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\OEmbed;
  28 use Friendica\Content\Smilies;
  29 use Friendica\Core\Hook;
  30 use Friendica\Core\Logger;
  31 use Friendica\Core\Protocol;
  32 use Friendica\Core\Renderer;
  33 use Friendica\Core\System;
  34 use Friendica\DI;
  35 use Friendica\Model\Contact;
  36 use Friendica\Model\Event;
  37 use Friendica\Model\Photo;
  38 use Friendica\Network\Probe;
  39 use Friendica\Object\Image;
  40 use Friendica\Protocol\Activity;
  41 use Friendica\Util\Images;
  42 use Friendica\Util\Map;
  43 use Friendica\Util\Network;
  44 use Friendica\Util\ParseUrl;
  45 use Friendica\Util\Proxy as ProxyUtils;
  46 use Friendica\Util\Strings;
  47 use Friendica\Util\XML;
  48
  49 class BBCode
  50 {
  51         const INTERNAL = 0;
  52         const API = 2;
  53         const DIASPORA = 3;
  54         const CONNECTORS = 4;
  55         const OSTATUS = 7;
  56         const TWITTER = 8;
  57         const BACKLINK = 8;
  58         const ACTIVITYPUB = 9;
  59
  60         /**
  61          * Fetches attachment data that were generated the old way
  62          *
  63          * @param string $body Message body
  64          * @return array
  65          *                     'type' -> Message type ('link', 'video', 'photo')
  66          *                     'text' -> Text before the shared message
  67          *                     'after' -> Text after the shared message
  68          *                     'image' -> Preview image of the message
  69          *                     'url' -> Url to the attached message
  70          *                     'title' -> Title of the attachment
  71          *                     'description' -> Description of the attachment
  72          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  73          */
  74         private static function getOldAttachmentData($body)
  75         {
  76                 $post = [];
  77
  78                 // Simplify image codes
  79                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  80
  81                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  82                         foreach ($attached as $data) {
  83                                 if (!in_array($data[1], ['type-link', 'type-video', 'type-photo'])) {
  84                                         continue;
  85                                 }
  86
  87                                 $post['type'] = substr($data[1], 5);
  88
  89                                 $pos = strpos($body, $data[0]);
  90                                 if ($pos > 0) {
  91                                         $post['text'] = trim(substr($body, 0, $pos));
  92                                         $post['after'] = trim(substr($body, $pos + strlen($data[0])));
  93                                 } else {
  94                                         $post['text'] = trim(str_replace($data[0], '', $body));
  95                                         $post['after'] = '';
  96                                 }
  97
  98                                 $attacheddata = $data[2];
  99
 100                                 if (preg_match("/\[img\](.*?)\[\/img\]/ism", $attacheddata, $matches)) {
 101
 102                                         $picturedata = Images::getInfoFromURLCached($matches[1]);
 103
 104                                         if ($picturedata) {
 105                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
 106                                                         $post['image'] = $matches[1];
 107                                                 } else {
 108                                                         $post['preview'] = $matches[1];
 109                                                 }
 110                                         }
 111                                 }
 112
 113                                 if (preg_match("/\[bookmark\=(.*?)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
 114                                         $post['url'] = $matches[1];
 115                                         $post['title'] = $matches[2];
 116                                 }
 117                                 if (!empty($post['url']) && (in_array($post['type'], ['link', 'video']))
 118                                         && preg_match("/\[url\=(.*?)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
 119                                         $post['url'] = $matches[1];
 120                                 }
 121
 122                                 // Search for description
 123                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
 124                                         $post['description'] = $matches[1];
 125                                 }
 126                         }
 127                 }
 128                 return $post;
 129         }
 130
 131         /**
 132          * Fetches attachment data that were generated with the "attachment" element
 133          *
 134          * @param string $body Message body
 135          * @return array
 136          *                     'type' -> Message type ('link', 'video', 'photo')
 137          *                     'text' -> Text before the shared message
 138          *                     'after' -> Text after the shared message
 139          *                     'image' -> Preview image of the message
 140          *                     'url' -> Url to the attached message
 141          *                     'title' -> Title of the attachment
 142          *                     'description' -> Description of the attachment
 143          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 144          */
 145         public static function getAttachmentData($body)
 146         {
 147                 $data = [
 148                         'type'        => '',
 149                         'text'        => '',
 150                         'after'       => '',
 151                         'image'       => null,
 152                         'url'         => '',
 153                         'title'       => '',
 154                         'description' => '',
 155                 ];
 156
 157                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 158                         return self::getOldAttachmentData($body);
 159                 }
 160
 161                 $attributes = $match[2];
 162
 163                 $data['text'] = trim($match[1]);
 164
 165                 $type = '';
 166                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 167                 if (!empty($matches[1])) {
 168                         $type = strtolower($matches[1]);
 169                 }
 170
 171                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 172                 if (!empty($matches[1])) {
 173                         $type = strtolower($matches[1]);
 174                 }
 175
 176                 if ($type == '') {
 177                         return [];
 178                 }
 179
 180                 if (!in_array($type, ['link', 'audio', 'photo', 'video'])) {
 181                         return [];
 182                 }
 183
 184                 if ($type != '') {
 185                         $data['type'] = $type;
 186                 }
 187
 188                 $url = '';
 189                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 190                 if (!empty($matches[1])) {
 191                         $url = $matches[1];
 192                 }
 193
 194                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 195                 if (!empty($matches[1])) {
 196                         $url = $matches[1];
 197                 }
 198
 199                 if ($url != '') {
 200                         $data['url'] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 201                 }
 202
 203                 $title = '';
 204                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 205                 if (!empty($matches[1])) {
 206                         $title = $matches[1];
 207                 }
 208
 209                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 210                 if (!empty($matches[1])) {
 211                         $title = $matches[1];
 212                 }
 213
 214                 if ($title != '') {
 215                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 216                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 217                         $title = str_replace(['[', ']'], ['&#91;', '&#93;'], $title);
 218                         $data['title'] = $title;
 219                 }
 220
 221                 $image = '';
 222                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 223                 if (!empty($matches[1])) {
 224                         $image = $matches[1];
 225                 }
 226
 227                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 228                 if (!empty($matches[1])) {
 229                         $image = $matches[1];
 230                 }
 231
 232                 if ($image != '') {
 233                         $data['image'] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 234                 }
 235
 236                 $preview = '';
 237                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 238                 if (!empty($matches[1])) {
 239                         $preview = $matches[1];
 240                 }
 241
 242                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 243                 if (!empty($matches[1])) {
 244                         $preview = $matches[1];
 245                 }
 246
 247                 if ($preview != '') {
 248                         $data['preview'] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 249                 }
 250
 251                 $data['description'] = trim($match[3]);
 252
 253                 $data['after'] = trim($match[4]);
 254
 255                 return $data;
 256         }
 257
 258         public static function getAttachedData($body, $item = [])
 259         {
 260                 /*
 261                 - text:
 262                 - type: link, video, photo
 263                 - title:
 264                 - url:
 265                 - image:
 266                 - description:
 267                 - (thumbnail)
 268                 */
 269
 270                 $has_title = !empty($item['title']);
 271                 $plink = $item['plink'] ?? '';
 272                 $post = self::getAttachmentData($body);
 273
 274                 // Get all linked images with alternative image description
 275                 if (preg_match_all("/\[img=([^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 276                         foreach ($pictures as $picture) {
 277                                 if (Photo::isLocal($picture[1])) {
 278                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]];
 279                                 }
 280                         }
 281                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 282                                 $post['image_description'] = $post['images'][0]['description'];
 283                         }
 284                 }
 285
 286                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 287                         foreach ($pictures as $picture) {
 288                                 if (Photo::isLocal($picture[1])) {
 289                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
 290                                 }
 291                         }
 292                 }
 293
 294                 // if nothing is found, it maybe having an image.
 295                 if (!isset($post['type'])) {
 296                         // Simplify image codes
 297                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 298                         $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 299                         $post['text'] = $body;
 300
 301                         if (preg_match_all("#\[url=([^\]]+?)\]\s*\[img\]([^\[]+?)\[/img\]\s*\[/url\]#ism", $body, $pictures, PREG_SET_ORDER)) {
 302                                 if ((count($pictures) == 1) && !$has_title) {
 303                                         if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) {
 304                                                 // Replace the preview picture with the real picture
 305                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
 306                                                 $data = ['url' => $url, 'type' => 'photo'];
 307                                         } else {
 308                                                 // Checking, if the link goes to a picture
 309                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 310                                         }
 311
 312                                         // Workaround:
 313                                         // Sometimes photo posts to the own album are not detected at the start.
 314                                         // So we seem to cannot use the cache for these cases. That's strange.
 315                                         if (($data['type'] != 'photo') && strstr($pictures[0][1], "/photos/")) {
 316                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 317                                         }
 318
 319                                         if ($data['type'] == 'photo') {
 320                                                 $post['type'] = 'photo';
 321                                                 if (isset($data['images'][0])) {
 322                                                         $post['image'] = $data['images'][0]['src'];
 323                                                         $post['url'] = $data['url'];
 324                                                 } else {
 325                                                         $post['image'] = $data['url'];
 326                                                 }
 327
 328                                                 $post['preview'] = $pictures[0][2];
 329                                                 $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 330                                         } else {
 331                                                 $imgdata = Images::getInfoFromURLCached($pictures[0][1]);
 332                                                 if ($imgdata && substr($imgdata['mime'], 0, 6) == 'image/') {
 333                                                         $post['type'] = 'photo';
 334                                                         $post['image'] = $pictures[0][1];
 335                                                         $post['preview'] = $pictures[0][2];
 336                                                         $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 337                                                 }
 338                                         }
 339                                 } elseif (count($pictures) > 0) {
 340                                         $post['type'] = 'link';
 341                                         $post['url'] = $plink;
 342                                         $post['image'] = $pictures[0][2];
 343                                         $post['text'] = $body;
 344
 345                                         foreach ($pictures as $picture) {
 346                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 347                                         }
 348                                 }
 349                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 350                                 if ((count($pictures) == 1) && !$has_title) {
 351                                         $post['type'] = 'photo';
 352                                         $post['image'] = $pictures[0][1];
 353                                         $post['text'] = str_replace($pictures[0][0], '', $body);
 354                                 } elseif (count($pictures) > 0) {
 355                                         $post['type'] = 'link';
 356                                         $post['url'] = $plink;
 357                                         $post['image'] = $pictures[0][1];
 358                                         $post['text'] = $body;
 359
 360                                         foreach ($pictures as $picture) {
 361                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 362                                         }
 363                                 }
 364                         }
 365
 366                         // Test for the external links
 367                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER);
 368                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER);
 369
 370                         $links = array_merge($links1, $links2);
 371
 372                         // If there is only a single one, then use it.
 373                         // This should cover link posts via API.
 374                         if ((count($links) == 1) && !isset($post['preview']) && !$has_title) {
 375                                 $post['type'] = 'link';
 376                                 $post['url'] = $links[0][1];
 377                         }
 378
 379                         // Simplify "video" element
 380                         $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']);
 381
 382                         // Now count the number of external media links
 383                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
 384                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
 385                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
 386                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
 387
 388                         // Add them to the other external links
 389                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 390
 391                         // Are there more than one?
 392                         if (count($links) > 1) {
 393                                 // The post will be the type "text", which means a blog post
 394                                 unset($post['type']);
 395                                 $post['url'] = $plink;
 396                         }
 397
 398                         if (!isset($post['type'])) {
 399                                 $post['type'] = "text";
 400                                 $post['text'] = trim($body);
 401                         }
 402                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
 403                         $data = ParseUrl::getSiteinfoCached($post['url'], true);
 404
 405                         if (isset($data['images'][0])) {
 406                                 $post['image'] = $data['images'][0]['src'];
 407                         }
 408                 }
 409
 410                 return $post;
 411         }
 412
 413         /**
 414          * Remove [attachment] BBCode and replaces it with a regular [url]
 415          *
 416          * @param string  $body
 417          * @param boolean $no_link_desc No link description
 418          *
 419          * @return string with replaced body
 420          */
 421         public static function removeAttachment($body, $no_link_desc = false)
 422         {
 423                 return preg_replace_callback("/\s*\[attachment (.*)\](.*?)\[\/attachment\]\s*/ism",
 424                         function ($match) use ($no_link_desc) {
 425                                 $attach_data = self::getAttachmentData($match[0]);
 426                                 if (empty($attach_data['url'])) {
 427                                         return $match[0];
 428                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
 429                                         return "\n[url]" . $attach_data['url'] . "[/url]\n";
 430                                 } else {
 431                                         return "\n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
 432                                 }
 433                 }, $body);
 434         }
 435
 436         /**
 437          * Converts a BBCode text into plaintext
 438          *
 439          * @param      $text
 440          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 441          *
 442          * @return string
 443          */
 444         public static function toPlaintext($text, $keep_urls = true)
 445         {
 446                 $naked_text = HTML::toPlaintext(self::convert($text, false, 0, true), 0, !$keep_urls);
 447
 448                 return $naked_text;
 449         }
 450
 451         private static function proxyUrl($image, $simplehtml = self::INTERNAL)
 452         {
 453                 // Only send proxied pictures to API and for internal display
 454                 if (in_array($simplehtml, [self::INTERNAL, self::API])) {
 455                         return ProxyUtils::proxifyUrl($image);
 456                 } else {
 457                         return $image;
 458                 }
 459         }
 460
 461         /**
 462          * This function changing the visual size (not the real size) of images.
 463          * The function does not work for pictures with an alternate text description.
 464          * This could only be changed by using some new "img" BBCode format.
 465          *
 466          * @param string $srctext The body with images
 467          * @return string The body with possibly scaled images
 468          */
 469         public static function scaleExternalImages(string $srctext)
 470         {
 471                 $s = $srctext;
 472
 473                 // Simplify image links
 474                 $s = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $s);
 475
 476                 $matches = null;
 477                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 478                 if ($c) {
 479                         foreach ($matches as $mtch) {
 480                                 Logger::log('scale_external_image: ' . $mtch[1]);
 481
 482                                 $hostname = str_replace('www.', '', substr(DI::baseUrl(), strpos(DI::baseUrl(), '://') + 3));
 483                                 if (stristr($mtch[1], $hostname)) {
 484                                         continue;
 485                                 }
 486
 487                                 $curlResult = Network::curl($mtch[1], true);
 488                                 if (!$curlResult->isSuccess()) {
 489                                         continue;
 490                                 }
 491
 492                                 $i = $curlResult->getBody();
 493                                 $type = $curlResult->getContentType();
 494                                 $type = Images::getMimeTypeByData($i, $mtch[1], $type);
 495
 496                                 if ($i) {
 497                                         $Image = new Image($i, $type);
 498                                         if ($Image->isValid()) {
 499                                                 $orig_width = $Image->getWidth();
 500                                                 $orig_height = $Image->getHeight();
 501
 502                                                 if ($orig_width > 640 || $orig_height > 640) {
 503                                                         $Image->scaleDown(640);
 504                                                         $new_width = $Image->getWidth();
 505                                                         $new_height = $Image->getHeight();
 506                                                         Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG);
 507                                                         $s = str_replace(
 508                                                                 $mtch[0],
 509                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]'
 510                                                                 . "\n",
 511                                                                 $s
 512                                                         );
 513                                                         Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG);
 514                                                 }
 515                                         }
 516                                 }
 517                         }
 518                 }
 519
 520                 return $s;
 521         }
 522
 523         /**
 524          * Truncates imported message body string length to max_import_size
 525          *
 526          * The purpose of this function is to apply system message length limits to
 527          * imported messages without including any embedded photos in the length
 528          *
 529          * @param string $body
 530          * @return string
 531          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 532          */
 533         public static function limitBodySize($body)
 534         {
 535                 $maxlen = DI::config()->get('config', 'max_import_size', 0);
 536
 537                 // If the length of the body, including the embedded images, is smaller
 538                 // than the maximum, then don't waste time looking for the images
 539                 if ($maxlen && (strlen($body) > $maxlen)) {
 540
 541                         Logger::log('the total body length exceeds the limit', Logger::DEBUG);
 542
 543                         $orig_body = $body;
 544                         $new_body = '';
 545                         $textlen = 0;
 546
 547                         $img_start = strpos($orig_body, '[img');
 548                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 549                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 550                         while (($img_st_close !== false) && ($img_end !== false)) {
 551
 552                                 $img_st_close++; // make it point to AFTER the closing bracket
 553                                 $img_end += $img_start;
 554                                 $img_end += strlen('[/img]');
 555
 556                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 557                                         // This is an embedded image
 558
 559                                         if (($textlen + $img_start) > $maxlen) {
 560                                                 if ($textlen < $maxlen) {
 561                                                         Logger::log('the limit happens before an embedded image', Logger::DEBUG);
 562                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 563                                                         $textlen = $maxlen;
 564                                                 }
 565                                         } else {
 566                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 567                                                 $textlen += $img_start;
 568                                         }
 569
 570                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 571                                 } else {
 572
 573                                         if (($textlen + $img_end) > $maxlen) {
 574                                                 if ($textlen < $maxlen) {
 575                                                         Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG);
 576                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 577                                                         $textlen = $maxlen;
 578                                                 }
 579                                         } else {
 580                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 581                                                 $textlen += $img_end;
 582                                         }
 583                                 }
 584                                 $orig_body = substr($orig_body, $img_end);
 585
 586                                 if ($orig_body === false) {
 587                                         // in case the body ends on a closing image tag
 588                                         $orig_body = '';
 589                                 }
 590
 591                                 $img_start = strpos($orig_body, '[img');
 592                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 593                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 594                         }
 595
 596                         if (($textlen + strlen($orig_body)) > $maxlen) {
 597                                 if ($textlen < $maxlen) {
 598                                         Logger::log('the limit happens after the end of the last image', Logger::DEBUG);
 599                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 600                                 }
 601                         } else {
 602                                 Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG);
 603                                 $new_body = $new_body . $orig_body;
 604                         }
 605
 606                         return $new_body;
 607                 } else {
 608                         return $body;
 609                 }
 610         }
 611
 612         /**
 613          * Processes [attachment] tags
 614          *
 615          * Note: Can produce a [bookmark] tag in the returned string
 616          *
 617          * @param string  $text
 618          * @param integer $simplehtml
 619          * @param bool    $tryoembed
 620          * @return string
 621          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 622          */
 623         private static function convertAttachment($text, $simplehtml = self::INTERNAL, $tryoembed = true)
 624         {
 625                 $data = self::getAttachmentData($text);
 626                 if (empty($data) || empty($data['url'])) {
 627                         return $text;
 628                 }
 629
 630                 if (isset($data['title'])) {
 631                         $data['title'] = strip_tags($data['title']);
 632                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 633                 } else {
 634                         $data['title'] = null;
 635                 }
 636
 637                 if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) {
 638                         $data['preview'] = $data['image'];
 639                         $data['image'] = '';
 640                 }
 641
 642                 $return = '';
 643                 try {
 644                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 645                                 $return = OEmbed::getHTML($data['url'], $data['title']);
 646                         } else {
 647                                 throw new Exception('OEmbed is disabled for this attachment.');
 648                         }
 649                 } catch (Exception $e) {
 650                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 651
 652                         if ($simplehtml != self::CONNECTORS) {
 653                                 $return = sprintf('<div class="type-%s">', $data['type']);
 654                         }
 655
 656                         if (!empty($data['title']) && !empty($data['url'])) {
 657                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 658                                         $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), $data['title']);
 659                                 } else {
 660                                         if (!empty($data['image'])) {
 661                                                 $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), $data['title']);
 662                                         } elseif (!empty($data['preview'])) {
 663                                                 $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']);
 664                                         }
 665                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 666                                 }
 667                         }
 668
 669                         if (!empty($data['description']) && $data['description'] != $data['title']) {
 670                                 // Sanitize the HTML by converting it to BBCode
 671                                 $bbcode = HTML::toBBCode($data['description']);
 672                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 673                         }
 674
 675                         if (!empty($data['url'])) {
 676                                 $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 677                         }
 678
 679                         if ($simplehtml != self::CONNECTORS) {
 680                                 $return .= '</div>';
 681                         }
 682                 }
 683
 684                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 685         }
 686
 687         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 688         {
 689                 $data = self::getAttachmentData($Text);
 690
 691                 if (!$data) {
 692                         return $Text;
 693                 } elseif ($nolink) {
 694                         return $data['text'] . ($data['after'] ?? '');
 695                 }
 696
 697                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 698                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 699                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 700                         $data['title'] = $data['url'];
 701                 } elseif (($text != '') && strstr($title, $text)) {
 702                         $data['text'] = $data['title'];
 703                         $data['title'] = $data['url'];
 704                 }
 705
 706                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 707                         return $data['title'] . $data['after'];
 708                 }
 709
 710                 // If the link already is included in the post, don't add it again
 711                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 712                         return $data['text'] . $data['after'];
 713                 }
 714
 715                 $text = $data['text'];
 716
 717                 if (!empty($data['url']) && !empty($data['title'])) {
 718                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 719                 } elseif (!empty($data['url'])) {
 720                         $text .= "\n[url]" . $data['url'] . '[/url]';
 721                 }
 722
 723                 return $text . "\n" . $data['after'];
 724         }
 725
 726         /**
 727          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 728          *
 729          * @param array $match Array with the matching values
 730          * @return string reformatted link including HTML codes
 731          */
 732         private static function convertUrlForActivityPubCallback($match)
 733         {
 734                 $url = $match[1];
 735
 736                 if (isset($match[2]) && ($match[1] != $match[2])) {
 737                         return $match[0];
 738                 }
 739
 740                 $parts = parse_url($url);
 741                 if (!isset($parts['scheme'])) {
 742                         return $match[0];
 743                 }
 744
 745                 return self::convertUrlForActivityPub($url);
 746         }
 747
 748         /**
 749          * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 750          *
 751          * @param string $url URL that is about to be reformatted
 752          * @return string reformatted link including HTML codes
 753          */
 754         private static function convertUrlForActivityPub($url)
 755         {
 756                 $html = '<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>';
 757                 return sprintf($html, $url, self::getStyledURL($url));
 758         }
 759
 760         /**
 761          * Converts an URL in a nicer format (without the scheme and possibly shortened)
 762          *
 763          * @param string $url URL that is about to be reformatted
 764          * @return string reformatted link
 765          */
 766         private static function getStyledURL($url)
 767         {
 768                 $parts = parse_url($url);
 769                 $scheme = $parts['scheme'] . '://';
 770                 $styled_url = str_replace($scheme, '', $url);
 771
 772                 if (strlen($styled_url) > 30) {
 773                         $styled_url = substr($styled_url, 0, 30) . "…";
 774                 }
 775
 776                 return $styled_url;
 777         }
 778
 779         /*
 780          * [noparse][i]italic[/i][/noparse] turns into
 781          * [noparse][ i ]italic[ /i ][/noparse],
 782          * to hide them from parser.
 783          */
 784         private static function escapeNoparseCallback($match)
 785         {
 786                 $whole_match = $match[0];
 787                 $captured = $match[1];
 788                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 789                 $new_str = str_replace($captured, $spacefied, $whole_match);
 790                 return $new_str;
 791         }
 792
 793         /*
 794          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 795          * now turns back and the [noparse] tags are trimed
 796          * returning [i]italic[/i]
 797          */
 798         private static function unescapeNoparseCallback($match)
 799         {
 800                 $captured = $match[1];
 801                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 802                 return $unspacefied;
 803         }
 804
 805         /**
 806          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 807          * occurrences
 808          *
 809          * @param string $text        Text to search
 810          * @param string $name        Tag name
 811          * @param int    $occurrences Number of first occurrences to skip
 812          * @return boolean|array
 813          */
 814         public static function getTagPosition($text, $name, $occurrences = 0)
 815         {
 816                 if ($occurrences < 0) {
 817                         $occurrences = 0;
 818                 }
 819
 820                 $start_open = -1;
 821                 for ($i = 0; $i <= $occurrences; $i++) {
 822                         if ($start_open !== false) {
 823                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 824                         }
 825                 }
 826
 827                 if ($start_open === false) {
 828                         return false;
 829                 }
 830
 831                 $start_equal = strpos($text, '=', $start_open);
 832                 $start_close = strpos($text, ']', $start_open);
 833
 834                 if ($start_close === false) {
 835                         return false;
 836                 }
 837
 838                 $start_close++;
 839
 840                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 841
 842                 if ($end_open === false) {
 843                         return false;
 844                 }
 845
 846                 $res = [
 847                         'start' => [
 848                                 'open' => $start_open,
 849                                 'close' => $start_close
 850                         ],
 851                         'end' => [
 852                                 'open' => $end_open,
 853                                 'close' => $end_open + strlen('[/' . $name . ']')
 854                         ],
 855                 ];
 856
 857                 if ($start_equal !== false) {
 858                         $res['start']['equal'] = $start_equal + 1;
 859                 }
 860
 861                 return $res;
 862         }
 863
 864         /**
 865          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 866          *
 867          * @param string $pattern Preg pattern string
 868          * @param string $replace Preg replace string
 869          * @param string $name    BBCode tag name
 870          * @param string $text    Text to search
 871          * @return string
 872          */
 873         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 874         {
 875                 $occurrences = 0;
 876                 $pos = self::getTagPosition($text, $name, $occurrences);
 877                 while ($pos !== false && $occurrences++ < 1000) {
 878                         $start = substr($text, 0, $pos['start']['open']);
 879                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 880                         $end = substr($text, $pos['end']['close']);
 881                         if ($end === false) {
 882                                 $end = '';
 883                         }
 884
 885                         $subject = preg_replace($pattern, $replace, $subject);
 886                         $text = $start . $subject . $end;
 887
 888                         $pos = self::getTagPosition($text, $name, $occurrences);
 889                 }
 890
 891                 return $text;
 892         }
 893
 894         private static function extractImagesFromItemBody($body)
 895         {
 896                 $saved_image = [];
 897                 $orig_body = $body;
 898                 $new_body = '';
 899
 900                 $cnt = 0;
 901                 $img_start = strpos($orig_body, '[img');
 902                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 903                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 904                 while (($img_st_close !== false) && ($img_end !== false)) {
 905                         $img_st_close++; // make it point to AFTER the closing bracket
 906                         $img_end += $img_start;
 907
 908                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 909                                 // This is an embedded image
 910                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 911                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 912
 913                                 $cnt++;
 914                         } else {
 915                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 916                         }
 917
 918                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 919
 920                         if ($orig_body === false) {
 921                                 // in case the body ends on a closing image tag
 922                                 $orig_body = '';
 923                         }
 924
 925                         $img_start = strpos($orig_body, '[img');
 926                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 927                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 928                 }
 929
 930                 $new_body = $new_body . $orig_body;
 931
 932                 return ['body' => $new_body, 'images' => $saved_image];
 933         }
 934
 935         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 936         {
 937                 $newbody = $body;
 938
 939                 $cnt = 0;
 940                 foreach ($images as $image) {
 941                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 942                         // it loops over the array starting from the first element and going sequentially
 943                         // to the last element
 944                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 945                                 '<img src="' . self::proxyUrl($image) . '" alt="' . DI::l10n()->t('Image/photo') . '" />', $newbody);
 946                         $cnt++;
 947                 }
 948
 949                 return $newbody;
 950         }
 951
 952         /**
 953          * This function converts a [share] block to text according to a provided callback function whose signature is:
 954          *
 955          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 956          *
 957          * Where:
 958          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 959          * data lookup
 960          * - $author_contact is a contact record array
 961          * - $content is the inner content of the [share] block
 962          * - $is_quote_share indicates whether there's any content before the [share] block
 963          * - Return value is the string that should replace the [share] block in the provided text
 964          *
 965          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 966          *
 967          * @param  string   $text     A BBCode string
 968          * @param  callable $callback
 969          * @return string The BBCode string with all [share] blocks replaced
 970          */
 971         public static function convertShare($text, callable $callback)
 972         {
 973                 $return = preg_replace_callback(
 974                         "/(.*?)\[share(.*?)\](.*)\[\/share\]/ism",
 975                         function ($match) use ($callback) {
 976                                 $attribute_string = $match[2];
 977                                 $attributes = [];
 978                                 foreach (['author', 'profile', 'avatar', 'link', 'posted'] as $field) {
 979                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 980                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 981                                 }
 982
 983                                 // We only call this so that a previously unknown contact can be added.
 984                                 // This is important for the function "Model\Contact::getDetailsByURL()".
 985                                 // This function then can fetch an entry from the contact table.
 986                                 $default['url'] = $attributes['profile'];
 987
 988                                 if (!empty($attributes['author'])) {
 989                                         $default['name'] = $attributes['author'];
 990                                 }
 991
 992                                 if (!empty($attributes['avatar'])) {
 993                                         $default['photo'] = $attributes['avatar'];
 994                                 }
 995
 996                                 Contact::getIdForURL($attributes['profile'], 0, true, $default);
 997
 998                                 $author_contact = Contact::getDetailsByURL($attributes['profile']);
 999                                 $author_contact['url'] = ($author_contact['url'] ?? $attributes['profile']);
1000                                 $author_contact['addr'] = ($author_contact['addr'] ?? '') ?: Protocol::getAddrFromProfileUrl($attributes['profile']);
1001
1002                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
1003                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
1004                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
1005
1006                                 if ($attributes['avatar']) {
1007                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
1008                                 }
1009
1010                                 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
1011                         },
1012                         $text
1013                 );
1014
1015                 return $return;
1016         }
1017
1018         /**
1019          * Default [share] tag conversion callback
1020          *
1021          * Note: Can produce a [bookmark] tag in the output
1022          *
1023          * @see BBCode::convertShare()
1024          * @param array   $attributes     [share] block attribute values
1025          * @param array   $author_contact Contact row of the shared author
1026          * @param string  $content        Inner content of the [share] block
1027          * @param boolean $is_quote_share Whether there is content before the [share] block
1028          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
1029          * @return string
1030          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1031          */
1032         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
1033         {
1034                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
1035
1036                 switch ($simplehtml) {
1037                         case self::API:
1038                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1039                                 break;
1040                         case self::DIASPORA:
1041                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1042                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1043                                 } else {
1044                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1045
1046                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1047                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1048                                         }
1049
1050                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1051
1052                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1053                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1054                                         }
1055                                 }
1056
1057                                 break;
1058                         case self::CONNECTORS:
1059                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1060                                 $headline .= DI::l10n()->t('<a href="%1$s" target="_blank" rel="noopener noreferrer">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1061                                 $headline .= ':</b></p>' . "\n";
1062
1063                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1064
1065                                 break;
1066                         case self::OSTATUS:
1067                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1068                                 break;
1069                         case self::ACTIVITYPUB:
1070                                 $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>';
1071                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1072                                 break;
1073                         default:
1074                                 $text = ($is_quote_share? "\n" : '');
1075
1076                                 $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1077                                 $text .= Renderer::replaceMacros($tpl, [
1078                                         '$profile' => $attributes['profile'],
1079                                         '$avatar'  => $attributes['avatar'],
1080                                         '$author'  => $attributes['author'],
1081                                         '$link'    => $attributes['link'],
1082                                         '$posted'  => $attributes['posted'],
1083                                         '$content' => trim($content)
1084                                 ]);
1085                                 break;
1086                 }
1087
1088                 return $text;
1089         }
1090
1091         private static function removePictureLinksCallback($match)
1092         {
1093                 $cache_key = 'remove:' . $match[1];
1094                 $text = DI::cache()->get($cache_key);
1095
1096                 if (is_null($text)) {
1097                         $a = DI::app();
1098
1099                         $stamp1 = microtime(true);
1100
1101                         $ch = @curl_init($match[1]);
1102                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1103                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1104                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1105                         @curl_exec($ch);
1106                         $curl_info = @curl_getinfo($ch);
1107
1108                         DI::profiler()->saveTimestamp($stamp1, "network", System::callstack());
1109
1110                         if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1111                                 $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]";
1112                         } else {
1113                                 $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]";
1114
1115                                 // if its not a picture then look if its a page that contains a picture link
1116                                 $body = Network::fetchUrl($match[1]);
1117
1118                                 $doc = new DOMDocument();
1119                                 @$doc->loadHTML($body);
1120                                 $xpath = new DOMXPath($doc);
1121                                 $list = $xpath->query("//meta[@name]");
1122                                 foreach ($list as $node) {
1123                                         $attr = [];
1124
1125                                         if ($node->attributes->length) {
1126                                                 foreach ($node->attributes as $attribute) {
1127                                                         $attr[$attribute->name] = $attribute->value;
1128                                                 }
1129                                         }
1130
1131                                         if (strtolower($attr['name']) == 'twitter:image') {
1132                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1133                                         }
1134                                 }
1135                         }
1136                         DI::cache()->set($cache_key, $text);
1137                 }
1138
1139                 return $text;
1140         }
1141
1142         private static function expandLinksCallback($match)
1143         {
1144                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1145                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1146                 } else {
1147                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1148                 }
1149         }
1150
1151         private static function cleanPictureLinksCallback($match)
1152         {
1153                 $a = DI::app();
1154
1155                 // When the picture link is the own photo path then we can avoid fetching the link
1156                 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()->get()) . '/photos/');
1157                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1158                         if (!empty($match[3])) {
1159                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1160                         } else {
1161                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1162                         }
1163                         return $text;
1164                 }
1165
1166                 $cache_key = 'clean:' . $match[1];
1167                 $text = DI::cache()->get($cache_key);
1168                 if (!is_null($text)) {
1169                         return $text;
1170                 }
1171
1172                 // Only fetch the header, not the content
1173                 $stamp1 = microtime(true);
1174
1175                 $ch = @curl_init($match[1]);
1176                 @curl_setopt($ch, CURLOPT_NOBODY, true);
1177                 @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1178                 @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1179                 @curl_exec($ch);
1180                 $curl_info = @curl_getinfo($ch);
1181
1182                 DI::profiler()->saveTimestamp($stamp1, "network", System::callstack());
1183
1184                 // if its a link to a picture then embed this picture
1185                 if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1186                         $text = '[img]' . $match[1] . '[/img]';
1187                 } else {
1188                         if (!empty($match[3])) {
1189                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1190                         } else {
1191                                 $text = '[img]' . $match[2] . '[/img]';
1192                         }
1193
1194                         // if its not a picture then look if its a page that contains a picture link
1195                         $body = Network::fetchUrl($match[1]);
1196
1197                         $doc = new DOMDocument();
1198                         @$doc->loadHTML($body);
1199                         $xpath = new DOMXPath($doc);
1200                         $list = $xpath->query("//meta[@name]");
1201                         foreach ($list as $node) {
1202                                 $attr = [];
1203                                 if ($node->attributes->length) {
1204                                         foreach ($node->attributes as $attribute) {
1205                                                 $attr[$attribute->name] = $attribute->value;
1206                                         }
1207                                 }
1208
1209                                 if (strtolower($attr['name']) == "twitter:image") {
1210                                         if (!empty($match[3])) {
1211                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1212                                         } else {
1213                                                 $text = "[img]" . $attr['content'] . "[/img]";
1214                                         }
1215                                 }
1216                         }
1217                 }
1218                 DI::cache()->set($cache_key, $text);
1219
1220                 return $text;
1221         }
1222
1223         public static function cleanPictureLinks($text)
1224         {
1225                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1226                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1227                 return $return;
1228         }
1229
1230         /**
1231          * Converts a BBCode message to HTML message
1232          *
1233          * BBcode 2 HTML was written by WAY2WEB.net
1234          * extended to work with Mistpark/Friendica - Mike Macgirvin
1235          *
1236          * Simple HTML values meaning:
1237          * - 0: Friendica display
1238          * - 1: Unused
1239          * - 2: Used for Windows Phone push, Friendica API
1240          * - 3: Used before converting to Markdown in bb2diaspora.php
1241          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1242          * - 5: Unused
1243          * - 6: Unused
1244          * - 7: Used for dfrn, OStatus
1245          * - 8: Used for Twitter, WP backlink text setting
1246          * - 9: ActivityPub
1247          *
1248          * @param string $text
1249          * @param bool   $try_oembed
1250          * @param int    $simple_html
1251          * @param bool   $for_plaintext
1252          * @return string
1253          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1254          */
1255         public static function convert($text, $try_oembed = true, $simple_html = self::INTERNAL, $for_plaintext = false)
1256         {
1257                 $a = DI::app();
1258
1259                 /*
1260                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1261                  *
1262                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1263                  * $match[1] = $url
1264                  * $match[2] = $title or absent
1265                  */
1266                 $try_oembed_callback = function ($match)
1267                 {
1268                         $url = $match[1];
1269                         $title = $match[2] ?? null;
1270
1271                         try {
1272                                 $return = OEmbed::getHTML($url, $title);
1273                         } catch (Exception $ex) {
1274                                 $return = $match[0];
1275                         }
1276
1277                         return $return;
1278                 };
1279
1280                 // Extracting code blocks before the whitespace processing and the autolinker
1281                 $codeblocks = [];
1282
1283                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1284                         function ($matches) use (&$codeblocks) {
1285                                 $return = '#codeblock-' . count($codeblocks) . '#';
1286                                 if (strpos($matches[2], "\n") !== false) {
1287                                         $codeblocks[] = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlspecialchars(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
1288                                 } else {
1289                                         $codeblocks[] = '<code>' . htmlspecialchars($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
1290                                 }
1291
1292                                 return $return;
1293                         },
1294                         $text
1295                 );
1296
1297                 // Hide all [noparse] contained bbtags by spacefying them
1298                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1299
1300                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1301                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1302                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1303
1304                 // Remove the abstract element. It is a non visible element.
1305                 $text = self::stripAbstract($text);
1306
1307                 // Move all spaces out of the tags
1308                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1309                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1310
1311                 // Extract the private images which use data urls since preg has issues with
1312                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1313                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1314
1315                 $extracted = self::extractImagesFromItemBody($text);
1316                 $text = $extracted['body'];
1317                 $saved_image = $extracted['images'];
1318
1319                 // If we find any event code, turn it into an event.
1320                 // After we're finished processing the bbcode we'll
1321                 // replace all of the event code with a reformatted version.
1322
1323                 $ev = Event::fromBBCode($text);
1324
1325                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1326                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1327
1328                 $text = str_replace("<", "&lt;", $text);
1329                 $text = str_replace(">", "&gt;", $text);
1330
1331                 // remove some newlines before the general conversion
1332                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1333                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1334
1335                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1336                 if (!$try_oembed) {
1337                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1338                 }
1339
1340                 // Convert new line chars to html <br /> tags
1341
1342                 // nlbr seems to be hopelessly messed up
1343                 //      $Text = nl2br($Text);
1344
1345                 // We'll emulate it.
1346
1347                 $text = trim($text);
1348                 $text = str_replace("\r\n", "\n", $text);
1349
1350                 // Remove linefeeds inside of the table elements. See issue #6799
1351                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1352                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1353                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1354                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1355                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1356                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1357                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1358                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1359                 do {
1360                         $oldtext = $text;
1361                         $text = str_replace($search, $replace, $text);
1362                 } while ($oldtext != $text);
1363
1364                 // Replace these here only once
1365                 $search = ["\n[table]", "[/table]\n"];
1366                 $replace = ["[table]", "[/table]"];
1367                 $text = str_replace($search, $replace, $text);
1368
1369                 // removing multiplicated newlines
1370                 if (DI::config()->get('system', 'remove_multiplicated_lines')) {
1371                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1372                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1373                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1374                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1375                         do {
1376                                 $oldtext = $text;
1377                                 $text = str_replace($search, $replace, $text);
1378                         } while ($oldtext != $text);
1379                 }
1380
1381                 /// @todo Have a closer look at the different html modes
1382                 // Handle attached links or videos
1383                 if ($simple_html == self::ACTIVITYPUB) {
1384                         $text = self::removeAttachment($text);
1385                 } elseif (!in_array($simple_html, [self::INTERNAL, self::CONNECTORS])) {
1386                         $text = self::removeAttachment($text, true);
1387                 } else {
1388                         $text = self::convertAttachment($text, $simple_html, $try_oembed);
1389                 }
1390
1391                 // leave open the posibility of [map=something]
1392                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1393                 if (strpos($text, '[/map]') !== false) {
1394                         $text = preg_replace_callback(
1395                                 "/\[map\](.*?)\[\/map\]/ism",
1396                                 function ($match) use ($simple_html) {
1397                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1398                                 },
1399                                 $text
1400                         );
1401                 }
1402
1403                 if (strpos($text, '[map=') !== false) {
1404                         $text = preg_replace_callback(
1405                                 "/\[map=(.*?)\]/ism",
1406                                 function ($match) use ($simple_html) {
1407                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1408                                 },
1409                                 $text
1410                         );
1411                 }
1412
1413                 if (strpos($text, '[map]') !== false) {
1414                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1415                 }
1416
1417                 // Check for headers
1418                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1419                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1420                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1421                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1422                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1423                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1424
1425                 // Check for paragraph
1426                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1427
1428                 // Check for bold text
1429                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1430
1431                 // Check for Italics text
1432                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1433
1434                 // Check for Underline text
1435                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1436
1437                 // Check for strike-through text
1438                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1439
1440                 // Check for over-line text
1441                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1442
1443                 // Check for colored text
1444                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1445
1446                 // Check for sized text
1447                 // [size=50] --> font-size: 50px (with the unit).
1448                 if ($simple_html != self::DIASPORA) {
1449                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1450                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1451                 } else {
1452                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1453                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1454                 }
1455
1456
1457                 // Check for centered text
1458                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1459
1460                 // Check for list text
1461                 $text = str_replace("[*]", "<li>", $text);
1462
1463                 // Check for style sheet commands
1464                 $text = preg_replace_callback(
1465                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1466                         function ($match) {
1467                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1468                         },
1469                         $text
1470                 );
1471
1472                 // Check for CSS classes
1473                 $text = preg_replace_callback(
1474                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1475                         function ($match) {
1476                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1477                         },
1478                         $text
1479                 );
1480
1481                 // handle nested lists
1482                 $endlessloop = 0;
1483
1484                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1485                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1486                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1487                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1488                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1489                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1490                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1491                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1492                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1493                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1494                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1495                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1496                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1497                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1498                 }
1499
1500                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1501                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1502                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1503                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1504
1505                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1506                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1507
1508                 $text = str_replace('[hr]', '<hr />', $text);
1509
1510                 if (!$for_plaintext) {
1511                         $escaped = [];
1512
1513                         // Escaping BBCodes susceptible to contain rogue URL we don'' want the autolinker to catch
1514                         $text = preg_replace_callback('#\[(url|img|audio|video|youtube|vimeo|share|attachment|iframe|bookmark).+?\[/\1\]#ism',
1515                                 function ($matches) use (&$escaped) {
1516                                         $return = '{escaped-' . count($escaped) . '}';
1517                                         $escaped[] = $matches[0];
1518
1519                                         return $return;
1520                                 },
1521                                 $text
1522                         );
1523
1524                         // Autolinker for isolated URLs
1525                         $text = preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1526
1527                         // Restoring escaped blocks
1528                         $text = preg_replace_callback('/{escaped-([0-9]+)}/iU',
1529                                 function ($matches) use ($escaped) {
1530                                         return $escaped[intval($matches[1])] ?? $matches[0];
1531                                 },
1532                                 $text
1533                         );
1534                 }
1535
1536                 // This is actually executed in Item::prepareBody()
1537
1538                 $nosmile = strpos($text, '[nosmile]') !== false;
1539                 $text = str_replace('[nosmile]', '', $text);
1540
1541                 // Check for font change text
1542                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1543
1544                 // Declare the format for [spoiler] layout
1545                 $SpoilerLayout = '<details class="spoiler"><summary>' . DI::l10n()->t('Click to open/close') . '</summary>$1</details>';
1546
1547                 // Check for [spoiler] text
1548                 // handle nested quotes
1549                 $endlessloop = 0;
1550                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1551                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1552                 }
1553
1554                 // Check for [spoiler=Title] text
1555
1556                 // handle nested quotes
1557                 $endlessloop = 0;
1558                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1559                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1560                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1561                                 $text);
1562                 }
1563
1564                 // Declare the format for [quote] layout
1565                 $QuoteLayout = '<blockquote>$1</blockquote>';
1566
1567                 // Check for [quote] text
1568                 // handle nested quotes
1569                 $endlessloop = 0;
1570                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1571                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1572                 }
1573
1574                 // Check for [quote=Author] text
1575
1576                 $t_wrote = DI::l10n()->t('$1 wrote:');
1577
1578                 // handle nested quotes
1579                 $endlessloop = 0;
1580                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1581                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1582                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1583                                 $text);
1584                 }
1585
1586
1587                 // [img=widthxheight]image source[/img]
1588                 $text = preg_replace_callback(
1589                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1590                         function ($matches) use ($simple_html) {
1591                                 if (strpos($matches[3], "data:image/") === 0) {
1592                                         return $matches[0];
1593                                 }
1594
1595                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1596                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1597                         },
1598                         $text
1599                 );
1600
1601                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1602                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1603
1604                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1605                         function ($matches) use ($simple_html) {
1606                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1607                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1608                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1609                         },
1610                         $text);
1611
1612                 // Images
1613                 // [img]pathtoimage[/img]
1614                 $text = preg_replace_callback(
1615                         "/\[img\](.*?)\[\/img\]/ism",
1616                         function ($matches) use ($simple_html) {
1617                                 if (strpos($matches[1], "data:image/") === 0) {
1618                                         return $matches[0];
1619                                 }
1620
1621                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1622                                 return "[img]" . $matches[1] . "[/img]";
1623                         },
1624                         $text
1625                 );
1626
1627                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1628                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1629
1630                 $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);
1631                 $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 />', $text);
1632                 //$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);
1633
1634                 // Simplify "video" element
1635                 $text = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1636
1637                 // Try to Oembed
1638                 if ($try_oembed) {
1639                         $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);
1640                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '<audio src="$1" controls="controls"><a href="$1">$1</a></audio>', $text);
1641
1642                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1643                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1644                 } else {
1645                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1646                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1647                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1648                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1649                 }
1650
1651                 // html5 video and audio
1652
1653
1654                 if ($try_oembed) {
1655                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1656                 } else {
1657                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1658                 }
1659
1660                 // Youtube extensions
1661                 if ($try_oembed) {
1662                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1663                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1664                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1665                 }
1666
1667                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1668                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1669                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1670
1671                 if ($try_oembed) {
1672                         $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);
1673                 } else {
1674                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1675                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank" rel="noopener noreferrer">https://www.youtube.com/watch?v=$1</a>', $text);
1676                 }
1677
1678                 if ($try_oembed) {
1679                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1680                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1681                 }
1682
1683                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1684                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1685
1686                 if ($try_oembed) {
1687                         $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);
1688                 } else {
1689                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1690                                 '<a href="https://vimeo.com/$1" target="_blank" rel="noopener noreferrer">https://vimeo.com/$1</a>', $text);
1691                 }
1692
1693                 // oembed tag
1694                 $text = OEmbed::BBCode2HTML($text);
1695
1696                 // Avoid triple linefeeds through oembed
1697                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1698
1699                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1700                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1701                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1702                 // start which is always required). Allow desc with a missing summary for compatibility.
1703
1704                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1705                         $sub = Event::getHTML($ev, $simple_html);
1706
1707                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1708                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1709                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1710                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1711                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1712                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1713                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1714                 }
1715
1716                 // Replace non graphical smilies for external posts
1717                 if (!$nosmile && !$for_plaintext) {
1718                         $text = Smilies::replace($text);
1719                 }
1720
1721                 if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA)) {
1722                         $conv = html_entity_decode(str_replace([' ', "\n", "\r"], '', $text));
1723                         // Emojis are always 4 byte Unicode characters
1724                         if (!empty($conv) && (strlen($conv) / mb_strlen($conv) == 4)) {
1725                                 $text = '<span style="font-size: xx-large; line-height: initial;">' . $text . '</span>';
1726                         }
1727                 }
1728
1729                 if (!$for_plaintext) {
1730                         if (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1731                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1732                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1733                         }
1734                 } else {
1735                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1736                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1737                 }
1738
1739                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1740
1741                 // Remove all hashtag addresses
1742                 if ($simple_html && !in_array($simple_html, [self::DIASPORA, self::OSTATUS, self::ACTIVITYPUB])) {
1743                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1744                 } elseif ($simple_html == self::DIASPORA) {
1745                         // The ! is converted to @ since Diaspora only understands the @
1746                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1747                                 '@<a href="$2">$3</a>',
1748                                 $text);
1749                 } elseif (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1750                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1751                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1752                                 $text);
1753                 } elseif (!$simple_html) {
1754                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1755                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1756                                 $text);
1757                 }
1758
1759                 // Bookmarks in red - will be converted to bookmarks in friendica
1760                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1761                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1762                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1763                                         "[bookmark=$1]$2[/bookmark]", $text);
1764
1765                 if (in_array($simple_html, [self::API, self::OSTATUS, self::TWITTER])) {
1766                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1767                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1768                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1769                 }
1770
1771                 // Perform URL Search
1772                 if ($try_oembed) {
1773                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1774                 }
1775
1776                 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1777
1778                 // Handle Diaspora posts
1779                 $text = preg_replace_callback(
1780                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1781                         function ($match) {
1782                                 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1783                         }, $text
1784                 );
1785
1786                 $text = preg_replace_callback(
1787                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1788                         function ($match) {
1789                                 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1790                         }, $text
1791                 );
1792
1793                 // Server independent link to posts and comments
1794                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1795                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1796                 $text = preg_replace($expression, DI::baseUrl()."/display/$1", $text);
1797
1798                 /* Tag conversion
1799                  * Supports:
1800                  * - #[url=<anything>]<term>[/url]
1801                  * - [url=<anything>]#<term>[/url]
1802                  */
1803                 $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) use ($simple_html) {
1804                         if ($simple_html == BBCode::ACTIVITYPUB) {
1805                                 return '<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1806                                         . '" data-tag="' . XML::escape($matches[1]) . '" rel="tag ugc">#'
1807                                         . XML::escape($matches[1]) . '</a>';
1808                         } else {
1809                                 return '#<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1810                                         . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1811                                         . XML::escape($matches[1]) . '</a>';
1812                         }
1813                 }, $text);
1814
1815                 // We need no target="_blank" rel="noopener noreferrer" for local links
1816                 // convert links start with DI::baseUrl() as local link without the target="_blank" rel="noopener noreferrer" attribute
1817                 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
1818                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1819                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1820
1821                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1822                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1823
1824                 // Red compatibility, though the link can't be authenticated on Friendica
1825                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1826
1827
1828                 // we may need to restrict this further if it picks up too many strays
1829                 // link acct:user@host to a webfinger profile redirector
1830
1831                 $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);
1832
1833                 // Perform MAIL Search
1834                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1835                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1836
1837                 // Unhide all [noparse] contained bbtags unspacefying them
1838                 // and triming the [noparse] tag.
1839
1840                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1841                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1842                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1843
1844                 /// @todo What is the meaning of these lines?
1845                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1846                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1847
1848                 // Currently deactivated, it made problems with " inside of alt texts.
1849                 //$text = preg_replace('/\&quot\;/', '"', $text);
1850
1851                 // fix any escaped ampersands that may have been converted into links
1852                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1853
1854                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1855                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1856
1857                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1858
1859                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1860                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">', $text);
1861
1862                 // sanitize href attributes (only whitelisted protocols URLs)
1863                 // default value for backward compatibility
1864                 $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
1865
1866                 // Always allowed protocol even if config isn't set or not including it
1867                 $allowed_link_protocols[] = '//';
1868                 $allowed_link_protocols[] = 'http://';
1869                 $allowed_link_protocols[] = 'https://';
1870                 $allowed_link_protocols[] = 'redir/';
1871
1872                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1873
1874                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1875                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text);
1876
1877                 // Shared content
1878                 $text = self::convertShare(
1879                         $text,
1880                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1881                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1882                         }
1883                 );
1884
1885                 if ($saved_image) {
1886                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1887                 }
1888
1889                 // Restore code blocks
1890                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1891                         function ($matches) use ($codeblocks) {
1892                                 $return = $matches[0];
1893                                 if (isset($codeblocks[intval($matches[1])])) {
1894                                         $return = $codeblocks[$matches[1]];
1895                                 }
1896                                 return $return;
1897                         },
1898                         $text
1899                 );
1900
1901                 // Clean up the HTML by loading and saving the HTML with the DOM.
1902                 // Bad structured html can break a whole page.
1903                 // For performance reasons do it only with activated item cache or at export.
1904                 if (!$try_oembed || (get_itemcachepath() != '')) {
1905                         $doc = new DOMDocument();
1906                         $doc->preserveWhiteSpace = false;
1907
1908                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1909
1910                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1911                         $encoding = '<?xml encoding="UTF-8">';
1912                         @$doc->loadHTML($encoding . $doctype . '<html><body>' . $text . '</body></html>');
1913                         $doc->encoding = 'UTF-8';
1914                         $text = $doc->saveHTML();
1915                         $text = str_replace(['<html><body>', '</body></html>', $doctype, $encoding], ['', '', '', ''], $text);
1916
1917                         $text = str_replace('<br></li>', '</li>', $text);
1918
1919                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1920                 }
1921
1922                 // Clean up some useless linebreaks in lists
1923                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1924                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1925                 //$Text = str_replace('</li><br />', '</li>', $Text);
1926                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1927                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1928
1929                 Hook::callAll('bbcode', $text);
1930
1931                 return trim($text);
1932         }
1933
1934         /**
1935          * Strips the "abstract" tag from the provided text
1936          *
1937          * @param string $text The text with BBCode
1938          * @return string The same text - but without "abstract" element
1939          */
1940         public static function stripAbstract($text)
1941         {
1942                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1943                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1944
1945                 return $text;
1946         }
1947
1948         /**
1949          * Returns the value of the "abstract" element
1950          *
1951          * @param string $text The text that maybe contains the element
1952          * @param string $addon The addon for which the abstract is meant for
1953          * @return string The abstract
1954          */
1955         public static function getAbstract($text, $addon = '')
1956         {
1957                 $abstract = '';
1958                 $abstracts = [];
1959                 $addon = strtolower($addon);
1960
1961                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1962                         foreach ($results AS $result) {
1963                                 $abstracts[strtolower($result[1])] = $result[2];
1964                         }
1965                 }
1966
1967                 if (isset($abstracts[$addon])) {
1968                         $abstract = $abstracts[$addon];
1969                 }
1970
1971                 if ($abstract == '' && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1972                         $abstract = $result[1];
1973                 }
1974
1975                 return $abstract;
1976         }
1977
1978         /**
1979          * Callback function to replace a Friendica style mention in a mention for Diaspora
1980          *
1981          * @param array $match Matching values for the callback
1982          *                     [1] = Mention type (! or @)
1983          *                     [2] = Name
1984          *                     [3] = Address
1985          * @return string Replaced mention
1986          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1987          * @throws \ImagickException
1988          */
1989         private static function bbCodeMention2DiasporaCallback($match)
1990         {
1991                 $contact = Contact::getDetailsByURL($match[3]);
1992
1993                 if (empty($contact['addr'])) {
1994                         $contact = Probe::uri($match[3]);
1995                 }
1996
1997                 if (empty($contact['addr'])) {
1998                         return $match[0];
1999                 }
2000
2001                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
2002                 return $mention;
2003         }
2004
2005         /**
2006          * Converts a BBCode text into Markdown
2007          *
2008          * This function converts a BBCode item body to be sent to Markdown-enabled
2009          * systems like Diaspora and Libertree
2010          *
2011          * @param string $text
2012          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
2013          * @return string
2014          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2015          */
2016         public static function toMarkdown($text, $for_diaspora = true)
2017         {
2018                 $original_text = $text;
2019
2020                 // Since Diaspora is creating a summary for links, this function removes them before posting
2021                 if ($for_diaspora) {
2022                         $text = self::removeShareInformation($text);
2023                 }
2024
2025                 /**
2026                  * Transform #tags, strip off the [url] and replace spaces with underscore
2027                  */
2028                 $url_search_string = "^\[\]";
2029                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2030                         function ($matches) {
2031                                 return '#' . str_replace(' ', '_', $matches[2]);
2032                         },
2033                         $text
2034                 );
2035
2036                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2037                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2038
2039                 // Convert it to HTML - don't try oembed
2040                 if ($for_diaspora) {
2041                         $text = self::convert($text, false, self::DIASPORA);
2042
2043                         // Add all tags that maybe were removed
2044                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2045                                 $tagline = '';
2046                                 foreach ($tags[2] as $tag) {
2047                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2048                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2049                                                 $tagline .= '#' . $tag . ' ';
2050                                         }
2051                                 }
2052                                 $text = $text . " " . $tagline;
2053                         }
2054                 } else {
2055                         $text = self::convert($text, false, self::CONNECTORS);
2056                 }
2057
2058                 // If a link is followed by a quote then there should be a newline before it
2059                 // Maybe we should make this newline at every time before a quote.
2060                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2061
2062                 $stamp1 = microtime(true);
2063
2064                 // Now convert HTML to Markdown
2065                 $text = HTML::toMarkdown($text);
2066
2067                 DI::profiler()->saveTimestamp($stamp1, "parser", System::callstack());
2068
2069                 // Libertree has a problem with escaped hashtags.
2070                 $text = str_replace(['\#'], ['#'], $text);
2071
2072                 // Remove any leading or trailing whitespace, as this will mess up
2073                 // the Diaspora signature verification and cause the item to disappear
2074                 $text = trim($text);
2075
2076                 if ($for_diaspora) {
2077                         $url_search_string = "^\[\]";
2078                         $text = preg_replace_callback(
2079                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2080                                 ['self', 'bbCodeMention2DiasporaCallback'],
2081                                 $text
2082                         );
2083                 }
2084
2085                 Hook::callAll('bb2diaspora', $text);
2086
2087                 return $text;
2088         }
2089
2090         /**
2091          * Pull out all #hashtags and @person tags from $string.
2092          *
2093          * We also get @person@domain.com - which would make
2094          * the regex quite complicated as tags can also
2095          * end a sentence. So we'll run through our results
2096          * and strip the period from any tags which end with one.
2097          * Returns array of tags found, or empty array.
2098          *
2099          * @param string $string Post content
2100          *
2101          * @return array List of tag and person names
2102          */
2103         public static function getTags($string)
2104         {
2105                 $ret = [];
2106
2107                 // Convert hashtag links to hashtags
2108                 $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2 ', $string);
2109
2110                 // ignore anything in a code block
2111                 $string = preg_replace('/\[code.*?\].*?\[\/code\]/sm', '', $string);
2112
2113                 // Force line feeds at bbtags
2114                 $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2115
2116                 // ignore anything in a bbtag
2117                 $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2118
2119                 // Match full names against @tags including the space between first and last
2120                 // We will look these up afterward to see if they are full names or not recognisable.
2121
2122                 if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2123                         foreach ($matches[1] as $match) {
2124                                 if (strstr($match, ']')) {
2125                                         // we might be inside a bbcode color tag - leave it alone
2126                                         continue;
2127                                 }
2128
2129                                 if (substr($match, -1, 1) === '.') {
2130                                         $ret[] = substr($match, 0, -1);
2131                                 } else {
2132                                         $ret[] = $match;
2133                                 }
2134                         }
2135                 }
2136
2137                 // Otherwise pull out single word tags. These can be @nickname, @first_last
2138                 // and #hash tags.
2139
2140                 if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?]+)([ \x0D\x0A,;:?]|$)/', $string, $matches)) {
2141                         foreach ($matches[1] as $match) {
2142                                 if (strstr($match, ']')) {
2143                                         // we might be inside a bbcode color tag - leave it alone
2144                                         continue;
2145                                 }
2146
2147                                 if (substr($match, -1, 1) === '.') {
2148                                         $match = substr($match,0,-1);
2149                                 }
2150
2151                                 // ignore strictly numeric tags like #1
2152                                 if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
2153                                         continue;
2154                                 }
2155
2156                                 // try not to catch url fragments
2157                                 if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2158                                         continue;
2159                                 }
2160                                 $ret[] = $match;
2161                         }
2162                 }
2163
2164                 return $ret;
2165         }
2166 }