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 = '2021-04-07';
  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         /**
  65          * Fetches attachment data that were generated the old way
  66          *
  67          * @param string $body Message body
  68          * @return array
  69          *                     'type' -> Message type ('link', 'video', 'photo')
  70          *                     'text' -> Text before the shared message
  71          *                     'after' -> Text after the shared message
  72          *                     'image' -> Preview image of the message
  73          *                     'url' -> Url to the attached message
  74          *                     'title' -> Title of the attachment
  75          *                     'description' -> Description of the attachment
  76          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  77          */
  78         private static function getOldAttachmentData($body)
  79         {
  80                 $post = [];
  81
  82                 // Simplify image codes
  83                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  84
  85                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  86                         foreach ($attached as $data) {
  87                                 if (!in_array($data[1], ['type-link', 'type-video', 'type-photo'])) {
  88                                         continue;
  89                                 }
  90
  91                                 $post['type'] = substr($data[1], 5);
  92
  93                                 $pos = strpos($body, $data[0]);
  94                                 if ($pos > 0) {
  95                                         $post['text'] = trim(substr($body, 0, $pos));
  96                                         $post['after'] = trim(substr($body, $pos + strlen($data[0])));
  97                                 } else {
  98                                         $post['text'] = trim(str_replace($data[0], '', $body));
  99                                         $post['after'] = '';
 100                                 }
 101
 102                                 $attacheddata = $data[2];
 103
 104                                 if (preg_match("/\[img\](.*?)\[\/img\]/ism", $attacheddata, $matches)) {
 105
 106                                         $picturedata = Images::getInfoFromURLCached($matches[1]);
 107
 108                                         if ($picturedata) {
 109                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
 110                                                         $post['image'] = $matches[1];
 111                                                 } else {
 112                                                         $post['preview'] = $matches[1];
 113                                                 }
 114                                         }
 115                                 }
 116
 117                                 if (preg_match("/\[bookmark\=(.*?)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
 118                                         $post['url'] = $matches[1];
 119                                         $post['title'] = $matches[2];
 120                                 }
 121                                 if (!empty($post['url']) && (in_array($post['type'], ['link', 'video']))
 122                                         && preg_match("/\[url\=(.*?)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
 123                                         $post['url'] = $matches[1];
 124                                 }
 125
 126                                 // Search for description
 127                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
 128                                         $post['description'] = $matches[1];
 129                                 }
 130                         }
 131                 }
 132                 return $post;
 133         }
 134
 135         /**
 136          * Fetches attachment data that were generated with the "attachment" element
 137          *
 138          * @param string $body Message body
 139          * @return array
 140          *                     'type' -> Message type ('link', 'video', 'photo')
 141          *                     'text' -> Text before the shared message
 142          *                     'after' -> Text after the shared message
 143          *                     'image' -> Preview image of the message
 144          *                     'url' -> Url to the attached message
 145          *                     'title' -> Title of the attachment
 146          *                     'description' -> Description of the attachment
 147          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 148          */
 149         public static function getAttachmentData($body)
 150         {
 151                 $data = [
 152                         'type'          => '',
 153                         'text'          => '',
 154                         'after'         => '',
 155                         'image'         => null,
 156                         'url'           => '',
 157                         'author_name'   => '',
 158                         'provider_name' => '',
 159                         'provider_url'  => '',
 160                         'title'         => '',
 161                         'description'   => '',
 162                 ];
 163
 164                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 165                         return self::getOldAttachmentData($body);
 166                 }
 167
 168                 $attributes = $match[2];
 169
 170                 $data['text'] = trim($match[1]);
 171
 172                 foreach (['type', 'url', 'title', 'image', 'preview', 'publisher_name', 'publisher_url', 'author_name', 'author_url'] as $field) {
 173                         preg_match('/' . preg_quote($field, '/') . '=("|\')(.*?)\1/ism', $attributes, $matches);
 174                         $value = $matches[2] ?? '';
 175
 176                         if ($value != '') {
 177                                 switch ($field) {
 178                                         case 'publisher_name':
 179                                                 $data['provider_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 180                                                 break;
 181                                         case 'publisher_url':
 182                                                 $data['provider_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 183                                                 break;
 184                                         case 'author_name':
 185                                                 $data['author_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 186                                                 if ($data['provider_name'] == $data['author_name']) {
 187                                                         $data['author_name'] = '';
 188                                                 }
 189                                                 break;
 190                                         case 'author_url':
 191                                                 $data['author_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 192                                                 if ($data['provider_url'] == $data['author_url']) {
 193                                                         $data['author_url'] = '';
 194                                                 }
 195                                                 break;
 196                                         case 'title':
 197                                                 $value = self::convert(html_entity_decode($value, ENT_QUOTES, 'UTF-8'), false, true);
 198                                                 $value = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 199                                                 $value = str_replace(['[', ']'], ['&#91;', '&#93;'], $value);
 200                                                 $data['title'] = $value;
 201                                         default:
 202                                                 $data[$field] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
 203                                                 break;
 204                                 }
 205                         }
 206                 }
 207
 208                 if (!in_array($data['type'], ['link', 'audio', 'photo', 'video'])) {
 209                         return [];
 210                 }
 211
 212                 $data['description'] = trim($match[3]);
 213
 214                 $data['after'] = trim($match[4]);
 215
 216                 $parts = parse_url($data['url']);
 217                 if (!empty($parts['scheme']) && !empty($parts['host'])) {
 218                         if (empty($data['provider_name'])) {
 219                                 $data['provider_name'] = $parts['host'];
 220                         }
 221                         if (empty($data['provider_url']) || empty(parse_url($data['provider_url'], PHP_URL_SCHEME))) {
 222                                 $data['provider_url'] = $parts['scheme'] . '://' . $parts['host'];
 223
 224                                 if (!empty($parts['port'])) {
 225                                         $data['provider_url'] .= ':' . $parts['port'];
 226                                 }
 227                         }
 228                 }
 229
 230                 return $data;
 231         }
 232
 233         public static function getAttachedData($body, $item = [])
 234         {
 235                 /*
 236                 - text:
 237                 - type: link, video, photo
 238                 - title:
 239                 - url:
 240                 - image:
 241                 - description:
 242                 - (thumbnail)
 243                 */
 244
 245                 $has_title = !empty($item['title']);
 246                 $plink = $item['plink'] ?? '';
 247                 $post = self::getAttachmentData($body);
 248
 249                 // Get all linked images with alternative image description
 250                 if (preg_match_all("/\[img=([^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 251                         foreach ($pictures as $picture) {
 252                                 if (Photo::isLocal($picture[1])) {
 253                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]];
 254                                 } else {
 255                                         $post['remote_images'][] = ['url' => $picture[1], 'description' => $picture[2]];
 256                                 }
 257                         }
 258                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 259                                 $post['image_description'] = $post['images'][0]['description'];
 260                         }
 261                 }
 262
 263                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 264                         foreach ($pictures as $picture) {
 265                                 if (Photo::isLocal($picture[1])) {
 266                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
 267                                 } else {
 268                                         $post['remote_images'][] = ['url' => $picture[1], 'description' => ''];
 269                                 }
 270                         }
 271                 }
 272
 273                 // if nothing is found, it maybe having an image.
 274                 if (!isset($post['type'])) {
 275                         // Simplify image codes
 276                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 277                         $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 278                         $post['text'] = $body;
 279
 280                         if (preg_match_all("#\[url=([^\]]+?)\]\s*\[img\]([^\[]+?)\[/img\]\s*\[/url\]#ism", $body, $pictures, PREG_SET_ORDER)) {
 281                                 if ((count($pictures) == 1) && !$has_title) {
 282                                         if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) {
 283                                                 // Replace the preview picture with the real picture
 284                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
 285                                                 $data = ['url' => $url, 'type' => 'photo'];
 286                                         } else {
 287                                                 // Checking, if the link goes to a picture
 288                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1]);
 289                                         }
 290
 291                                         // Workaround:
 292                                         // Sometimes photo posts to the own album are not detected at the start.
 293                                         // So we seem to cannot use the cache for these cases. That's strange.
 294                                         if (($data['type'] != 'photo') && strstr($pictures[0][1], "/photos/")) {
 295                                                 $data = ParseUrl::getSiteinfo($pictures[0][1]);
 296                                         }
 297
 298                                         if ($data['type'] == 'photo') {
 299                                                 $post['type'] = 'photo';
 300                                                 if (isset($data['images'][0])) {
 301                                                         $post['image'] = $data['images'][0]['src'];
 302                                                         $post['url'] = $data['url'];
 303                                                 } else {
 304                                                         $post['image'] = $data['url'];
 305                                                 }
 306
 307                                                 $post['preview'] = $pictures[0][2];
 308                                                 $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 309                                         } else {
 310                                                 $imgdata = Images::getInfoFromURLCached($pictures[0][1]);
 311                                                 if ($imgdata && substr($imgdata['mime'], 0, 6) == 'image/') {
 312                                                         $post['type'] = 'photo';
 313                                                         $post['image'] = $pictures[0][1];
 314                                                         $post['preview'] = $pictures[0][2];
 315                                                         $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 316                                                 }
 317                                         }
 318                                 } elseif (count($pictures) > 0) {
 319                                         $post['type'] = 'link';
 320                                         $post['url'] = $plink;
 321                                         $post['image'] = $pictures[0][2];
 322                                         $post['text'] = $body;
 323
 324                                         foreach ($pictures as $picture) {
 325                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 326                                         }
 327                                 }
 328                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 329                                 if ($has_title) {
 330                                         $post['type'] = 'link';
 331                                         $post['url'] = $plink;
 332                                 } else {
 333                                         $post['type'] = 'photo';
 334                                 }
 335
 336                                 $post['image'] = $pictures[0][1];
 337                                 $post['text'] = $body;
 338                                 foreach ($pictures as $picture) {
 339                                         $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 340                                 }
 341                         }
 342
 343                         // Test for the external links
 344                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER);
 345                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER);
 346
 347                         $links = array_merge($links1, $links2);
 348
 349                         // If there is only a single one, then use it.
 350                         // This should cover link posts via API.
 351                         if ((count($links) == 1) && !isset($post['preview']) && !$has_title) {
 352                                 $post['type'] = 'link';
 353                                 $post['url'] = $links[0][1];
 354                         }
 355
 356                         // Simplify "video" element
 357                         $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']);
 358
 359                         // Now count the number of external media links
 360                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
 361                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
 362                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
 363                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
 364
 365                         // Add them to the other external links
 366                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 367
 368                         // Are there more than one?
 369                         if (count($links) > 1) {
 370                                 // The post will be the type "text", which means a blog post
 371                                 unset($post['type']);
 372                                 $post['url'] = $plink;
 373                         }
 374
 375                         if (!isset($post['type'])) {
 376                                 $post['type'] = "text";
 377                                 $post['text'] = trim($body);
 378                         }
 379
 380                         if (($post['type'] == 'photo') && empty($post['images']) && !empty($post['remote_images'])) {
 381                                 $post['images'] = $post['remote_images'];
 382                                 $post['image'] = $post['images'][0]['url'];
 383                                 if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 384                                         $post['image_description'] = $post['images'][0]['description'];
 385                                 }
 386                         }
 387                         unset($post['remote_images']);
 388                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
 389                         $data = ParseUrl::getSiteinfoCached($post['url']);
 390
 391                         if (isset($data['images'][0])) {
 392                                 $post['image'] = $data['images'][0]['src'];
 393                         }
 394                 }
 395
 396                 return $post;
 397         }
 398
 399         /**
 400          * Remove [attachment] BBCode and replaces it with a regular [url]
 401          *
 402          * @param string  $body
 403          * @param boolean $no_link_desc No link description
 404          *
 405          * @return string with replaced body
 406          */
 407         public static function removeAttachment($body, $no_link_desc = false)
 408         {
 409                 return preg_replace_callback("/\s*\[attachment (.*)\](.*?)\[\/attachment\]\s*/ism",
 410                         function ($match) use ($no_link_desc) {
 411                                 $attach_data = self::getAttachmentData($match[0]);
 412                                 if (empty($attach_data['url'])) {
 413                                         return $match[0];
 414                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
 415                                         return "\n[url]" . $attach_data['url'] . "[/url]\n";
 416                                 } else {
 417                                         return "\n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
 418                                 }
 419                 }, $body);
 420         }
 421
 422         /**
 423          * Converts a BBCode text into plaintext
 424          *
 425          * @param      $text
 426          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 427          *
 428          * @return string
 429          */
 430         public static function toPlaintext($text, $keep_urls = true)
 431         {
 432                 $naked_text = HTML::toPlaintext(self::convert($text, false, 0, true), 0, !$keep_urls);
 433
 434                 return $naked_text;
 435         }
 436
 437         private static function proxyUrl($image, $simplehtml = self::INTERNAL)
 438         {
 439                 // Only send proxied pictures to API and for internal display
 440                 if (in_array($simplehtml, [self::INTERNAL, self::API])) {
 441                         return ProxyUtils::proxifyUrl($image);
 442                 } else {
 443                         return $image;
 444                 }
 445         }
 446
 447         /**
 448          * This function changing the visual size (not the real size) of images.
 449          * The function does not work for pictures with an alternate text description.
 450          * This could only be changed by using some new "img" BBCode format.
 451          *
 452          * @param string $srctext The body with images
 453          * @return string The body with possibly scaled images
 454          */
 455         public static function scaleExternalImages(string $srctext)
 456         {
 457                 $s = $srctext;
 458
 459                 // Simplify image links
 460                 $s = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $s);
 461
 462                 $matches = null;
 463                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 464                 if ($c) {
 465                         foreach ($matches as $mtch) {
 466                                 Logger::info('scale_external_image', ['image' => $mtch[1]]);
 467
 468                                 $hostname = str_replace('www.', '', substr(DI::baseUrl(), strpos(DI::baseUrl(), '://') + 3));
 469                                 if (stristr($mtch[1], $hostname)) {
 470                                         continue;
 471                                 }
 472
 473                                 $curlResult = DI::httpRequest()->get($mtch[1]);
 474                                 if (!$curlResult->isSuccess()) {
 475                                         continue;
 476                                 }
 477
 478                                 $i = $curlResult->getBody();
 479                                 $type = $curlResult->getContentType();
 480                                 $type = Images::getMimeTypeByData($i, $mtch[1], $type);
 481
 482                                 if ($i) {
 483                                         $Image = new Image($i, $type);
 484                                         if ($Image->isValid()) {
 485                                                 $orig_width = $Image->getWidth();
 486                                                 $orig_height = $Image->getHeight();
 487
 488                                                 if ($orig_width > 640 || $orig_height > 640) {
 489                                                         $Image->scaleDown(640);
 490                                                         $new_width = $Image->getWidth();
 491                                                         $new_height = $Image->getHeight();
 492                                                         Logger::info('External images scaled', ['orig_width' => $orig_width, 'new_width' => $new_width, 'orig_height' => $orig_height, 'new_height' => $new_height, 'match' => $mtch[0]]);
 493                                                         $s = str_replace(
 494                                                                 $mtch[0],
 495                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]'
 496                                                                 . "\n",
 497                                                                 $s
 498                                                         );
 499                                                         Logger::info('New string', ['image' => $s]);
 500                                                 }
 501                                         }
 502                                 }
 503                         }
 504                 }
 505
 506                 return $s;
 507         }
 508
 509         /**
 510          * Truncates imported message body string length to max_import_size
 511          *
 512          * The purpose of this function is to apply system message length limits to
 513          * imported messages without including any embedded photos in the length
 514          *
 515          * @param string $body
 516          * @return string
 517          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 518          */
 519         public static function limitBodySize($body)
 520         {
 521                 $maxlen = DI::config()->get('config', 'max_import_size', 0);
 522
 523                 // If the length of the body, including the embedded images, is smaller
 524                 // than the maximum, then don't waste time looking for the images
 525                 if ($maxlen && (strlen($body) > $maxlen)) {
 526
 527                         Logger::info('the total body length exceeds the limit', ['maxlen' => $maxlen, 'body_len' => strlen($body)]);
 528
 529                         $orig_body = $body;
 530                         $new_body = '';
 531                         $textlen = 0;
 532
 533                         $img_start = strpos($orig_body, '[img');
 534                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 535                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 536                         while (($img_st_close !== false) && ($img_end !== false)) {
 537
 538                                 $img_st_close++; // make it point to AFTER the closing bracket
 539                                 $img_end += $img_start;
 540                                 $img_end += strlen('[/img]');
 541
 542                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 543                                         // This is an embedded image
 544
 545                                         if (($textlen + $img_start) > $maxlen) {
 546                                                 if ($textlen < $maxlen) {
 547                                                         Logger::info('the limit happens before an embedded image');
 548                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 549                                                         $textlen = $maxlen;
 550                                                 }
 551                                         } else {
 552                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 553                                                 $textlen += $img_start;
 554                                         }
 555
 556                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 557                                 } else {
 558
 559                                         if (($textlen + $img_end) > $maxlen) {
 560                                                 if ($textlen < $maxlen) {
 561                                                         Logger::info('the limit happens before the end of a non-embedded image');
 562                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 563                                                         $textlen = $maxlen;
 564                                                 }
 565                                         } else {
 566                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 567                                                 $textlen += $img_end;
 568                                         }
 569                                 }
 570                                 $orig_body = substr($orig_body, $img_end);
 571
 572                                 if ($orig_body === false) {
 573                                         // in case the body ends on a closing image tag
 574                                         $orig_body = '';
 575                                 }
 576
 577                                 $img_start = strpos($orig_body, '[img');
 578                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 579                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 580                         }
 581
 582                         if (($textlen + strlen($orig_body)) > $maxlen) {
 583                                 if ($textlen < $maxlen) {
 584                                         Logger::info('the limit happens after the end of the last image');
 585                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 586                                 }
 587                         } else {
 588                                 Logger::info('the text size with embedded images extracted did not violate the limit');
 589                                 $new_body = $new_body . $orig_body;
 590                         }
 591
 592                         return $new_body;
 593                 } else {
 594                         return $body;
 595                 }
 596         }
 597
 598         /**
 599          * Processes [attachment] tags
 600          *
 601          * Note: Can produce a [bookmark] tag in the returned string
 602          *
 603          * @param string  $text
 604          * @param integer $simplehtml
 605          * @param bool    $tryoembed
 606          * @return string
 607          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 608          */
 609         private static function convertAttachment($text, $simplehtml = self::INTERNAL, $tryoembed = true)
 610         {
 611                 $data = self::getAttachmentData($text);
 612                 if (empty($data) || empty($data['url'])) {
 613                         return $text;
 614                 }
 615
 616                 if (isset($data['title'])) {
 617                         $data['title'] = strip_tags($data['title']);
 618                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 619                 } else {
 620                         $data['title'] = null;
 621                 }
 622
 623                 if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) {
 624                         $data['preview'] = $data['image'];
 625                         $data['image'] = '';
 626                 }
 627
 628                 $return = '';
 629                 try {
 630                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 631                                 $return = OEmbed::getHTML($data['url'], $data['title']);
 632                         } else {
 633                                 throw new Exception('OEmbed is disabled for this attachment.');
 634                         }
 635                 } catch (Exception $e) {
 636                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 637
 638                         if ($simplehtml != self::CONNECTORS) {
 639                                 $return = sprintf('<div class="type-%s">', $data['type']);
 640                         }
 641
 642                         if (!empty($data['title']) && !empty($data['url'])) {
 643                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 644                                         $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), $data['title']);
 645                                 } else {
 646                                         if (!empty($data['image'])) {
 647                                                 $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), $data['title']);
 648                                         } elseif (!empty($data['preview'])) {
 649                                                 $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']);
 650                                         }
 651                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 652                                 }
 653                         }
 654
 655                         if (!empty($data['description']) && $data['description'] != $data['title']) {
 656                                 // Sanitize the HTML by converting it to BBCode
 657                                 $bbcode = HTML::toBBCode($data['description']);
 658                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 659                         }
 660
 661                         if (!empty($data['provider_url']) && !empty($data['provider_name'])) {
 662                                 if (!empty($data['author_name'])) {
 663                                         $return .= sprintf('<sup><a href="%s">%s (%s)</a></sup>', $data['provider_url'], $data['author_name'], $data['provider_name']);
 664                                 } else {
 665                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['provider_url'], $data['provider_name']);
 666                                 }
 667                         }
 668
 669                         if ($simplehtml != self::CONNECTORS) {
 670                                 $return .= '</div>';
 671                         }
 672                 }
 673
 674                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 675         }
 676
 677         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 678         {
 679                 $data = self::getAttachmentData($Text);
 680
 681                 if (!$data) {
 682                         return $Text;
 683                 } elseif ($nolink) {
 684                         return $data['text'] . ($data['after'] ?? '');
 685                 }
 686
 687                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 688                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 689                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 690                         $data['title'] = $data['url'];
 691                 } elseif (($text != '') && strstr($title, $text)) {
 692                         $data['text'] = $data['title'];
 693                         $data['title'] = $data['url'];
 694                 }
 695
 696                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 697                         return $data['title'] . $data['after'];
 698                 }
 699
 700                 // If the link already is included in the post, don't add it again
 701                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 702                         return $data['text'] . $data['after'];
 703                 }
 704
 705                 $text = $data['text'];
 706
 707                 if (!empty($data['url']) && !empty($data['title'])) {
 708                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 709                 } elseif (!empty($data['url'])) {
 710                         $text .= "\n[url]" . $data['url'] . '[/url]';
 711                 }
 712
 713                 return $text . "\n" . $data['after'];
 714         }
 715
 716         /**
 717          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 718          *
 719          * @param array $match Array with the matching values
 720          * @return string reformatted link including HTML codes
 721          */
 722         private static function convertUrlForActivityPubCallback($match)
 723         {
 724                 $url = $match[1];
 725
 726                 if (isset($match[2]) && ($match[1] != $match[2])) {
 727                         return $match[0];
 728                 }
 729
 730                 $parts = parse_url($url);
 731                 if (!isset($parts['scheme'])) {
 732                         return $match[0];
 733                 }
 734
 735                 return self::convertUrlForActivityPub($url);
 736         }
 737
 738         /**
 739          * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 740          *
 741          * @param string $url URL that is about to be reformatted
 742          * @return string reformatted link including HTML codes
 743          */
 744         private static function convertUrlForActivityPub($url)
 745         {
 746                 $html = '<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>';
 747                 return sprintf($html, $url, self::getStyledURL($url));
 748         }
 749
 750         /**
 751          * Converts an URL in a nicer format (without the scheme and possibly shortened)
 752          *
 753          * @param string $url URL that is about to be reformatted
 754          * @return string reformatted link
 755          */
 756         private static function getStyledURL($url)
 757         {
 758                 $parts = parse_url($url);
 759                 $scheme = $parts['scheme'] . '://';
 760                 $styled_url = str_replace($scheme, '', $url);
 761
 762                 if (strlen($styled_url) > 30) {
 763                         $styled_url = substr($styled_url, 0, 30) . "…";
 764                 }
 765
 766                 return $styled_url;
 767         }
 768
 769         /*
 770          * [noparse][i]italic[/i][/noparse] turns into
 771          * [noparse][ i ]italic[ /i ][/noparse],
 772          * to hide them from parser.
 773          */
 774         private static function escapeNoparseCallback($match)
 775         {
 776                 $whole_match = $match[0];
 777                 $captured = $match[1];
 778                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 779                 $new_str = str_replace($captured, $spacefied, $whole_match);
 780                 return $new_str;
 781         }
 782
 783         /*
 784          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 785          * now turns back and the [noparse] tags are trimed
 786          * returning [i]italic[/i]
 787          */
 788         private static function unescapeNoparseCallback($match)
 789         {
 790                 $captured = $match[1];
 791                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 792                 return $unspacefied;
 793         }
 794
 795         /**
 796          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 797          * occurrences
 798          *
 799          * @param string $text        Text to search
 800          * @param string $name        Tag name
 801          * @param int    $occurrences Number of first occurrences to skip
 802          * @return boolean|array
 803          */
 804         public static function getTagPosition($text, $name, $occurrences = 0)
 805         {
 806                 if ($occurrences < 0) {
 807                         $occurrences = 0;
 808                 }
 809
 810                 $start_open = -1;
 811                 for ($i = 0; $i <= $occurrences; $i++) {
 812                         if ($start_open !== false) {
 813                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 814                         }
 815                 }
 816
 817                 if ($start_open === false) {
 818                         return false;
 819                 }
 820
 821                 $start_equal = strpos($text, '=', $start_open);
 822                 $start_close = strpos($text, ']', $start_open);
 823
 824                 if ($start_close === false) {
 825                         return false;
 826                 }
 827
 828                 $start_close++;
 829
 830                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 831
 832                 if ($end_open === false) {
 833                         return false;
 834                 }
 835
 836                 $res = [
 837                         'start' => [
 838                                 'open' => $start_open,
 839                                 'close' => $start_close
 840                         ],
 841                         'end' => [
 842                                 'open' => $end_open,
 843                                 'close' => $end_open + strlen('[/' . $name . ']')
 844                         ],
 845                 ];
 846
 847                 if ($start_equal !== false) {
 848                         $res['start']['equal'] = $start_equal + 1;
 849                 }
 850
 851                 return $res;
 852         }
 853
 854         /**
 855          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 856          *
 857          * @param string $pattern Preg pattern string
 858          * @param string $replace Preg replace string
 859          * @param string $name    BBCode tag name
 860          * @param string $text    Text to search
 861          * @return string
 862          */
 863         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 864         {
 865                 $occurrences = 0;
 866                 $pos = self::getTagPosition($text, $name, $occurrences);
 867                 while ($pos !== false && $occurrences++ < 1000) {
 868                         $start = substr($text, 0, $pos['start']['open']);
 869                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 870                         $end = substr($text, $pos['end']['close']);
 871                         if ($end === false) {
 872                                 $end = '';
 873                         }
 874
 875                         $subject = preg_replace($pattern, $replace, $subject);
 876                         $text = $start . $subject . $end;
 877
 878                         $pos = self::getTagPosition($text, $name, $occurrences);
 879                 }
 880
 881                 return $text;
 882         }
 883
 884         private static function extractImagesFromItemBody($body)
 885         {
 886                 $saved_image = [];
 887                 $orig_body = $body;
 888                 $new_body = '';
 889
 890                 $cnt = 0;
 891                 $img_start = strpos($orig_body, '[img');
 892                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 893                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 894                 while (($img_st_close !== false) && ($img_end !== false)) {
 895                         $img_st_close++; // make it point to AFTER the closing bracket
 896                         $img_end += $img_start;
 897
 898                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 899                                 // This is an embedded image
 900                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 901                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 902
 903                                 $cnt++;
 904                         } else {
 905                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 906                         }
 907
 908                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 909
 910                         if ($orig_body === false) {
 911                                 // in case the body ends on a closing image tag
 912                                 $orig_body = '';
 913                         }
 914
 915                         $img_start = strpos($orig_body, '[img');
 916                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 917                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 918                 }
 919
 920                 $new_body = $new_body . $orig_body;
 921
 922                 return ['body' => $new_body, 'images' => $saved_image];
 923         }
 924
 925         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 926         {
 927                 $newbody = $body;
 928
 929                 $cnt = 0;
 930                 foreach ($images as $image) {
 931                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 932                         // it loops over the array starting from the first element and going sequentially
 933                         // to the last element
 934                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 935                                 '<img src="' . self::proxyUrl($image) . '" alt="' . DI::l10n()->t('Image/photo') . '" />', $newbody);
 936                         $cnt++;
 937                 }
 938
 939                 return $newbody;
 940         }
 941
 942         /**
 943          * This function converts a [share] block to text according to a provided callback function whose signature is:
 944          *
 945          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 946          *
 947          * Where:
 948          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 949          * data lookup
 950          * - $author_contact is a contact record array
 951          * - $content is the inner content of the [share] block
 952          * - $is_quote_share indicates whether there's any content before the [share] block
 953          * - Return value is the string that should replace the [share] block in the provided text
 954          *
 955          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 956          *
 957          * @param  string   $text     A BBCode string
 958          * @param  callable $callback
 959          * @return string The BBCode string with all [share] blocks replaced
 960          */
 961         public static function convertShare($text, callable $callback)
 962         {
 963                 $return = preg_replace_callback(
 964                         "/(.*?)\[share(.*?)\](.*)\[\/share\]/ism",
 965                         function ($match) use ($callback) {
 966                                 $attribute_string = $match[2];
 967                                 $attributes = [];
 968                                 foreach (['author', 'profile', 'avatar', 'link', 'posted', 'guid'] as $field) {
 969                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 970                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 971                                 }
 972
 973                                 $author_contact = Contact::getByURL($attributes['profile'], false, ['url', 'addr', 'name', 'micro']);
 974                                 $author_contact['url'] = ($author_contact['url'] ?? $attributes['profile']);
 975                                 $author_contact['addr'] = ($author_contact['addr'] ?? '') ?: Protocol::getAddrFromProfileUrl($attributes['profile']);
 976
 977                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
 978                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
 979                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
 980
 981                                 if ($attributes['avatar']) {
 982                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
 983                                 }
 984
 985                                 $content = preg_replace(Strings::autoLinkRegEx(), '<a href="$1">$1</a>', $match[3]);
 986
 987                                 return $match[1] . $callback($attributes, $author_contact, $content, trim($match[1]) != '');
 988                         },
 989                         $text
 990                 );
 991
 992                 return $return;
 993         }
 994
 995         /**
 996          * Default [share] tag conversion callback
 997          *
 998          * Note: Can produce a [bookmark] tag in the output
 999          *
1000          * @see BBCode::convertShare()
1001          * @param array   $attributes     [share] block attribute values
1002          * @param array   $author_contact Contact row of the shared author
1003          * @param string  $content        Inner content of the [share] block
1004          * @param boolean $is_quote_share Whether there is content before the [share] block
1005          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
1006          * @return string
1007          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1008          */
1009         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
1010         {
1011                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
1012
1013                 switch ($simplehtml) {
1014                         case self::API:
1015                                 $text = ($is_quote_share? '<br>' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1016                                 break;
1017                         case self::DIASPORA:
1018                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1019                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1020                                 } else {
1021                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1022
1023                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1024                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1025                                         }
1026
1027                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1028
1029                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1030                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1031                                         }
1032                                 }
1033
1034                                 break;
1035                         case self::CONNECTORS:
1036                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1037                                 $headline .= DI::l10n()->t('<a href="%1$s" target="_blank" rel="noopener noreferrer">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1038                                 $headline .= ':</b></p>' . "\n";
1039
1040                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1041
1042                                 break;
1043                         case self::OSTATUS:
1044                                 $text = ($is_quote_share? '<br>' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1045                                 break;
1046                         case self::ACTIVITYPUB:
1047                                 $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>';
1048                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1049                                 break;
1050                         default:
1051                                 $text = ($is_quote_share? "\n" : '');
1052
1053                                 $contact = Contact::getByURL($attributes['profile'], false, ['network']);
1054                                 $network = $contact['network'] ?? Protocol::PHANTOM;
1055
1056                                 $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1057                                 $text .= Renderer::replaceMacros($tpl, [
1058                                         '$profile'      => $attributes['profile'],
1059                                         '$avatar'       => $attributes['avatar'],
1060                                         '$author'       => $attributes['author'],
1061                                         '$link'         => $attributes['link'],
1062                                         '$link_title'   => DI::l10n()->t('link to source'),
1063                                         '$posted'       => $attributes['posted'],
1064                                         '$guid'         => $attributes['guid'],
1065                                         '$network_name' => ContactSelector::networkToName($network, $attributes['profile']),
1066                                         '$network_icon' => ContactSelector::networkToIcon($network, $attributes['profile']),
1067                                         '$content'      => self::setMentions(trim($content), 0, $network),
1068                                 ]);
1069                                 break;
1070                 }
1071
1072                 return $text;
1073         }
1074
1075         private static function removePictureLinksCallback($match)
1076         {
1077                 $cache_key = 'remove:' . $match[1];
1078                 $text = DI::cache()->get($cache_key);
1079
1080                 if (is_null($text)) {
1081                         $a = DI::app();
1082
1083                         $stamp1 = microtime(true);
1084
1085                         $ch = @curl_init($match[1]);
1086                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1087                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1088                         @curl_setopt($ch, CURLOPT_USERAGENT, DI::httpRequest()->getUserAgent());
1089                         @curl_exec($ch);
1090                         $curl_info = @curl_getinfo($ch);
1091
1092                         DI::profiler()->saveTimestamp($stamp1, "network");
1093
1094                         if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1095                                 $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]";
1096                         } else {
1097                                 $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]";
1098
1099                                 // if its not a picture then look if its a page that contains a picture link
1100                                 $body = DI::httpRequest()->fetch($match[1]);
1101                                 if (empty($body)) {
1102                                         DI::cache()->set($cache_key, $text);
1103                                         return $text;
1104                                 }
1105
1106                                 $doc = new DOMDocument();
1107                                 @$doc->loadHTML($body);
1108                                 $xpath = new DOMXPath($doc);
1109                                 $list = $xpath->query("//meta[@name]");
1110                                 foreach ($list as $node) {
1111                                         $attr = [];
1112
1113                                         if ($node->attributes->length) {
1114                                                 foreach ($node->attributes as $attribute) {
1115                                                         $attr[$attribute->name] = $attribute->value;
1116                                                 }
1117                                         }
1118
1119                                         if (strtolower($attr['name']) == 'twitter:image') {
1120                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1121                                         }
1122                                 }
1123                         }
1124                         DI::cache()->set($cache_key, $text);
1125                 }
1126
1127                 return $text;
1128         }
1129
1130         private static function expandLinksCallback($match)
1131         {
1132                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1133                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1134                 } else {
1135                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1136                 }
1137         }
1138
1139         private static function cleanPictureLinksCallback($match)
1140         {
1141                 // When the picture link is the own photo path then we can avoid fetching the link
1142                 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()->get()) . '/photos/');
1143                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1144                         if (!empty($match[3])) {
1145                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1146                         } else {
1147                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1148                         }
1149                         return $text;
1150                 }
1151
1152                 $cache_key = 'clean:' . $match[1];
1153                 $text = DI::cache()->get($cache_key);
1154                 if (!is_null($text)) {
1155                         return $text;
1156                 }
1157
1158                 // Only fetch the header, not the content
1159                 $stamp1 = microtime(true);
1160
1161                 $ch = @curl_init($match[1]);
1162                 @curl_setopt($ch, CURLOPT_NOBODY, true);
1163                 @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1164                 @curl_setopt($ch, CURLOPT_USERAGENT, DI::httpRequest()->getUserAgent());
1165                 @curl_exec($ch);
1166                 $curl_info = @curl_getinfo($ch);
1167
1168                 DI::profiler()->saveTimestamp($stamp1, "network");
1169
1170                 // if its a link to a picture then embed this picture
1171                 if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1172                         $text = '[img]' . $match[1] . '[/img]';
1173                 } else {
1174                         if (!empty($match[3])) {
1175                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1176                         } else {
1177                                 $text = '[img]' . $match[2] . '[/img]';
1178                         }
1179
1180                         // if its not a picture then look if its a page that contains a picture link
1181                         $body = DI::httpRequest()->fetch($match[1]);
1182                         if (empty($body)) {
1183                                 DI::cache()->set($cache_key, $text);
1184                                 return $text;
1185                         }
1186
1187                         $doc = new DOMDocument();
1188                         @$doc->loadHTML($body);
1189                         $xpath = new DOMXPath($doc);
1190                         $list = $xpath->query("//meta[@name]");
1191                         foreach ($list as $node) {
1192                                 $attr = [];
1193                                 if ($node->attributes->length) {
1194                                         foreach ($node->attributes as $attribute) {
1195                                                 $attr[$attribute->name] = $attribute->value;
1196                                         }
1197                                 }
1198
1199                                 if (strtolower($attr['name']) == "twitter:image") {
1200                                         if (!empty($match[3])) {
1201                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1202                                         } else {
1203                                                 $text = "[img]" . $attr['content'] . "[/img]";
1204                                         }
1205                                 }
1206                         }
1207                 }
1208                 DI::cache()->set($cache_key, $text);
1209
1210                 return $text;
1211         }
1212
1213         public static function cleanPictureLinks($text)
1214         {
1215                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1216                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1217                 return $return;
1218         }
1219
1220         public static function removeLinks(string $bbcode)
1221         {
1222                 $bbcode = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $1 ', $bbcode);
1223                 $bbcode = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $bbcode);
1224
1225                 $bbcode = preg_replace('/[@!#]\[url\=.*?\].*?\[\/url\]/ism', '', $bbcode);
1226                 $bbcode = preg_replace("/\[url=[^\[\]]*\](.*)\[\/url\]/Usi", ' $1 ', $bbcode);
1227                 $bbcode = preg_replace('/[@!#]?\[url.*?\[\/url\]/ism', '', $bbcode);
1228                 return $bbcode;
1229         }
1230
1231         /**
1232          * Converts a BBCode message to HTML message
1233          *
1234          * BBcode 2 HTML was written by WAY2WEB.net
1235          * extended to work with Mistpark/Friendica - Mike Macgirvin
1236          *
1237          * Simple HTML values meaning:
1238          * - 0: Friendica display
1239          * - 1: Unused
1240          * - 2: Used for Windows Phone push, Friendica API
1241          * - 3: Used before converting to Markdown in bb2diaspora.php
1242          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1243          * - 5: Unused
1244          * - 6: Unused
1245          * - 7: Used for dfrn, OStatus
1246          * - 8: Used for Twitter, WP backlink text setting
1247          * - 9: ActivityPub
1248          *
1249          * @param string $text
1250          * @param bool   $try_oembed
1251          * @param int    $simple_html
1252          * @param bool   $for_plaintext
1253          * @return string
1254          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1255          */
1256         public static function convert(string $text = null, $try_oembed = true, $simple_html = self::INTERNAL, $for_plaintext = false)
1257         {
1258                 // Accounting for null default column values
1259                 if (is_null($text) || $text === '') {
1260                         return '';
1261                 }
1262
1263                 Hook::callAll('bbcode', $text);
1264
1265                 $a = DI::app();
1266
1267                 $text = self::performWithEscapedTags($text, ['code'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) {
1268                         $text = self::performWithEscapedTags($text, ['noparse', 'nobb', 'pre'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) {
1269                                 /*
1270                                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1271                                  *
1272                                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1273                                  * $match[1] = $url
1274                                  * $match[2] = $title or absent
1275                                  */
1276                                 $try_oembed_callback = function ($match)
1277                                 {
1278                                         $url = $match[1];
1279                                         $title = $match[2] ?? null;
1280
1281                                         try {
1282                                                 $return = OEmbed::getHTML($url, $title);
1283                                         } catch (Exception $ex) {
1284                                                 $return = $match[0];
1285                                         }
1286
1287                                         return $return;
1288                                 };
1289
1290                                 // Remove the abstract element. It is a non visible element.
1291                                 $text = self::stripAbstract($text);
1292
1293                                 // Line ending normalisation
1294                                 $text = str_replace("\r\n", "\n", $text);
1295
1296                                 // Move new lines outside of tags
1297                                 $text = preg_replace("#\[(\w*)](\n*)#ism", '$2[$1]', $text);
1298                                 $text = preg_replace("#(\n*)\[/(\w*)]#ism", '[/$2]$1', $text);
1299
1300                                 // Extract the private images which use data urls since preg has issues with
1301                                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1302                                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1303
1304                                 $extracted = self::extractImagesFromItemBody($text);
1305                                 $text = $extracted['body'];
1306                                 $saved_image = $extracted['images'];
1307
1308                                 // If we find any event code, turn it into an event.
1309                                 // After we're finished processing the bbcode we'll
1310                                 // replace all of the event code with a reformatted version.
1311
1312                                 $ev = Event::fromBBCode($text);
1313
1314                                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1315                                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1316
1317                                 $text = str_replace("<", "&lt;", $text);
1318                                 $text = str_replace(">", "&gt;", $text);
1319
1320                                 // remove some newlines before the general conversion
1321                                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1322                                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1323
1324                                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1325                                 if (!$try_oembed) {
1326                                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1327                                 }
1328
1329                                 // Remove linefeeds inside of the table elements. See issue #6799
1330                                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1331                                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1332                                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1333                                         "\n[hr]", "[hr]\n", " [hr]", "[hr] ",
1334                                         "\n[attachment ", " [attachment ", "\n[/attachment]", "[/attachment]\n", " [/attachment]", "[/attachment] ",
1335                                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1336                                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1337                                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1338                                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1339                                         "[hr]", "[hr]", "[hr]", "[hr]",
1340                                         "[attachment ", "[attachment ", "[/attachment]", "[/attachment]", "[/attachment]", "[/attachment]",
1341                                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1342                                 do {
1343                                         $oldtext = $text;
1344                                         $text = str_replace($search, $replace, $text);
1345                                 } while ($oldtext != $text);
1346
1347                                 // Replace these here only once
1348                                 $search = ["\n[table]", "[/table]\n"];
1349                                 $replace = ["[table]", "[/table]"];
1350                                 $text = str_replace($search, $replace, $text);
1351
1352                                 // Trim new lines regardless of the system.remove_multiplicated_lines config value
1353                                 $text = trim($text, "\n");
1354
1355                                 // removing multiplicated newlines
1356                                 if (DI::config()->get('system', 'remove_multiplicated_lines')) {
1357                                         $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",
1358                                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1359                                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[*]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1360                                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1361                                         do {
1362                                                 $oldtext = $text;
1363                                                 $text = str_replace($search, $replace, $text);
1364                                         } while ($oldtext != $text);
1365                                 }
1366
1367                                 // Add HTML new lines
1368                                 $text = str_replace("\n", '<br>', $text);
1369
1370                                 /// @todo Have a closer look at the different html modes
1371                                 // Handle attached links or videos
1372                                 if ($simple_html == self::ACTIVITYPUB) {
1373                                         $text = self::removeAttachment($text);
1374                                 } elseif (!in_array($simple_html, [self::INTERNAL, self::CONNECTORS])) {
1375                                         $text = self::removeAttachment($text, true);
1376                                 } else {
1377                                         $text = self::convertAttachment($text, $simple_html, $try_oembed);
1378                                 }
1379
1380                                 $nosmile = strpos($text, '[nosmile]') !== false;
1381                                 $text = str_replace('[nosmile]', '', $text);
1382
1383                                 // Replace non graphical smilies for external posts
1384                                 if (!$nosmile && !$for_plaintext) {
1385                                         $text = self::performWithEscapedTags($text, ['img'], function ($text) {
1386                                                 return Smilies::replace($text);
1387                                         });
1388                                 }
1389
1390                                 // leave open the posibility of [map=something]
1391                                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1392                                 if (strpos($text, '[/map]') !== false) {
1393                                         $text = preg_replace_callback(
1394                                                 "/\[map\](.*?)\[\/map\]/ism",
1395                                                 function ($match) use ($simple_html) {
1396                                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1397                                                 },
1398                                                 $text
1399                                         );
1400                                 }
1401
1402                                 if (strpos($text, '[map=') !== false) {
1403                                         $text = preg_replace_callback(
1404                                                 "/\[map=(.*?)\]/ism",
1405                                                 function ($match) use ($simple_html) {
1406                                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1407                                                 },
1408                                                 $text
1409                                         );
1410                                 }
1411
1412                                 if (strpos($text, '[map]') !== false) {
1413                                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1414                                 }
1415
1416                                 // Check for headers
1417                                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1418                                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1419                                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1420                                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1421                                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1422                                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1423
1424                                 // Check for paragraph
1425                                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1426
1427                                 // Check for bold text
1428                                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1429
1430                                 // Check for Italics text
1431                                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1432
1433                                 // Check for Underline text
1434                                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1435
1436                                 // Check for strike-through text
1437                                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1438
1439                                 // Check for over-line text
1440                                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1441
1442                                 // Check for colored text
1443                                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1444
1445                                 // Check for sized text
1446                                 // [size=50] --> font-size: 50px (with the unit).
1447                                 if ($simple_html != self::DIASPORA) {
1448                                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1px;line-height:normal;">$2</span>', $text);
1449                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1;line-height:normal;">$2</span>', $text);
1450                                 } else {
1451                                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1452                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1453                                 }
1454
1455
1456                                 // Check for centered text
1457                                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", '<div style="text-align:center;">$1</div>', $text);
1458
1459                                 // Check for list text
1460                                 $text = str_replace("[*]", "<li>", $text);
1461
1462                                 // Check for style sheet commands
1463                                 $text = preg_replace("(\[style=(.*?)\](.*?)\[\/style\])ism", '<span style="$1">$2</span>', $text);
1464
1465                                 // Check for CSS classes
1466                                 $text = preg_replace("(\[class=(.*?)\](.*?)\[\/class\])ism", '<span class="$1">$2</span>', $text);
1467
1468                                 // handle nested lists
1469                                 $endlessloop = 0;
1470
1471                                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1472                                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1473                                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1474                                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1475                                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1476                                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1477                                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1478                                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1479                                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1480                                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1481                                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1482                                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1483                                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1484                                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1485                                 }
1486
1487                                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1488                                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1489                                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1490                                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1491
1492                                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1493                                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1494
1495                                 $text = str_replace('[hr]', '<hr />', $text);
1496
1497                                 if (!$for_plaintext) {
1498                                         $text = self::performWithEscapedTags($text, ['url', 'img', 'audio', 'video', 'youtube', 'vimeo', 'share', 'attachment', 'iframe', 'bookmark'], function ($text) {
1499                                                 return preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1500                                         });
1501                                 }
1502
1503                                 // Check for font change text
1504                                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1505
1506                                 // Declare the format for [spoiler] layout
1507                                 $SpoilerLayout = '<details class="spoiler"><summary>' . DI::l10n()->t('Click to open/close') . '</summary>$1</details>';
1508
1509                                 // Check for [spoiler] text
1510                                 // handle nested quotes
1511                                 $endlessloop = 0;
1512                                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1513                                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1514                                 }
1515
1516                                 // Check for [spoiler=Title] text
1517
1518                                 // handle nested quotes
1519                                 $endlessloop = 0;
1520                                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1521                                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1522                                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1523                                                 $text);
1524                                 }
1525
1526                                 // Declare the format for [quote] layout
1527                                 $QuoteLayout = '<blockquote>$1</blockquote>';
1528
1529                                 // Check for [quote] text
1530                                 // handle nested quotes
1531                                 $endlessloop = 0;
1532                                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1533                                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1534                                 }
1535
1536                                 // Check for [quote=Author] text
1537
1538                                 $t_wrote = DI::l10n()->t('$1 wrote:');
1539
1540                                 // handle nested quotes
1541                                 $endlessloop = 0;
1542                                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1543                                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1544                                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1545                                                 $text);
1546                                 }
1547
1548
1549                                 // [img=widthxheight]image source[/img]
1550                                 $text = preg_replace_callback(
1551                                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1552                                         function ($matches) use ($simple_html) {
1553                                                 if (strpos($matches[3], "data:image/") === 0) {
1554                                                         return $matches[0];
1555                                                 }
1556
1557                                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1558                                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1559                                         },
1560                                         $text
1561                                 );
1562
1563                                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1564                                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1565
1566                                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1567                                         function ($matches) use ($simple_html) {
1568                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1569                                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1570                                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1571                                         },
1572                                         $text);
1573
1574                                 // Images
1575                                 // [img]pathtoimage[/img]
1576                                 $text = preg_replace_callback(
1577                                         "/\[img\](.*?)\[\/img\]/ism",
1578                                         function ($matches) use ($simple_html) {
1579                                                 if (strpos($matches[1], "data:image/") === 0) {
1580                                                         return $matches[0];
1581                                                 }
1582
1583                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1584                                                 return "[img]" . $matches[1] . "[/img]";
1585                                         },
1586                                         $text
1587                                 );
1588
1589                                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1590                                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1591
1592                                 $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);
1593                                 $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>', $text);
1594                                 //$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);
1595
1596                                 // Simplify "video" element
1597                                 $text = preg_replace('(\[video[^\]]*?\ssrc\s?=\s?([^\s\]]+)[^\]]*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1598
1599                                 if ($try_oembed) {
1600                                         // html5 video and audio
1601                                         $text = preg_replace("/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4).*?)\[\/video\]/ism",
1602                                                 '<video src="$1" controls width="100%" height="auto"><a href="$1">$1</a></video>', $text);
1603
1604                                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1605                                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1606
1607                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1608                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1609                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '<audio src="$1" controls><a href="$1">$1</a></audio>', $text);
1610                                 } else {
1611                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1612                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1613                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1614                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1615                                 }
1616
1617                                 // Backward compatibility, [iframe] support has been removed in version 2020.12
1618                                 $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1619
1620                                 // Youtube extensions
1621                                 if ($try_oembed) {
1622                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1623                                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1624                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1625                                 }
1626
1627                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1628                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1629                                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1630
1631                                 if ($try_oembed) {
1632                                         $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);
1633                                 } else {
1634                                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1635                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank" rel="noopener noreferrer">https://www.youtube.com/watch?v=$1</a>', $text);
1636                                 }
1637
1638                                 if ($try_oembed) {
1639                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1640                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1641                                 }
1642
1643                                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1644                                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1645
1646                                 if ($try_oembed) {
1647                                         $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);
1648                                 } else {
1649                                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1650                                                 '<a href="https://vimeo.com/$1" target="_blank" rel="noopener noreferrer">https://vimeo.com/$1</a>', $text);
1651                                 }
1652
1653                                 // oembed tag
1654                                 $text = OEmbed::BBCode2HTML($text);
1655
1656                                 // Avoid triple linefeeds through oembed
1657                                 $text = str_replace("<br style='clear:left'></span><br><br>", "<br style='clear:left'></span><br>", $text);
1658
1659                                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1660                                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1661                                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1662                                 // start which is always required). Allow desc with a missing summary for compatibility.
1663
1664                                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1665                                         $sub = Event::getHTML($ev, $simple_html);
1666
1667                                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1668                                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1669                                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1670                                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1671                                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1672                                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1673                                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1674                                 }
1675
1676                                 if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA)) {
1677                                         $conv = html_entity_decode(str_replace([' ', "\n", "\r"], '', $text));
1678                                         // Emojis are always 4 byte Unicode characters
1679                                         if (!empty($conv) && (strlen($conv) / mb_strlen($conv) == 4)) {
1680                                                 $text = '<span style="font-size: xx-large; line-height: normal;">' . $text . '</span>';
1681                                         }
1682                                 }
1683
1684                                 if (!$for_plaintext) {
1685                                         if (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1686                                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1687                                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1688                                         }
1689                                 } else {
1690                                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1691                                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1692                                 }
1693
1694                                 // Remove all hashtag addresses
1695                                 if ($simple_html && !in_array($simple_html, [self::DIASPORA, self::OSTATUS, self::ACTIVITYPUB])) {
1696                                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1697                                 } elseif ($simple_html == self::DIASPORA) {
1698                                         // The ! is converted to @ since Diaspora only understands the @
1699                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1700                                                 '@<a href="$2">$3</a>',
1701                                                 $text);
1702                                 } elseif (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1703                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1704                                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1705                                                 $text);
1706                                 } elseif (!$simple_html) {
1707                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1708                                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1709                                                 $text);
1710                                 }
1711
1712                                 // Bookmarks in red - will be converted to bookmarks in friendica
1713                                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1714                                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1715                                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1716                                                         "[bookmark=$1]$2[/bookmark]", $text);
1717
1718                                 if (in_array($simple_html, [self::API, self::OSTATUS, self::TWITTER])) {
1719                                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1720                                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1721                                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1722                                 }
1723
1724                                 // Perform URL Search
1725                                 if ($try_oembed) {
1726                                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1727                                 }
1728
1729                                 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1730
1731                                 // Handle Diaspora posts
1732                                 $text = preg_replace_callback(
1733                                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1734                                         function ($match) {
1735                                                 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1736                                         }, $text
1737                                 );
1738
1739                                 $text = preg_replace_callback(
1740                                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1741                                         function ($match) {
1742                                                 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1743                                         }, $text
1744                                 );
1745
1746                                 // Server independent link to posts and comments
1747                                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1748                                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1749                                 $text = preg_replace($expression, DI::baseUrl()."/display/$1", $text);
1750
1751                                 /* Tag conversion
1752                                  * Supports:
1753                                  * - #[url=<anything>]<term>[/url]
1754                                  * - [url=<anything>]#<term>[/url]
1755                                  */
1756                                 $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) use ($simple_html) {
1757                                         if ($simple_html == BBCode::ACTIVITYPUB) {
1758                                                 return '<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1759                                                         . '" data-tag="' . XML::escape($matches[1]) . '" rel="tag ugc">#'
1760                                                         . XML::escape($matches[1]) . '</a>';
1761                                         } else {
1762                                                 return '#<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1763                                                         . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1764                                                         . XML::escape($matches[1]) . '</a>';
1765                                         }
1766                                 }, $text);
1767
1768                                 // We need no target="_blank" rel="noopener noreferrer" for local links
1769                                 // convert links start with DI::baseUrl() as local link without the target="_blank" rel="noopener noreferrer" attribute
1770                                 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
1771                                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1772                                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1773
1774                                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1775                                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1776
1777                                 // Red compatibility, though the link can't be authenticated on Friendica
1778                                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1779
1780
1781                                 // we may need to restrict this further if it picks up too many strays
1782                                 // link acct:user@host to a webfinger profile redirector
1783
1784                                 $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);
1785
1786                                 // Perform MAIL Search
1787                                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1788                                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1789
1790                                 /// @todo What is the meaning of these lines?
1791                                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1792                                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1793
1794                                 // Currently deactivated, it made problems with " inside of alt texts.
1795                                 //$text = preg_replace('/\&quot\;/', '"', $text);
1796
1797                                 // fix any escaped ampersands that may have been converted into links
1798                                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1799
1800                                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1801                                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1802
1803                                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1804
1805                                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1806                                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">', $text);
1807
1808                                 // sanitize href attributes (only allowlisted protocols URLs)
1809                                 // default value for backward compatibility
1810                                 $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
1811
1812                                 // Always allowed protocol even if config isn't set or not including it
1813                                 $allowed_link_protocols[] = '//';
1814                                 $allowed_link_protocols[] = 'http://';
1815                                 $allowed_link_protocols[] = 'https://';
1816                                 $allowed_link_protocols[] = 'redir/';
1817
1818                                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1819
1820                                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1821                                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text);
1822
1823                                 // Shared content
1824                                 $text = self::convertShare(
1825                                         $text,
1826                                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1827                                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1828                                         }
1829                                 );
1830
1831                                 $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1832
1833                                 return $text;
1834                         }); // Escaped noparse, nobb, pre
1835
1836                         // Remove escaping tags and replace new lines that remain
1837                         $text = preg_replace_callback('/\[(noparse|nobb)](.*?)\[\/\1]/ism', function ($match) {
1838                                 return str_replace("\n", "<br>", $match[2]);
1839                         }, $text);
1840
1841                         // Additionally, [pre] tags preserve spaces
1842                         $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", function ($match) {
1843                                 return str_replace([' ', "\n"], ['&nbsp;', "<br>"], htmlentities($match[1], ENT_NOQUOTES,'UTF-8'));
1844                         }, $text);
1845
1846                         return $text;
1847                 }); // Escaped code
1848
1849                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1850                         function ($matches) {
1851                                 if (strpos($matches[2], "\n") !== false) {
1852                                         $return = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlentities(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
1853                                 } else {
1854                                         $return = '<code>' . htmlentities($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
1855                                 }
1856
1857                                 return $return;
1858                         },
1859                         $text
1860                 );
1861
1862                 // Default iframe allowed domains/path
1863                 $allowedIframeDomains = [
1864                         DI::baseUrl()->getHostname()
1865                         . (DI::baseUrl()->getUrlPath() ? '/' . DI::baseUrl()->getUrlPath() : '')
1866                         . '/oembed/', # The path part has to change with the source in Content\Oembed::iframe
1867                         'www.youtube.com/embed/',
1868                         'player.vimeo.com/video/',
1869                 ];
1870
1871                 $allowedIframeDomains = array_merge(
1872                         $allowedIframeDomains,
1873                         DI::config()->get('system', 'allowed_oembed') ?
1874                                 explode(',', DI::config()->get('system', 'allowed_oembed'))
1875                                 : []
1876                 );
1877
1878                 $text = HTML::purify($text, $allowedIframeDomains);
1879
1880                 return $text;
1881         }
1882
1883         /**
1884          * Strips the "abstract" tag from the provided text
1885          *
1886          * @param string $text The text with BBCode
1887          * @return string The same text - but without "abstract" element
1888          */
1889         public static function stripAbstract($text)
1890         {
1891                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1892                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1893
1894                 return $text;
1895         }
1896
1897         /**
1898          * Returns the value of the "abstract" element
1899          *
1900          * @param string $text The text that maybe contains the element
1901          * @param string $addon The addon for which the abstract is meant for
1902          * @return string The abstract
1903          */
1904         public static function getAbstract($text, $addon = '')
1905         {
1906                 $abstract = '';
1907                 $abstracts = [];
1908                 $addon = strtolower($addon);
1909
1910                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1911                         foreach ($results AS $result) {
1912                                 $abstracts[strtolower($result[1])] = $result[2];
1913                         }
1914                 }
1915
1916                 if (isset($abstracts[$addon])) {
1917                         $abstract = $abstracts[$addon];
1918                 }
1919
1920                 if ($abstract == '' && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1921                         $abstract = $result[1];
1922                 }
1923
1924                 return $abstract;
1925         }
1926
1927         /**
1928          * Callback function to replace a Friendica style mention in a mention for Diaspora
1929          *
1930          * @param array $match Matching values for the callback
1931          *                     [1] = Mention type (! or @)
1932          *                     [2] = Name
1933          *                     [3] = Address
1934          * @return string Replaced mention
1935          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1936          * @throws \ImagickException
1937          */
1938         private static function bbCodeMention2DiasporaCallback($match)
1939         {
1940                 $contact = Contact::getByURL($match[3], false, ['addr']);
1941                 if (empty($contact['addr'])) {
1942                         return $match[0];
1943                 }
1944
1945                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
1946                 return $mention;
1947         }
1948
1949         /**
1950          * Converts a BBCode text into Markdown
1951          *
1952          * This function converts a BBCode item body to be sent to Markdown-enabled
1953          * systems like Diaspora and Libertree
1954          *
1955          * @param string $text
1956          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1957          * @return string
1958          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1959          */
1960         public static function toMarkdown($text, $for_diaspora = true)
1961         {
1962                 $original_text = $text;
1963
1964                 // Since Diaspora is creating a summary for links, this function removes them before posting
1965                 if ($for_diaspora) {
1966                         $text = self::removeShareInformation($text);
1967                 }
1968
1969                 /**
1970                  * Transform #tags, strip off the [url] and replace spaces with underscore
1971                  */
1972                 $url_search_string = "^\[\]";
1973                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
1974                         function ($matches) {
1975                                 return '#' . str_replace(' ', '_', $matches[2]);
1976                         },
1977                         $text
1978                 );
1979
1980                 // Converting images with size parameters to simple images. Markdown doesn't know it.
1981                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
1982
1983                 // Convert it to HTML - don't try oembed
1984                 if ($for_diaspora) {
1985                         $text = self::convert($text, false, self::DIASPORA);
1986
1987                         // Add all tags that maybe were removed
1988                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
1989                                 $tagline = '';
1990                                 foreach ($tags[2] as $tag) {
1991                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
1992                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
1993                                                 $tagline .= '#' . $tag . ' ';
1994                                         }
1995                                 }
1996                                 $text = $text . " " . $tagline;
1997                         }
1998                 } else {
1999                         $text = self::convert($text, false, self::CONNECTORS);
2000                 }
2001
2002                 // If a link is followed by a quote then there should be a newline before it
2003                 // Maybe we should make this newline at every time before a quote.
2004                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2005
2006                 $stamp1 = microtime(true);
2007
2008                 // Now convert HTML to Markdown
2009                 $text = HTML::toMarkdown($text);
2010
2011                 DI::profiler()->saveTimestamp($stamp1, "parser");
2012
2013                 // Libertree has a problem with escaped hashtags.
2014                 $text = str_replace(['\#'], ['#'], $text);
2015
2016                 // Remove any leading or trailing whitespace, as this will mess up
2017                 // the Diaspora signature verification and cause the item to disappear
2018                 $text = trim($text);
2019
2020                 if ($for_diaspora) {
2021                         $url_search_string = "^\[\]";
2022                         $text = preg_replace_callback(
2023                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2024                                 ['self', 'bbCodeMention2DiasporaCallback'],
2025                                 $text
2026                         );
2027                 }
2028
2029                 Hook::callAll('bb2diaspora', $text);
2030
2031                 return $text;
2032         }
2033
2034         /**
2035          * Pull out all #hashtags and @person tags from $string.
2036          *
2037          * We also get @person@domain.com - which would make
2038          * the regex quite complicated as tags can also
2039          * end a sentence. So we'll run through our results
2040          * and strip the period from any tags which end with one.
2041          * Returns array of tags found, or empty array.
2042          *
2043          * @param string $string Post content
2044          *
2045          * @return array List of tag and person names
2046          */
2047         public static function getTags($string)
2048         {
2049                 $ret = [];
2050
2051                 BBCode::performWithEscapedTags($string, ['noparse', 'pre', 'code', 'img'], function ($string) use (&$ret) {
2052                         // Convert hashtag links to hashtags
2053                         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2 ', $string);
2054
2055                         // Force line feeds at bbtags
2056                         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2057
2058                         // ignore anything in a bbtag
2059                         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2060
2061                         // Match full names against @tags including the space between first and last
2062                         // We will look these up afterward to see if they are full names or not recognisable.
2063
2064                         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2065                                 foreach ($matches[1] as $match) {
2066                                         if (strstr($match, ']')) {
2067                                                 // we might be inside a bbcode color tag - leave it alone
2068                                                 continue;
2069                                         }
2070
2071                                         if (substr($match, -1, 1) === '.') {
2072                                                 $ret[] = substr($match, 0, -1);
2073                                         } else {
2074                                                 $ret[] = $match;
2075                                         }
2076                                 }
2077                         }
2078
2079                         // Otherwise pull out single word tags. These can be @nickname, @first_last
2080                         // and #hash tags.
2081
2082                         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?\']*[^\^ \x0D\x0A,;:?!\'.])/', $string, $matches)) {
2083                                 foreach ($matches[1] as $match) {
2084                                         if (strstr($match, ']')) {
2085                                                 // we might be inside a bbcode color tag - leave it alone
2086                                                 continue;
2087                                         }
2088
2089                                         // try not to catch url fragments
2090                                         if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2091                                                 continue;
2092                                         }
2093
2094                                         $ret[] = $match;
2095                                 }
2096                         }
2097                 });
2098
2099                 return array_unique($ret);
2100         }
2101
2102         /**
2103          * Perform a custom function on a text after having escaped blocks enclosed in the provided tag list.
2104          *
2105          * @param string   $text
2106          * @param array    $tagList A list of tag names, e.g ['noparse', 'nobb', 'pre']
2107          * @param callable $callback
2108          * @return string
2109          * @throws Exception
2110          *@see Strings::performWithEscapedBlocks
2111          *
2112          */
2113         public static function performWithEscapedTags(string $text, array $tagList, callable $callback)
2114         {
2115                 $tagList = array_map('preg_quote', $tagList);
2116
2117                 return Strings::performWithEscapedBlocks($text, '#\[(?:' . implode('|', $tagList) . ').*?\[/(?:' . implode('|', $tagList) . ')]#ism', $callback);
2118         }
2119
2120         /**
2121          * Replaces mentions in the provided message body for the provided user and network if any
2122          *
2123          * @param $body
2124          * @param $profile_uid
2125          * @param $network
2126          * @return string
2127          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2128          * @throws \ImagickException
2129          */
2130         public static function setMentions($body, $profile_uid = 0, $network = '')
2131         {
2132                 BBCode::performWithEscapedTags($body, ['noparse', 'pre', 'code', 'img'], function ($body) use ($profile_uid, $network) {
2133                         $tags = BBCode::getTags($body);
2134
2135                         $tagged = [];
2136                         $inform = '';
2137
2138                         foreach ($tags as $tag) {
2139                                 $tag_type = substr($tag, 0, 1);
2140
2141                                 if ($tag_type == Tag::TAG_CHARACTER[Tag::HASHTAG]) {
2142                                         continue;
2143                                 }
2144
2145                                 /*
2146                                  * If we already tagged 'Robert Johnson', don't try and tag 'Robert'.
2147                                  * Robert Johnson should be first in the $tags array
2148                                  */
2149                                 foreach ($tagged as $nextTag) {
2150                                         if (stristr($nextTag, $tag . ' ')) {
2151                                                 continue 2;
2152                                         }
2153                                 }
2154
2155                                 $success = Item::replaceTag($body, $inform, $profile_uid, $tag, $network);
2156
2157                                 if ($success['replaced']) {
2158                                         $tagged[] = $tag;
2159                                 }
2160                         }
2161
2162                         return $body;
2163                 });
2164
2165                 return $body;
2166         }
2167
2168         /**
2169          * @param string      $author  Author display name
2170          * @param string      $profile Author profile URL
2171          * @param string      $avatar  Author profile picture URL
2172          * @param string      $link    Post source URL
2173          * @param string      $posted  Post created date
2174          * @param string|null $guid    Post guid (if any)
2175          * @return string
2176          * @TODO Rewrite to handle over whole record array
2177          */
2178         public static function getShareOpeningTag(string $author, string $profile, string $avatar, string $link, string $posted, string $guid = null)
2179         {
2180                 $header = "[share author='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $author) .
2181                         "' profile='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $profile) .
2182                         "' avatar='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $avatar) .
2183                         "' link='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $link) .
2184                         "' posted='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $posted);
2185
2186                 if ($guid) {
2187                         $header .= "' guid='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $guid);
2188                 }
2189
2190                 $header  .= "']";
2191
2192                 return $header;
2193         }
2194
2195         /**
2196          * Returns the BBCode relevant to embed the provided URL in a post body.
2197          * For media type, it will return [img], [video] and [audio] tags.
2198          * For regular web pages, it will either output a [bookmark] tag if title and description were provided,
2199          * an [attachment] tag or a simple [url] tag depending on $tryAttachment.
2200          *
2201          * @param string      $url
2202          * @param bool        $tryAttachment
2203          * @param string|null $title
2204          * @param string|null $description
2205          * @param string|null $tags
2206          * @return string
2207          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2208          *@see ParseUrl::getSiteinfoCached
2209          *
2210          */
2211         public static function embedURL(string $url, bool $tryAttachment = true, string $title = null, string $description = null, string $tags = null): string
2212         {
2213                 DI::logger()->info($url);
2214
2215                 // If there is already some content information submitted we don't
2216                 // need to parse the url for content.
2217                 if (!empty($title) && !empty($description)) {
2218                         $title = str_replace(["\r", "\n"], ['', ''], $title);
2219
2220                         $description = '[quote]' . trim($description) . '[/quote]' . "\n";
2221
2222                         $str_tags = '';
2223                         if (!empty($tags)) {
2224                                 $arr_tags = ParseUrl::convertTagsToArray($tags);
2225                                 if (count($arr_tags)) {
2226                                         $str_tags = "\n" . implode(' ', $arr_tags) . "\n";
2227                                 }
2228                         }
2229
2230                         $result = sprintf('[bookmark=%s]%s[/bookmark]%s', $url, ($title) ? $title : $url, $description) . $str_tags;
2231
2232                         DI::logger()->info('(unparsed): returns: ' . $result);
2233
2234                         return $result;
2235                 }
2236
2237                 $siteinfo = ParseUrl::getSiteinfoCached($url);
2238
2239                 if (in_array($siteinfo['type'], ['image', 'video', 'audio'])) {
2240                         switch ($siteinfo['type']) {
2241                                 case 'video':
2242                                         $bbcode = "\n" . '[video]' . $url . '[/video]' . "\n";
2243                                         break;
2244                                 case 'audio':
2245                                         $bbcode = "\n" . '[audio]' . $url . '[/audio]' . "\n";
2246                                         break;
2247                                 default:
2248                                         $bbcode = "\n" . '[img]' . $url . '[/img]' . "\n";
2249                                         break;
2250                         }
2251
2252                         return $bbcode;
2253                 }
2254
2255                 unset($siteinfo['keywords']);
2256
2257                 // Bypass attachment if parse url for a comment
2258                 if (!$tryAttachment) {
2259                         return "\n" . '[url=' . $url . ']' . $siteinfo['title'] . '[/url]';
2260                 }
2261
2262                 // Format it as BBCode attachment
2263                 return "\n" . PageInfo::getFooterFromData($siteinfo);
2264         }
2265 }