src/Content/Text/BBCode.php

   1 <?php
   2 /**
   3  * @copyright Copyright (C) 2020, Friendica
   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\Smilies;
  31 use Friendica\Core\Hook;
  32 use Friendica\Core\Logger;
  33 use Friendica\Core\Protocol;
  34 use Friendica\Core\Renderer;
  35 use Friendica\Core\System;
  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         const INTERNAL = 0;
  53         const API = 2;
  54         const DIASPORA = 3;
  55         const CONNECTORS = 4;
  56         const OSTATUS = 7;
  57         const TWITTER = 8;
  58         const BACKLINK = 8;
  59         const ACTIVITYPUB = 9;
  60
  61         /**
  62          * Fetches attachment data that were generated the old way
  63          *
  64          * @param string $body Message body
  65          * @return array
  66          *                     'type' -> Message type ('link', 'video', 'photo')
  67          *                     'text' -> Text before the shared message
  68          *                     'after' -> Text after the shared message
  69          *                     'image' -> Preview image of the message
  70          *                     'url' -> Url to the attached message
  71          *                     'title' -> Title of the attachment
  72          *                     'description' -> Description of the attachment
  73          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  74          */
  75         private static function getOldAttachmentData($body)
  76         {
  77                 $post = [];
  78
  79                 // Simplify image codes
  80                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  81
  82                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  83                         foreach ($attached as $data) {
  84                                 if (!in_array($data[1], ['type-link', 'type-video', 'type-photo'])) {
  85                                         continue;
  86                                 }
  87
  88                                 $post['type'] = substr($data[1], 5);
  89
  90                                 $pos = strpos($body, $data[0]);
  91                                 if ($pos > 0) {
  92                                         $post['text'] = trim(substr($body, 0, $pos));
  93                                         $post['after'] = trim(substr($body, $pos + strlen($data[0])));
  94                                 } else {
  95                                         $post['text'] = trim(str_replace($data[0], '', $body));
  96                                         $post['after'] = '';
  97                                 }
  98
  99                                 $attacheddata = $data[2];
 100
 101                                 if (preg_match("/\[img\](.*?)\[\/img\]/ism", $attacheddata, $matches)) {
 102
 103                                         $picturedata = Images::getInfoFromURLCached($matches[1]);
 104
 105                                         if ($picturedata) {
 106                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
 107                                                         $post['image'] = $matches[1];
 108                                                 } else {
 109                                                         $post['preview'] = $matches[1];
 110                                                 }
 111                                         }
 112                                 }
 113
 114                                 if (preg_match("/\[bookmark\=(.*?)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
 115                                         $post['url'] = $matches[1];
 116                                         $post['title'] = $matches[2];
 117                                 }
 118                                 if (!empty($post['url']) && (in_array($post['type'], ['link', 'video']))
 119                                         && preg_match("/\[url\=(.*?)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
 120                                         $post['url'] = $matches[1];
 121                                 }
 122
 123                                 // Search for description
 124                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
 125                                         $post['description'] = $matches[1];
 126                                 }
 127                         }
 128                 }
 129                 return $post;
 130         }
 131
 132         /**
 133          * Fetches attachment data that were generated with the "attachment" element
 134          *
 135          * @param string $body Message body
 136          * @return array
 137          *                     'type' -> Message type ('link', 'video', 'photo')
 138          *                     'text' -> Text before the shared message
 139          *                     'after' -> Text after the shared message
 140          *                     'image' -> Preview image of the message
 141          *                     'url' -> Url to the attached message
 142          *                     'title' -> Title of the attachment
 143          *                     'description' -> Description of the attachment
 144          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 145          */
 146         public static function getAttachmentData($body)
 147         {
 148                 $data = [
 149                         'type'        => '',
 150                         'text'        => '',
 151                         'after'       => '',
 152                         'image'       => null,
 153                         'url'         => '',
 154                         'title'       => '',
 155                         'description' => '',
 156                 ];
 157
 158                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 159                         return self::getOldAttachmentData($body);
 160                 }
 161
 162                 $attributes = $match[2];
 163
 164                 $data['text'] = trim($match[1]);
 165
 166                 $type = '';
 167                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 168                 if (!empty($matches[1])) {
 169                         $type = strtolower($matches[1]);
 170                 }
 171
 172                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 173                 if (!empty($matches[1])) {
 174                         $type = strtolower($matches[1]);
 175                 }
 176
 177                 if ($type == '') {
 178                         return [];
 179                 }
 180
 181                 if (!in_array($type, ['link', 'audio', 'photo', 'video'])) {
 182                         return [];
 183                 }
 184
 185                 if ($type != '') {
 186                         $data['type'] = $type;
 187                 }
 188
 189                 $url = '';
 190                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 191                 if (!empty($matches[1])) {
 192                         $url = $matches[1];
 193                 }
 194
 195                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 196                 if (!empty($matches[1])) {
 197                         $url = $matches[1];
 198                 }
 199
 200                 if ($url != '') {
 201                         $data['url'] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 202                 }
 203
 204                 $title = '';
 205                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 206                 if (!empty($matches[1])) {
 207                         $title = $matches[1];
 208                 }
 209
 210                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 211                 if (!empty($matches[1])) {
 212                         $title = $matches[1];
 213                 }
 214
 215                 if ($title != '') {
 216                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 217                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 218                         $title = str_replace(['[', ']'], ['&#91;', '&#93;'], $title);
 219                         $data['title'] = $title;
 220                 }
 221
 222                 $image = '';
 223                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 224                 if (!empty($matches[1])) {
 225                         $image = $matches[1];
 226                 }
 227
 228                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 229                 if (!empty($matches[1])) {
 230                         $image = $matches[1];
 231                 }
 232
 233                 if ($image != '') {
 234                         $data['image'] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 235                 }
 236
 237                 $preview = '';
 238                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 239                 if (!empty($matches[1])) {
 240                         $preview = $matches[1];
 241                 }
 242
 243                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 244                 if (!empty($matches[1])) {
 245                         $preview = $matches[1];
 246                 }
 247
 248                 if ($preview != '') {
 249                         $data['preview'] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 250                 }
 251
 252                 $data['description'] = trim($match[3]);
 253
 254                 $data['after'] = trim($match[4]);
 255
 256                 return $data;
 257         }
 258
 259         public static function getAttachedData($body, $item = [])
 260         {
 261                 /*
 262                 - text:
 263                 - type: link, video, photo
 264                 - title:
 265                 - url:
 266                 - image:
 267                 - description:
 268                 - (thumbnail)
 269                 */
 270
 271                 $has_title = !empty($item['title']);
 272                 $plink = $item['plink'] ?? '';
 273                 $post = self::getAttachmentData($body);
 274
 275                 // Get all linked images with alternative image description
 276                 if (preg_match_all("/\[img=([^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 277                         foreach ($pictures as $picture) {
 278                                 if (Photo::isLocal($picture[1])) {
 279                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]];
 280                                 }
 281                         }
 282                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 283                                 $post['image_description'] = $post['images'][0]['description'];
 284                         }
 285                 }
 286
 287                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 288                         foreach ($pictures as $picture) {
 289                                 if (Photo::isLocal($picture[1])) {
 290                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
 291                                 }
 292                         }
 293                 }
 294
 295                 // if nothing is found, it maybe having an image.
 296                 if (!isset($post['type'])) {
 297                         // Simplify image codes
 298                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 299                         $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 300                         $post['text'] = $body;
 301
 302                         if (preg_match_all("#\[url=([^\]]+?)\]\s*\[img\]([^\[]+?)\[/img\]\s*\[/url\]#ism", $body, $pictures, PREG_SET_ORDER)) {
 303                                 if ((count($pictures) == 1) && !$has_title) {
 304                                         if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) {
 305                                                 // Replace the preview picture with the real picture
 306                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
 307                                                 $data = ['url' => $url, 'type' => 'photo'];
 308                                         } else {
 309                                                 // Checking, if the link goes to a picture
 310                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 311                                         }
 312
 313                                         // Workaround:
 314                                         // Sometimes photo posts to the own album are not detected at the start.
 315                                         // So we seem to cannot use the cache for these cases. That's strange.
 316                                         if (($data['type'] != 'photo') && strstr($pictures[0][1], "/photos/")) {
 317                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 318                                         }
 319
 320                                         if ($data['type'] == 'photo') {
 321                                                 $post['type'] = 'photo';
 322                                                 if (isset($data['images'][0])) {
 323                                                         $post['image'] = $data['images'][0]['src'];
 324                                                         $post['url'] = $data['url'];
 325                                                 } else {
 326                                                         $post['image'] = $data['url'];
 327                                                 }
 328
 329                                                 $post['preview'] = $pictures[0][2];
 330                                                 $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 331                                         } else {
 332                                                 $imgdata = Images::getInfoFromURLCached($pictures[0][1]);
 333                                                 if ($imgdata && substr($imgdata['mime'], 0, 6) == 'image/') {
 334                                                         $post['type'] = 'photo';
 335                                                         $post['image'] = $pictures[0][1];
 336                                                         $post['preview'] = $pictures[0][2];
 337                                                         $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 338                                                 }
 339                                         }
 340                                 } elseif (count($pictures) > 0) {
 341                                         $post['type'] = 'link';
 342                                         $post['url'] = $plink;
 343                                         $post['image'] = $pictures[0][2];
 344                                         $post['text'] = $body;
 345
 346                                         foreach ($pictures as $picture) {
 347                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 348                                         }
 349                                 }
 350                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 351                                 if ((count($pictures) == 1) && !$has_title) {
 352                                         $post['type'] = 'photo';
 353                                         $post['image'] = $pictures[0][1];
 354                                         $post['text'] = str_replace($pictures[0][0], '', $body);
 355                                 } elseif (count($pictures) > 0) {
 356                                         $post['type'] = 'link';
 357                                         $post['url'] = $plink;
 358                                         $post['image'] = $pictures[0][1];
 359                                         $post['text'] = $body;
 360
 361                                         foreach ($pictures as $picture) {
 362                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 363                                         }
 364                                 }
 365                         }
 366
 367                         // Test for the external links
 368                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER);
 369                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER);
 370
 371                         $links = array_merge($links1, $links2);
 372
 373                         // If there is only a single one, then use it.
 374                         // This should cover link posts via API.
 375                         if ((count($links) == 1) && !isset($post['preview']) && !$has_title) {
 376                                 $post['type'] = 'link';
 377                                 $post['url'] = $links[0][1];
 378                         }
 379
 380                         // Simplify "video" element
 381                         $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']);
 382
 383                         // Now count the number of external media links
 384                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
 385                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
 386                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
 387                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
 388
 389                         // Add them to the other external links
 390                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 391
 392                         // Are there more than one?
 393                         if (count($links) > 1) {
 394                                 // The post will be the type "text", which means a blog post
 395                                 unset($post['type']);
 396                                 $post['url'] = $plink;
 397                         }
 398
 399                         if (!isset($post['type'])) {
 400                                 $post['type'] = "text";
 401                                 $post['text'] = trim($body);
 402                         }
 403                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
 404                         $data = ParseUrl::getSiteinfoCached($post['url'], true);
 405
 406                         if (isset($data['images'][0])) {
 407                                 $post['image'] = $data['images'][0]['src'];
 408                         }
 409                 }
 410
 411                 return $post;
 412         }
 413
 414         /**
 415          * Remove [attachment] BBCode and replaces it with a regular [url]
 416          *
 417          * @param string  $body
 418          * @param boolean $no_link_desc No link description
 419          *
 420          * @return string with replaced body
 421          */
 422         public static function removeAttachment($body, $no_link_desc = false)
 423         {
 424                 return preg_replace_callback("/\s*\[attachment (.*)\](.*?)\[\/attachment\]\s*/ism",
 425                         function ($match) use ($no_link_desc) {
 426                                 $attach_data = self::getAttachmentData($match[0]);
 427                                 if (empty($attach_data['url'])) {
 428                                         return $match[0];
 429                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
 430                                         return "\n[url]" . $attach_data['url'] . "[/url]\n";
 431                                 } else {
 432                                         return "\n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
 433                                 }
 434                 }, $body);
 435         }
 436
 437         /**
 438          * Converts a BBCode text into plaintext
 439          *
 440          * @param      $text
 441          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 442          *
 443          * @return string
 444          */
 445         public static function toPlaintext($text, $keep_urls = true)
 446         {
 447                 $naked_text = HTML::toPlaintext(self::convert($text, false, 0, true), 0, !$keep_urls);
 448
 449                 return $naked_text;
 450         }
 451
 452         private static function proxyUrl($image, $simplehtml = self::INTERNAL)
 453         {
 454                 // Only send proxied pictures to API and for internal display
 455                 if (in_array($simplehtml, [self::INTERNAL, self::API])) {
 456                         return ProxyUtils::proxifyUrl($image);
 457                 } else {
 458                         return $image;
 459                 }
 460         }
 461
 462         /**
 463          * This function changing the visual size (not the real size) of images.
 464          * The function does not work for pictures with an alternate text description.
 465          * This could only be changed by using some new "img" BBCode format.
 466          *
 467          * @param string $srctext The body with images
 468          * @return string The body with possibly scaled images
 469          */
 470         public static function scaleExternalImages(string $srctext)
 471         {
 472                 $s = $srctext;
 473
 474                 // Simplify image links
 475                 $s = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $s);
 476
 477                 $matches = null;
 478                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 479                 if ($c) {
 480                         foreach ($matches as $mtch) {
 481                                 Logger::log('scale_external_image: ' . $mtch[1]);
 482
 483                                 $hostname = str_replace('www.', '', substr(DI::baseUrl(), strpos(DI::baseUrl(), '://') + 3));
 484                                 if (stristr($mtch[1], $hostname)) {
 485                                         continue;
 486                                 }
 487
 488                                 $curlResult = DI::httpRequest()->get($mtch[1], true);
 489                                 if (!$curlResult->isSuccess()) {
 490                                         continue;
 491                                 }
 492
 493                                 $i = $curlResult->getBody();
 494                                 $type = $curlResult->getContentType();
 495                                 $type = Images::getMimeTypeByData($i, $mtch[1], $type);
 496
 497                                 if ($i) {
 498                                         $Image = new Image($i, $type);
 499                                         if ($Image->isValid()) {
 500                                                 $orig_width = $Image->getWidth();
 501                                                 $orig_height = $Image->getHeight();
 502
 503                                                 if ($orig_width > 640 || $orig_height > 640) {
 504                                                         $Image->scaleDown(640);
 505                                                         $new_width = $Image->getWidth();
 506                                                         $new_height = $Image->getHeight();
 507                                                         Logger::info('External images scaled', ['orig_width' => $orig_width, 'new_width' => $new_width, 'orig_height' => $orig_height, 'new_height' => $new_height, 'match' => $mtch[0]]);
 508                                                         $s = str_replace(
 509                                                                 $mtch[0],
 510                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]'
 511                                                                 . "\n",
 512                                                                 $s
 513                                                         );
 514                                                         Logger::info('New string', ['image' => $s]);
 515                                                 }
 516                                         }
 517                                 }
 518                         }
 519                 }
 520
 521                 return $s;
 522         }
 523
 524         /**
 525          * Truncates imported message body string length to max_import_size
 526          *
 527          * The purpose of this function is to apply system message length limits to
 528          * imported messages without including any embedded photos in the length
 529          *
 530          * @param string $body
 531          * @return string
 532          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 533          */
 534         public static function limitBodySize($body)
 535         {
 536                 $maxlen = DI::config()->get('config', 'max_import_size', 0);
 537
 538                 // If the length of the body, including the embedded images, is smaller
 539                 // than the maximum, then don't waste time looking for the images
 540                 if ($maxlen && (strlen($body) > $maxlen)) {
 541
 542                         Logger::info('the total body length exceeds the limit', ['maxlen' => $maxlen, 'body_len' => strlen($body)]);
 543
 544                         $orig_body = $body;
 545                         $new_body = '';
 546                         $textlen = 0;
 547
 548                         $img_start = strpos($orig_body, '[img');
 549                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 550                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 551                         while (($img_st_close !== false) && ($img_end !== false)) {
 552
 553                                 $img_st_close++; // make it point to AFTER the closing bracket
 554                                 $img_end += $img_start;
 555                                 $img_end += strlen('[/img]');
 556
 557                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 558                                         // This is an embedded image
 559
 560                                         if (($textlen + $img_start) > $maxlen) {
 561                                                 if ($textlen < $maxlen) {
 562                                                         Logger::info('the limit happens before an embedded image');
 563                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 564                                                         $textlen = $maxlen;
 565                                                 }
 566                                         } else {
 567                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 568                                                 $textlen += $img_start;
 569                                         }
 570
 571                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 572                                 } else {
 573
 574                                         if (($textlen + $img_end) > $maxlen) {
 575                                                 if ($textlen < $maxlen) {
 576                                                         Logger::info('the limit happens before the end of a non-embedded image');
 577                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 578                                                         $textlen = $maxlen;
 579                                                 }
 580                                         } else {
 581                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 582                                                 $textlen += $img_end;
 583                                         }
 584                                 }
 585                                 $orig_body = substr($orig_body, $img_end);
 586
 587                                 if ($orig_body === false) {
 588                                         // in case the body ends on a closing image tag
 589                                         $orig_body = '';
 590                                 }
 591
 592                                 $img_start = strpos($orig_body, '[img');
 593                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 594                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 595                         }
 596
 597                         if (($textlen + strlen($orig_body)) > $maxlen) {
 598                                 if ($textlen < $maxlen) {
 599                                         Logger::info('the limit happens after the end of the last image');
 600                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 601                                 }
 602                         } else {
 603                                 Logger::info('the text size with embedded images extracted did not violate the limit');
 604                                 $new_body = $new_body . $orig_body;
 605                         }
 606
 607                         return $new_body;
 608                 } else {
 609                         return $body;
 610                 }
 611         }
 612
 613         /**
 614          * Processes [attachment] tags
 615          *
 616          * Note: Can produce a [bookmark] tag in the returned string
 617          *
 618          * @param string  $text
 619          * @param integer $simplehtml
 620          * @param bool    $tryoembed
 621          * @return string
 622          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 623          */
 624         private static function convertAttachment($text, $simplehtml = self::INTERNAL, $tryoembed = true)
 625         {
 626                 $data = self::getAttachmentData($text);
 627                 if (empty($data) || empty($data['url'])) {
 628                         return $text;
 629                 }
 630
 631                 if (isset($data['title'])) {
 632                         $data['title'] = strip_tags($data['title']);
 633                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 634                 } else {
 635                         $data['title'] = null;
 636                 }
 637
 638                 if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) {
 639                         $data['preview'] = $data['image'];
 640                         $data['image'] = '';
 641                 }
 642
 643                 $return = '';
 644                 try {
 645                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 646                                 $return = OEmbed::getHTML($data['url'], $data['title']);
 647                         } else {
 648                                 throw new Exception('OEmbed is disabled for this attachment.');
 649                         }
 650                 } catch (Exception $e) {
 651                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 652
 653                         if ($simplehtml != self::CONNECTORS) {
 654                                 $return = sprintf('<div class="type-%s">', $data['type']);
 655                         }
 656
 657                         if (!empty($data['title']) && !empty($data['url'])) {
 658                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 659                                         $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), $data['title']);
 660                                 } else {
 661                                         if (!empty($data['image'])) {
 662                                                 $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), $data['title']);
 663                                         } elseif (!empty($data['preview'])) {
 664                                                 $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']);
 665                                         }
 666                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 667                                 }
 668                         }
 669
 670                         if (!empty($data['description']) && $data['description'] != $data['title']) {
 671                                 // Sanitize the HTML by converting it to BBCode
 672                                 $bbcode = HTML::toBBCode($data['description']);
 673                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 674                         }
 675
 676                         if (!empty($data['url'])) {
 677                                 $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 678                         }
 679
 680                         if ($simplehtml != self::CONNECTORS) {
 681                                 $return .= '</div>';
 682                         }
 683                 }
 684
 685                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 686         }
 687
 688         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 689         {
 690                 $data = self::getAttachmentData($Text);
 691
 692                 if (!$data) {
 693                         return $Text;
 694                 } elseif ($nolink) {
 695                         return $data['text'] . ($data['after'] ?? '');
 696                 }
 697
 698                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 699                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 700                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 701                         $data['title'] = $data['url'];
 702                 } elseif (($text != '') && strstr($title, $text)) {
 703                         $data['text'] = $data['title'];
 704                         $data['title'] = $data['url'];
 705                 }
 706
 707                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 708                         return $data['title'] . $data['after'];
 709                 }
 710
 711                 // If the link already is included in the post, don't add it again
 712                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 713                         return $data['text'] . $data['after'];
 714                 }
 715
 716                 $text = $data['text'];
 717
 718                 if (!empty($data['url']) && !empty($data['title'])) {
 719                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 720                 } elseif (!empty($data['url'])) {
 721                         $text .= "\n[url]" . $data['url'] . '[/url]';
 722                 }
 723
 724                 return $text . "\n" . $data['after'];
 725         }
 726
 727         /**
 728          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 729          *
 730          * @param array $match Array with the matching values
 731          * @return string reformatted link including HTML codes
 732          */
 733         private static function convertUrlForActivityPubCallback($match)
 734         {
 735                 $url = $match[1];
 736
 737                 if (isset($match[2]) && ($match[1] != $match[2])) {
 738                         return $match[0];
 739                 }
 740
 741                 $parts = parse_url($url);
 742                 if (!isset($parts['scheme'])) {
 743                         return $match[0];
 744                 }
 745
 746                 return self::convertUrlForActivityPub($url);
 747         }
 748
 749         /**
 750          * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 751          *
 752          * @param string $url URL that is about to be reformatted
 753          * @return string reformatted link including HTML codes
 754          */
 755         private static function convertUrlForActivityPub($url)
 756         {
 757                 $html = '<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>';
 758                 return sprintf($html, $url, self::getStyledURL($url));
 759         }
 760
 761         /**
 762          * Converts an URL in a nicer format (without the scheme and possibly shortened)
 763          *
 764          * @param string $url URL that is about to be reformatted
 765          * @return string reformatted link
 766          */
 767         private static function getStyledURL($url)
 768         {
 769                 $parts = parse_url($url);
 770                 $scheme = $parts['scheme'] . '://';
 771                 $styled_url = str_replace($scheme, '', $url);
 772
 773                 if (strlen($styled_url) > 30) {
 774                         $styled_url = substr($styled_url, 0, 30) . "…";
 775                 }
 776
 777                 return $styled_url;
 778         }
 779
 780         /*
 781          * [noparse][i]italic[/i][/noparse] turns into
 782          * [noparse][ i ]italic[ /i ][/noparse],
 783          * to hide them from parser.
 784          */
 785         private static function escapeNoparseCallback($match)
 786         {
 787                 $whole_match = $match[0];
 788                 $captured = $match[1];
 789                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 790                 $new_str = str_replace($captured, $spacefied, $whole_match);
 791                 return $new_str;
 792         }
 793
 794         /*
 795          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 796          * now turns back and the [noparse] tags are trimed
 797          * returning [i]italic[/i]
 798          */
 799         private static function unescapeNoparseCallback($match)
 800         {
 801                 $captured = $match[1];
 802                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 803                 return $unspacefied;
 804         }
 805
 806         /**
 807          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 808          * occurrences
 809          *
 810          * @param string $text        Text to search
 811          * @param string $name        Tag name
 812          * @param int    $occurrences Number of first occurrences to skip
 813          * @return boolean|array
 814          */
 815         public static function getTagPosition($text, $name, $occurrences = 0)
 816         {
 817                 if ($occurrences < 0) {
 818                         $occurrences = 0;
 819                 }
 820
 821                 $start_open = -1;
 822                 for ($i = 0; $i <= $occurrences; $i++) {
 823                         if ($start_open !== false) {
 824                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 825                         }
 826                 }
 827
 828                 if ($start_open === false) {
 829                         return false;
 830                 }
 831
 832                 $start_equal = strpos($text, '=', $start_open);
 833                 $start_close = strpos($text, ']', $start_open);
 834
 835                 if ($start_close === false) {
 836                         return false;
 837                 }
 838
 839                 $start_close++;
 840
 841                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 842
 843                 if ($end_open === false) {
 844                         return false;
 845                 }
 846
 847                 $res = [
 848                         'start' => [
 849                                 'open' => $start_open,
 850                                 'close' => $start_close
 851                         ],
 852                         'end' => [
 853                                 'open' => $end_open,
 854                                 'close' => $end_open + strlen('[/' . $name . ']')
 855                         ],
 856                 ];
 857
 858                 if ($start_equal !== false) {
 859                         $res['start']['equal'] = $start_equal + 1;
 860                 }
 861
 862                 return $res;
 863         }
 864
 865         /**
 866          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 867          *
 868          * @param string $pattern Preg pattern string
 869          * @param string $replace Preg replace string
 870          * @param string $name    BBCode tag name
 871          * @param string $text    Text to search
 872          * @return string
 873          */
 874         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 875         {
 876                 $occurrences = 0;
 877                 $pos = self::getTagPosition($text, $name, $occurrences);
 878                 while ($pos !== false && $occurrences++ < 1000) {
 879                         $start = substr($text, 0, $pos['start']['open']);
 880                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 881                         $end = substr($text, $pos['end']['close']);
 882                         if ($end === false) {
 883                                 $end = '';
 884                         }
 885
 886                         $subject = preg_replace($pattern, $replace, $subject);
 887                         $text = $start . $subject . $end;
 888
 889                         $pos = self::getTagPosition($text, $name, $occurrences);
 890                 }
 891
 892                 return $text;
 893         }
 894
 895         private static function extractImagesFromItemBody($body)
 896         {
 897                 $saved_image = [];
 898                 $orig_body = $body;
 899                 $new_body = '';
 900
 901                 $cnt = 0;
 902                 $img_start = strpos($orig_body, '[img');
 903                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 904                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 905                 while (($img_st_close !== false) && ($img_end !== false)) {
 906                         $img_st_close++; // make it point to AFTER the closing bracket
 907                         $img_end += $img_start;
 908
 909                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 910                                 // This is an embedded image
 911                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 912                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 913
 914                                 $cnt++;
 915                         } else {
 916                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 917                         }
 918
 919                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 920
 921                         if ($orig_body === false) {
 922                                 // in case the body ends on a closing image tag
 923                                 $orig_body = '';
 924                         }
 925
 926                         $img_start = strpos($orig_body, '[img');
 927                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 928                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 929                 }
 930
 931                 $new_body = $new_body . $orig_body;
 932
 933                 return ['body' => $new_body, 'images' => $saved_image];
 934         }
 935
 936         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 937         {
 938                 $newbody = $body;
 939
 940                 $cnt = 0;
 941                 foreach ($images as $image) {
 942                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 943                         // it loops over the array starting from the first element and going sequentially
 944                         // to the last element
 945                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 946                                 '<img src="' . self::proxyUrl($image) . '" alt="' . DI::l10n()->t('Image/photo') . '" />', $newbody);
 947                         $cnt++;
 948                 }
 949
 950                 return $newbody;
 951         }
 952
 953         /**
 954          * This function converts a [share] block to text according to a provided callback function whose signature is:
 955          *
 956          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 957          *
 958          * Where:
 959          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 960          * data lookup
 961          * - $author_contact is a contact record array
 962          * - $content is the inner content of the [share] block
 963          * - $is_quote_share indicates whether there's any content before the [share] block
 964          * - Return value is the string that should replace the [share] block in the provided text
 965          *
 966          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 967          *
 968          * @param  string   $text     A BBCode string
 969          * @param  callable $callback
 970          * @return string The BBCode string with all [share] blocks replaced
 971          */
 972         public static function convertShare($text, callable $callback)
 973         {
 974                 $return = preg_replace_callback(
 975                         "/(.*?)\[share(.*?)\](.*)\[\/share\]/ism",
 976                         function ($match) use ($callback) {
 977                                 $attribute_string = $match[2];
 978                                 $attributes = [];
 979                                 foreach (['author', 'profile', 'avatar', 'link', 'posted', 'guid'] as $field) {
 980                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 981                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 982                                 }
 983
 984                                 $author_contact = Contact::getByURL($attributes['profile'], false, ['url', 'addr', 'name', 'micro']);
 985                                 $author_contact['url'] = ($author_contact['url'] ?? $attributes['profile']);
 986                                 $author_contact['addr'] = ($author_contact['addr'] ?? '') ?: Protocol::getAddrFromProfileUrl($attributes['profile']);
 987
 988                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
 989                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
 990                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
 991
 992                                 if ($attributes['avatar']) {
 993                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
 994                                 }
 995
 996                                 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
 997                         },
 998                         $text
 999                 );
1000
1001                 return $return;
1002         }
1003
1004         /**
1005          * Default [share] tag conversion callback
1006          *
1007          * Note: Can produce a [bookmark] tag in the output
1008          *
1009          * @see BBCode::convertShare()
1010          * @param array   $attributes     [share] block attribute values
1011          * @param array   $author_contact Contact row of the shared author
1012          * @param string  $content        Inner content of the [share] block
1013          * @param boolean $is_quote_share Whether there is content before the [share] block
1014          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
1015          * @return string
1016          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1017          */
1018         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
1019         {
1020                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
1021
1022                 switch ($simplehtml) {
1023                         case self::API:
1024                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1025                                 break;
1026                         case self::DIASPORA:
1027                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1028                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1029                                 } else {
1030                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1031
1032                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1033                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1034                                         }
1035
1036                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1037
1038                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1039                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1040                                         }
1041                                 }
1042
1043                                 break;
1044                         case self::CONNECTORS:
1045                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1046                                 $headline .= DI::l10n()->t('<a href="%1$s" target="_blank" rel="noopener noreferrer">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1047                                 $headline .= ':</b></p>' . "\n";
1048
1049                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1050
1051                                 break;
1052                         case self::OSTATUS:
1053                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1054                                 break;
1055                         case self::ACTIVITYPUB:
1056                                 $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>';
1057                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1058                                 break;
1059                         default:
1060                                 $text = ($is_quote_share? "\n" : '');
1061
1062                                 $contact = Contact::getByURL($attributes['profile'], false, ['network']);
1063                                 $network = $contact['network'] ?? Protocol::PHANTOM;
1064
1065                                 $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1066                                 $text .= Renderer::replaceMacros($tpl, [
1067                                         '$profile'      => $attributes['profile'],
1068                                         '$avatar'       => $attributes['avatar'],
1069                                         '$author'       => $attributes['author'],
1070                                         '$link'         => $attributes['link'],
1071                                         '$link_title'   => DI::l10n()->t('link to source'),
1072                                         '$posted'       => $attributes['posted'],
1073                                         '$guid'         => $attributes['guid'],
1074                                         '$network_name' => ContactSelector::networkToName($network, $attributes['profile']),
1075                                         '$network_icon' => ContactSelector::networkToIcon($network, $attributes['profile']),
1076                                         '$content'      => self::setMentions(trim($content), 0, $network),
1077                                 ]);
1078                                 break;
1079                 }
1080
1081                 return $text;
1082         }
1083
1084         private static function removePictureLinksCallback($match)
1085         {
1086                 $cache_key = 'remove:' . $match[1];
1087                 $text = DI::cache()->get($cache_key);
1088
1089                 if (is_null($text)) {
1090                         $a = DI::app();
1091
1092                         $stamp1 = microtime(true);
1093
1094                         $ch = @curl_init($match[1]);
1095                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1096                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1097                         @curl_setopt($ch, CURLOPT_USERAGENT, DI::httpRequest()->getUserAgent());
1098                         @curl_exec($ch);
1099                         $curl_info = @curl_getinfo($ch);
1100
1101                         DI::profiler()->saveTimestamp($stamp1, "network");
1102
1103                         if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1104                                 $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]";
1105                         } else {
1106                                 $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]";
1107
1108                                 // if its not a picture then look if its a page that contains a picture link
1109                                 $body = DI::httpRequest()->fetch($match[1]);
1110
1111                                 $doc = new DOMDocument();
1112                                 @$doc->loadHTML($body);
1113                                 $xpath = new DOMXPath($doc);
1114                                 $list = $xpath->query("//meta[@name]");
1115                                 foreach ($list as $node) {
1116                                         $attr = [];
1117
1118                                         if ($node->attributes->length) {
1119                                                 foreach ($node->attributes as $attribute) {
1120                                                         $attr[$attribute->name] = $attribute->value;
1121                                                 }
1122                                         }
1123
1124                                         if (strtolower($attr['name']) == 'twitter:image') {
1125                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1126                                         }
1127                                 }
1128                         }
1129                         DI::cache()->set($cache_key, $text);
1130                 }
1131
1132                 return $text;
1133         }
1134
1135         private static function expandLinksCallback($match)
1136         {
1137                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1138                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1139                 } else {
1140                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1141                 }
1142         }
1143
1144         private static function cleanPictureLinksCallback($match)
1145         {
1146                 $a = DI::app();
1147
1148                 // When the picture link is the own photo path then we can avoid fetching the link
1149                 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()->get()) . '/photos/');
1150                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1151                         if (!empty($match[3])) {
1152                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1153                         } else {
1154                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1155                         }
1156                         return $text;
1157                 }
1158
1159                 $cache_key = 'clean:' . $match[1];
1160                 $text = DI::cache()->get($cache_key);
1161                 if (!is_null($text)) {
1162                         return $text;
1163                 }
1164
1165                 // Only fetch the header, not the content
1166                 $stamp1 = microtime(true);
1167
1168                 $ch = @curl_init($match[1]);
1169                 @curl_setopt($ch, CURLOPT_NOBODY, true);
1170                 @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1171                 @curl_setopt($ch, CURLOPT_USERAGENT, DI::httpRequest()->getUserAgent());
1172                 @curl_exec($ch);
1173                 $curl_info = @curl_getinfo($ch);
1174
1175                 DI::profiler()->saveTimestamp($stamp1, "network");
1176
1177                 // if its a link to a picture then embed this picture
1178                 if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1179                         $text = '[img]' . $match[1] . '[/img]';
1180                 } else {
1181                         if (!empty($match[3])) {
1182                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1183                         } else {
1184                                 $text = '[img]' . $match[2] . '[/img]';
1185                         }
1186
1187                         // if its not a picture then look if its a page that contains a picture link
1188                         $body = DI::httpRequest()->fetch($match[1]);
1189
1190                         $doc = new DOMDocument();
1191                         @$doc->loadHTML($body);
1192                         $xpath = new DOMXPath($doc);
1193                         $list = $xpath->query("//meta[@name]");
1194                         foreach ($list as $node) {
1195                                 $attr = [];
1196                                 if ($node->attributes->length) {
1197                                         foreach ($node->attributes as $attribute) {
1198                                                 $attr[$attribute->name] = $attribute->value;
1199                                         }
1200                                 }
1201
1202                                 if (strtolower($attr['name']) == "twitter:image") {
1203                                         if (!empty($match[3])) {
1204                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1205                                         } else {
1206                                                 $text = "[img]" . $attr['content'] . "[/img]";
1207                                         }
1208                                 }
1209                         }
1210                 }
1211                 DI::cache()->set($cache_key, $text);
1212
1213                 return $text;
1214         }
1215
1216         public static function cleanPictureLinks($text)
1217         {
1218                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1219                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1220                 return $return;
1221         }
1222
1223         public static function removeLinks(string $bbcode)
1224         {
1225                 $bbcode = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $1 ', $bbcode);
1226                 $bbcode = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $bbcode);
1227
1228                 $bbcode = preg_replace('/[@!#]\[url\=.*?\].*?\[\/url\]/ism', '', $bbcode);
1229                 $bbcode = preg_replace("/\[url=[^\[\]]*\](.*)\[\/url\]/Usi", ' $1 ', $bbcode);
1230                 $bbcode = preg_replace('/[@!#]?\[url.*?\[\/url\]/ism', '', $bbcode);
1231                 return $bbcode;
1232         }
1233
1234         /**
1235          * Converts a BBCode message to HTML message
1236          *
1237          * BBcode 2 HTML was written by WAY2WEB.net
1238          * extended to work with Mistpark/Friendica - Mike Macgirvin
1239          *
1240          * Simple HTML values meaning:
1241          * - 0: Friendica display
1242          * - 1: Unused
1243          * - 2: Used for Windows Phone push, Friendica API
1244          * - 3: Used before converting to Markdown in bb2diaspora.php
1245          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1246          * - 5: Unused
1247          * - 6: Unused
1248          * - 7: Used for dfrn, OStatus
1249          * - 8: Used for Twitter, WP backlink text setting
1250          * - 9: ActivityPub
1251          *
1252          * @param string $text
1253          * @param bool   $try_oembed
1254          * @param int    $simple_html
1255          * @param bool   $for_plaintext
1256          * @return string
1257          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1258          */
1259         public static function convert(string $text = null, $try_oembed = true, $simple_html = self::INTERNAL, $for_plaintext = false)
1260         {
1261                 // Accounting for null default column values
1262                 if (is_null($text) || $text === '') {
1263                         return '';
1264                 }
1265
1266                 $a = DI::app();
1267
1268                 $text = self::performWithEscapedTags($text, ['code'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) {
1269                         $text = self::performWithEscapedTags($text, ['noparse', 'nobb', 'pre'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) {
1270                                 /*
1271                                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1272                                  *
1273                                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1274                                  * $match[1] = $url
1275                                  * $match[2] = $title or absent
1276                                  */
1277                                 $try_oembed_callback = function ($match)
1278                                 {
1279                                         $url = $match[1];
1280                                         $title = $match[2] ?? null;
1281
1282                                         try {
1283                                                 $return = OEmbed::getHTML($url, $title);
1284                                         } catch (Exception $ex) {
1285                                                 $return = $match[0];
1286                                         }
1287
1288                                         return $return;
1289                                 };
1290
1291
1292
1293                                 // Remove the abstract element. It is a non visible element.
1294                                 $text = self::stripAbstract($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                                 // Convert new line chars to html <br /> tags
1330
1331                                 // nlbr seems to be hopelessly messed up
1332                                 //      $Text = nl2br($Text);
1333
1334                                 // We'll emulate it.
1335
1336                                 $text = trim($text);
1337                                 $text = str_replace("\r\n", "\n", $text);
1338
1339                                 // Remove linefeeds inside of the table elements. See issue #6799
1340                                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1341                                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1342                                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1343                                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1344                                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1345                                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1346                                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1347                                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1348                                 do {
1349                                         $oldtext = $text;
1350                                         $text = str_replace($search, $replace, $text);
1351                                 } while ($oldtext != $text);
1352
1353                                 // Replace these here only once
1354                                 $search = ["\n[table]", "[/table]\n"];
1355                                 $replace = ["[table]", "[/table]"];
1356                                 $text = str_replace($search, $replace, $text);
1357
1358                                 // removing multiplicated newlines
1359                                 if (DI::config()->get('system', 'remove_multiplicated_lines')) {
1360                                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1361                                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1362                                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1363                                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1364                                         do {
1365                                                 $oldtext = $text;
1366                                                 $text = str_replace($search, $replace, $text);
1367                                         } while ($oldtext != $text);
1368                                 }
1369
1370                                 /// @todo Have a closer look at the different html modes
1371                                 // Handle attached links or videos
1372                                 if ($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, $try_oembed);
1378                                 }
1379
1380                                 // leave open the posibility of [map=something]
1381                                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1382                                 if (strpos($text, '[/map]') !== false) {
1383                                         $text = preg_replace_callback(
1384                                                 "/\[map\](.*?)\[\/map\]/ism",
1385                                                 function ($match) use ($simple_html) {
1386                                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1387                                                 },
1388                                                 $text
1389                                         );
1390                                 }
1391
1392                                 if (strpos($text, '[map=') !== false) {
1393                                         $text = preg_replace_callback(
1394                                                 "/\[map=(.*?)\]/ism",
1395                                                 function ($match) use ($simple_html) {
1396                                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1397                                                 },
1398                                                 $text
1399                                         );
1400                                 }
1401
1402                                 if (strpos($text, '[map]') !== false) {
1403                                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1404                                 }
1405
1406                                 // Check for headers
1407                                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1408                                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1409                                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1410                                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1411                                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1412                                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1413
1414                                 // Check for paragraph
1415                                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1416
1417                                 // Check for bold text
1418                                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1419
1420                                 // Check for Italics text
1421                                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1422
1423                                 // Check for Underline text
1424                                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1425
1426                                 // Check for strike-through text
1427                                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1428
1429                                 // Check for over-line text
1430                                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1431
1432                                 // Check for colored text
1433                                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1434
1435                                 // Check for sized text
1436                                 // [size=50] --> font-size: 50px (with the unit).
1437                                 if ($simple_html != self::DIASPORA) {
1438                                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1439                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1440                                 } else {
1441                                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1442                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1443                                 }
1444
1445
1446                                 // Check for centered text
1447                                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1448
1449                                 // Check for list text
1450                                 $text = str_replace("[*]", "<li>", $text);
1451
1452                                 // Check for style sheet commands
1453                                 $text = preg_replace_callback(
1454                                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1455                                         function ($match) {
1456                                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1457                                         },
1458                                         $text
1459                                 );
1460
1461                                 // Check for CSS classes
1462                                 $text = preg_replace_callback(
1463                                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1464                                         function ($match) {
1465                                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1466                                         },
1467                                         $text
1468                                 );
1469
1470                                 // handle nested lists
1471                                 $endlessloop = 0;
1472
1473                                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1474                                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1475                                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1476                                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1477                                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1478                                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1479                                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1480                                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1481                                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1482                                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1483                                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1484                                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1485                                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1486                                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1487                                 }
1488
1489                                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1490                                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1491                                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1492                                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1493
1494                                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1495                                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1496
1497                                 $text = str_replace('[hr]', '<hr />', $text);
1498
1499                                 if (!$for_plaintext) {
1500                                         $escaped = [];
1501
1502                                         // Escaping BBCodes susceptible to contain rogue URL we don'' want the autolinker to catch
1503                                         $text = preg_replace_callback('#\[(url|img|audio|video|youtube|vimeo|share|attachment|iframe|bookmark).+?\[/\1\]#ism',
1504                                                 function ($matches) use (&$escaped) {
1505                                                         $return = '{escaped-' . count($escaped) . '}';
1506                                                         $escaped[] = $matches[0];
1507
1508                                                         return $return;
1509                                                 },
1510                                                 $text
1511                                         );
1512
1513                                         // Autolinker for isolated URLs
1514                                         $text = preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1515
1516                                         // Restoring escaped blocks
1517                                         $text = preg_replace_callback('/{escaped-([0-9]+)}/iU',
1518                                                 function ($matches) use ($escaped) {
1519                                                         return $escaped[intval($matches[1])] ?? $matches[0];
1520                                                 },
1521                                                 $text
1522                                         );
1523                                 }
1524
1525                                 // This is actually executed in Item::prepareBody()
1526
1527                                 $nosmile = strpos($text, '[nosmile]') !== false;
1528                                 $text = str_replace('[nosmile]', '', $text);
1529
1530                                 // Check for font change text
1531                                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1532
1533                                 // Declare the format for [spoiler] layout
1534                                 $SpoilerLayout = '<details class="spoiler"><summary>' . DI::l10n()->t('Click to open/close') . '</summary>$1</details>';
1535
1536                                 // Check for [spoiler] text
1537                                 // handle nested quotes
1538                                 $endlessloop = 0;
1539                                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1540                                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1541                                 }
1542
1543                                 // Check for [spoiler=Title] text
1544
1545                                 // handle nested quotes
1546                                 $endlessloop = 0;
1547                                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1548                                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1549                                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1550                                                 $text);
1551                                 }
1552
1553                                 // Declare the format for [quote] layout
1554                                 $QuoteLayout = '<blockquote>$1</blockquote>';
1555
1556                                 // Check for [quote] text
1557                                 // handle nested quotes
1558                                 $endlessloop = 0;
1559                                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1560                                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1561                                 }
1562
1563                                 // Check for [quote=Author] text
1564
1565                                 $t_wrote = DI::l10n()->t('$1 wrote:');
1566
1567                                 // handle nested quotes
1568                                 $endlessloop = 0;
1569                                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1570                                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1571                                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1572                                                 $text);
1573                                 }
1574
1575
1576                                 // [img=widthxheight]image source[/img]
1577                                 $text = preg_replace_callback(
1578                                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1579                                         function ($matches) use ($simple_html) {
1580                                                 if (strpos($matches[3], "data:image/") === 0) {
1581                                                         return $matches[0];
1582                                                 }
1583
1584                                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1585                                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1586                                         },
1587                                         $text
1588                                 );
1589
1590                                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1591                                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1592
1593                                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1594                                         function ($matches) use ($simple_html) {
1595                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1596                                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1597                                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1598                                         },
1599                                         $text);
1600
1601                                 // Images
1602                                 // [img]pathtoimage[/img]
1603                                 $text = preg_replace_callback(
1604                                         "/\[img\](.*?)\[\/img\]/ism",
1605                                         function ($matches) use ($simple_html) {
1606                                                 if (strpos($matches[1], "data:image/") === 0) {
1607                                                         return $matches[0];
1608                                                 }
1609
1610                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1611                                                 return "[img]" . $matches[1] . "[/img]";
1612                                         },
1613                                         $text
1614                                 );
1615
1616                                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1617                                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1618
1619                                 $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);
1620                                 $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 />', $text);
1621                                 //$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);
1622
1623                                 // Simplify "video" element
1624                                 $text = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1625
1626                                 // Try to Oembed
1627                                 if ($try_oembed) {
1628                                         $text = preg_replace("/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4).*?)\[\/video\]/ism", '<video src="$1" controls="controls" width="' . $a->videowidth . '" height="' . $a->videoheight . '" loop="true"><a href="$1">$1</a></video>', $text);
1629                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '<audio src="$1" controls="controls"><a href="$1">$1</a></audio>', $text);
1630
1631                                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1632                                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1633                                 } else {
1634                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1635                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1636                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1637                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1638                                 }
1639
1640                                 // html5 video and audio
1641
1642
1643                                 if ($try_oembed) {
1644                                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1645                                 } else {
1646                                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1647                                 }
1648
1649                                 // Youtube extensions
1650                                 if ($try_oembed) {
1651                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1652                                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1653                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1654                                 }
1655
1656                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1657                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1658                                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1659
1660                                 if ($try_oembed) {
1661                                         $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);
1662                                 } else {
1663                                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1664                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank" rel="noopener noreferrer">https://www.youtube.com/watch?v=$1</a>', $text);
1665                                 }
1666
1667                                 if ($try_oembed) {
1668                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1669                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1670                                 }
1671
1672                                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1673                                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1674
1675                                 if ($try_oembed) {
1676                                         $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);
1677                                 } else {
1678                                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1679                                                 '<a href="https://vimeo.com/$1" target="_blank" rel="noopener noreferrer">https://vimeo.com/$1</a>', $text);
1680                                 }
1681
1682                                 // oembed tag
1683                                 $text = OEmbed::BBCode2HTML($text);
1684
1685                                 // Avoid triple linefeeds through oembed
1686                                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1687
1688                                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1689                                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1690                                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1691                                 // start which is always required). Allow desc with a missing summary for compatibility.
1692
1693                                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1694                                         $sub = Event::getHTML($ev, $simple_html);
1695
1696                                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1697                                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1698                                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1699                                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1700                                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1701                                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1702                                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1703                                 }
1704
1705                                 // Replace non graphical smilies for external posts
1706                                 if (!$nosmile && !$for_plaintext) {
1707                                         $text = Smilies::replace($text);
1708                                 }
1709
1710                                 if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA)) {
1711                                         $conv = html_entity_decode(str_replace([' ', "\n", "\r"], '', $text));
1712                                         // Emojis are always 4 byte Unicode characters
1713                                         if (!empty($conv) && (strlen($conv) / mb_strlen($conv) == 4)) {
1714                                                 $text = '<span style="font-size: xx-large; line-height: initial;">' . $text . '</span>';
1715                                         }
1716                                 }
1717
1718                                 if (!$for_plaintext) {
1719                                         if (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1720                                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1721                                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1722                                         }
1723                                 } else {
1724                                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1725                                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1726                                 }
1727
1728                                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1729
1730                                 // Remove all hashtag addresses
1731                                 if ($simple_html && !in_array($simple_html, [self::DIASPORA, self::OSTATUS, self::ACTIVITYPUB])) {
1732                                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1733                                 } elseif ($simple_html == self::DIASPORA) {
1734                                         // The ! is converted to @ since Diaspora only understands the @
1735                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1736                                                 '@<a href="$2">$3</a>',
1737                                                 $text);
1738                                 } elseif (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1739                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1740                                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1741                                                 $text);
1742                                 } elseif (!$simple_html) {
1743                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1744                                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1745                                                 $text);
1746                                 }
1747
1748                                 // Bookmarks in red - will be converted to bookmarks in friendica
1749                                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1750                                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1751                                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1752                                                         "[bookmark=$1]$2[/bookmark]", $text);
1753
1754                                 if (in_array($simple_html, [self::API, self::OSTATUS, self::TWITTER])) {
1755                                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1756                                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1757                                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1758                                 }
1759
1760                                 // Perform URL Search
1761                                 if ($try_oembed) {
1762                                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1763                                 }
1764
1765                                 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1766
1767                                 // Handle Diaspora posts
1768                                 $text = preg_replace_callback(
1769                                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1770                                         function ($match) {
1771                                                 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1772                                         }, $text
1773                                 );
1774
1775                                 $text = preg_replace_callback(
1776                                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1777                                         function ($match) {
1778                                                 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1779                                         }, $text
1780                                 );
1781
1782                                 // Server independent link to posts and comments
1783                                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1784                                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1785                                 $text = preg_replace($expression, DI::baseUrl()."/display/$1", $text);
1786
1787                                 /* Tag conversion
1788                                  * Supports:
1789                                  * - #[url=<anything>]<term>[/url]
1790                                  * - [url=<anything>]#<term>[/url]
1791                                  */
1792                                 $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) use ($simple_html) {
1793                                         if ($simple_html == BBCode::ACTIVITYPUB) {
1794                                                 return '<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1795                                                         . '" data-tag="' . XML::escape($matches[1]) . '" rel="tag ugc">#'
1796                                                         . XML::escape($matches[1]) . '</a>';
1797                                         } else {
1798                                                 return '#<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1799                                                         . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1800                                                         . XML::escape($matches[1]) . '</a>';
1801                                         }
1802                                 }, $text);
1803
1804                                 // We need no target="_blank" rel="noopener noreferrer" for local links
1805                                 // convert links start with DI::baseUrl() as local link without the target="_blank" rel="noopener noreferrer" attribute
1806                                 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
1807                                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1808                                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1809
1810                                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1811                                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1812
1813                                 // Red compatibility, though the link can't be authenticated on Friendica
1814                                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1815
1816
1817                                 // we may need to restrict this further if it picks up too many strays
1818                                 // link acct:user@host to a webfinger profile redirector
1819
1820                                 $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);
1821
1822                                 // Perform MAIL Search
1823                                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1824                                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1825
1826                                 /// @todo What is the meaning of these lines?
1827                                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1828                                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1829
1830                                 // Currently deactivated, it made problems with " inside of alt texts.
1831                                 //$text = preg_replace('/\&quot\;/', '"', $text);
1832
1833                                 // fix any escaped ampersands that may have been converted into links
1834                                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1835
1836                                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1837                                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1838
1839                                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1840
1841                                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1842                                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">', $text);
1843
1844                                 // sanitize href attributes (only allowlisted protocols URLs)
1845                                 // default value for backward compatibility
1846                                 $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
1847
1848                                 // Always allowed protocol even if config isn't set or not including it
1849                                 $allowed_link_protocols[] = '//';
1850                                 $allowed_link_protocols[] = 'http://';
1851                                 $allowed_link_protocols[] = 'https://';
1852                                 $allowed_link_protocols[] = 'redir/';
1853
1854                                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1855
1856                                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1857                                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text);
1858
1859                                 // Shared content
1860                                 $text = self::convertShare(
1861                                         $text,
1862                                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1863                                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1864                                         }
1865                                 );
1866
1867                                 $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1868
1869                                 return $text;
1870                         }); // Escaped noparse, nobb, pre
1871
1872                         // Remove escaping tags
1873                         $text = preg_replace("/\[noparse\](.*?)\[\/noparse\]/ism", '\1', $text);
1874                         $text = preg_replace("/\[nobb\](.*?)\[\/nobb\]/ism", '\1', $text);
1875
1876                         // Additionally, [pre] tags preserve spaces
1877                         $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", function ($match) {
1878                                 return str_replace(' ', '&nbsp;', $match[1]);
1879                         }, $text);
1880
1881                         return $text;
1882                 }); // Escaped code
1883
1884                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1885                         function ($matches) {
1886                                 if (strpos($matches[2], "\n") !== false) {
1887                                         $return = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlspecialchars(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
1888                                 } else {
1889                                         $return = '<code>' . htmlspecialchars($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
1890                                 }
1891
1892                                 return $return;
1893                         },
1894                         $text
1895                 );
1896
1897                 // Clean up the HTML by loading and saving the HTML with the DOM.
1898                 // Bad structured html can break a whole page.
1899                 // For performance reasons do it only with activated item cache or at export.
1900                 if (!$try_oembed || (get_itemcachepath() != '')) {
1901                         $doc = new DOMDocument();
1902                         $doc->preserveWhiteSpace = false;
1903
1904                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1905
1906                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1907                         $encoding = '<?xml encoding="UTF-8">';
1908                         @$doc->loadHTML($encoding . $doctype . '<html><body>' . $text . '</body></html>');
1909                         $doc->encoding = 'UTF-8';
1910                         $text = $doc->saveHTML();
1911                         $text = str_replace(['<html><body>', '</body></html>', $doctype, $encoding], ['', '', '', ''], $text);
1912
1913                         $text = str_replace('<br></li>', '</li>', $text);
1914
1915                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1916                 }
1917
1918                 // Clean up some useless linebreaks in lists
1919                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1920                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1921                 //$Text = str_replace('</li><br />', '</li>', $Text);
1922                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1923                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1924
1925                 Hook::callAll('bbcode', $text);
1926
1927                 return trim($text);
1928         }
1929
1930         /**
1931          * Strips the "abstract" tag from the provided text
1932          *
1933          * @param string $text The text with BBCode
1934          * @return string The same text - but without "abstract" element
1935          */
1936         public static function stripAbstract($text)
1937         {
1938                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1939                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1940
1941                 return $text;
1942         }
1943
1944         /**
1945          * Returns the value of the "abstract" element
1946          *
1947          * @param string $text The text that maybe contains the element
1948          * @param string $addon The addon for which the abstract is meant for
1949          * @return string The abstract
1950          */
1951         public static function getAbstract($text, $addon = '')
1952         {
1953                 $abstract = '';
1954                 $abstracts = [];
1955                 $addon = strtolower($addon);
1956
1957                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1958                         foreach ($results AS $result) {
1959                                 $abstracts[strtolower($result[1])] = $result[2];
1960                         }
1961                 }
1962
1963                 if (isset($abstracts[$addon])) {
1964                         $abstract = $abstracts[$addon];
1965                 }
1966
1967                 if ($abstract == '' && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1968                         $abstract = $result[1];
1969                 }
1970
1971                 return $abstract;
1972         }
1973
1974         /**
1975          * Callback function to replace a Friendica style mention in a mention for Diaspora
1976          *
1977          * @param array $match Matching values for the callback
1978          *                     [1] = Mention type (! or @)
1979          *                     [2] = Name
1980          *                     [3] = Address
1981          * @return string Replaced mention
1982          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1983          * @throws \ImagickException
1984          */
1985         private static function bbCodeMention2DiasporaCallback($match)
1986         {
1987                 $contact = Contact::getByURL($match[3], false, ['addr']);
1988                 if (empty($contact['addr'])) {
1989                         return $match[0];
1990                 }
1991
1992                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
1993                 return $mention;
1994         }
1995
1996         /**
1997          * Converts a BBCode text into Markdown
1998          *
1999          * This function converts a BBCode item body to be sent to Markdown-enabled
2000          * systems like Diaspora and Libertree
2001          *
2002          * @param string $text
2003          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
2004          * @return string
2005          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2006          */
2007         public static function toMarkdown($text, $for_diaspora = true)
2008         {
2009                 $original_text = $text;
2010
2011                 // Since Diaspora is creating a summary for links, this function removes them before posting
2012                 if ($for_diaspora) {
2013                         $text = self::removeShareInformation($text);
2014                 }
2015
2016                 /**
2017                  * Transform #tags, strip off the [url] and replace spaces with underscore
2018                  */
2019                 $url_search_string = "^\[\]";
2020                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2021                         function ($matches) {
2022                                 return '#' . str_replace(' ', '_', $matches[2]);
2023                         },
2024                         $text
2025                 );
2026
2027                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2028                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2029
2030                 // Convert it to HTML - don't try oembed
2031                 if ($for_diaspora) {
2032                         $text = self::convert($text, false, self::DIASPORA);
2033
2034                         // Add all tags that maybe were removed
2035                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2036                                 $tagline = '';
2037                                 foreach ($tags[2] as $tag) {
2038                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2039                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2040                                                 $tagline .= '#' . $tag . ' ';
2041                                         }
2042                                 }
2043                                 $text = $text . " " . $tagline;
2044                         }
2045                 } else {
2046                         $text = self::convert($text, false, self::CONNECTORS);
2047                 }
2048
2049                 // If a link is followed by a quote then there should be a newline before it
2050                 // Maybe we should make this newline at every time before a quote.
2051                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2052
2053                 $stamp1 = microtime(true);
2054
2055                 // Now convert HTML to Markdown
2056                 $text = HTML::toMarkdown($text);
2057
2058                 DI::profiler()->saveTimestamp($stamp1, "parser");
2059
2060                 // Libertree has a problem with escaped hashtags.
2061                 $text = str_replace(['\#'], ['#'], $text);
2062
2063                 // Remove any leading or trailing whitespace, as this will mess up
2064                 // the Diaspora signature verification and cause the item to disappear
2065                 $text = trim($text);
2066
2067                 if ($for_diaspora) {
2068                         $url_search_string = "^\[\]";
2069                         $text = preg_replace_callback(
2070                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2071                                 ['self', 'bbCodeMention2DiasporaCallback'],
2072                                 $text
2073                         );
2074                 }
2075
2076                 Hook::callAll('bb2diaspora', $text);
2077
2078                 return $text;
2079         }
2080
2081         /**
2082          * Pull out all #hashtags and @person tags from $string.
2083          *
2084          * We also get @person@domain.com - which would make
2085          * the regex quite complicated as tags can also
2086          * end a sentence. So we'll run through our results
2087          * and strip the period from any tags which end with one.
2088          * Returns array of tags found, or empty array.
2089          *
2090          * @param string $string Post content
2091          *
2092          * @return array List of tag and person names
2093          */
2094         public static function getTags($string)
2095         {
2096                 $ret = [];
2097
2098                 BBCode::performWithEscapedTags($string, ['noparse', 'pre', 'code'], function ($string) use (&$ret) {
2099                         // Convert hashtag links to hashtags
2100                         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2 ', $string);
2101
2102                         // Force line feeds at bbtags
2103                         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2104
2105                         // ignore anything in a bbtag
2106                         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2107
2108                         // Match full names against @tags including the space between first and last
2109                         // We will look these up afterward to see if they are full names or not recognisable.
2110
2111                         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2112                                 foreach ($matches[1] as $match) {
2113                                         if (strstr($match, ']')) {
2114                                                 // we might be inside a bbcode color tag - leave it alone
2115                                                 continue;
2116                                         }
2117
2118                                         if (substr($match, -1, 1) === '.') {
2119                                                 $ret[] = substr($match, 0, -1);
2120                                         } else {
2121                                                 $ret[] = $match;
2122                                         }
2123                                 }
2124                         }
2125
2126                         // Otherwise pull out single word tags. These can be @nickname, @first_last
2127                         // and #hash tags.
2128
2129                         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?\']*[^\^ \x0D\x0A,;:?!\'.])/', $string, $matches)) {
2130                                 foreach ($matches[1] as $match) {
2131                                         if (strstr($match, ']')) {
2132                                                 // we might be inside a bbcode color tag - leave it alone
2133                                                 continue;
2134                                         }
2135
2136                                         // ignore strictly numeric tags like #1
2137                                         if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
2138                                                 continue;
2139                                         }
2140
2141                                         // try not to catch url fragments
2142                                         if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2143                                                 continue;
2144                                         }
2145
2146                                         $ret[] = $match;
2147                                 }
2148                         }
2149                 });
2150
2151                 return array_unique($ret);
2152         }
2153
2154         /**
2155          * Perform a custom function on a text after having escaped blocks enclosed in the provided tag list.
2156          *
2157          * @param string   $text
2158          * @param array    $tagList A list of tag names, e.g ['noparse', 'nobb', 'pre']
2159          * @param callable $callback
2160          * @return string
2161          * @throws Exception
2162          *@see Strings::performWithEscapedBlocks
2163          *
2164          */
2165         public static function performWithEscapedTags(string $text, array $tagList, callable $callback)
2166         {
2167                 $tagList = array_map('preg_quote', $tagList);
2168
2169                 return Strings::performWithEscapedBlocks($text, '#\[(?:' . implode('|', $tagList) . ').*?\[/(?:' . implode('|', $tagList) . ')]#ism', $callback);
2170         }
2171
2172         /**
2173          * Replaces mentions in the provided message body for the provided user and network if any
2174          *
2175          * @param $body
2176          * @param $profile_uid
2177          * @param $network
2178          * @return string
2179          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2180          * @throws \ImagickException
2181          */
2182         public static function setMentions($body, $profile_uid = 0, $network = '')
2183         {
2184                 BBCode::performWithEscapedTags($body, ['noparse', 'pre', 'code', 'img'], function ($body) use ($profile_uid, $network) {
2185                         $tags = BBCode::getTags($body);
2186
2187                         $tagged = [];
2188                         $inform = '';
2189
2190                         foreach ($tags as $tag) {
2191                                 $tag_type = substr($tag, 0, 1);
2192
2193                                 if ($tag_type == Tag::TAG_CHARACTER[Tag::HASHTAG]) {
2194                                         continue;
2195                                 }
2196
2197                                 /*
2198                                  * If we already tagged 'Robert Johnson', don't try and tag 'Robert'.
2199                                  * Robert Johnson should be first in the $tags array
2200                                  */
2201                                 foreach ($tagged as $nextTag) {
2202                                         if (stristr($nextTag, $tag . ' ')) {
2203                                                 continue 2;
2204                                         }
2205                                 }
2206
2207                                 $success = Item::replaceTag($body, $inform, $profile_uid, $tag, $network);
2208
2209                                 if ($success['replaced']) {
2210                                         $tagged[] = $tag;
2211                                 }
2212                         }
2213
2214                         return $body;
2215                 });
2216
2217                 return $body;
2218         }
2219
2220         /**
2221          * @param string      $author  Author display name
2222          * @param string      $profile Author profile URL
2223          * @param string      $avatar  Author profile picture URL
2224          * @param string      $link    Post source URL
2225          * @param string      $posted  Post created date
2226          * @param string|null $guid    Post guid (if any)
2227          * @return string
2228          * @TODO Rewrite to handle over whole record array
2229          */
2230         public static function getShareOpeningTag(string $author, string $profile, string $avatar, string $link, string $posted, string $guid = null)
2231         {
2232                 $header = "[share author='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $author) .
2233                         "' profile='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $profile) .
2234                         "' avatar='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $avatar) .
2235                         "' link='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $link) .
2236                         "' posted='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $posted);
2237
2238                 if ($guid) {
2239                         $header .= "' guid='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $guid);
2240                 }
2241
2242                 $header  .= "']";
2243
2244                 return $header;
2245         }
2246 }