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\Tag;
  41 use Friendica\Object\Image;
  42 use Friendica\Protocol\Activity;
  43 use Friendica\Util\Images;
  44 use Friendica\Util\Map;
  45 use Friendica\Util\ParseUrl;
  46 use Friendica\Util\Proxy as ProxyUtils;
  47 use Friendica\Util\Strings;
  48 use Friendica\Util\XML;
  49
  50 class BBCode
  51 {
  52         // Update this value to the current date whenever changes are made to BBCode::convert
  53         const VERSION = '2021-04-24';
  54
  55         const INTERNAL = 0;
  56         const API = 2;
  57         const DIASPORA = 3;
  58         const CONNECTORS = 4;
  59         const OSTATUS = 7;
  60         const TWITTER = 8;
  61         const BACKLINK = 8;
  62         const ACTIVITYPUB = 9;
  63
  64         const ANCHOR = '<br class="anchor">';
  65         /**
  66          * Fetches attachment data that were generated the old way
  67          *
  68          * @param string $body Message body
  69          * @return array
  70          *                     'type' -> Message type ('link', 'video', 'photo')
  71          *                     'text' -> Text before the shared message
  72          *                     'after' -> Text after the shared message
  73          *                     'image' -> Preview image of the message
  74          *                     'url' -> Url to the attached message
  75          *                     'title' -> Title of the attachment
  76          *                     'description' -> Description of the attachment
  77          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  78          */
  79         private static function getOldAttachmentData($body)
  80         {
  81                 $post = [];
  82
  83                 // Simplify image codes
  84                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  85
  86                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  87                         foreach ($attached as $data) {
  88                                 if (!in_array($data[1], ['type-link', 'type-video', 'type-photo'])) {
  89                                         continue;
  90                                 }
  91
  92                                 $post['type'] = substr($data[1], 5);
  93
  94                                 $pos = strpos($body, $data[0]);
  95                                 if ($pos > 0) {
  96                                         $post['text'] = trim(substr($body, 0, $pos));
  97                                         $post['after'] = trim(substr($body, $pos + strlen($data[0])));
  98                                 } else {
  99                                         $post['text'] = trim(str_replace($data[0], '', $body));
 100                                         $post['after'] = '';
 101                                 }
 102
 103                                 $attacheddata = $data[2];
 104
 105                                 if (preg_match("/\[img\](.*?)\[\/img\]/ism", $attacheddata, $matches)) {
 106
 107                                         $picturedata = Images::getInfoFromURLCached($matches[1]);
 108
 109                                         if ($picturedata) {
 110                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
 111                                                         $post['image'] = $matches[1];
 112                                                 } else {
 113                                                         $post['preview'] = $matches[1];
 114                                                 }
 115                                         }
 116                                 }
 117
 118                                 if (preg_match("/\[bookmark\=(.*?)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
 119                                         $post['url'] = $matches[1];
 120                                         $post['title'] = $matches[2];
 121                                 }
 122                                 if (!empty($post['url']) && (in_array($post['type'], ['link', 'video']))
 123                                         && preg_match("/\[url\=(.*?)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
 124                                         $post['url'] = $matches[1];
 125                                 }
 126
 127                                 // Search for description
 128                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
 129                                         $post['description'] = $matches[1];
 130                                 }
 131                         }
 132                 }
 133                 return $post;
 134         }
 135
 136         /**
 137          * Fetches attachment data that were generated with the "attachment" element
 138          *
 139          * @param string $body Message body
 140          * @return array
 141          *                     'type' -> Message type ('link', 'video', 'photo')
 142          *                     'text' -> Text before the shared message
 143          *                     'after' -> Text after the shared message
 144          *                     'image' -> Preview image of the message
 145          *                     'url' -> Url to the attached message
 146          *                     'title' -> Title of the attachment
 147          *                     'description' -> Description of the attachment
 148          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 149          */
 150         public static function getAttachmentData($body)
 151         {
 152                 $data = [
 153                         'type'          => '',
 154                         'text'          => '',
 155                         'after'         => '',
 156                         'image'         => null,
 157                         'url'           => '',
 158                         'author_name'   => '',
 159                         'author_url'    => '',
 160                         'provider_name' => '',
 161                         'provider_url'  => '',
 162                         'title'         => '',
 163                         'description'   => '',
 164                 ];
 165
 166                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 167                         return self::getOldAttachmentData($body);
 168                 }
 169
 170                 $attributes = $match[2];
 171
 172                 $data['text'] = trim($match[1]);
 173
 174                 foreach (['type', 'url', 'title', 'image', 'preview', 'publisher_name', 'publisher_url', 'author_name', 'author_url'] as $field) {
 175                         preg_match('/' . preg_quote($field, '/') . '=("|\')(.*?)\1/ism', $attributes, $matches);
 176                         $value = $matches[2] ?? '';
 177
 178                         if ($value != '') {
 179                                 switch ($field) {
 180                                         case 'publisher_name':
 181                                                 $data['provider_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 182                                                 break;
 183                                         case 'publisher_url':
 184                                                 $data['provider_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 185                                                 break;
 186                                         case 'author_name':
 187                                                 $data['author_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 188                                                 if ($data['provider_name'] == $data['author_name']) {
 189                                                         $data['author_name'] = '';
 190                                                 }
 191                                                 break;
 192                                         case 'author_url':
 193                                                 $data['author_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 194                                                 if ($data['provider_url'] == $data['author_url']) {
 195                                                         $data['author_url'] = '';
 196                                                 }
 197                                                 break;
 198                                         case 'title':
 199                                                 $value = self::convert(html_entity_decode($value, ENT_QUOTES, 'UTF-8'), false, true);
 200                                                 $value = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 201                                                 $value = str_replace(['[', ']'], ['&#91;', '&#93;'], $value);
 202                                                 $data['title'] = $value;
 203                                         default:
 204                                                 $data[$field] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 205                                                 break;
 206                                 }
 207                         }
 208                 }
 209
 210                 if (!in_array($data['type'], ['link', 'audio', 'photo', 'video'])) {
 211                         return [];
 212                 }
 213
 214                 $data['description'] = trim($match[3]);
 215
 216                 $data['after'] = trim($match[4]);
 217
 218                 $parts = parse_url($data['url']);
 219                 if (!empty($parts['scheme']) && !empty($parts['host'])) {
 220                         if (empty($data['provider_name'])) {
 221                                 $data['provider_name'] = $parts['host'];
 222                         }
 223                         if (empty($data['provider_url']) || empty(parse_url($data['provider_url'], PHP_URL_SCHEME))) {
 224                                 $data['provider_url'] = $parts['scheme'] . '://' . $parts['host'];
 225
 226                                 if (!empty($parts['port'])) {
 227                                         $data['provider_url'] .= ':' . $parts['port'];
 228                                 }
 229                         }
 230                 }
 231
 232                 return $data;
 233         }
 234
 235         public static function getAttachedData($body, $item = [])
 236         {
 237                 /*
 238                 - text:
 239                 - type: link, video, photo
 240                 - title:
 241                 - url:
 242                 - image:
 243                 - description:
 244                 - (thumbnail)
 245                 */
 246
 247                 $has_title = !empty($item['title']);
 248                 $plink = $item['plink'] ?? '';
 249                 $post = self::getAttachmentData($body);
 250
 251                 // Get all linked images with alternative image description
 252                 if (preg_match_all("/\[img=([^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 253                         foreach ($pictures as $picture) {
 254                                 if (Photo::isLocal($picture[1])) {
 255                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]];
 256                                 } else {
 257                                         $post['remote_images'][] = ['url' => $picture[1], 'description' => $picture[2]];
 258                                 }
 259                         }
 260                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 261                                 $post['image_description'] = $post['images'][0]['description'];
 262                         }
 263                 }
 264
 265                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 266                         foreach ($pictures as $picture) {
 267                                 if (Photo::isLocal($picture[1])) {
 268                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
 269                                 } else {
 270                                         $post['remote_images'][] = ['url' => $picture[1], 'description' => ''];
 271                                 }
 272                         }
 273                 }
 274
 275                 // if nothing is found, it maybe having an image.
 276                 if (!isset($post['type'])) {
 277                         // Simplify image codes
 278                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 279                         $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 280                         $post['text'] = $body;
 281
 282                         if (preg_match_all("#\[url=([^\]]+?)\]\s*\[img\]([^\[]+?)\[/img\]\s*\[/url\]#ism", $body, $pictures, PREG_SET_ORDER)) {
 283                                 if ((count($pictures) == 1) && !$has_title) {
 284                                         if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) {
 285                                                 // Replace the preview picture with the real picture
 286                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
 287                                                 $data = ['url' => $url, 'type' => 'photo'];
 288                                         } else {
 289                                                 // Checking, if the link goes to a picture
 290                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1]);
 291                                         }
 292
 293                                         // Workaround:
 294                                         // Sometimes photo posts to the own album are not detected at the start.
 295                                         // So we seem to cannot use the cache for these cases. That's strange.
 296                                         if (($data['type'] != 'photo') && strstr($pictures[0][1], "/photos/")) {
 297                                                 $data = ParseUrl::getSiteinfo($pictures[0][1]);
 298                                         }
 299
 300                                         if ($data['type'] == 'photo') {
 301                                                 $post['type'] = 'photo';
 302                                                 if (isset($data['images'][0])) {
 303                                                         $post['image'] = $data['images'][0]['src'];
 304                                                         $post['url'] = $data['url'];
 305                                                 } else {
 306                                                         $post['image'] = $data['url'];
 307                                                 }
 308
 309                                                 $post['preview'] = $pictures[0][2];
 310                                                 $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 311                                         } else {
 312                                                 $imgdata = Images::getInfoFromURLCached($pictures[0][1]);
 313                                                 if ($imgdata && substr($imgdata['mime'], 0, 6) == 'image/') {
 314                                                         $post['type'] = 'photo';
 315                                                         $post['image'] = $pictures[0][1];
 316                                                         $post['preview'] = $pictures[0][2];
 317                                                         $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 318                                                 }
 319                                         }
 320                                 } elseif (count($pictures) > 0) {
 321                                         $post['type'] = 'link';
 322                                         $post['url'] = $plink;
 323                                         $post['image'] = $pictures[0][2];
 324                                         $post['text'] = $body;
 325
 326                                         foreach ($pictures as $picture) {
 327                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 328                                         }
 329                                 }
 330                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 331                                 if ($has_title) {
 332                                         $post['type'] = 'link';
 333                                         $post['url'] = $plink;
 334                                 } else {
 335                                         $post['type'] = 'photo';
 336                                 }
 337
 338                                 $post['image'] = $pictures[0][1];
 339                                 $post['text'] = $body;
 340                                 foreach ($pictures as $picture) {
 341                                         $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 342                                 }
 343                         }
 344
 345                         // Test for the external links
 346                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER);
 347                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER);
 348
 349                         $links = array_merge($links1, $links2);
 350
 351                         // If there is only a single one, then use it.
 352                         // This should cover link posts via API.
 353                         if ((count($links) == 1) && !isset($post['preview']) && !$has_title) {
 354                                 $post['type'] = 'link';
 355                                 $post['url'] = $links[0][1];
 356                         }
 357
 358                         // Simplify "video" element
 359                         $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']);
 360
 361                         // Now count the number of external media links
 362                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
 363                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
 364                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
 365                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
 366
 367                         // Add them to the other external links
 368                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 369
 370                         // Are there more than one?
 371                         if (count($links) > 1) {
 372                                 // The post will be the type "text", which means a blog post
 373                                 unset($post['type']);
 374                                 $post['url'] = $plink;
 375                         }
 376
 377                         if (!isset($post['type'])) {
 378                                 $post['type'] = "text";
 379                                 $post['text'] = trim($body);
 380                         }
 381
 382                         if (($post['type'] == 'photo') && empty($post['images']) && !empty($post['remote_images'])) {
 383                                 $post['images'] = $post['remote_images'];
 384                                 $post['image'] = $post['images'][0]['url'];
 385                                 if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 386                                         $post['image_description'] = $post['images'][0]['description'];
 387                                 }
 388                         }
 389                         unset($post['remote_images']);
 390                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
 391                         $data = ParseUrl::getSiteinfoCached($post['url']);
 392
 393                         if (isset($data['images'][0])) {
 394                                 $post['image'] = $data['images'][0]['src'];
 395                         }
 396                 }
 397
 398                 return $post;
 399         }
 400
 401         /**
 402          * Remove [attachment] BBCode and replaces it with a regular [url]
 403          *
 404          * @param string  $body
 405          * @param boolean $no_link_desc No link description
 406          *
 407          * @return string with replaced body
 408          */
 409         public static function removeAttachment($body, $no_link_desc = false)
 410         {
 411                 return preg_replace_callback("/\s*\[attachment (.*)\](.*?)\[\/attachment\]\s*/ism",
 412                         function ($match) use ($no_link_desc) {
 413                                 $attach_data = self::getAttachmentData($match[0]);
 414                                 if (empty($attach_data['url'])) {
 415                                         return $match[0];
 416                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
 417                                         return "\n[url]" . $attach_data['url'] . "[/url]\n";
 418                                 } else {
 419                                         return "\n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
 420                                 }
 421                 }, $body);
 422         }
 423
 424         /**
 425          * Converts a BBCode text into plaintext
 426          *
 427          * @param      $text
 428          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 429          *
 430          * @return string
 431          */
 432         public static function toPlaintext($text, $keep_urls = true)
 433         {
 434                 $naked_text = HTML::toPlaintext(self::convert($text, false, 0, true), 0, !$keep_urls);
 435
 436                 return $naked_text;
 437         }
 438
 439         private static function proxyUrl($image, $simplehtml = self::INTERNAL)
 440         {
 441                 // Only send proxied pictures to API and for internal display
 442                 if (in_array($simplehtml, [self::INTERNAL, self::API])) {
 443                         return ProxyUtils::proxifyUrl($image);
 444                 } else {
 445                         return $image;
 446                 }
 447         }
 448
 449         /**
 450          * This function changing the visual size (not the real size) of images.
 451          * The function does not work for pictures with an alternate text description.
 452          * This could only be changed by using some new "img" BBCode format.
 453          *
 454          * @param string $srctext The body with images
 455          * @return string The body with possibly scaled images
 456          */
 457         public static function scaleExternalImages(string $srctext)
 458         {
 459                 $s = $srctext;
 460
 461                 // Simplify image links
 462                 $s = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $s);
 463
 464                 $matches = null;
 465                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 466                 if ($c) {
 467                         foreach ($matches as $mtch) {
 468                                 Logger::info('scale_external_image', ['image' => $mtch[1]]);
 469
 470                                 $hostname = str_replace('www.', '', substr(DI::baseUrl(), strpos(DI::baseUrl(), '://') + 3));
 471                                 if (stristr($mtch[1], $hostname)) {
 472                                         continue;
 473                                 }
 474
 475                                 $curlResult = DI::httpRequest()->get($mtch[1]);
 476                                 if (!$curlResult->isSuccess()) {
 477                                         continue;
 478                                 }
 479
 480                                 $i = $curlResult->getBody();
 481                                 $type = $curlResult->getContentType();
 482                                 $type = Images::getMimeTypeByData($i, $mtch[1], $type);
 483
 484                                 if ($i) {
 485                                         $Image = new Image($i, $type);
 486                                         if ($Image->isValid()) {
 487                                                 $orig_width = $Image->getWidth();
 488                                                 $orig_height = $Image->getHeight();
 489
 490                                                 if ($orig_width > 640 || $orig_height > 640) {
 491                                                         $Image->scaleDown(640);
 492                                                         $new_width = $Image->getWidth();
 493                                                         $new_height = $Image->getHeight();
 494                                                         Logger::info('External images scaled', ['orig_width' => $orig_width, 'new_width' => $new_width, 'orig_height' => $orig_height, 'new_height' => $new_height, 'match' => $mtch[0]]);
 495                                                         $s = str_replace(
 496                                                                 $mtch[0],
 497                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]'
 498                                                                 . "\n",
 499                                                                 $s
 500                                                         );
 501                                                         Logger::info('New string', ['image' => $s]);
 502                                                 }
 503                                         }
 504                                 }
 505                         }
 506                 }
 507
 508                 return $s;
 509         }
 510
 511         /**
 512          * Truncates imported message body string length to max_import_size
 513          *
 514          * The purpose of this function is to apply system message length limits to
 515          * imported messages without including any embedded photos in the length
 516          *
 517          * @param string $body
 518          * @return string
 519          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 520          */
 521         public static function limitBodySize($body)
 522         {
 523                 $maxlen = DI::config()->get('config', 'max_import_size', 0);
 524
 525                 // If the length of the body, including the embedded images, is smaller
 526                 // than the maximum, then don't waste time looking for the images
 527                 if ($maxlen && (strlen($body) > $maxlen)) {
 528
 529                         Logger::info('the total body length exceeds the limit', ['maxlen' => $maxlen, 'body_len' => strlen($body)]);
 530
 531                         $orig_body = $body;
 532                         $new_body = '';
 533                         $textlen = 0;
 534
 535                         $img_start = strpos($orig_body, '[img');
 536                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 537                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 538                         while (($img_st_close !== false) && ($img_end !== false)) {
 539
 540                                 $img_st_close++; // make it point to AFTER the closing bracket
 541                                 $img_end += $img_start;
 542                                 $img_end += strlen('[/img]');
 543
 544                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 545                                         // This is an embedded image
 546
 547                                         if (($textlen + $img_start) > $maxlen) {
 548                                                 if ($textlen < $maxlen) {
 549                                                         Logger::info('the limit happens before an embedded image');
 550                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 551                                                         $textlen = $maxlen;
 552                                                 }
 553                                         } else {
 554                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 555                                                 $textlen += $img_start;
 556                                         }
 557
 558                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 559                                 } else {
 560
 561                                         if (($textlen + $img_end) > $maxlen) {
 562                                                 if ($textlen < $maxlen) {
 563                                                         Logger::info('the limit happens before the end of a non-embedded image');
 564                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 565                                                         $textlen = $maxlen;
 566                                                 }
 567                                         } else {
 568                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 569                                                 $textlen += $img_end;
 570                                         }
 571                                 }
 572                                 $orig_body = substr($orig_body, $img_end);
 573
 574                                 if ($orig_body === false) {
 575                                         // in case the body ends on a closing image tag
 576                                         $orig_body = '';
 577                                 }
 578
 579                                 $img_start = strpos($orig_body, '[img');
 580                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 581                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 582                         }
 583
 584                         if (($textlen + strlen($orig_body)) > $maxlen) {
 585                                 if ($textlen < $maxlen) {
 586                                         Logger::info('the limit happens after the end of the last image');
 587                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 588                                 }
 589                         } else {
 590                                 Logger::info('the text size with embedded images extracted did not violate the limit');
 591                                 $new_body = $new_body . $orig_body;
 592                         }
 593
 594                         return $new_body;
 595                 } else {
 596                         return $body;
 597                 }
 598         }
 599
 600         /**
 601          * Processes [attachment] tags
 602          *
 603          * Note: Can produce a [bookmark] tag in the returned string
 604          *
 605          * @param string  $text
 606          * @param integer $simplehtml
 607          * @param bool    $tryoembed
 608          * @param array   $data
 609          * @return string
 610          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 611          */
 612         public static function convertAttachment($text, $simplehtml = self::INTERNAL, $tryoembed = true, array $data = [])
 613         {
 614                 $data = $data ?: self::getAttachmentData($text);
 615                 if (empty($data) || empty($data['url'])) {
 616                         return $text;
 617                 }
 618
 619                 $stamp1 = microtime(true);
 620
 621                 if (isset($data['title'])) {
 622                         $data['title'] = strip_tags($data['title']);
 623                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 624                 } else {
 625                         $data['title'] = null;
 626                 }
 627
 628                 if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) {
 629                         $data['preview'] = $data['image'];
 630                         $data['image'] = '';
 631                 }
 632
 633                 $return = '';
 634                 try {
 635                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 636                                 $return = OEmbed::getHTML($data['url'], $data['title']);
 637                         } else {
 638                                 throw new Exception('OEmbed is disabled for this attachment.');
 639                         }
 640                 } catch (Exception $e) {
 641                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 642
 643                         if ($simplehtml != self::CONNECTORS) {
 644                                 $return = sprintf('<div class="type-%s">', $data['type']);
 645                         }
 646
 647                         if (!empty($data['title']) && !empty($data['url'])) {
 648                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 649                                         $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), $data['title']);
 650                                 } else {
 651                                         if (!empty($data['image'])) {
 652                                                 $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), $data['title']);
 653                                         } elseif (!empty($data['preview'])) {
 654                                                 $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), $data['title']);
 655                                         }
 656                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 657                                 }
 658                         }
 659
 660                         if (!empty($data['description']) && $data['description'] != $data['title']) {
 661                                 // Sanitize the HTML
 662                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(HTML::purify($data['description'])));
 663                         }
 664
 665                         if (!empty($data['provider_url']) && !empty($data['provider_name'])) {
 666                                 if (!empty($data['author_name'])) {
 667                                         $return .= sprintf('<sup><a href="%s">%s (%s)</a></sup>', $data['provider_url'], $data['author_name'], $data['provider_name']);
 668                                 } else {
 669                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['provider_url'], $data['provider_name']);
 670                                 }
 671                         }
 672
 673                         if ($simplehtml != self::CONNECTORS) {
 674                                 $return .= '</div>';
 675                         }
 676                 }
 677
 678                 DI::profiler()->saveTimestamp($stamp1, 'rendering');
 679                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 680         }
 681
 682         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 683         {
 684                 $data = self::getAttachmentData($Text);
 685
 686                 if (!$data) {
 687                         return $Text;
 688                 } elseif ($nolink) {
 689                         return $data['text'] . ($data['after'] ?? '');
 690                 }
 691
 692                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 693                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 694                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 695                         $data['title'] = $data['url'];
 696                 } elseif (($text != '') && strstr($title, $text)) {
 697                         $data['text'] = $data['title'];
 698                         $data['title'] = $data['url'];
 699                 }
 700
 701                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 702                         return $data['title'] . $data['after'];
 703                 }
 704
 705                 // If the link already is included in the post, don't add it again
 706                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 707                         return $data['text'] . $data['after'];
 708                 }
 709
 710                 $text = $data['text'];
 711
 712                 if (!empty($data['url']) && !empty($data['title'])) {
 713                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 714                 } elseif (!empty($data['url'])) {
 715                         $text .= "\n[url]" . $data['url'] . '[/url]';
 716                 }
 717
 718                 return $text . "\n" . $data['after'];
 719         }
 720
 721         /**
 722          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 723          *
 724          * @param array $match Array with the matching values
 725          * @return string reformatted link including HTML codes
 726          */
 727         private static function convertUrlForActivityPubCallback($match)
 728         {
 729                 $url = $match[1];
 730
 731                 if (isset($match[2]) && ($match[1] != $match[2])) {
 732                         return $match[0];
 733                 }
 734
 735                 $parts = parse_url($url);
 736                 if (!isset($parts['scheme'])) {
 737                         return $match[0];
 738                 }
 739
 740                 return self::convertUrlForActivityPub($url);
 741         }
 742
 743         /**
 744          * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 745          *
 746          * @param string $url URL that is about to be reformatted
 747          * @return string reformatted link including HTML codes
 748          */
 749         private static function convertUrlForActivityPub($url)
 750         {
 751                 $html = '<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>';
 752                 return sprintf($html, $url, self::getStyledURL($url));
 753         }
 754
 755         /**
 756          * Converts an URL in a nicer format (without the scheme and possibly shortened)
 757          *
 758          * @param string $url URL that is about to be reformatted
 759          * @return string reformatted link
 760          */
 761         private static function getStyledURL($url)
 762         {
 763                 $parts = parse_url($url);
 764                 $scheme = $parts['scheme'] . '://';
 765                 $styled_url = str_replace($scheme, '', $url);
 766
 767                 if (strlen($styled_url) > 30) {
 768                         $styled_url = substr($styled_url, 0, 30) . "…";
 769                 }
 770
 771                 return $styled_url;
 772         }
 773
 774         /*
 775          * [noparse][i]italic[/i][/noparse] turns into
 776          * [noparse][ i ]italic[ /i ][/noparse],
 777          * to hide them from parser.
 778          */
 779         private static function escapeNoparseCallback($match)
 780         {
 781                 $whole_match = $match[0];
 782                 $captured = $match[1];
 783                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 784                 $new_str = str_replace($captured, $spacefied, $whole_match);
 785                 return $new_str;
 786         }
 787
 788         /*
 789          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 790          * now turns back and the [noparse] tags are trimed
 791          * returning [i]italic[/i]
 792          */
 793         private static function unescapeNoparseCallback($match)
 794         {
 795                 $captured = $match[1];
 796                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 797                 return $unspacefied;
 798         }
 799
 800         /**
 801          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 802          * occurrences
 803          *
 804          * @param string $text        Text to search
 805          * @param string $name        Tag name
 806          * @param int    $occurrences Number of first occurrences to skip
 807          * @return boolean|array
 808          */
 809         public static function getTagPosition($text, $name, $occurrences = 0)
 810         {
 811                 if ($occurrences < 0) {
 812                         $occurrences = 0;
 813                 }
 814
 815                 $start_open = -1;
 816                 for ($i = 0; $i <= $occurrences; $i++) {
 817                         if ($start_open !== false) {
 818                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 819                         }
 820                 }
 821
 822                 if ($start_open === false) {
 823                         return false;
 824                 }
 825
 826                 $start_equal = strpos($text, '=', $start_open);
 827                 $start_close = strpos($text, ']', $start_open);
 828
 829                 if ($start_close === false) {
 830                         return false;
 831                 }
 832
 833                 $start_close++;
 834
 835                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 836
 837                 if ($end_open === false) {
 838                         return false;
 839                 }
 840
 841                 $res = [
 842                         'start' => [
 843                                 'open' => $start_open,
 844                                 'close' => $start_close
 845                         ],
 846                         'end' => [
 847                                 'open' => $end_open,
 848                                 'close' => $end_open + strlen('[/' . $name . ']')
 849                         ],
 850                 ];
 851
 852                 if ($start_equal !== false) {
 853                         $res['start']['equal'] = $start_equal + 1;
 854                 }
 855
 856                 return $res;
 857         }
 858
 859         /**
 860          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 861          *
 862          * @param string $pattern Preg pattern string
 863          * @param string $replace Preg replace string
 864          * @param string $name    BBCode tag name
 865          * @param string $text    Text to search
 866          * @return string
 867          */
 868         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 869         {
 870                 $occurrences = 0;
 871                 $pos = self::getTagPosition($text, $name, $occurrences);
 872                 while ($pos !== false && $occurrences++ < 1000) {
 873                         $start = substr($text, 0, $pos['start']['open']);
 874                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 875                         $end = substr($text, $pos['end']['close']);
 876                         if ($end === false) {
 877                                 $end = '';
 878                         }
 879
 880                         $subject = preg_replace($pattern, $replace, $subject);
 881                         $text = $start . $subject . $end;
 882
 883                         $pos = self::getTagPosition($text, $name, $occurrences);
 884                 }
 885
 886                 return $text;
 887         }
 888
 889         private static function extractImagesFromItemBody($body)
 890         {
 891                 $saved_image = [];
 892                 $orig_body = $body;
 893                 $new_body = '';
 894
 895                 $cnt = 0;
 896                 $img_start = strpos($orig_body, '[img');
 897                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 898                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 899                 while (($img_st_close !== false) && ($img_end !== false)) {
 900                         $img_st_close++; // make it point to AFTER the closing bracket
 901                         $img_end += $img_start;
 902
 903                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 904                                 // This is an embedded image
 905                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 906                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 907
 908                                 $cnt++;
 909                         } else {
 910                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 911                         }
 912
 913                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 914
 915                         if ($orig_body === false) {
 916                                 // in case the body ends on a closing image tag
 917                                 $orig_body = '';
 918                         }
 919
 920                         $img_start = strpos($orig_body, '[img');
 921                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 922                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 923                 }
 924
 925                 $new_body = $new_body . $orig_body;
 926
 927                 return ['body' => $new_body, 'images' => $saved_image];
 928         }
 929
 930         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 931         {
 932                 $newbody = $body;
 933
 934                 $cnt = 0;
 935                 foreach ($images as $image) {
 936                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 937                         // it loops over the array starting from the first element and going sequentially
 938                         // to the last element
 939                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 940                                 '<img src="' . self::proxyUrl($image) . '" alt="' . DI::l10n()->t('Image/photo') . '" />', $newbody);
 941                         $cnt++;
 942                 }
 943
 944                 return $newbody;
 945         }
 946
 947         /**
 948          *
 949          * @param  string   $text     A BBCode string
 950          * @return array share attributes
 951          */
 952         public static function fetchShareAttributes($text)
 953         {
 954                 $attributes = [];
 955                 if (!preg_match("/(.*?)\[share(.*?)\](.*)\[\/share\]/ism", $text, $matches)) {
 956                         return $attributes;
 957                 }
 958
 959                 $attribute_string = $matches[2];
 960                 foreach (['author', 'profile', 'avatar', 'link', 'posted', 'guid'] as $field) {
 961                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 962                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 963                 }
 964                 return $attributes;
 965         }
 966
 967         /**
 968          * This function converts a [share] block to text according to a provided callback function whose signature is:
 969          *
 970          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 971          *
 972          * Where:
 973          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 974          * data lookup
 975          * - $author_contact is a contact record array
 976          * - $content is the inner content of the [share] block
 977          * - $is_quote_share indicates whether there's any content before the [share] block
 978          * - Return value is the string that should replace the [share] block in the provided text
 979          *
 980          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 981          *
 982          * @param  string   $text     A BBCode string
 983          * @param  callable $callback
 984          * @return string The BBCode string with all [share] blocks replaced
 985          */
 986         public static function convertShare($text, callable $callback)
 987         {
 988                 $return = preg_replace_callback(
 989                         "/(.*?)\[share(.*?)\](.*)\[\/share\]/ism",
 990                         function ($match) use ($callback) {
 991                                 $attribute_string = $match[2];
 992                                 $attributes = [];
 993                                 foreach (['author', 'profile', 'avatar', 'link', 'posted', 'guid'] as $field) {
 994                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 995                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 996                                 }
 997
 998                                 $author_contact = Contact::getByURL($attributes['profile'], false, ['url', 'addr', 'name', 'micro']);
 999                                 $author_contact['url'] = ($author_contact['url'] ?? $attributes['profile']);
1000                                 $author_contact['addr'] = ($author_contact['addr'] ?? '') ?: Protocol::getAddrFromProfileUrl($attributes['profile']);
1001
1002                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
1003                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
1004                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
1005
1006                                 if ($attributes['avatar']) {
1007                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
1008                                 }
1009
1010                                 $content = preg_replace(Strings::autoLinkRegEx(), '<a href="$1">$1</a>', $match[3]);
1011
1012                                 return $match[1] . $callback($attributes, $author_contact, $content, trim($match[1]) != '');
1013                         },
1014                         $text
1015                 );
1016
1017                 return $return;
1018         }
1019
1020         /**
1021          * Default [share] tag conversion callback
1022          *
1023          * Note: Can produce a [bookmark] tag in the output
1024          *
1025          * @see BBCode::convertShare()
1026          * @param array   $attributes     [share] block attribute values
1027          * @param array   $author_contact Contact row of the shared author
1028          * @param string  $content        Inner content of the [share] block
1029          * @param boolean $is_quote_share Whether there is content before the [share] block
1030          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
1031          * @return string
1032          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1033          */
1034         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
1035         {
1036                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
1037
1038                 switch ($simplehtml) {
1039                         case self::API:
1040                                 $text = ($is_quote_share? '<br>' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1041                                 break;
1042                         case self::DIASPORA:
1043                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1044                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1045                                 } else {
1046                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1047
1048                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1049                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1050                                         }
1051
1052                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1053
1054                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1055                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1056                                         }
1057                                 }
1058
1059                                 break;
1060                         case self::CONNECTORS:
1061                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1062                                 $headline .= DI::l10n()->t('<a href="%1$s" target="_blank" rel="noopener noreferrer">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1063                                 $headline .= ':</b></p>' . "\n";
1064
1065                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1066
1067                                 break;
1068                         case self::OSTATUS:
1069                                 $text = ($is_quote_share? '<br>' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1070                                 break;
1071                         case self::ACTIVITYPUB:
1072                                 $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>';
1073                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1074                                 break;
1075                         default:
1076                                 $text = ($is_quote_share? "\n" : '');
1077
1078                                 $contact = Contact::getByURL($attributes['profile'], false, ['network']);
1079                                 $network = $contact['network'] ?? Protocol::PHANTOM;
1080
1081                                 $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1082                                 $text .= Renderer::replaceMacros($tpl, [
1083                                         '$profile'      => $attributes['profile'],
1084                                         '$avatar'       => $attributes['avatar'],
1085                                         '$author'       => $attributes['author'],
1086                                         '$link'         => $attributes['link'],
1087                                         '$link_title'   => DI::l10n()->t('link to source'),
1088                                         '$posted'       => $attributes['posted'],
1089                                         '$guid'         => $attributes['guid'],
1090                                         '$network_name' => ContactSelector::networkToName($network, $attributes['profile']),
1091                                         '$network_icon' => ContactSelector::networkToIcon($network, $attributes['profile']),
1092                                         '$content'      => self::setMentions(trim($content), 0, $network) . self::ANCHOR,
1093                                 ]);
1094                                 break;
1095                 }
1096
1097                 return $text;
1098         }
1099
1100         private static function removePictureLinksCallback($match)
1101         {
1102                 $cache_key = 'remove:' . $match[1];
1103                 $text = DI::cache()->get($cache_key);
1104
1105                 if (is_null($text)) {
1106                         $a = DI::app();
1107
1108                         $stamp1 = microtime(true);
1109
1110                         $ch = @curl_init($match[1]);
1111                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1112                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1113                         @curl_setopt($ch, CURLOPT_USERAGENT, DI::httpRequest()->getUserAgent());
1114                         @curl_exec($ch);
1115                         $curl_info = @curl_getinfo($ch);
1116
1117                         DI::profiler()->saveTimestamp($stamp1, "network");
1118
1119                         if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1120                                 $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]";
1121                         } else {
1122                                 $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]";
1123
1124                                 // if its not a picture then look if its a page that contains a picture link
1125                                 $body = DI::httpRequest()->fetch($match[1]);
1126                                 if (empty($body)) {
1127                                         DI::cache()->set($cache_key, $text);
1128                                         return $text;
1129                                 }
1130
1131                                 $doc = new DOMDocument();
1132                                 @$doc->loadHTML($body);
1133                                 $xpath = new DOMXPath($doc);
1134                                 $list = $xpath->query("//meta[@name]");
1135                                 foreach ($list as $node) {
1136                                         $attr = [];
1137
1138                                         if ($node->attributes->length) {
1139                                                 foreach ($node->attributes as $attribute) {
1140                                                         $attr[$attribute->name] = $attribute->value;
1141                                                 }
1142                                         }
1143
1144                                         if (strtolower($attr['name']) == 'twitter:image') {
1145                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1146                                         }
1147                                 }
1148                         }
1149                         DI::cache()->set($cache_key, $text);
1150                 }
1151
1152                 return $text;
1153         }
1154
1155         private static function expandLinksCallback($match)
1156         {
1157                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1158                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1159                 } else {
1160                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1161                 }
1162         }
1163
1164         private static function cleanPictureLinksCallback($match)
1165         {
1166                 // When the picture link is the own photo path then we can avoid fetching the link
1167                 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()->get()) . '/photos/');
1168                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1169                         if (!empty($match[3])) {
1170                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1171                         } else {
1172                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1173                         }
1174                         return $text;
1175                 }
1176
1177                 $cache_key = 'clean:' . $match[1];
1178                 $text = DI::cache()->get($cache_key);
1179                 if (!is_null($text)) {
1180                         return $text;
1181                 }
1182
1183                 // Only fetch the header, not the content
1184                 $stamp1 = microtime(true);
1185
1186                 $ch = @curl_init($match[1]);
1187                 @curl_setopt($ch, CURLOPT_NOBODY, true);
1188                 @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1189                 @curl_setopt($ch, CURLOPT_USERAGENT, DI::httpRequest()->getUserAgent());
1190                 @curl_exec($ch);
1191                 $curl_info = @curl_getinfo($ch);
1192
1193                 DI::profiler()->saveTimestamp($stamp1, "network");
1194
1195                 // if its a link to a picture then embed this picture
1196                 if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1197                         $text = '[img]' . $match[1] . '[/img]';
1198                 } else {
1199                         if (!empty($match[3])) {
1200                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1201                         } else {
1202                                 $text = '[img]' . $match[2] . '[/img]';
1203                         }
1204
1205                         // if its not a picture then look if its a page that contains a picture link
1206                         $body = DI::httpRequest()->fetch($match[1]);
1207                         if (empty($body)) {
1208                                 DI::cache()->set($cache_key, $text);
1209                                 return $text;
1210                         }
1211
1212                         $doc = new DOMDocument();
1213                         @$doc->loadHTML($body);
1214                         $xpath = new DOMXPath($doc);
1215                         $list = $xpath->query("//meta[@name]");
1216                         foreach ($list as $node) {
1217                                 $attr = [];
1218                                 if ($node->attributes->length) {
1219                                         foreach ($node->attributes as $attribute) {
1220                                                 $attr[$attribute->name] = $attribute->value;
1221                                         }
1222                                 }
1223
1224                                 if (strtolower($attr['name']) == "twitter:image") {
1225                                         if (!empty($match[3])) {
1226                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1227                                         } else {
1228                                                 $text = "[img]" . $attr['content'] . "[/img]";
1229                                         }
1230                                 }
1231                         }
1232                 }
1233                 DI::cache()->set($cache_key, $text);
1234
1235                 return $text;
1236         }
1237
1238         public static function cleanPictureLinks($text)
1239         {
1240                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1241                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1242                 return $return;
1243         }
1244
1245         public static function removeLinks(string $bbcode)
1246         {
1247                 $bbcode = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $1 ', $bbcode);
1248                 $bbcode = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $bbcode);
1249
1250                 $bbcode = preg_replace('/[@!#]\[url\=.*?\].*?\[\/url\]/ism', '', $bbcode);
1251                 $bbcode = preg_replace("/\[url=[^\[\]]*\](.*)\[\/url\]/Usi", ' $1 ', $bbcode);
1252                 $bbcode = preg_replace('/[@!#]?\[url.*?\[\/url\]/ism', '', $bbcode);
1253                 return $bbcode;
1254         }
1255
1256         /**
1257          * Converts a BBCode message to HTML message
1258          *
1259          * BBcode 2 HTML was written by WAY2WEB.net
1260          * extended to work with Mistpark/Friendica - Mike Macgirvin
1261          *
1262          * Simple HTML values meaning:
1263          * - 0: Friendica display
1264          * - 1: Unused
1265          * - 2: Used for Windows Phone push, Friendica API
1266          * - 3: Used before converting to Markdown in bb2diaspora.php
1267          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1268          * - 5: Unused
1269          * - 6: Unused
1270          * - 7: Used for dfrn, OStatus
1271          * - 8: Used for Twitter, WP backlink text setting
1272          * - 9: ActivityPub
1273          *
1274          * @param string $text
1275          * @param bool   $try_oembed
1276          * @param int    $simple_html
1277          * @param bool   $for_plaintext
1278          * @return string
1279          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1280          */
1281         public static function convert(string $text = null, $try_oembed = true, $simple_html = self::INTERNAL, $for_plaintext = false)
1282         {
1283                 // Accounting for null default column values
1284                 if (is_null($text) || $text === '') {
1285                         return '';
1286                 }
1287
1288                 Hook::callAll('bbcode', $text);
1289
1290                 $a = DI::app();
1291
1292                 $text = self::performWithEscapedTags($text, ['code'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) {
1293                         $text = self::performWithEscapedTags($text, ['noparse', 'nobb', 'pre'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) {
1294                                 /*
1295                                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1296                                  *
1297                                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1298                                  * $match[1] = $url
1299                                  * $match[2] = $title or absent
1300                                  */
1301                                 $try_oembed_callback = function ($match)
1302                                 {
1303                                         $url = $match[1];
1304                                         $title = $match[2] ?? null;
1305
1306                                         try {
1307                                                 $return = OEmbed::getHTML($url, $title);
1308                                         } catch (Exception $ex) {
1309                                                 $return = $match[0];
1310                                         }
1311
1312                                         return $return;
1313                                 };
1314
1315                                 // Remove the abstract element. It is a non visible element.
1316                                 $text = self::stripAbstract($text);
1317
1318                                 // Line ending normalisation
1319                                 $text = str_replace("\r\n", "\n", $text);
1320
1321                                 // Move new lines outside of tags
1322                                 $text = preg_replace("#\[(\w*)](\n*)#ism", '$2[$1]', $text);
1323                                 $text = preg_replace("#(\n*)\[/(\w*)]#ism", '[/$2]$1', $text);
1324
1325                                 // Extract the private images which use data urls since preg has issues with
1326                                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1327                                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1328
1329                                 $extracted = self::extractImagesFromItemBody($text);
1330                                 $text = $extracted['body'];
1331                                 $saved_image = $extracted['images'];
1332
1333                                 // If we find any event code, turn it into an event.
1334                                 // After we're finished processing the bbcode we'll
1335                                 // replace all of the event code with a reformatted version.
1336
1337                                 $ev = Event::fromBBCode($text);
1338
1339                                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1340                                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1341
1342                                 $text = str_replace("<", "&lt;", $text);
1343                                 $text = str_replace(">", "&gt;", $text);
1344
1345                                 // remove some newlines before the general conversion
1346                                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1347                                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1348
1349                                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1350                                 if (!$try_oembed) {
1351                                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1352                                 }
1353
1354                                 // Remove linefeeds inside of the table elements. See issue #6799
1355                                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1356                                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1357                                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1358                                         "\n[hr]", "[hr]\n", " [hr]", "[hr] ",
1359                                         "\n[attachment ", " [attachment ", "\n[/attachment]", "[/attachment]\n", " [/attachment]", "[/attachment] ",
1360                                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1361                                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1362                                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1363                                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1364                                         "[hr]", "[hr]", "[hr]", "[hr]",
1365                                         "[attachment ", "[attachment ", "[/attachment]", "[/attachment]", "[/attachment]", "[/attachment]",
1366                                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1367                                 do {
1368                                         $oldtext = $text;
1369                                         $text = str_replace($search, $replace, $text);
1370                                 } while ($oldtext != $text);
1371
1372                                 // Replace these here only once
1373                                 $search = ["\n[table]", "[/table]\n"];
1374                                 $replace = ["[table]", "[/table]"];
1375                                 $text = str_replace($search, $replace, $text);
1376
1377                                 // Trim new lines regardless of the system.remove_multiplicated_lines config value
1378                                 $text = trim($text, "\n");
1379
1380                                 // removing multiplicated newlines
1381                                 if (DI::config()->get('system', 'remove_multiplicated_lines')) {
1382                                         $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",
1383                                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1384                                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[*]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1385                                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1386                                         do {
1387                                                 $oldtext = $text;
1388                                                 $text = str_replace($search, $replace, $text);
1389                                         } while ($oldtext != $text);
1390                                 }
1391
1392                                 // Add HTML new lines
1393                                 $text = str_replace("\n", '<br>', $text);
1394
1395                                 /// @todo Have a closer look at the different html modes
1396                                 // Handle attached links or videos
1397                                 if ($simple_html == self::ACTIVITYPUB) {
1398                                         $text = self::removeAttachment($text);
1399                                 } elseif (!in_array($simple_html, [self::INTERNAL, self::CONNECTORS])) {
1400                                         $text = self::removeAttachment($text, true);
1401                                 } else {
1402                                         $text = self::convertAttachment($text, $simple_html, $try_oembed);
1403                                 }
1404
1405                                 $nosmile = strpos($text, '[nosmile]') !== false;
1406                                 $text = str_replace('[nosmile]', '', $text);
1407
1408                                 // Replace non graphical smilies for external posts
1409                                 if (!$nosmile && !$for_plaintext) {
1410                                         $text = self::performWithEscapedTags($text, ['img'], function ($text) {
1411                                                 return Smilies::replace($text);
1412                                         });
1413                                 }
1414
1415                                 // leave open the posibility of [map=something]
1416                                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1417                                 if (strpos($text, '[/map]') !== false) {
1418                                         $text = preg_replace_callback(
1419                                                 "/\[map\](.*?)\[\/map\]/ism",
1420                                                 function ($match) use ($simple_html) {
1421                                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1422                                                 },
1423                                                 $text
1424                                         );
1425                                 }
1426
1427                                 if (strpos($text, '[map=') !== false) {
1428                                         $text = preg_replace_callback(
1429                                                 "/\[map=(.*?)\]/ism",
1430                                                 function ($match) use ($simple_html) {
1431                                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1432                                                 },
1433                                                 $text
1434                                         );
1435                                 }
1436
1437                                 if (strpos($text, '[map]') !== false) {
1438                                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1439                                 }
1440
1441                                 // Check for headers
1442                                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1443                                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1444                                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1445                                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1446                                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1447                                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1448
1449                                 // Check for paragraph
1450                                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1451
1452                                 // Check for bold text
1453                                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1454
1455                                 // Check for Italics text
1456                                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1457
1458                                 // Check for Underline text
1459                                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1460
1461                                 // Check for strike-through text
1462                                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1463
1464                                 // Check for over-line text
1465                                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1466
1467                                 // Check for colored text
1468                                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1469
1470                                 // Check for sized text
1471                                 // [size=50] --> font-size: 50px (with the unit).
1472                                 if ($simple_html != self::DIASPORA) {
1473                                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1px;line-height:normal;">$2</span>', $text);
1474                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1;line-height:normal;">$2</span>', $text);
1475                                 } else {
1476                                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1477                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1478                                 }
1479
1480
1481                                 // Check for centered text
1482                                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", '<div style="text-align:center;">$1</div>', $text);
1483
1484                                 // Check for list text
1485                                 $text = str_replace("[*]", "<li>", $text);
1486
1487                                 // Check for style sheet commands
1488                                 $text = preg_replace("(\[style=(.*?)\](.*?)\[\/style\])ism", '<span style="$1">$2</span>', $text);
1489
1490                                 // Check for CSS classes
1491                                 $text = preg_replace("(\[class=(.*?)\](.*?)\[\/class\])ism", '<span class="$1">$2</span>', $text);
1492
1493                                 // handle nested lists
1494                                 $endlessloop = 0;
1495
1496                                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1497                                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1498                                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1499                                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1500                                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1501                                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1502                                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1503                                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1504                                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1505                                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1506                                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1507                                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1508                                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1509                                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1510                                 }
1511
1512                                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1513                                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1514                                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1515                                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1516
1517                                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1518                                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1519
1520                                 $text = str_replace('[hr]', '<hr />', $text);
1521
1522                                 if (!$for_plaintext) {
1523                                         $text = self::performWithEscapedTags($text, ['url', 'img', 'audio', 'video', 'youtube', 'vimeo', 'share', 'attachment', 'iframe', 'bookmark'], function ($text) {
1524                                                 return preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1525                                         });
1526                                 }
1527
1528                                 // Check for font change text
1529                                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1530
1531                                 // Declare the format for [spoiler] layout
1532                                 $SpoilerLayout = '<details class="spoiler"><summary>' . DI::l10n()->t('Click to open/close') . '</summary>$1</details>';
1533
1534                                 // Check for [spoiler] text
1535                                 // handle nested quotes
1536                                 $endlessloop = 0;
1537                                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1538                                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1539                                 }
1540
1541                                 // Check for [spoiler=Title] text
1542
1543                                 // handle nested quotes
1544                                 $endlessloop = 0;
1545                                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1546                                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1547                                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1548                                                 $text);
1549                                 }
1550
1551                                 // Declare the format for [quote] layout
1552                                 $QuoteLayout = '<blockquote>$1</blockquote>';
1553
1554                                 // Check for [quote] text
1555                                 // handle nested quotes
1556                                 $endlessloop = 0;
1557                                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1558                                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1559                                 }
1560
1561                                 // Check for [quote=Author] text
1562
1563                                 $t_wrote = DI::l10n()->t('$1 wrote:');
1564
1565                                 // handle nested quotes
1566                                 $endlessloop = 0;
1567                                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1568                                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1569                                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1570                                                 $text);
1571                                 }
1572
1573
1574                                 // [img=widthxheight]image source[/img]
1575                                 $text = preg_replace_callback(
1576                                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1577                                         function ($matches) use ($simple_html) {
1578                                                 if (strpos($matches[3], "data:image/") === 0) {
1579                                                         return $matches[0];
1580                                                 }
1581
1582                                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1583                                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1584                                         },
1585                                         $text
1586                                 );
1587
1588                                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1589                                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1590
1591                                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1592                                         function ($matches) use ($simple_html) {
1593                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1594                                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1595                                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1596                                         },
1597                                         $text);
1598
1599                                 // Images
1600                                 // [img]pathtoimage[/img]
1601                                 $text = preg_replace_callback(
1602                                         "/\[img\](.*?)\[\/img\]/ism",
1603                                         function ($matches) use ($simple_html) {
1604                                                 if (strpos($matches[1], "data:image/") === 0) {
1605                                                         return $matches[0];
1606                                                 }
1607
1608                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1609                                                 return "[img]" . $matches[1] . "[/img]";
1610                                         },
1611                                         $text
1612                                 );
1613
1614                                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1615                                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1616
1617                                 $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);
1618                                 $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>', $text);
1619                                 //$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);
1620
1621                                 // Simplify "video" element
1622                                 $text = preg_replace('(\[video[^\]]*?\ssrc\s?=\s?([^\s\]]+)[^\]]*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1623
1624                                 if ($try_oembed) {
1625                                         // html5 video and audio
1626                                         $text = preg_replace("/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4).*?)\[\/video\]/ism",
1627                                                 '<video src="$1" controls width="100%" height="auto"><a href="$1">$1</a></video>', $text);
1628
1629                                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1630                                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1631
1632                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1633                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1634                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '<audio src="$1" controls><a href="$1">$1</a></audio>', $text);
1635                                 } else {
1636                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1637                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1638                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1639                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1640                                 }
1641
1642                                 // Backward compatibility, [iframe] support has been removed in version 2020.12
1643                                 $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1644
1645                                 // Youtube extensions
1646                                 if ($try_oembed) {
1647                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1648                                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1649                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1650                                 }
1651
1652                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1653                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1654                                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1655
1656                                 if ($try_oembed) {
1657                                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism", '<iframe width="' . $a->videowidth . '" height="' . $a->videoheight . '" src="https://www.youtube.com/embed/$1" frameborder="0" ></iframe>', $text);
1658                                 } else {
1659                                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1660                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank" rel="noopener noreferrer">https://www.youtube.com/watch?v=$1</a>', $text);
1661                                 }
1662
1663                                 if ($try_oembed) {
1664                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1665                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1666                                 }
1667
1668                                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1669                                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1670
1671                                 if ($try_oembed) {
1672                                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism", '<iframe width="' . $a->videowidth . '" height="' . $a->videoheight . '" src="https://player.vimeo.com/video/$1" frameborder="0" ></iframe>', $text);
1673                                 } else {
1674                                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1675                                                 '<a href="https://vimeo.com/$1" target="_blank" rel="noopener noreferrer">https://vimeo.com/$1</a>', $text);
1676                                 }
1677
1678                                 // oembed tag
1679                                 $text = OEmbed::BBCode2HTML($text);
1680
1681                                 // Avoid triple linefeeds through oembed
1682                                 $text = str_replace("<br style='clear:left'></span><br><br>", "<br style='clear:left'></span><br>", $text);
1683
1684                                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1685                                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1686                                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1687                                 // start which is always required). Allow desc with a missing summary for compatibility.
1688
1689                                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1690                                         $sub = Event::getHTML($ev, $simple_html);
1691
1692                                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1693                                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1694                                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1695                                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1696                                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1697                                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1698                                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1699                                 }
1700
1701                                 if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA)) {
1702                                         $conv = html_entity_decode(str_replace([' ', "\n", "\r"], '', $text));
1703                                         // Emojis are always 4 byte Unicode characters
1704                                         if (!empty($conv) && (strlen($conv) / mb_strlen($conv) == 4)) {
1705                                                 $text = '<span style="font-size: xx-large; line-height: normal;">' . $text . '</span>';
1706                                         }
1707                                 }
1708
1709                                 if (!$for_plaintext) {
1710                                         if (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1711                                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1712                                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1713                                         }
1714                                 } else {
1715                                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1716                                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1717                                 }
1718
1719                                 // Remove all hashtag addresses
1720                                 if ($simple_html && !in_array($simple_html, [self::DIASPORA, self::OSTATUS, self::ACTIVITYPUB])) {
1721                                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1722                                 } elseif ($simple_html == self::DIASPORA) {
1723                                         // The ! is converted to @ since Diaspora only understands the @
1724                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1725                                                 '@<a href="$2">$3</a>',
1726                                                 $text);
1727                                 } elseif (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1728                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1729                                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1730                                                 $text);
1731                                 } elseif (!$simple_html) {
1732                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1733                                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1734                                                 $text);
1735                                 }
1736
1737                                 // Bookmarks in red - will be converted to bookmarks in friendica
1738                                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1739                                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1740                                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1741                                                         "[bookmark=$1]$2[/bookmark]", $text);
1742
1743                                 if (in_array($simple_html, [self::API, self::OSTATUS, self::TWITTER])) {
1744                                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1745                                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1746                                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1747                                 }
1748
1749                                 // Perform URL Search
1750                                 if ($try_oembed) {
1751                                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1752                                 }
1753
1754                                 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1755
1756                                 // Handle Diaspora posts
1757                                 $text = preg_replace_callback(
1758                                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1759                                         function ($match) {
1760                                                 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1761                                         }, $text
1762                                 );
1763
1764                                 $text = preg_replace_callback(
1765                                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1766                                         function ($match) {
1767                                                 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1768                                         }, $text
1769                                 );
1770
1771                                 // Server independent link to posts and comments
1772                                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1773                                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1774                                 $text = preg_replace($expression, DI::baseUrl()."/display/$1", $text);
1775
1776                                 /* Tag conversion
1777                                  * Supports:
1778                                  * - #[url=<anything>]<term>[/url]
1779                                  * - [url=<anything>]#<term>[/url]
1780                                  */
1781                                 $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) use ($simple_html) {
1782                                         if ($simple_html == BBCode::ACTIVITYPUB) {
1783                                                 return '<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1784                                                         . '" data-tag="' . XML::escape($matches[1]) . '" rel="tag ugc">#'
1785                                                         . XML::escape($matches[1]) . '</a>';
1786                                         } else {
1787                                                 return '#<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1788                                                         . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1789                                                         . XML::escape($matches[1]) . '</a>';
1790                                         }
1791                                 }, $text);
1792
1793                                 // We need no target="_blank" rel="noopener noreferrer" for local links
1794                                 // convert links start with DI::baseUrl() as local link without the target="_blank" rel="noopener noreferrer" attribute
1795                                 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
1796                                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1797                                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1798
1799                                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1800                                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1801
1802                                 // Red compatibility, though the link can't be authenticated on Friendica
1803                                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1804
1805
1806                                 // we may need to restrict this further if it picks up too many strays
1807                                 // link acct:user@host to a webfinger profile redirector
1808
1809                                 $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);
1810
1811                                 // Perform MAIL Search
1812                                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1813                                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1814
1815                                 /// @todo What is the meaning of these lines?
1816                                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1817                                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1818
1819                                 // Currently deactivated, it made problems with " inside of alt texts.
1820                                 //$text = preg_replace('/\&quot\;/', '"', $text);
1821
1822                                 // fix any escaped ampersands that may have been converted into links
1823                                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1824
1825                                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1826                                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1827
1828                                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1829
1830                                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1831                                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">', $text);
1832
1833                                 // sanitize href attributes (only allowlisted protocols URLs)
1834                                 // default value for backward compatibility
1835                                 $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
1836
1837                                 // Always allowed protocol even if config isn't set or not including it
1838                                 $allowed_link_protocols[] = '//';
1839                                 $allowed_link_protocols[] = 'http://';
1840                                 $allowed_link_protocols[] = 'https://';
1841                                 $allowed_link_protocols[] = 'redir/';
1842
1843                                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1844
1845                                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1846                                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text);
1847
1848                                 // Shared content
1849                                 $text = self::convertShare(
1850                                         $text,
1851                                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1852                                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1853                                         }
1854                                 );
1855
1856                                 $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1857
1858                                 return $text;
1859                         }); // Escaped noparse, nobb, pre
1860
1861                         // Remove escaping tags and replace new lines that remain
1862                         $text = preg_replace_callback('/\[(noparse|nobb)](.*?)\[\/\1]/ism', function ($match) {
1863                                 return str_replace("\n", "<br>", $match[2]);
1864                         }, $text);
1865
1866                         // Additionally, [pre] tags preserve spaces
1867                         $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", function ($match) {
1868                                 return str_replace([' ', "\n"], ['&nbsp;', "<br>"], htmlentities($match[1], ENT_NOQUOTES,'UTF-8'));
1869                         }, $text);
1870
1871                         return $text;
1872                 }); // Escaped code
1873
1874                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1875                         function ($matches) {
1876                                 if (strpos($matches[2], "\n") !== false) {
1877                                         $return = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlentities(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
1878                                 } else {
1879                                         $return = '<code>' . htmlentities($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
1880                                 }
1881
1882                                 return $return;
1883                         },
1884                         $text
1885                 );
1886
1887                 // Default iframe allowed domains/path
1888                 $allowedIframeDomains = [
1889                         DI::baseUrl()->getHostname()
1890                         . (DI::baseUrl()->getUrlPath() ? '/' . DI::baseUrl()->getUrlPath() : '')
1891                         . '/oembed/', # The path part has to change with the source in Content\Oembed::iframe
1892                         'www.youtube.com/embed/',
1893                         'player.vimeo.com/video/',
1894                 ];
1895
1896                 $allowedIframeDomains = array_merge(
1897                         $allowedIframeDomains,
1898                         DI::config()->get('system', 'allowed_oembed') ?
1899                                 explode(',', DI::config()->get('system', 'allowed_oembed'))
1900                                 : []
1901                 );
1902
1903                 $text = HTML::purify($text, $allowedIframeDomains);
1904
1905                 return $text;
1906         }
1907
1908         /**
1909          * Strips the "abstract" tag from the provided text
1910          *
1911          * @param string $text The text with BBCode
1912          * @return string The same text - but without "abstract" element
1913          */
1914         public static function stripAbstract($text)
1915         {
1916                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1917                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1918
1919                 return $text;
1920         }
1921
1922         /**
1923          * Returns the value of the "abstract" element
1924          *
1925          * @param string $text The text that maybe contains the element
1926          * @param string $addon The addon for which the abstract is meant for
1927          * @return string The abstract
1928          */
1929         public static function getAbstract($text, $addon = '')
1930         {
1931                 $abstract = '';
1932                 $abstracts = [];
1933                 $addon = strtolower($addon);
1934
1935                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1936                         foreach ($results AS $result) {
1937                                 $abstracts[strtolower($result[1])] = $result[2];
1938                         }
1939                 }
1940
1941                 if (isset($abstracts[$addon])) {
1942                         $abstract = $abstracts[$addon];
1943                 }
1944
1945                 if ($abstract == '' && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1946                         $abstract = $result[1];
1947                 }
1948
1949                 return $abstract;
1950         }
1951
1952         /**
1953          * Callback function to replace a Friendica style mention in a mention for Diaspora
1954          *
1955          * @param array $match Matching values for the callback
1956          *                     [1] = Mention type (! or @)
1957          *                     [2] = Name
1958          *                     [3] = Address
1959          * @return string Replaced mention
1960          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1961          * @throws \ImagickException
1962          */
1963         private static function bbCodeMention2DiasporaCallback($match)
1964         {
1965                 $contact = Contact::getByURL($match[3], false, ['addr']);
1966                 if (empty($contact['addr'])) {
1967                         return $match[0];
1968                 }
1969
1970                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
1971                 return $mention;
1972         }
1973
1974         /**
1975          * Converts a BBCode text into Markdown
1976          *
1977          * This function converts a BBCode item body to be sent to Markdown-enabled
1978          * systems like Diaspora and Libertree
1979          *
1980          * @param string $text
1981          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1982          * @return string
1983          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1984          */
1985         public static function toMarkdown($text, $for_diaspora = true)
1986         {
1987                 $original_text = $text;
1988
1989                 // Since Diaspora is creating a summary for links, this function removes them before posting
1990                 if ($for_diaspora) {
1991                         $text = self::removeShareInformation($text);
1992                 }
1993
1994                 /**
1995                  * Transform #tags, strip off the [url] and replace spaces with underscore
1996                  */
1997                 $url_search_string = "^\[\]";
1998                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
1999                         function ($matches) {
2000                                 return '#' . str_replace(' ', '_', $matches[2]);
2001                         },
2002                         $text
2003                 );
2004
2005                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2006                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2007
2008                 // Convert it to HTML - don't try oembed
2009                 if ($for_diaspora) {
2010                         $text = self::convert($text, false, self::DIASPORA);
2011
2012                         // Add all tags that maybe were removed
2013                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2014                                 $tagline = '';
2015                                 foreach ($tags[2] as $tag) {
2016                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2017                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2018                                                 $tagline .= '#' . $tag . ' ';
2019                                         }
2020                                 }
2021                                 $text = $text . " " . $tagline;
2022                         }
2023                 } else {
2024                         $text = self::convert($text, false, self::CONNECTORS);
2025                 }
2026
2027                 // If a link is followed by a quote then there should be a newline before it
2028                 // Maybe we should make this newline at every time before a quote.
2029                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2030
2031                 $stamp1 = microtime(true);
2032
2033                 // Now convert HTML to Markdown
2034                 $text = HTML::toMarkdown($text);
2035
2036                 DI::profiler()->saveTimestamp($stamp1, "parser");
2037
2038                 // Libertree has a problem with escaped hashtags.
2039                 $text = str_replace(['\#'], ['#'], $text);
2040
2041                 // Remove any leading or trailing whitespace, as this will mess up
2042                 // the Diaspora signature verification and cause the item to disappear
2043                 $text = trim($text);
2044
2045                 if ($for_diaspora) {
2046                         $url_search_string = "^\[\]";
2047                         $text = preg_replace_callback(
2048                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2049                                 ['self', 'bbCodeMention2DiasporaCallback'],
2050                                 $text
2051                         );
2052                 }
2053
2054                 Hook::callAll('bb2diaspora', $text);
2055
2056                 return $text;
2057         }
2058
2059         /**
2060          * Pull out all #hashtags and @person tags from $string.
2061          *
2062          * We also get @person@domain.com - which would make
2063          * the regex quite complicated as tags can also
2064          * end a sentence. So we'll run through our results
2065          * and strip the period from any tags which end with one.
2066          * Returns array of tags found, or empty array.
2067          *
2068          * @param string $string Post content
2069          *
2070          * @return array List of tag and person names
2071          */
2072         public static function getTags($string)
2073         {
2074                 $ret = [];
2075
2076                 BBCode::performWithEscapedTags($string, ['noparse', 'pre', 'code', 'img'], function ($string) use (&$ret) {
2077                         // Convert hashtag links to hashtags
2078                         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2 ', $string);
2079
2080                         // Force line feeds at bbtags
2081                         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2082
2083                         // ignore anything in a bbtag
2084                         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2085
2086                         // Match full names against @tags including the space between first and last
2087                         // We will look these up afterward to see if they are full names or not recognisable.
2088
2089                         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2090                                 foreach ($matches[1] as $match) {
2091                                         if (strstr($match, ']')) {
2092                                                 // we might be inside a bbcode color tag - leave it alone
2093                                                 continue;
2094                                         }
2095
2096                                         if (substr($match, -1, 1) === '.') {
2097                                                 $ret[] = substr($match, 0, -1);
2098                                         } else {
2099                                                 $ret[] = $match;
2100                                         }
2101                                 }
2102                         }
2103
2104                         // Otherwise pull out single word tags. These can be @nickname, @first_last
2105                         // and #hash tags.
2106
2107                         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?\']*[^\^ \x0D\x0A,;:?!\'.])/', $string, $matches)) {
2108                                 foreach ($matches[1] as $match) {
2109                                         if (strstr($match, ']')) {
2110                                                 // we might be inside a bbcode color tag - leave it alone
2111                                                 continue;
2112                                         }
2113
2114                                         // try not to catch url fragments
2115                                         if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2116                                                 continue;
2117                                         }
2118
2119                                         $ret[] = $match;
2120                                 }
2121                         }
2122                 });
2123
2124                 return array_unique($ret);
2125         }
2126
2127         /**
2128          * Perform a custom function on a text after having escaped blocks enclosed in the provided tag list.
2129          *
2130          * @param string   $text
2131          * @param array    $tagList A list of tag names, e.g ['noparse', 'nobb', 'pre']
2132          * @param callable $callback
2133          * @return string
2134          * @throws Exception
2135          *@see Strings::performWithEscapedBlocks
2136          *
2137          */
2138         public static function performWithEscapedTags(string $text, array $tagList, callable $callback)
2139         {
2140                 $tagList = array_map('preg_quote', $tagList);
2141
2142                 return Strings::performWithEscapedBlocks($text, '#\[(?:' . implode('|', $tagList) . ').*?\[/(?:' . implode('|', $tagList) . ')]#ism', $callback);
2143         }
2144
2145         /**
2146          * Replaces mentions in the provided message body for the provided user and network if any
2147          *
2148          * @param $body
2149          * @param $profile_uid
2150          * @param $network
2151          * @return string
2152          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2153          * @throws \ImagickException
2154          */
2155         public static function setMentions($body, $profile_uid = 0, $network = '')
2156         {
2157                 BBCode::performWithEscapedTags($body, ['noparse', 'pre', 'code', 'img'], function ($body) use ($profile_uid, $network) {
2158                         $tags = BBCode::getTags($body);
2159
2160                         $tagged = [];
2161                         $inform = '';
2162
2163                         foreach ($tags as $tag) {
2164                                 $tag_type = substr($tag, 0, 1);
2165
2166                                 if ($tag_type == Tag::TAG_CHARACTER[Tag::HASHTAG]) {
2167                                         continue;
2168                                 }
2169
2170                                 /*
2171                                  * If we already tagged 'Robert Johnson', don't try and tag 'Robert'.
2172                                  * Robert Johnson should be first in the $tags array
2173                                  */
2174                                 foreach ($tagged as $nextTag) {
2175                                         if (stristr($nextTag, $tag . ' ')) {
2176                                                 continue 2;
2177                                         }
2178                                 }
2179
2180                                 $success = Item::replaceTag($body, $inform, $profile_uid, $tag, $network);
2181
2182                                 if ($success['replaced']) {
2183                                         $tagged[] = $tag;
2184                                 }
2185                         }
2186
2187                         return $body;
2188                 });
2189
2190                 return $body;
2191         }
2192
2193         /**
2194          * @param string      $author  Author display name
2195          * @param string      $profile Author profile URL
2196          * @param string      $avatar  Author profile picture URL
2197          * @param string      $link    Post source URL
2198          * @param string      $posted  Post created date
2199          * @param string|null $guid    Post guid (if any)
2200          * @return string
2201          * @TODO Rewrite to handle over whole record array
2202          */
2203         public static function getShareOpeningTag(string $author, string $profile, string $avatar, string $link, string $posted, string $guid = null)
2204         {
2205                 $header = "[share author='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $author) .
2206                         "' profile='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $profile) .
2207                         "' avatar='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $avatar) .
2208                         "' link='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $link) .
2209                         "' posted='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $posted);
2210
2211                 if ($guid) {
2212                         $header .= "' guid='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $guid);
2213                 }
2214
2215                 $header  .= "']";
2216
2217                 return $header;
2218         }
2219
2220         /**
2221          * Returns the BBCode relevant to embed the provided URL in a post body.
2222          * For media type, it will return [img], [video] and [audio] tags.
2223          * For regular web pages, it will either output a [bookmark] tag if title and description were provided,
2224          * an [attachment] tag or a simple [url] tag depending on $tryAttachment.
2225          *
2226          * @param string      $url
2227          * @param bool        $tryAttachment
2228          * @param string|null $title
2229          * @param string|null $description
2230          * @param string|null $tags
2231          * @return string
2232          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2233          *@see ParseUrl::getSiteinfoCached
2234          *
2235          */
2236         public static function embedURL(string $url, bool $tryAttachment = true, string $title = null, string $description = null, string $tags = null): string
2237         {
2238                 DI::logger()->info($url);
2239
2240                 // If there is already some content information submitted we don't
2241                 // need to parse the url for content.
2242                 if (!empty($title) && !empty($description)) {
2243                         $title = str_replace(["\r", "\n"], ['', ''], $title);
2244
2245                         $description = '[quote]' . trim($description) . '[/quote]' . "\n";
2246
2247                         $str_tags = '';
2248                         if (!empty($tags)) {
2249                                 $arr_tags = ParseUrl::convertTagsToArray($tags);
2250                                 if (count($arr_tags)) {
2251                                         $str_tags = "\n" . implode(' ', $arr_tags) . "\n";
2252                                 }
2253                         }
2254
2255                         $result = sprintf('[bookmark=%s]%s[/bookmark]%s', $url, ($title) ? $title : $url, $description) . $str_tags;
2256
2257                         DI::logger()->info('(unparsed): returns: ' . $result);
2258
2259                         return $result;
2260                 }
2261
2262                 $siteinfo = ParseUrl::getSiteinfoCached($url);
2263
2264                 if (in_array($siteinfo['type'], ['image', 'video', 'audio'])) {
2265                         switch ($siteinfo['type']) {
2266                                 case 'video':
2267                                         $bbcode = "\n" . '[video]' . $url . '[/video]' . "\n";
2268                                         break;
2269                                 case 'audio':
2270                                         $bbcode = "\n" . '[audio]' . $url . '[/audio]' . "\n";
2271                                         break;
2272                                 default:
2273                                         $bbcode = "\n" . '[img]' . $url . '[/img]' . "\n";
2274                                         break;
2275                         }
2276
2277                         return $bbcode;
2278                 }
2279
2280                 unset($siteinfo['keywords']);
2281
2282                 // Bypass attachment if parse url for a comment
2283                 if (!$tryAttachment) {
2284                         return "\n" . '[url=' . $url . ']' . $siteinfo['title'] . '[/url]';
2285                 }
2286
2287                 // Format it as BBCode attachment
2288                 return "\n" . PageInfo::getFooterFromData($siteinfo);
2289         }
2290 }