src/Content/Text/BBCode.php

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