImagesFromItemBody(string $body): array
952         {
953                 $saved_image = [];
954                 $orig_body = $body;
955                 $new_body = '';
956
957                 $cnt = 0;
958                 $img_start = strpos($orig_body, '[img');
959                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
960                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
961                 while (($img_st_close !== false) && ($img_end !== false)) {
962                         $img_st_close++; // make it point to AFTER the closing bracket
963                         $img_end += $img_start;
964
965                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
966                                 // This is an embedded image
967                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
968                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
969
970                                 $cnt++;
971                         } else {
972                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
973                         }
974
975                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
976
977                         if ($orig_body === false) {
978                                 // in case the body ends on a closing image tag
979                                 $orig_body = '';
980                         }
981
982                         $img_start = strpos($orig_body, '[img');
983                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
984                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
985                 }
986
987                 $new_body = $new_body . $orig_body;
988
989                 return ['body' => $new_body, 'images' => $saved_image];
990         }
991
992         private static function interpolateSavedImagesIntoItemBody(int $uriid, string $body, array $images): string
993         {
994                 $newbody = $body;
995
996                 $cnt = 0;
997                 foreach ($images as $image) {
998                         // We're depending on the property of 'foreach' (specified on the PHP website) that
999                         // it loops over the array starting from the first element and going sequentially
1000                         // to the last element
1001                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
1002                                 '<img src="' . self::proxyUrl($image, self::INTERNAL, $uriid) . '" alt="' . DI::l10n()->t('Image/photo') . '" />', $newbody);
1003                         $cnt++;
1004                 }
1005
1006                 return $newbody;
1007         }
1008
1009         /**
1010          * @param string $text A BBCode string
1011          * @return array Empty array if no share tag is present or the following array, missing attributes end up empty strings:
1012          *               - comment: Text before the opening share tag
1013          *               - shared : Text inside the share tags
1014          *               - author : (Optional) Display name of the shared author
1015          *               - profile: (Optional) Profile page URL of the shared author
1016          *               - avatar : (Optional) Profile picture URL of the shared author
1017          *               - link   : (Optional) Canonical URL of the shared post
1018          *               - posted : (Optional) Date the shared post was initially posted ("Y-m-d H:i:s" in GMT)
1019          *               - guid   : (Optional) Shared post GUID if any
1020          */
1021         public static function fetchShareAttributes(string $text): array
1022         {
1023                 DI::profiler()->startRecording('rendering');
1024                 // See Issue https://github.com/friendica/friendica/issues/10454
1025                 // Hashtags in usernames are expanded to links. This here is a quick fix.
1026                 $text = preg_replace('~([@!#])\[url=.*?](.*?)\[/url]~ism', '$1$2', $text);
1027
1028                 if (!preg_match('~(.*?)\[share(.*?)](.*)\[/share]~ism', $text, $matches)) {
1029                         DI::profiler()->stopRecording();
1030                         return [];
1031                 }
1032
1033                 $attributes = self::extractShareAttributes($matches[2]);
1034
1035                 $attributes['comment'] = trim($matches[1]);
1036                 $attributes['shared'] = trim($matches[3]);
1037
1038                 DI::profiler()->stopRecording();
1039                 return $attributes;
1040         }
1041
1042         /**
1043          * @see BBCode::fetchShareAttributes()
1044          * @param string $shareString Internal opening share tag string matched by the regular expression
1045          * @return array A fixed attribute array where missing attribute are represented by empty strings
1046          */
1047         private static function extractShareAttributes(string $shareString): array
1048         {
1049                 $attributes = [];
1050                 foreach (['author', 'profile', 'avatar', 'link', 'posted', 'guid', 'message_id'] as $field) {
1051                         preg_match("/$field=(['\"])(.+?)\\1/ism", $shareString, $matches);
1052                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
1053                 }
1054
1055                 return $attributes;
1056         }
1057
1058         /**
1059          * This function converts a [share] block to text according to a provided callback function whose signature is:
1060          *
1061          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
1062          *
1063          * Where:
1064          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
1065          * data lookup
1066          * - $author_contact is a contact record array
1067          * - $content is the inner content of the [share] block
1068          * - $is_quote_share indicates whether there's any content before the [share] block
1069          * - Return value is the string that should replace the [share] block in the provided text
1070          *
1071          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
1072          *
1073          * @param  string   $text     A BBCode string
1074          * @param  callable $callback
1075          * @return string The BBCode string with all [share] blocks replaced
1076          */
1077         public static function convertShare(string $text, callable $callback, int $uriid = 0): string
1078         {
1079                 DI::profiler()->startRecording('rendering');
1080                 $return = preg_replace_callback(
1081                         '~(.*?)\[share(.*?)](.*)\[/share]~ism',
1082                         function ($match) use ($callback, $uriid) {
1083                                 $attributes = self::extractShareAttributes($match[2]);
1084
1085                                 $author_contact = Contact::getByURL($attributes['profile'], false, ['id', 'url', 'addr', 'name', 'micro']);
1086                                 $author_contact['url'] = ($author_contact['url'] ?? $attributes['profile']);
1087                                 $author_contact['addr'] = ($author_contact['addr'] ?? '');
1088
1089                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
1090                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
1091                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
1092
1093                                 if (!empty($author_contact['id'])) {
1094                                         $attributes['avatar'] = Contact::getAvatarUrlForId($author_contact['id'], Proxy::SIZE_THUMB);
1095                                 } elseif ($attributes['avatar']) {
1096                                         $attributes['avatar'] = self::proxyUrl($attributes['avatar'], self::INTERNAL, $uriid, Proxy::SIZE_THUMB);
1097                                 }
1098
1099                                 $content = preg_replace(Strings::autoLinkRegEx(), '<a href="$1">$1</a>', $match[3]);
1100
1101                                 return $match[1] . $callback($attributes, $author_contact, $content, trim($match[1]) != '');
1102                         },
1103                         $text
1104                 );
1105
1106                 DI::profiler()->stopRecording();
1107                 return $return;
1108         }
1109
1110         /**
1111          * Convert complex IMG and ZMG elements
1112          *
1113          * @param [type] $text
1114          * @param integer $simplehtml
1115          * @param integer $uriid
1116          * @return string
1117          */
1118         private static function convertImages(string $text, int $simplehtml, int $uriid = 0): string
1119         {
1120                 DI::profiler()->startRecording('rendering');
1121                 $return = preg_replace_callback(
1122                         "/\[[zi]mg(.*?)\]([^\[\]]*)\[\/[zi]mg\]/ism",
1123                         function ($match) use ($simplehtml, $uriid) {
1124                                 $attribute_string = $match[1];
1125                                 $attributes = [];
1126                                 foreach (['alt', 'width', 'height'] as $field) {
1127                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
1128                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
1129                                 }
1130
1131                                 $img_str = '<img src="' .
1132                                 self::proxyUrl($match[2], $simplehtml, $uriid) . '"';
1133                                 foreach ($attributes as $key => $value) {
1134                                         if (!empty($value)) {
1135                                                 $img_str .= ' ' . $key . '="' . htmlspecialchars($value, ENT_COMPAT) . '"';
1136                                         }
1137                                 }
1138                                 return $img_str . '>';
1139                         },
1140                         $text
1141                 );
1142
1143                 DI::profiler()->stopRecording();
1144                 return $return;
1145         }
1146
1147         /**
1148          * Default [share] tag conversion callback
1149          *
1150          * Note: Can produce a [bookmark] tag in the output
1151          *
1152          * @see BBCode::convertShare()
1153          * @param array   $attributes     [share] block attribute values
1154          * @param array   $author_contact Contact row of the shared author
1155          * @param string  $content        Inner content of the [share] block
1156          * @param boolean $is_quote_share Whether there is content before the [share] block
1157          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
1158          * @return string
1159          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1160          */
1161         private static function convertShareCallback(array $attributes, array $author_contact, string $content, bool $is_quote_share, int $simplehtml): string
1162         {
1163                 DI::profiler()->startRecording('rendering');
1164                 $mention = $attributes['author'] . ' (' . ($author_contact['addr'] ?? '') . ')';
1165
1166                 switch ($simplehtml) {
1167                         case self::API:
1168                                 $text = ($is_quote_share? '<br>' : '') .
1169                                 '<b><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . "</a>:</b><br>\n" .
1170                                 '<blockquote class="shared_content" dir="auto">' . $content . '</blockquote>';
1171                                 break;
1172                         case self::DIASPORA:
1173                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1174                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1175                                 } else {
1176                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1177
1178                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1179                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1180                                         }
1181
1182                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1183
1184                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1185                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1186                                         }
1187                                 }
1188
1189                                 break;
1190                         case self::CONNECTORS:
1191                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1192                                 $headline .= DI::l10n()->t('<a href="%1$s" target="_blank" rel="noopener noreferrer">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1193                                 $headline .= ':</b></p>' . "\n";
1194
1195                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content" dir="auto">' . trim($content) . '</blockquote>' . "\n";
1196
1197                                 break;
1198                         case self::OSTATUS:
1199                                 $text = ($is_quote_share? '<br>' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1200                                 break;
1201                         case self::ACTIVITYPUB:
1202                                 $author = '@<span class="vcard"><a href="' . $author_contact['url'] . '" class="url u-url mention" title="' . $author_contact['addr'] . '"><span class="fn nickname mention">' . $author_contact['addr'] . '</span></a>:</span>';
1203                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1204                                 break;
1205                         default:
1206                                 $text = ($is_quote_share? "\n" : '');
1207
1208                                 $contact = Contact::getByURL($attributes['profile'], false, ['network']);
1209                                 $network = $contact['network'] ?? Protocol::PHANTOM;
1210
1211                                 $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1212                                 $text .= Renderer::replaceMacros($tpl, [
1213                                         '$profile'      => $attributes['profile'],
1214                                         '$avatar'       => $attributes['avatar'],
1215                                         '$author'       => $attributes['author'],
1216                                         '$link'         => $attributes['link'],
1217                                         '$link_title'   => DI::l10n()->t('Link to source'),
1218                                         '$posted'       => $attributes['posted'],
1219                                         '$guid'         => $attributes['guid'],
1220                                         '$network_name' => ContactSelector::networkToName($network, $attributes['profile']),
1221                                         '$network_icon' => ContactSelector::networkToIcon($network, $attributes['profile']),
1222                                         '$content'      => self::TOP_ANCHOR . self::setMentions(trim($content), 0, $network) . self::BOTTOM_ANCHOR,
1223                                 ]);
1224                                 break;
1225                 }
1226
1227                 return $text;
1228         }
1229
1230         private static function removePictureLinksCallback(array $match): string
1231         {
1232                 $cache_key = 'remove:' . $match[1];
1233                 $text = DI::cache()->get($cache_key);
1234
1235                 if (is_null($text)) {
1236                         $curlResult = DI::httpClient()->head($match[1], [HttpClientOptions::TIMEOUT => DI::config()->get('system', 'xrd_timeout')]);
1237                         if ($curlResult->isSuccess()) {
1238                                 $mimetype = $curlResult->getHeader('Content-Type')[0] ?? '';
1239                         } else {
1240                                 $mimetype = '';
1241                         }
1242
1243                         if (substr($mimetype, 0, 6) == 'image/') {
1244                                 $text = '[url=' . $match[1] . ']' . $match[1] . '[/url]';
1245                         } else {
1246                                 $text = '[url=' . $match[2] . ']' . $match[2] . '[/url]';
1247
1248                                 // if its not a picture then look if its a page that contains a picture link
1249                                 $body = DI::httpClient()->fetch($match[1], HttpClientAccept::HTML, 0);
1250                                 if (empty($body)) {
1251                                         DI::cache()->set($cache_key, $text);
1252                                         return $text;
1253                                 }
1254
1255                                 $doc = new DOMDocument();
1256                                 @$doc->loadHTML($body);
1257                                 $xpath = new DOMXPath($doc);
1258                                 $list = $xpath->query('//meta[@name]');
1259                                 foreach ($list as $node) {
1260                                         $attr = [];
1261
1262                                         if ($node->attributes->length) {
1263                                                 foreach ($node->attributes as $attribute) {
1264                                                         $attr[$attribute->name] = $attribute->value;
1265                                                 }
1266                                         }
1267
1268                                         if (strtolower($attr['name']) == 'twitter:image') {
1269                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1270                                         }
1271                                 }
1272                         }
1273                         DI::cache()->set($cache_key, $text);
1274                 }
1275
1276                 return $text;
1277         }
1278
1279         /**
1280          * Callback: Expands links from given $match array
1281          *
1282          * @param arrat $match Array with link match
1283          * @return string BBCode
1284          */
1285         private static function expandLinksCallback(array $match): string
1286         {
1287                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1288                         return ($match[1] . '[url]' . $match[2] . '[/url]');
1289                 } else {
1290                         return ($match[1] . $match[3] . ' [url]' . $match[2] . '[/url]');
1291                 }
1292         }
1293
1294         /**
1295          * Callback: Cleans picture links
1296          *
1297          * @param arrat $match Array with link match
1298          * @return string BBCode
1299          */
1300         private static function cleanPictureLinksCallback(array $match): string
1301         {
1302                 // When the picture link is the own photo path then we can avoid fetching the link
1303                 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()->get()) . '/photos/');
1304                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1305                         if (!empty($match[3])) {
1306                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1307                         } else {
1308                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1309                         }
1310                         return $text;
1311                 }
1312
1313                 $cache_key = 'clean:' . $match[1];
1314                 $text = DI::cache()->get($cache_key);
1315                 if (!is_null($text)) {
1316                         return $text;
1317                 }
1318
1319                 $curlResult = DI::httpClient()->head($match[1], [HttpClientOptions::TIMEOUT => DI::config()->get('system', 'xrd_timeout')]);
1320                 if ($curlResult->isSuccess()) {
1321                         $mimetype = $curlResult->getHeader('Content-Type')[0] ?? '';
1322                 } else {
1323                         $mimetype = '';
1324                 }
1325
1326                 // if its a link to a picture then embed this picture
1327                 if (substr($mimetype, 0, 6) == 'image/') {
1328                         $text = '[img]' . $match[1] . '[/img]';
1329                 } else {
1330                         if (!empty($match[3])) {
1331                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1332                         } else {
1333                                 $text = '[img]' . $match[2] . '[/img]';
1334                         }
1335
1336                         // if its not a picture then look if its a page that contains a picture link
1337                         $body = DI::httpClient()->fetch($match[1], HttpClientAccept::HTML, 0);
1338                         if (empty($body)) {
1339                                 DI::cache()->set($cache_key, $text);
1340                                 return $text;
1341                         }
1342
1343                         $doc = new DOMDocument();
1344                         @$doc->loadHTML($body);
1345                         $xpath = new DOMXPath($doc);
1346                         $list = $xpath->query('//meta[@name]');
1347                         foreach ($list as $node) {
1348                                 $attr = [];
1349                                 if ($node->attributes->length) {
1350                                         foreach ($node->attributes as $attribute) {
1351                                                 $attr[$attribute->name] = $attribute->value;
1352                                         }
1353                                 }
1354
1355                                 if (strtolower($attr['name']) == "twitter:image") {
1356                                         if (!empty($match[3])) {
1357                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1358                                         } else {
1359                                                 $text = "[img]" . $attr['content'] . "[/img]";
1360                                         }
1361                                 }
1362                         }
1363                 }
1364                 DI::cache()->set($cache_key, $text);
1365
1366                 return $text;
1367         }
1368
1369         /**
1370          * Cleans picture links
1371          *
1372          * @param string $text HTML/BBCode string
1373          * @return string Cleaned HTML/BBCode
1374          */
1375         public static function cleanPictureLinks(string $text): string
1376         {
1377                 DI::profiler()->startRecording('rendering');
1378                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1379                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1380                 DI::profiler()->stopRecording();
1381                 return $return;
1382         }
1383
1384         /**
1385          * Removes links
1386          *
1387          * @param string $text HTML/BBCode string
1388          * @return string Cleaned HTML/BBCode
1389          */
1390         public static function removeLinks(string $bbcode): string
1391         {
1392                 DI::profiler()->startRecording('rendering');
1393                 $bbcode = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $1 ', $bbcode);
1394                 $bbcode = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $bbcode);
1395
1396                 $bbcode = preg_replace('/[@!#]\[url\=.*?\].*?\[\/url\]/ism', '', $bbcode);
1397                 $bbcode = preg_replace("/\[url=[^\[\]]*\](.*)\[\/url\]/Usi", ' $1 ', $bbcode);
1398                 $bbcode = preg_replace('/[@!#]?\[url.*?\[\/url\]/ism', '', $bbcode);
1399                 DI::profiler()->stopRecording();
1400                 return $bbcode;
1401         }
1402
1403         /**
1404          * Replace names in mentions with nicknames
1405          *
1406          * @param string $body HTML/BBCode
1407          * @return string Body with replaced mentions
1408          */
1409         public static function setMentionsToNicknames(string $body): string
1410         {
1411                 DI::profiler()->startRecording('rendering');
1412                 $regexp = "/([@!])\[url\=([^\[\]]*)\].*?\[\/url\]/ism";
1413                 $body = preg_replace_callback($regexp, ['self', 'mentionCallback'], $body);
1414                 DI::profiler()->stopRecording();
1415                 return $body;
1416         }
1417
1418         /**
1419          * Callback function to replace a Friendica style mention in a mention with the nickname
1420          *
1421          * @param array $match Matching values for the callback
1422          * @return string Replaced mention or empty string
1423          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1424          */
1425         private static function mentionCallback(array $match): string
1426         {
1427                 if (empty($match[2])) {
1428                         return '';
1429                 }
1430
1431                 $data = Contact::getByURL($match[2], false, ['url', 'nick']);
1432                 if (empty($data['nick'])) {
1433                         return $match[0];
1434                 }
1435
1436                 return $match[1] . '[url=' . $data['url'] . ']' . $data['nick'] . '[/url]';
1437         }
1438
1439         /**
1440          * Converts a BBCode message for a given URI-ID to a HTML message
1441          *
1442          * BBcode 2 HTML was written by WAY2WEB.net
1443          * extended to work with Mistpark/Friendica - Mike Macgirvin
1444          *
1445          * Simple HTML values meaning:
1446          * - 0: Friendica display
1447          * - 1: Unused
1448          * - 2: Used for Windows Phone push, Friendica API
1449          * - 3: Used before converting to Markdown in bb2diaspora.php
1450          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1451          * - 5: Unused
1452          * - 6: Unused
1453          * - 7: Used for dfrn, OStatus
1454          * - 8: Used for Twitter, WP backlink text setting
1455          * - 9: ActivityPub
1456          *
1457          * @param int    $uriid
1458          * @param string $text
1459          * @param int    $simple_html
1460          * @return string
1461          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1462          */
1463         public static function convertForUriId(int $uriid = null, string $text = null, int $simple_html = self::INTERNAL): string
1464         {
1465                 $try_oembed = ($simple_html == self::INTERNAL);
1466
1467                 return self::convert($text ?? '', $try_oembed, $simple_html, false, $uriid ?? 0);
1468         }
1469
1470         /**
1471          * Converts a BBCode message to HTML message
1472          *
1473          * BBcode 2 HTML was written by WAY2WEB.net
1474          * extended to work with Mistpark/Friendica - Mike Macgirvin
1475          *
1476          * Simple HTML values meaning:
1477          * - 0: Friendica display
1478          * - 1: Unused
1479          * - 2: Used for Windows Phone push, Friendica API
1480          * - 3: Used before converting to Markdown in bb2diaspora.php
1481          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1482          * - 5: Unused
1483          * - 6: Unused
1484          * - 7: Used for dfrn, OStatus
1485          * - 8: Used for Twitter, WP backlink text setting
1486          * - 9: ActivityPub
1487          *
1488          * @param string $text
1489          * @param bool   $try_oembed
1490          * @param int    $simple_html
1491          * @param bool   $for_plaintext
1492          * @param int    $uriid
1493          * @return string Converted code or empty string
1494          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1495          */
1496         public static function convert(string $text = null, bool $try_oembed = true, int $simple_html = self::INTERNAL, bool $for_plaintext = false, int $uriid = 0): string
1497         {
1498                 // Accounting for null default column values
1499                 if (is_null($text) || $text === '') {
1500                         return '';
1501                 }
1502
1503                 DI::profiler()->startRecording('rendering');
1504
1505                 Hook::callAll('bbcode', $text);
1506
1507                 $a = DI::app();
1508
1509                 $text = self::performWithEscapedTags($text, ['code'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a, $uriid) {
1510                         $text = self::performWithEscapedTags($text, ['noparse', 'nobb', 'pre'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a, $uriid) {
1511                                 /*
1512                                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1513                                  *
1514                                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1515                                  * $match[1] = $url
1516                                  * $match[2] = $title or absent
1517                                  */
1518                                 $try_oembed_callback = function (array $match)
1519                                 {
1520                                         $url = $match[1];
1521                                         $title = $match[2] ?? '';
1522
1523                                         try {
1524                                                 $return = OEmbed::getHTML($url, $title);
1525                                         } catch (Exception $ex) {
1526                                                 $return = $match[0];
1527                                         }
1528
1529                                         return $return;
1530                                 };
1531
1532                                 // Remove the abstract element. It is a non visible element.
1533                                 $text = self::stripAbstract($text);
1534
1535                                 // Line ending normalisation
1536                                 $text = str_replace("\r\n", "\n", $text);
1537
1538                                 // Move new lines outside of tags
1539                                 $text = preg_replace("#\[(\w*)](\n*)#ism", '$2[$1]', $text);
1540                                 $text = preg_replace("#(\n*)\[/(\w*)]#ism", '[/$2]$1', $text);
1541
1542                                 // Extract the private images which use data urls since preg has issues with
1543                                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1544                                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1545
1546                                 $extracted = self::extractImagesFromItemBody($text);
1547                                 $text = $extracted['body'];
1548                                 $saved_image = $extracted['images'];
1549
1550                                 // If we find any event code, turn it into an event.
1551                                 // After we're finished processing the bbcode we'll
1552                                 // replace all of the event code with a reformatted version.
1553
1554                                 $ev = Event::fromBBCode($text);
1555
1556                                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1557                                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1558
1559                                 $text = str_replace("<", "&lt;", $text);
1560                                 $text = str_replace(">", "&gt;", $text);
1561
1562                                 // remove some newlines before the general conversion
1563                                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1564                                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1565
1566                                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1567                                 if (!$try_oembed) {
1568                                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1569                                 }
1570
1571                                 // Remove linefeeds inside of the table elements. See issue #6799
1572                                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1573                                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1574                                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1575                                         "\n[hr]", "[hr]\n", " [hr]", "[hr] ",
1576                                         "\n[attachment ", " [attachment ", "\n[/attachment]", "[/attachment]\n", " [/attachment]", "[/attachment] ",
1577                                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1578                                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1579                                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1580                                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1581                                         "[hr]", "[hr]", "[hr]", "[hr]",
1582                                         "[attachment ", "[attachment ", "[/attachment]", "[/attachment]", "[/attachment]", "[/attachment]",
1583                                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1584                                 do {
1585                                         $oldtext = $text;
1586                                         $text = str_replace($search, $replace, $text);
1587                                 } while ($oldtext != $text);
1588
1589                                 // Replace these here only once
1590                                 $search = ["\n[table]", "[/table]\n"];
1591                                 $replace = ["[table]", "[/table]"];
1592                                 $text = str_replace($search, $replace, $text);
1593
1594                                 // Trim new lines regardless of the system.remove_multiplicated_lines config value
1595                                 $text = trim($text, "\n");
1596
1597                                 // removing multiplicated newlines
1598                                 if (DI::config()->get('system', 'remove_multiplicated_lines')) {
1599                                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[*]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1600                                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1601                                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[*]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1602                                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1603                                         do {
1604                                                 $oldtext = $text;
1605                                                 $text = str_replace($search, $replace, $text);
1606                                         } while ($oldtext != $text);
1607                                 }
1608
1609                                 /// @todo Have a closer look at the different html modes
1610                                 // Handle attached links or videos
1611                                 if (in_array($simple_html, [self::API, self::ACTIVITYPUB])) {
1612                                         $text = self::removeAttachment($text);
1613                                 } elseif (!in_array($simple_html, [self::INTERNAL, self::EXTERNAL, self::CONNECTORS])) {
1614                                         $text = self::removeAttachment($text, true);
1615                                 } else {
1616                                         $text = self::convertAttachment($text, $simple_html, $try_oembed, [], $uriid);
1617                                 }
1618
1619                                 $nosmile = strpos($text, '[nosmile]') !== false;
1620                                 $text = str_replace('[nosmile]', '', $text);
1621
1622                                 // Replace non graphical smilies for external posts
1623                                 if (!$nosmile) {
1624                                         $text = self::performWithEscapedTags($text, ['img'], function ($text) use ($simple_html, $for_plaintext) {
1625                                                 return Smilies::replace($text, ($simple_html != self::INTERNAL) || $for_plaintext);
1626                                         });
1627                                 }
1628
1629                                 // leave open the posibility of [map=something]
1630                                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1631                                 if (strpos($text, '[/map]') !== false) {
1632                                         $text = preg_replace_callback(
1633                                                 "/\[map\](.*?)\[\/map\]/ism",
1634                                                 function ($match) use ($simple_html) {
1635                                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1636                                                 },
1637                                                 $text
1638                                         );
1639                                 }
1640
1641                                 if (strpos($text, '[map=') !== false) {
1642                                         $text = preg_replace_callback(
1643                                                 "/\[map=(.*?)\]/ism",
1644                                                 function ($match) use ($simple_html) {
1645                                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1646                                                 },
1647                                                 $text
1648                                         );
1649                                 }
1650
1651                                 if (strpos($text, '[map]') !== false) {
1652                                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1653                                 }
1654
1655                                 // Check for headers
1656                                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1657                                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1658                                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1659                                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1660                                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1661                                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1662
1663                                 // Check for paragraph
1664                                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1665
1666                                 // Check for bold text
1667                                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1668
1669                                 // Check for Italics text
1670                                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1671
1672                                 // Check for Underline text
1673                                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1674
1675                                 // Check for strike-through text
1676                                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1677
1678                                 // Check for over-line text
1679                                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1680
1681                                 // Check for colored text
1682                                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1683
1684                                 // Check for sized text
1685                                 // [size=50] --> font-size: 50px (with the unit).
1686                                 if ($simple_html != self::DIASPORA) {
1687                                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1px;line-height:normal;">$2</span>', $text);
1688                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1;line-height:normal;">$2</span>', $text);
1689                                 } else {
1690                                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1691                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1692                                 }
1693
1694
1695                                 // Check for centered text
1696                                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", '<div style="text-align:center;">$1</div>', $text);
1697
1698                                 // Check for list text
1699                                 $text = str_replace("[*]", "<li>", $text);
1700
1701                                 // Check for block-level custom CSS
1702                                 $text = preg_replace('#(?<=^|\n)\[style=(.*?)](.*?)\[/style](?:\n|$)#ism', '<div style="$1">$2</div>', $text);
1703
1704                                 // Check for inline custom CSS
1705                                 $text = preg_replace("(\[style=(.*?)\](.*?)\[\/style\])ism", '<span style="$1">$2</span>', $text);
1706
1707                                 // Mastodon Emoji (internal tag, do not document for users)
1708                                 $text = preg_replace("(\[emoji=(.*?)](.*?)\[/emoji])ism", '<span class="mastodon emoji"><img src="$1" alt="$2" title="$2"/></span>', $text);
1709
1710                                 // Check for CSS classes
1711                                 // @deprecated since 2021.12, left for backward-compatibility reasons
1712                                 $text = preg_replace("(\[class=(.*?)\](.*?)\[\/class\])ism", '<span class="$1">$2</span>', $text);
1713                                 // Add HTML new lines
1714                                 $text = str_replace("\n", '<br>', $text);
1715
1716                                 // handle nested lists
1717                                 $endlessloop = 0;
1718
1719                                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1720                                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1721                                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1722                                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1723                                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1724                                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1725                                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1726                                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1727                                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1728                                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1729                                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1730                                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1731                                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1732                                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1733                                 }
1734
1735                                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1736                                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1737                                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1738                                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1739
1740                                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1741                                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1742
1743                                 $text = str_replace('[hr]', '<hr />', $text);
1744
1745                                 if (!$for_plaintext) {
1746                                         $text = self::performWithEscapedTags($text, ['url', 'img', 'audio', 'video', 'youtube', 'vimeo', 'share', 'attachment', 'iframe', 'bookmark'], function ($text) {
1747                                                 return preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1748                                         });
1749                                 }
1750
1751                                 // Check for font change text
1752                                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1753
1754                                 // Declare the format for [spoiler] layout
1755                                 $SpoilerLayout = '<details class="spoiler"><summary>' . DI::l10n()->t('Click to open/close') . '</summary>$1</details>';
1756
1757                                 // Check for [spoiler] text
1758                                 // handle nested quotes
1759                                 $endlessloop = 0;
1760                                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1761                                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1762                                 }
1763
1764                                 // Check for [spoiler=Title] text
1765
1766                                 // handle nested quotes
1767                                 $endlessloop = 0;
1768                                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1769                                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1770                                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1771                                                 $text);
1772                                 }
1773
1774                                 // Declare the format for [quote] layout
1775                                 $QuoteLayout = '<blockquote>$1</blockquote>';
1776
1777                                 // Check for [quote] text
1778                                 // handle nested quotes
1779                                 $endlessloop = 0;
1780                                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1781                                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1782                                 }
1783
1784                                 // Check for [quote=Author] text
1785
1786                                 $t_wrote = DI::l10n()->t('$1 wrote:');
1787
1788                                 // handle nested quotes
1789                                 $endlessloop = 0;
1790                                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1791                                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1792                                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1793                                                 $text);
1794                                 }
1795
1796
1797                                 // [img=widthxheight]image source[/img]
1798                                 $text = preg_replace_callback(
1799                                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1800                                         function ($matches) use ($simple_html, $uriid) {
1801                                                 if (strpos($matches[3], "data:image/") === 0) {
1802                                                         return $matches[0];
1803                                                 }
1804
1805                                                 $matches[3] = self::proxyUrl($matches[3], $simple_html, $uriid);
1806                                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1807                                         },
1808                                         $text
1809                                 );
1810
1811                                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1812                                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1813
1814                                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1815                                         function ($matches) use ($simple_html, $uriid) {
1816                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html, $uriid);
1817                                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1818                                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1819                                         },
1820                                         $text);
1821
1822                                 // Images
1823                                 // [img]pathtoimage[/img]
1824                                 $text = preg_replace_callback(
1825                                         "/\[img\](.*?)\[\/img\]/ism",
1826                                         function ($matches) use ($simple_html, $uriid) {
1827                                                 if (strpos($matches[1], "data:image/") === 0) {
1828                                                         return $matches[0];
1829                                                 }
1830
1831                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html, $uriid);
1832                                                 return "[img]" . $matches[1] . "[/img]";
1833                                         },
1834                                         $text
1835                                 );
1836
1837                                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1838                                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1839
1840                                 $text = self::convertImages($text, $simple_html, $uriid);
1841
1842                                 $text = preg_replace("/\[crypt\](.*?)\[\/crypt\]/ism", '<br><img src="' .DI::baseUrl() . '/images/lock_icon.gif" alt="' . DI::l10n()->t('Encrypted content') . '" title="' . DI::l10n()->t('Encrypted content') . '" /><br>', $text);
1843                                 $text = preg_replace("/\[crypt(.*?)\](.*?)\[\/crypt\]/ism", '<br><img src="' .DI::baseUrl() . '/images/lock_icon.gif" alt="' . DI::l10n()->t('Encrypted content') . '" title="' . '$1' . ' ' . DI::l10n()->t('Encrypted content') . '" /><br>', $text);
1844                                 //$text = preg_replace("/\[crypt=(.*?)\](.*?)\[\/crypt\]/ism", '<br><img src="' .DI::baseUrl() . '/images/lock_icon.gif" alt="' . DI::l10n()->t('Encrypted content') . '" title="' . '$1' . ' ' . DI::l10n()->t('Encrypted content') . '" /><br>', $text);
1845
1846                                 // Simplify "video" element
1847                                 $text = preg_replace('(\[video[^\]]*?\ssrc\s?=\s?([^\s\]]+)[^\]]*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1848
1849                                 if ($try_oembed) {
1850                                         // html5 video and audio
1851                                         $text = preg_replace("/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4).*?)\[\/video\]/ism",
1852                                                 '<video src="$1" controls width="100%" height="auto"><a href="$1">$1</a></video>', $text);
1853
1854                                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1855                                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1856
1857                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1858                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1859                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '<audio src="$1" controls><a href="$1">$1</a></audio>', $text);
1860                                 } else {
1861                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1862                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1863                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1864                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1865                                 }
1866
1867                                 // Backward compatibility, [iframe] support has been removed in version 2020.12
1868                                 $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1869
1870                                 // Youtube extensions
1871                                 if ($try_oembed) {
1872                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1873                                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1874                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/shorts\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1875                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1876                                 }
1877
1878                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1879                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1880                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/shorts\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1881                                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1882
1883                                 if ($try_oembed) {
1884                                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism", '<iframe width="' . $a->getThemeInfoValue('videowidth') . '" height="' . $a->getThemeInfoValue('videoheight') . '" src="https://www.youtube.com/embed/$1" frameborder="0" ></iframe>', $text);
1885                                 } else {
1886                                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1887                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank" rel="noopener noreferrer">https://www.youtube.com/watch?v=$1</a>', $text);
1888                                 }
1889
1890                                 if ($try_oembed) {
1891                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1892                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1893                                 }
1894
1895                                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1896                                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1897
1898                                 if ($try_oembed) {
1899                                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism", '<iframe width="' . $a->getThemeInfoValue('videowidth') . '" height="' . $a->getThemeInfoValue('videoheight') . '" src="https://player.vimeo.com/video/$1" frameborder="0" ></iframe>', $text);
1900                                 } else {
1901                                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1902                                                 '<a href="https://vimeo.com/$1" target="_blank" rel="noopener noreferrer">https://vimeo.com/$1</a>', $text);
1903                                 }
1904
1905                                 // oembed tag
1906                                 $text = OEmbed::BBCode2HTML($text);
1907
1908                                 // Avoid triple linefeeds through oembed
1909                                 $text = str_replace("<br style='clear:left'></span><br><br>", "<br style='clear:left'></span><br>", $text);
1910
1911                                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1912                                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1913                                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1914                                 // start which is always required). Allow desc with a missing summary for compatibility.
1915
1916                                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1917                                         $sub = Event::getHTML($ev, $simple_html, $uriid);
1918
1919                                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1920                                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1921                                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1922                                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1923                                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1924                                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1925                                 }
1926
1927                                 if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA)) {
1928                                         $conv = html_entity_decode(str_replace([' ', "\n", "\r"], '', $text));
1929                                         // Emojis are always 4 byte Unicode characters
1930                                         if (!empty($conv) && (strlen($conv) / mb_strlen($conv) == 4)) {
1931                                                 $text = '<span style="font-size: xx-large; line-height: normal;">' . $text . '</span>';
1932                                         }
1933                                 }
1934
1935                                 // Handle mentions and hashtag links
1936                                 if ($simple_html == self::DIASPORA) {
1937                                         // The ! is converted to @ since Diaspora only understands the @
1938                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1939                                                 '@<a href="$2">$3</a>',
1940                                                 $text);
1941                                 } elseif (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1942                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1943                                                 '<span class="h-card"><a href="$2" class="u-url mention">$1<span>$3</span></a></span>',
1944                                                 $text);
1945                                         $text = preg_replace("/([#])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1946                                                 '<a href="$2" class="mention hashtag" rel="tag">$1<span>$3</span></a>',
1947                                                 $text);
1948                                 } elseif (in_array($simple_html, [self::INTERNAL, self::EXTERNAL, self::API])) {
1949                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1950                                                 '<bdi>$1<a href="$2" class="userinfo mention" title="$3">$3</a></bdi>',
1951                                                 $text);
1952                                 } else {
1953                                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1954                                 }
1955
1956                                 if (!$for_plaintext) {
1957                                         if (in_array($simple_html, [self::OSTATUS, self::API, self::ACTIVITYPUB])) {
1958                                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1959                                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1960                                         }
1961                                 } else {
1962                                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1963                                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1964                                 }
1965
1966                                 // Bookmarks in red - will be converted to bookmarks in friendica
1967                                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1968                                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1969                                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1970                                                         "[bookmark=$1]$2[/bookmark]", $text);
1971
1972                                 if (in_array($simple_html, [self::OSTATUS, self::TWITTER])) {
1973                                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1974                                         //$text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $text);
1975                                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]', $text);
1976                                 }
1977
1978                                 // Perform URL Search
1979                                 if ($try_oembed) {
1980                                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1981                                 }
1982
1983                                 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1984
1985                                 // Handle Diaspora posts
1986                                 $text = preg_replace_callback(
1987                                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1988                                         function ($match) {
1989                                                 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1990                                         }, $text
1991                                 );
1992
1993                                 $text = preg_replace_callback(
1994                                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1995                                         function ($match) {
1996                                                 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1997                                         }, $text
1998                                 );
1999
2000                                 // Server independent link to posts and comments
2001                                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
2002                                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
2003                                 $text = preg_replace($expression, DI::baseUrl()."/display/$1", $text);
2004
2005                                 /* Tag conversion
2006                                  * Supports:
2007                                  * - #[url=<anything>]<term>[/url]
2008                                  * - [url=<anything>]#<term>[/url]
2009                                  */
2010                                 self::performWithEscapedTags($text, ['url', 'share'], function ($text) use ($simple_html) {
2011                                         $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) use ($simple_html) {
2012                                                 if ($simple_html == self::ACTIVITYPUB) {
2013                                                         return '<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
2014                                                                 . '" data-tag="' . XML::escape($matches[1]) . '" rel="tag ugc">#'
2015                                                                 . XML::escape($matches[1]) . '</a>';
2016                                                 } else {
2017                                                         return '#<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
2018                                                                 . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
2019                                                                 . XML::escape($matches[1]) . '</a>';
2020                                                 }
2021                                         }, $text);
2022                                         return $text;
2023                                 });
2024
2025                                 // We need no target="_blank" rel="noopener noreferrer" for local links
2026                                 // convert links start with DI::baseUrl() as local link without the target="_blank" rel="noopener noreferrer" attribute
2027                                 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
2028                                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
2029                                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
2030
2031                                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
2032                                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
2033
2034                                 // Red compatibility, though the link can't be authenticated on Friendica
2035                                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
2036
2037
2038                                 // we may need to restrict this further if it picks up too many strays
2039                                 // link acct:user@host to a webfinger profile redirector
2040
2041                                 $text = preg_replace('/acct:([^@]+)@((?!\-)(?:[a-zA-Z\d\-]{0,62}[a-zA-Z\d]\.){1,126}(?!\d+)[a-zA-Z\d]{1,63})/', '<a href="' . DI::baseUrl() . '/acctlink?addr=$1@$2" target="extlink">acct:$1@$2</a>', $text);
2042
2043                                 // Perform MAIL Search
2044                                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
2045                                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
2046
2047                                 /// @todo What is the meaning of these lines?
2048                                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
2049                                 $text = preg_replace('/\&\#039\;/', '\'', $text);
2050
2051                                 // Currently deactivated, it made problems with " inside of alt texts.
2052                                 //$text = preg_replace('/\&quot\;/', '"', $text);
2053
2054                                 // fix any escaped ampersands that may have been converted into links
2055                                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
2056
2057                                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
2058                                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
2059
2060                                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
2061
2062                                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
2063                                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">', $text);
2064
2065                                 // sanitize href attributes (only allowlisted protocols URLs)
2066                                 // default value for backward compatibility
2067                                 $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
2068
2069                                 // Always allowed protocol even if config isn't set or not including it
2070                                 $allowed_link_protocols[] = '//';
2071                                 $allowed_link_protocols[] = 'http://';
2072                                 $allowed_link_protocols[] = 'https://';
2073                                 $allowed_link_protocols[] = 'redir/';
2074
2075                                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
2076
2077                                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
2078                                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text);
2079
2080                                 // Shared content
2081                                 $text = self::convertShare(
2082                                         $text,
2083                                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
2084                                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
2085                                         }, $uriid
2086                                 );
2087
2088                                 $text = self::interpolateSavedImagesIntoItemBody($uriid, $text, $saved_image);
2089
2090                                 return $text;
2091                         }); // Escaped noparse, nobb, pre
2092
2093                         // Remove escaping tags and replace new lines that remain
2094                         $text = preg_replace_callback('/\[(noparse|nobb)](.*?)\[\/\1]/ism', function ($match) {
2095                                 return str_replace("\n", "<br>", $match[2]);
2096                         }, $text);
2097
2098                         // Additionally, [pre] tags preserve spaces
2099                         $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", function ($match) {
2100                                 return str_replace([' ', "\n"], ['&nbsp;', "<br>"], htmlentities($match[1], ENT_NOQUOTES,'UTF-8'));
2101                         }, $text);
2102
2103                         return $text;
2104                 }); // Escaped code
2105
2106                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
2107                         function ($matches) {
2108                                 if (strpos($matches[2], "\n") !== false) {
2109                                         $return = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlentities(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
2110                                 } else {
2111                                         $return = '<code>' . htmlentities($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
2112                                 }
2113
2114                                 return $return;
2115                         },
2116                         $text
2117                 );
2118
2119                 // Default iframe allowed domains/path
2120                 $allowedIframeDomains = [
2121                         DI::baseUrl()->getHostname()
2122                         . (DI::baseUrl()->getUrlPath() ? '/' . DI::baseUrl()->getUrlPath() : '')
2123                         . '/oembed/', # The path part has to change with the source in Content\Oembed::iframe
2124                         'www.youtube.com/embed/',
2125                         'player.vimeo.com/video/',
2126                 ];
2127
2128                 $allowedIframeDomains = array_merge(
2129                         $allowedIframeDomains,
2130                         DI::config()->get('system', 'allowed_oembed') ?
2131                                 explode(',', DI::config()->get('system', 'allowed_oembed'))
2132                                 : []
2133                 );
2134
2135                 $text = HTML::purify($text, $allowedIframeDomains);
2136                 DI::profiler()->stopRecording();
2137
2138                 return trim($text);
2139         }
2140
2141         /**
2142          * Strips the "abstract" tag from the provided text
2143          *
2144          * @param string $text The text with BBCode
2145          * @return string The same text - but without "abstract" element
2146          */
2147         public static function stripAbstract(string $text): string
2148         {
2149                 DI::profiler()->startRecording('rendering');
2150
2151                 $text = BBCode::performWithEscapedTags($text, ['code', 'noparse', 'nobb', 'pre'], function ($text) {
2152                         $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", ' ', $text);
2153                         $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", ' ', $text);
2154                         return $text;
2155                 });
2156
2157                 DI::profiler()->stopRecording();
2158                 return $text;
2159         }
2160
2161         /**
2162          * Returns the value of the "abstract" element
2163          *
2164          * @param string $text  The text that maybe contains the element
2165          * @param string $addon The addon for which the abstract is meant for
2166          * @return string The abstract
2167          */
2168         public static function getAbstract(string $text, string $addon = ''): string
2169         {
2170                 DI::profiler()->startRecording('rendering');
2171                 $addon = strtolower($addon);
2172
2173                 $abstract = BBCode::performWithEscapedTags($text, ['code', 'noparse', 'nobb', 'pre'], function ($text) use ($addon) {
2174                         if ($addon && preg_match('#\[abstract=' . preg_quote($addon, '#') . '](.*?)\[/abstract]#ism', $text, $matches)) {
2175                                 return $matches[1];
2176                         }
2177
2178                         if (preg_match("#\[abstract](.*?)\[/abstract]#ism", $text, $matches)) {
2179                                 return $matches[1];
2180                         }
2181
2182                         return '';
2183                 });
2184
2185                 DI::profiler()->stopRecording();
2186                 return $abstract;
2187         }
2188
2189         /**
2190          * Callback function to replace a Friendica style mention in a mention for Diaspora
2191          *
2192          * @param array $match Matching values for the callback
2193          *                     [1] = Mention type (! or @)
2194          *                     [2] = Name
2195          *                     [3] = Address
2196          * @return string Replaced mention
2197          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2198          * @throws \ImagickException
2199          */
2200         private static function bbCodeMention2DiasporaCallback(array $match): string
2201         {
2202                 $contact = Contact::getByURL($match[3], false, ['addr']);
2203                 if (empty($contact['addr'])) {
2204                         return $match[0];
2205                 }
2206
2207                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
2208                 return $mention;
2209         }
2210
2211         /**
2212          * Converts a BBCode text into Markdown
2213          *
2214          * This function converts a BBCode item body to be sent to Markdown-enabled
2215          * systems like Diaspora and Libertree
2216          *
2217          * @param string $text
2218          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
2219          * @return string
2220          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2221          */
2222         public static function toMarkdown(string $text, bool $for_diaspora = true): string
2223         {
2224                 DI::profiler()->startRecording('rendering');
2225                 $original_text = $text;
2226
2227                 // Since Diaspora is creating a summary for links, this function removes them before posting
2228                 if ($for_diaspora) {
2229                         $text = self::removeShareInformation($text);
2230                 }
2231
2232                 /**
2233                  * Transform #tags, strip off the [url] and replace spaces with underscore
2234                  */
2235                 $url_search_string = "^\[\]";
2236                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2237                         function ($matches) {
2238                                 return '#' . str_replace(' ', '_', $matches[2]);
2239                         },
2240                         $text
2241                 );
2242
2243                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2244                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2245
2246                 // Convert it to HTML - don't try oembed
2247                 if ($for_diaspora) {
2248                         $text = self::convert($text, false, self::DIASPORA);
2249
2250                         // Add all tags that maybe were removed
2251                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2252                                 $tagline = '';
2253                                 foreach ($tags[2] as $tag) {
2254                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2255                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2256                                                 $tagline .= '#' . $tag . ' ';
2257                                         }
2258                                 }
2259                                 $text = $text . ' ' . $tagline;
2260                         }
2261                 } else {
2262                         $text = self::convert($text, false, self::CONNECTORS);
2263                 }
2264
2265                 // If a link is followed by a quote then there should be a newline before it
2266                 // Maybe we should make this newline at every time before a quote.
2267                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2268
2269                 // Now convert HTML to Markdown
2270                 $text = HTML::toMarkdown($text);
2271
2272                 // Libertree has a problem with escaped hashtags.
2273                 $text = str_replace(['\#'], ['#'], $text);
2274
2275                 // Remove any leading or trailing whitespace, as this will mess up
2276                 // the Diaspora signature verification and cause the item to disappear
2277                 $text = trim($text);
2278
2279                 if ($for_diaspora) {
2280                         $url_search_string = "^\[\]";
2281                         $text = preg_replace_callback(
2282                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2283                                 ['self', 'bbCodeMention2DiasporaCallback'],
2284                                 $text
2285                         );
2286                 }
2287
2288                 Hook::callAll('bb2diaspora', $text);
2289
2290                 DI::profiler()->stopRecording();
2291                 return $text;
2292         }
2293
2294         /**
2295          * Pull out all #hashtags and @person tags from $string.
2296          *
2297          * We also get @person@domain.com - which would make
2298          * the regex quite complicated as tags can also
2299          * end a sentence. So we'll run through our results
2300          * and strip the period from any tags which end with one.
2301          * Returns array of tags found, or empty array.
2302          *
2303          * @param string $string Post content
2304          * @return array List of tag and person names
2305          */
2306         public static function getTags(string $string): array
2307         {
2308                 DI::profiler()->startRecording('rendering');
2309                 $ret = [];
2310
2311                 self::performWithEscapedTags($string, ['noparse', 'pre', 'code', 'img'], function ($string) use (&$ret) {
2312                         // Convert hashtag links to hashtags
2313                         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2 ', $string);
2314
2315                         // Force line feeds at bbtags
2316                         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2317
2318                         // ignore anything in a bbtag
2319                         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2320
2321                         // Match full names against @tags including the space between first and last
2322                         // We will look these up afterward to see if they are full names or not recognisable.
2323
2324                         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2325                                 foreach ($matches[1] as $match) {
2326                                         if (strstr($match, ']')) {
2327                                                 // we might be inside a bbcode color tag - leave it alone
2328                                                 continue;
2329                                         }
2330
2331                                         if (substr($match, -1, 1) === '.') {
2332                                                 $ret[] = substr($match, 0, -1);
2333                                         } else {
2334                                                 $ret[] = $match;
2335                                         }
2336                                 }
2337                         }
2338
2339                         // Otherwise pull out single word tags. These can be @nickname, @first_last
2340                         // and #hash tags.
2341
2342                         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?\']*[^\^ \x0D\x0A,;:?!\'.])/', $string, $matches)) {
2343                                 foreach ($matches[1] as $match) {
2344                                         if (strstr($match, ']')) {
2345                                                 // we might be inside a bbcode color tag - leave it alone
2346                                                 continue;
2347                                         }
2348
2349                                         // try not to catch url fragments
2350                                         if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2351                                                 continue;
2352                                         }
2353
2354                                         $ret[] = $match;
2355                                 }
2356                         }
2357                 });
2358
2359                 DI::profiler()->stopRecording();
2360                 return array_unique($ret);
2361         }
2362
2363         /**
2364          * Expand tags to URLs, checks the tag is at the start of a line or preceded by a non-word character
2365          *
2366          * @param string $body HTML/BBCode
2367          * @return string body with expanded tags
2368          */
2369         public static function expandTags(string $body): string
2370         {
2371                 return preg_replace_callback("/(?<=\W|^)([!#@])([^\^ \x0D\x0A,;:?'\"]*[^\^ \x0D\x0A,;:?!'\".])/",
2372                         function (array $match) {
2373                                 switch ($match[1]) {
2374                                         case '!':
2375                                         case '@':
2376                                                 $contact = Contact::getByURL($match[2]);
2377                                                 if (!empty($contact)) {
2378                                                         return $match[1] . '[url=' . $contact['url'] . ']' . $contact['name'] . '[/url]';
2379                                                 } else {
2380                                                         return $match[1] . $match[2];
2381                                                 }
2382                                                 break;
2383
2384                                         case '#':
2385                                         default:
2386                                                 return $match[1] . '[url=' . DI::baseUrl() . '/search?tag=' . $match[2] . ']' . $match[2] . '[/url]';
2387                                 }
2388                         }, $body);
2389         }
2390
2391         /**
2392          * Perform a custom function on a text after having escaped blocks enclosed in the provided tag list.
2393          *
2394          * @param string   $text HTML/BBCode
2395          * @param array    $tagList A list of tag names, e.g ['noparse', 'nobb', 'pre']
2396          * @param callable $callback
2397          * @return string
2398          * @see Strings::performWithEscapedBlocks
2399          */
2400         public static function performWithEscapedTags(string $text, array $tagList, callable $callback): string
2401         {
2402                 $tagList = array_map('preg_quote', $tagList);
2403
2404                 return Strings::performWithEscapedBlocks($text, '#\[(?:' . implode('|', $tagList) . ').*?\[/(?:' . implode('|', $tagList) . ')]#ism', $callback);
2405         }
2406
2407         /**
2408          * Replaces mentions in the provided message body in BBCode links for the provided user and network if any
2409          *
2410          * @param string $body HTML/BBCode
2411          * @param int $profile_uid Profile user id
2412          * @param string $network Network name
2413          * @return string HTML/BBCode with inserted images
2414          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2415          * @throws \ImagickException
2416          */
2417         public static function setMentions(string $body, $profile_uid = 0, $network = '')
2418         {
2419                 DI::profiler()->startRecording('rendering');
2420                 $body = self::performWithEscapedTags($body, ['noparse', 'pre', 'code', 'img'], function ($body) use ($profile_uid, $network) {
2421                         $tags = self::getTags($body);
2422
2423                         $tagged = [];
2424
2425                         foreach ($tags as $tag) {
2426                                 $tag_type = substr($tag, 0, 1);
2427
2428                                 if ($tag_type == Tag::TAG_CHARACTER[Tag::HASHTAG]) {
2429                                         continue;
2430                                 }
2431
2432                                 /*
2433                                  * If we already tagged 'Robert Johnson', don't try and tag 'Robert'.
2434                                  * Robert Johnson should be first in the $tags array
2435                                  */
2436                                 foreach ($tagged as $nextTag) {
2437                                         if (stristr($nextTag, $tag . ' ')) {
2438                                                 continue 2;
2439                                         }
2440                                 }
2441
2442                                 if (($success = Item::replaceTag($body, $profile_uid, $tag, $network)) && $success['replaced']) {
2443                                         $tagged[] = $tag;
2444                                 }
2445                         }
2446
2447                         return $body;
2448                 });
2449
2450                 DI::profiler()->stopRecording();
2451                 return $body;
2452         }
2453
2454         /**
2455          * @param string      $author  Author display name
2456          * @param string      $profile Author profile URL
2457          * @param string      $avatar  Author profile picture URL
2458          * @param string      $link    Post source URL
2459          * @param string      $posted  Post created date
2460          * @param string|null $guid    Post guid (if any)
2461          * @param string|null $uri     Post uri (if any)
2462          * @return string
2463          * @TODO Rewrite to handle over whole record array
2464          */
2465         public static function getShareOpeningTag(string $author, string $profile, string $avatar, string $link, string $posted, string $guid = null, string $uri = null): string
2466         {
2467                 DI::profiler()->startRecording('rendering');
2468                 $header = "[share author='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $author) .
2469                         "' profile='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $profile) .
2470                         "' avatar='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $avatar) .
2471                         "' link='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $link) .
2472                         "' posted='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $posted);
2473
2474                 if ($guid) {
2475                         $header .= "' guid='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $guid);
2476                 }
2477
2478                 if ($uri) {
2479                         $header .= "' message_id='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $uri);
2480                 }
2481
2482                 $header  .= "']";
2483
2484                 DI::profiler()->stopRecording();
2485                 return $header;
2486         }
2487
2488         /**
2489          * Returns the BBCode relevant to embed the provided URL in a post body.
2490          * For media type, it will return [img], [video] and [audio] tags.
2491          * For regular web pages, it will either output a [bookmark] tag if title and description were provided,
2492          * an [attachment] tag or a simple [url] tag depending on $tryAttachment.
2493          *
2494          * @param string      $url
2495          * @param bool        $tryAttachment
2496          * @param string|null $title
2497          * @param string|null $description
2498          * @param string|null $tags
2499          * @return string
2500          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2501          * @see ParseUrl::getSiteinfoCached
2502          */
2503         public static function embedURL(string $url, bool $tryAttachment = true, string $title = null, string $description = null, string $tags = null): string
2504         {
2505                 DI::profiler()->startRecording('rendering');
2506                 DI::logger()->info($url);
2507
2508                 // If there is already some content information submitted we don't
2509                 // need to parse the url for content.
2510                 if (!empty($title) && !empty($description)) {
2511                         $title = str_replace(["\r", "\n"], ['', ''], $title);
2512
2513                         $description = '[quote]' . trim($description) . '[/quote]' . "\n";
2514
2515                         $str_tags = '';
2516                         if (!empty($tags)) {
2517                                 $arr_tags = ParseUrl::convertTagsToArray($tags);
2518                                 if (count($arr_tags)) {
2519                                         $str_tags = "\n" . implode(' ', $arr_tags) . "\n";
2520                                 }
2521                         }
2522
2523                         $result = sprintf('[bookmark=%s]%s[/bookmark]%s', $url, ($title) ? $title : $url, $description) . $str_tags;
2524
2525                         DI::logger()->info('(unparsed): returns: ' . $result);
2526
2527                         DI::profiler()->stopRecording();
2528                         return $result;
2529                 }
2530
2531                 $siteinfo = ParseUrl::getSiteinfoCached($url);
2532
2533                 if (in_array($siteinfo['type'], ['image', 'video', 'audio'])) {
2534                         switch ($siteinfo['type']) {
2535                                 case 'video':
2536                                         $bbcode = "\n" . '[video]' . $url . '[/video]' . "\n";
2537                                         break;
2538                                 case 'audio':
2539                                         $bbcode = "\n" . '[audio]' . $url . '[/audio]' . "\n";
2540                                         break;
2541                                 default:
2542                                         $bbcode = "\n" . '[img]' . $url . '[/img]' . "\n";
2543                                         break;
2544                         }
2545
2546                         DI::profiler()->stopRecording();
2547                         return $bbcode;
2548                 }
2549
2550                 unset($siteinfo['keywords']);
2551
2552                 // Bypass attachment if parse url for a comment
2553                 if (!$tryAttachment) {
2554                         DI::profiler()->stopRecording();
2555                         return "\n" . '[url=' . $url . ']' . $siteinfo['title'] . '[/url]';
2556                 }
2557
2558                 // Format it as BBCode attachment
2559                 $bbcode = "\n" . PageInfo::getFooterFromData($siteinfo);
2560                 DI::profiler()->stopRecording();
2561                 return $bbcode;
2562         }
2563 }