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         /**
  52          * Fetches attachment data that were generated the old way
  53          *
  54          * @param string $body Message body
  55          * @return array
  56          *                     'type' -> Message type ('link', 'video', 'photo')
  57          *                     'text' -> Text before the shared message
  58          *                     'after' -> Text after the shared message
  59          *                     'image' -> Preview image of the message
  60          *                     'url' -> Url to the attached message
  61          *                     'title' -> Title of the attachment
  62          *                     'description' -> Description of the attachment
  63          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  64          */
  65         private static function getOldAttachmentData($body)
  66         {
  67                 $post = [];
  68
  69                 // Simplify image codes
  70                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  71
  72                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  73                         foreach ($attached as $data) {
  74                                 if (!in_array($data[1], ['type-link', 'type-video', 'type-photo'])) {
  75                                         continue;
  76                                 }
  77
  78                                 $post['type'] = substr($data[1], 5);
  79
  80                                 $pos = strpos($body, $data[0]);
  81                                 if ($pos > 0) {
  82                                         $post['text'] = trim(substr($body, 0, $pos));
  83                                         $post['after'] = trim(substr($body, $pos + strlen($data[0])));
  84                                 } else {
  85                                         $post['text'] = trim(str_replace($data[0], '', $body));
  86                                         $post['after'] = '';
  87                                 }
  88
  89                                 $attacheddata = $data[2];
  90
  91                                 if (preg_match("/\[img\](.*?)\[\/img\]/ism", $attacheddata, $matches)) {
  92
  93                                         $picturedata = Images::getInfoFromURLCached($matches[1]);
  94
  95                                         if ($picturedata) {
  96                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
  97                                                         $post['image'] = $matches[1];
  98                                                 } else {
  99                                                         $post['preview'] = $matches[1];
 100                                                 }
 101                                         }
 102                                 }
 103
 104                                 if (preg_match("/\[bookmark\=(.*?)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
 105                                         $post['url'] = $matches[1];
 106                                         $post['title'] = $matches[2];
 107                                 }
 108                                 if (!empty($post['url']) && (in_array($post['type'], ['link', 'video']))
 109                                         && preg_match("/\[url\=(.*?)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
 110                                         $post['url'] = $matches[1];
 111                                 }
 112
 113                                 // Search for description
 114                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
 115                                         $post['description'] = $matches[1];
 116                                 }
 117                         }
 118                 }
 119                 return $post;
 120         }
 121
 122         /**
 123          * Fetches attachment data that were generated with the "attachment" element
 124          *
 125          * @param string $body Message body
 126          * @return array
 127          *                     'type' -> Message type ('link', 'video', 'photo')
 128          *                     'text' -> Text before the shared message
 129          *                     'after' -> Text after the shared message
 130          *                     'image' -> Preview image of the message
 131          *                     'url' -> Url to the attached message
 132          *                     'title' -> Title of the attachment
 133          *                     'description' -> Description of the attachment
 134          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 135          */
 136         public static function getAttachmentData($body)
 137         {
 138                 $data = [
 139                         'type'        => '',
 140                         'text'        => '',
 141                         'after'       => '',
 142                         'image'       => null,
 143                         'url'         => '',
 144                         'title'       => '',
 145                         'description' => '',
 146                 ];
 147
 148                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 149                         return self::getOldAttachmentData($body);
 150                 }
 151
 152                 $attributes = $match[2];
 153
 154                 $data['text'] = trim($match[1]);
 155
 156                 $type = '';
 157                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 158                 if (!empty($matches[1])) {
 159                         $type = strtolower($matches[1]);
 160                 }
 161
 162                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 163                 if (!empty($matches[1])) {
 164                         $type = strtolower($matches[1]);
 165                 }
 166
 167                 if ($type == '') {
 168                         return [];
 169                 }
 170
 171                 if (!in_array($type, ['link', 'audio', 'photo', 'video'])) {
 172                         return [];
 173                 }
 174
 175                 if ($type != '') {
 176                         $data['type'] = $type;
 177                 }
 178
 179                 $url = '';
 180                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 181                 if (!empty($matches[1])) {
 182                         $url = $matches[1];
 183                 }
 184
 185                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 186                 if (!empty($matches[1])) {
 187                         $url = $matches[1];
 188                 }
 189
 190                 if ($url != '') {
 191                         $data['url'] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 192                 }
 193
 194                 $title = '';
 195                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 196                 if (!empty($matches[1])) {
 197                         $title = $matches[1];
 198                 }
 199
 200                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 201                 if (!empty($matches[1])) {
 202                         $title = $matches[1];
 203                 }
 204
 205                 if ($title != '') {
 206                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 207                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 208                         $title = str_replace(['[', ']'], ['&#91;', '&#93;'], $title);
 209                         $data['title'] = $title;
 210                 }
 211
 212                 $image = '';
 213                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 214                 if (!empty($matches[1])) {
 215                         $image = $matches[1];
 216                 }
 217
 218                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 219                 if (!empty($matches[1])) {
 220                         $image = $matches[1];
 221                 }
 222
 223                 if ($image != '') {
 224                         $data['image'] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 225                 }
 226
 227                 $preview = '';
 228                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 229                 if (!empty($matches[1])) {
 230                         $preview = $matches[1];
 231                 }
 232
 233                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 234                 if (!empty($matches[1])) {
 235                         $preview = $matches[1];
 236                 }
 237
 238                 if ($preview != '') {
 239                         $data['preview'] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 240                 }
 241
 242                 $data['description'] = trim($match[3]);
 243
 244                 $data['after'] = trim($match[4]);
 245
 246                 return $data;
 247         }
 248
 249         public static function getAttachedData($body, $item = [])
 250         {
 251                 /*
 252                 - text:
 253                 - type: link, video, photo
 254                 - title:
 255                 - url:
 256                 - image:
 257                 - description:
 258                 - (thumbnail)
 259                 */
 260
 261                 $has_title = !empty($item['title']);
 262                 $plink = $item['plink'] ?? '';
 263                 $post = self::getAttachmentData($body);
 264
 265                 // Get all linked images with alternative image description
 266                 if (preg_match_all("/\[img=([^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 267                         foreach ($pictures as $picture) {
 268                                 if (Photo::isLocal($picture[1])) {
 269                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]];
 270                                 }
 271                         }
 272                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 273                                 $post['image_description'] = $post['images'][0]['description'];
 274                         }
 275                 }
 276
 277                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 278                         foreach ($pictures as $picture) {
 279                                 if (Photo::isLocal($picture[1])) {
 280                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
 281                                 }
 282                         }
 283                 }
 284
 285                 // if nothing is found, it maybe having an image.
 286                 if (!isset($post['type'])) {
 287                         // Simplify image codes
 288                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 289                         $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 290                         $post['text'] = $body;
 291
 292                         if (preg_match_all("(\[url=(.*?)\]\s*\[img\](.*?)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 293                                 if ((count($pictures) == 1) && !$has_title) {
 294                                         if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) {
 295                                                 // Replace the preview picture with the real picture
 296                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
 297                                                 $data = ['url' => $url, 'type' => 'photo'];
 298                                         } else {
 299                                                 // Checking, if the link goes to a picture
 300                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 301                                         }
 302
 303                                         // Workaround:
 304                                         // Sometimes photo posts to the own album are not detected at the start.
 305                                         // So we seem to cannot use the cache for these cases. That's strange.
 306                                         if (($data['type'] != 'photo') && strstr($pictures[0][1], "/photos/")) {
 307                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 308                                         }
 309
 310                                         if ($data['type'] == 'photo') {
 311                                                 $post['type'] = 'photo';
 312                                                 if (isset($data['images'][0])) {
 313                                                         $post['image'] = $data['images'][0]['src'];
 314                                                         $post['url'] = $data['url'];
 315                                                 } else {
 316                                                         $post['image'] = $data['url'];
 317                                                 }
 318
 319                                                 $post['preview'] = $pictures[0][2];
 320                                                 $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 321                                         } else {
 322                                                 $imgdata = Images::getInfoFromURLCached($pictures[0][1]);
 323                                                 if ($imgdata && substr($imgdata['mime'], 0, 6) == 'image/') {
 324                                                         $post['type'] = 'photo';
 325                                                         $post['image'] = $pictures[0][1];
 326                                                         $post['preview'] = $pictures[0][2];
 327                                                         $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 328                                                 }
 329                                         }
 330                                 } elseif (count($pictures) > 0) {
 331                                         $post['type'] = 'link';
 332                                         $post['url'] = $plink;
 333                                         $post['image'] = $pictures[0][2];
 334                                         $post['text'] = $body;
 335
 336                                         foreach ($pictures as $picture) {
 337                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 338                                         }
 339                                 }
 340                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 341                                 if ((count($pictures) == 1) && !$has_title) {
 342                                         $post['type'] = 'photo';
 343                                         $post['image'] = $pictures[0][1];
 344                                         $post['text'] = str_replace($pictures[0][0], '', $body);
 345                                 } elseif (count($pictures) > 0) {
 346                                         $post['type'] = 'link';
 347                                         $post['url'] = $plink;
 348                                         $post['image'] = $pictures[0][1];
 349                                         $post['text'] = $body;
 350
 351                                         foreach ($pictures as $picture) {
 352                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 353                                         }
 354                                 }
 355                         }
 356
 357                         // Test for the external links
 358                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER);
 359                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER);
 360
 361                         $links = array_merge($links1, $links2);
 362
 363                         // If there is only a single one, then use it.
 364                         // This should cover link posts via API.
 365                         if ((count($links) == 1) && !isset($post['preview']) && !$has_title) {
 366                                 $post['type'] = 'link';
 367                                 $post['url'] = $links[0][1];
 368                         }
 369
 370                         // Simplify "video" element
 371                         $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']);
 372
 373                         // Now count the number of external media links
 374                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
 375                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
 376                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
 377                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
 378
 379                         // Add them to the other external links
 380                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 381
 382                         // Are there more than one?
 383                         if (count($links) > 1) {
 384                                 // The post will be the type "text", which means a blog post
 385                                 unset($post['type']);
 386                                 $post['url'] = $plink;
 387                         }
 388
 389                         if (!isset($post['type'])) {
 390                                 $post['type'] = "text";
 391                                 $post['text'] = trim($body);
 392                         }
 393                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
 394                         $data = ParseUrl::getSiteinfoCached($post['url'], true);
 395
 396                         if (isset($data['images'][0])) {
 397                                 $post['image'] = $data['images'][0]['src'];
 398                         }
 399                 }
 400
 401                 return $post;
 402         }
 403
 404         /**
 405          * Remove [attachment] BBCode and replaces it with a regular [url]
 406          *
 407          * @param string  $body
 408          * @param boolean $no_link_desc No link description
 409          *
 410          * @return string with replaced body
 411          */
 412         public static function removeAttachment($body, $no_link_desc = false)
 413         {
 414                 return preg_replace_callback("/\s*\[attachment (.*)\](.*?)\[\/attachment\]\s*/ism",
 415                         function ($match) use ($no_link_desc) {
 416                                 $attach_data = self::getAttachmentData($match[0]);
 417                                 if (empty($attach_data['url'])) {
 418                                         return $match[0];
 419                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
 420                                         return "\n[url]" . $attach_data['url'] . "[/url]\n";
 421                                 } else {
 422                                         return "\n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
 423                                 }
 424                 }, $body);
 425         }
 426
 427         /**
 428          * Converts a BBCode text into plaintext
 429          *
 430          * @param      $text
 431          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 432          *
 433          * @return string
 434          */
 435         public static function toPlaintext($text, $keep_urls = true)
 436         {
 437                 $naked_text = HTML::toPlaintext(BBCode::convert($text, false, 0, true), 0, !$keep_urls);
 438
 439                 return $naked_text;
 440         }
 441
 442         private static function proxyUrl($image, $simplehtml = false)
 443         {
 444                 // Only send proxied pictures to API and for internal display
 445                 if (in_array($simplehtml, [false, 2])) {
 446                         return ProxyUtils::proxifyUrl($image);
 447                 } else {
 448                         return $image;
 449                 }
 450         }
 451
 452         public static function scaleExternalImages($srctext)
 453         {
 454                 $s = $srctext;
 455
 456                 $matches = null;
 457                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 458                 if ($c) {
 459                         foreach ($matches as $mtch) {
 460                                 Logger::log('scale_external_image: ' . $mtch[1]);
 461
 462                                 $hostname = str_replace('www.', '', substr(DI::baseUrl(), strpos(DI::baseUrl(), '://') + 3));
 463                                 if (stristr($mtch[1], $hostname)) {
 464                                         continue;
 465                                 }
 466
 467                                 $i = Network::fetchUrl($mtch[1]);
 468                                 if (!$i) {
 469                                         return $srctext;
 470                                 }
 471
 472                                 // guess mimetype from headers or filename
 473                                 $type = Images::guessType($mtch[1], true);
 474
 475                                 if ($i) {
 476                                         $Image = new Image($i, $type);
 477                                         if ($Image->isValid()) {
 478                                                 $orig_width = $Image->getWidth();
 479                                                 $orig_height = $Image->getHeight();
 480
 481                                                 if ($orig_width > 640 || $orig_height > 640) {
 482                                                         $Image->scaleDown(640);
 483                                                         $new_width = $Image->getWidth();
 484                                                         $new_height = $Image->getHeight();
 485                                                         Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG);
 486                                                         $s = str_replace(
 487                                                                 $mtch[0],
 488                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]'
 489                                                                 . "\n",
 490                                                                 $s
 491                                                         );
 492                                                         Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG);
 493                                                 }
 494                                         }
 495                                 }
 496                         }
 497                 }
 498
 499                 return $s;
 500         }
 501
 502         /**
 503          * Truncates imported message body string length to max_import_size
 504          *
 505          * The purpose of this function is to apply system message length limits to
 506          * imported messages without including any embedded photos in the length
 507          *
 508          * @param string $body
 509          * @return string
 510          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 511          */
 512         public static function limitBodySize($body)
 513         {
 514                 $maxlen = DI::config()->get('config', 'max_import_size', 0);
 515
 516                 // If the length of the body, including the embedded images, is smaller
 517                 // than the maximum, then don't waste time looking for the images
 518                 if ($maxlen && (strlen($body) > $maxlen)) {
 519
 520                         Logger::log('the total body length exceeds the limit', Logger::DEBUG);
 521
 522                         $orig_body = $body;
 523                         $new_body = '';
 524                         $textlen = 0;
 525
 526                         $img_start = strpos($orig_body, '[img');
 527                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 528                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 529                         while (($img_st_close !== false) && ($img_end !== false)) {
 530
 531                                 $img_st_close++; // make it point to AFTER the closing bracket
 532                                 $img_end += $img_start;
 533                                 $img_end += strlen('[/img]');
 534
 535                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 536                                         // This is an embedded image
 537
 538                                         if (($textlen + $img_start) > $maxlen) {
 539                                                 if ($textlen < $maxlen) {
 540                                                         Logger::log('the limit happens before an embedded image', Logger::DEBUG);
 541                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 542                                                         $textlen = $maxlen;
 543                                                 }
 544                                         } else {
 545                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 546                                                 $textlen += $img_start;
 547                                         }
 548
 549                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 550                                 } else {
 551
 552                                         if (($textlen + $img_end) > $maxlen) {
 553                                                 if ($textlen < $maxlen) {
 554                                                         Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG);
 555                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 556                                                         $textlen = $maxlen;
 557                                                 }
 558                                         } else {
 559                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 560                                                 $textlen += $img_end;
 561                                         }
 562                                 }
 563                                 $orig_body = substr($orig_body, $img_end);
 564
 565                                 if ($orig_body === false) {
 566                                         // in case the body ends on a closing image tag
 567                                         $orig_body = '';
 568                                 }
 569
 570                                 $img_start = strpos($orig_body, '[img');
 571                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 572                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 573                         }
 574
 575                         if (($textlen + strlen($orig_body)) > $maxlen) {
 576                                 if ($textlen < $maxlen) {
 577                                         Logger::log('the limit happens after the end of the last image', Logger::DEBUG);
 578                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 579                                 }
 580                         } else {
 581                                 Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG);
 582                                 $new_body = $new_body . $orig_body;
 583                         }
 584
 585                         return $new_body;
 586                 } else {
 587                         return $body;
 588                 }
 589         }
 590
 591         /**
 592          * Processes [attachment] tags
 593          *
 594          * Note: Can produce a [bookmark] tag in the returned string
 595          *
 596          * @param string   $text
 597          * @param bool|int $simplehtml
 598          * @param bool     $tryoembed
 599          * @return string
 600          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 601          */
 602         private static function convertAttachment($text, $simplehtml = false, $tryoembed = true)
 603         {
 604                 $data = self::getAttachmentData($text);
 605                 if (empty($data) || empty($data['url'])) {
 606                         return $text;
 607                 }
 608
 609                 if (isset($data['title'])) {
 610                         $data['title'] = strip_tags($data['title']);
 611                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 612                 } else {
 613                         $data['title'] = null;
 614                 }
 615
 616                 if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) {
 617                         $data['preview'] = $data['image'];
 618                         $data['image'] = '';
 619                 }
 620
 621                 $return = '';
 622                 try {
 623                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 624                                 $return = OEmbed::getHTML($data['url'], $data['title']);
 625                         } else {
 626                                 throw new Exception('OEmbed is disabled for this attachment.');
 627                         }
 628                 } catch (Exception $e) {
 629                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 630
 631                         if ($simplehtml != 4) {
 632                                 $return = sprintf('<div class="type-%s">', $data['type']);
 633                         }
 634
 635                         if (!empty($data['title']) && !empty($data['url'])) {
 636                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 637                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data['url'], self::proxyUrl($data['image'], $simplehtml), $data['title']);
 638                                 } else {
 639                                         if (!empty($data['image'])) {
 640                                                 $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a><br />', $data['url'], self::proxyUrl($data['image'], $simplehtml), $data['title']);
 641                                         } elseif (!empty($data['preview'])) {
 642                                                 $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br />', $data['url'], self::proxyUrl($data['preview'], $simplehtml), $data['title']);
 643                                         }
 644                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 645                                 }
 646                         }
 647
 648                         if (!empty($data['description']) && $data['description'] != $data['title']) {
 649                                 // Sanitize the HTML by converting it to BBCode
 650                                 $bbcode = HTML::toBBCode($data['description']);
 651                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 652                         }
 653
 654                         if (!empty($data['url'])) {
 655                                 $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 656                         }
 657
 658                         if ($simplehtml != 4) {
 659                                 $return .= '</div>';
 660                         }
 661                 }
 662
 663                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 664         }
 665
 666         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 667         {
 668                 $data = self::getAttachmentData($Text);
 669
 670                 if (!$data) {
 671                         return $Text;
 672                 } elseif ($nolink) {
 673                         return $data['text'] . ($data['after'] ?? '');
 674                 }
 675
 676                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 677                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 678                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 679                         $data['title'] = $data['url'];
 680                 } elseif (($text != '') && strstr($title, $text)) {
 681                         $data['text'] = $data['title'];
 682                         $data['title'] = $data['url'];
 683                 }
 684
 685                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 686                         return $data['title'] . $data['after'];
 687                 }
 688
 689                 // If the link already is included in the post, don't add it again
 690                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 691                         return $data['text'] . $data['after'];
 692                 }
 693
 694                 $text = $data['text'];
 695
 696                 if (!empty($data['url']) && !empty($data['title'])) {
 697                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 698                 } elseif (!empty($data['url'])) {
 699                         $text .= "\n[url]" . $data['url'] . '[/url]';
 700                 }
 701
 702                 return $text . "\n" . $data['after'];
 703         }
 704
 705         /**
 706          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 707          *
 708          * @param array $match Array with the matching values
 709          * @return string reformatted link including HTML codes
 710          */
 711         private static function convertUrlForActivityPubCallback($match)
 712         {
 713                 $url = $match[1];
 714
 715                 if (isset($match[2]) && ($match[1] != $match[2])) {
 716                         return $match[0];
 717                 }
 718
 719                 $parts = parse_url($url);
 720                 if (!isset($parts['scheme'])) {
 721                         return $match[0];
 722                 }
 723
 724                 return self::convertUrlForActivityPub($url);
 725         }
 726
 727         /**
 728          * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 729          *
 730          * @param string $url URL that is about to be reformatted
 731          * @return string reformatted link including HTML codes
 732          */
 733         private static function convertUrlForActivityPub($url)
 734         {
 735                 $html = '<a href="%s" target="_blank">%s</a>';
 736                 return sprintf($html, $url, self::getStyledURL($url));
 737         }
 738
 739         /**
 740          * Converts an URL in a nicer format (without the scheme and possibly shortened)
 741          *
 742          * @param string $url URL that is about to be reformatted
 743          * @return string reformatted link
 744          */
 745         private static function getStyledURL($url)
 746         {
 747                 $parts = parse_url($url);
 748                 $scheme = $parts['scheme'] . '://';
 749                 $styled_url = str_replace($scheme, '', $url);
 750
 751                 if (strlen($styled_url) > 30) {
 752                         $styled_url = substr($styled_url, 0, 30) . "…";
 753                 }
 754
 755                 return $styled_url;
 756         }
 757
 758         /*
 759          * [noparse][i]italic[/i][/noparse] turns into
 760          * [noparse][ i ]italic[ /i ][/noparse],
 761          * to hide them from parser.
 762          */
 763         private static function escapeNoparseCallback($match)
 764         {
 765                 $whole_match = $match[0];
 766                 $captured = $match[1];
 767                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 768                 $new_str = str_replace($captured, $spacefied, $whole_match);
 769                 return $new_str;
 770         }
 771
 772         /*
 773          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 774          * now turns back and the [noparse] tags are trimed
 775          * returning [i]italic[/i]
 776          */
 777         private static function unescapeNoparseCallback($match)
 778         {
 779                 $captured = $match[1];
 780                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 781                 return $unspacefied;
 782         }
 783
 784         /**
 785          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 786          * occurrences
 787          *
 788          * @param string $text        Text to search
 789          * @param string $name        Tag name
 790          * @param int    $occurrences Number of first occurrences to skip
 791          * @return boolean|array
 792          */
 793         public static function getTagPosition($text, $name, $occurrences = 0)
 794         {
 795                 if ($occurrences < 0) {
 796                         $occurrences = 0;
 797                 }
 798
 799                 $start_open = -1;
 800                 for ($i = 0; $i <= $occurrences; $i++) {
 801                         if ($start_open !== false) {
 802                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 803                         }
 804                 }
 805
 806                 if ($start_open === false) {
 807                         return false;
 808                 }
 809
 810                 $start_equal = strpos($text, '=', $start_open);
 811                 $start_close = strpos($text, ']', $start_open);
 812
 813                 if ($start_close === false) {
 814                         return false;
 815                 }
 816
 817                 $start_close++;
 818
 819                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 820
 821                 if ($end_open === false) {
 822                         return false;
 823                 }
 824
 825                 $res = [
 826                         'start' => [
 827                                 'open' => $start_open,
 828                                 'close' => $start_close
 829                         ],
 830                         'end' => [
 831                                 'open' => $end_open,
 832                                 'close' => $end_open + strlen('[/' . $name . ']')
 833                         ],
 834                 ];
 835
 836                 if ($start_equal !== false) {
 837                         $res['start']['equal'] = $start_equal + 1;
 838                 }
 839
 840                 return $res;
 841         }
 842
 843         /**
 844          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 845          *
 846          * @param string $pattern Preg pattern string
 847          * @param string $replace Preg replace string
 848          * @param string $name    BBCode tag name
 849          * @param string $text    Text to search
 850          * @return string
 851          */
 852         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 853         {
 854                 $occurrences = 0;
 855                 $pos = self::getTagPosition($text, $name, $occurrences);
 856                 while ($pos !== false && $occurrences++ < 1000) {
 857                         $start = substr($text, 0, $pos['start']['open']);
 858                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 859                         $end = substr($text, $pos['end']['close']);
 860                         if ($end === false) {
 861                                 $end = '';
 862                         }
 863
 864                         $subject = preg_replace($pattern, $replace, $subject);
 865                         $text = $start . $subject . $end;
 866
 867                         $pos = self::getTagPosition($text, $name, $occurrences);
 868                 }
 869
 870                 return $text;
 871         }
 872
 873         private static function extractImagesFromItemBody($body)
 874         {
 875                 $saved_image = [];
 876                 $orig_body = $body;
 877                 $new_body = '';
 878
 879                 $cnt = 0;
 880                 $img_start = strpos($orig_body, '[img');
 881                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 882                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 883                 while (($img_st_close !== false) && ($img_end !== false)) {
 884                         $img_st_close++; // make it point to AFTER the closing bracket
 885                         $img_end += $img_start;
 886
 887                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 888                                 // This is an embedded image
 889                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 890                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 891
 892                                 $cnt++;
 893                         } else {
 894                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 895                         }
 896
 897                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 898
 899                         if ($orig_body === false) {
 900                                 // in case the body ends on a closing image tag
 901                                 $orig_body = '';
 902                         }
 903
 904                         $img_start = strpos($orig_body, '[img');
 905                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 906                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 907                 }
 908
 909                 $new_body = $new_body . $orig_body;
 910
 911                 return ['body' => $new_body, 'images' => $saved_image];
 912         }
 913
 914         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 915         {
 916                 $newbody = $body;
 917
 918                 $cnt = 0;
 919                 foreach ($images as $image) {
 920                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 921                         // it loops over the array starting from the first element and going sequentially
 922                         // to the last element
 923                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 924                                 '<img src="' . self::proxyUrl($image) . '" alt="' . DI::l10n()->t('Image/photo') . '" />', $newbody);
 925                         $cnt++;
 926                 }
 927
 928                 return $newbody;
 929         }
 930
 931         /**
 932          * This function converts a [share] block to text according to a provided callback function whose signature is:
 933          *
 934          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 935          *
 936          * Where:
 937          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 938          * data lookup
 939          * - $author_contact is a contact record array
 940          * - $content is the inner content of the [share] block
 941          * - $is_quote_share indicates whether there's any content before the [share] block
 942          * - Return value is the string that should replace the [share] block in the provided text
 943          *
 944          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 945          *
 946          * @param  string   $text     A BBCode string
 947          * @param  callable $callback
 948          * @return string The BBCode string with all [share] blocks replaced
 949          */
 950         public static function convertShare($text, callable $callback)
 951         {
 952                 $return = preg_replace_callback(
 953                         "/(.*?)\[share(.*?)\](.*)\[\/share\]/ism",
 954                         function ($match) use ($callback) {
 955                                 $attribute_string = $match[2];
 956                                 $attributes = [];
 957                                 foreach (['author', 'profile', 'avatar', 'link', 'posted'] as $field) {
 958                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 959                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 960                                 }
 961
 962                                 // We only call this so that a previously unknown contact can be added.
 963                                 // This is important for the function "Model\Contact::getDetailsByURL()".
 964                                 // This function then can fetch an entry from the contact table.
 965                                 $default['url'] = $attributes['profile'];
 966
 967                                 if (!empty($attributes['author'])) {
 968                                         $default['name'] = $attributes['author'];
 969                                 }
 970
 971                                 if (!empty($attributes['avatar'])) {
 972                                         $default['photo'] = $attributes['avatar'];
 973                                 }
 974
 975                                 Contact::getIdForURL($attributes['profile'], 0, true, $default);
 976
 977                                 $author_contact = Contact::getDetailsByURL($attributes['profile']);
 978                                 $author_contact['addr'] = ($author_contact['addr']  ?? '') ?: Protocol::getAddrFromProfileUrl($attributes['profile']);
 979
 980                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
 981                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
 982                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
 983
 984                                 if ($attributes['avatar']) {
 985                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
 986                                 }
 987
 988                                 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
 989                         },
 990                         $text
 991                 );
 992
 993                 return $return;
 994         }
 995
 996         /**
 997          * Default [share] tag conversion callback
 998          *
 999          * Note: Can produce a [bookmark] tag in the output
1000          *
1001          * @see BBCode::convertShare()
1002          * @param array   $attributes     [share] block attribute values
1003          * @param array   $author_contact Contact row of the shared author
1004          * @param string  $content        Inner content of the [share] block
1005          * @param boolean $is_quote_share Whether there is content before the [share] block
1006          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
1007          * @return string
1008          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1009          */
1010         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
1011         {
1012                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
1013
1014                 switch ($simplehtml) {
1015                         case 1:
1016                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' <a href="' . $attributes['profile'] . '">' . $mention . '</a>: </p>' . "\n" . '«' . $content . '»';
1017                                 break;
1018                         case 2:
1019                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1020                                 break;
1021                         case 3: // Diaspora
1022                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1023                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1024                                 } else {
1025                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1026
1027                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1028                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1029                                         }
1030
1031                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1032
1033                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1034                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1035                                         }
1036                                 }
1037
1038                                 break;
1039                         case 4:
1040                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1041                                 $headline .= DI::l10n()->t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1042                                 $headline .= ':</b></p>' . "\n";
1043
1044                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1045
1046                                 break;
1047                         case 5:
1048                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1049                                 break;
1050                         case 7: // statusnet/GNU Social
1051                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1052                                 break;
1053                         case 9: // ActivityPub
1054                                 $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>';
1055                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1056                                 break;
1057                         default:
1058                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1059                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($attributes['link'])) {
1060                                         try {
1061                                                 $text = ($is_quote_share? '<br />' : '') . OEmbed::getHTML($attributes['link']);
1062                                         } catch (Exception $e) {
1063                                                 $text = ($is_quote_share? '<br />' : '') . sprintf('[bookmark=%s]%s[/bookmark]', $attributes['link'], $content);
1064                                         }
1065                                 } else {
1066                                         $text = ($is_quote_share? "\n" : '');
1067
1068                                         $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1069                                         $text .= Renderer::replaceMacros($tpl, [
1070                                                 '$profile' => $attributes['profile'],
1071                                                 '$avatar'  => $attributes['avatar'],
1072                                                 '$author'  => $attributes['author'],
1073                                                 '$link'    => $attributes['link'],
1074                                                 '$posted'  => $attributes['posted'],
1075                                                 '$content' => trim($content)
1076                                         ]);
1077                                 }
1078                                 break;
1079                 }
1080
1081                 return $text;
1082         }
1083
1084         private static function removePictureLinksCallback($match)
1085         {
1086                 $cache_key = 'remove:' . $match[1];
1087                 $text = DI::cache()->get($cache_key);
1088
1089                 if (is_null($text)) {
1090                         $a = DI::app();
1091
1092                         $stamp1 = microtime(true);
1093
1094                         $ch = @curl_init($match[1]);
1095                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1096                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1097                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1098                         @curl_exec($ch);
1099                         $curl_info = @curl_getinfo($ch);
1100
1101                         DI::profiler()->saveTimestamp($stamp1, "network", System::callstack());
1102
1103                         if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1104                                 $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]";
1105                         } else {
1106                                 $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]";
1107
1108                                 // if its not a picture then look if its a page that contains a picture link
1109                                 $body = Network::fetchUrl($match[1]);
1110
1111                                 $doc = new DOMDocument();
1112                                 @$doc->loadHTML($body);
1113                                 $xpath = new DOMXPath($doc);
1114                                 $list = $xpath->query("//meta[@name]");
1115                                 foreach ($list as $node) {
1116                                         $attr = [];
1117
1118                                         if ($node->attributes->length) {
1119                                                 foreach ($node->attributes as $attribute) {
1120                                                         $attr[$attribute->name] = $attribute->value;
1121                                                 }
1122                                         }
1123
1124                                         if (strtolower($attr['name']) == 'twitter:image') {
1125                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1126                                         }
1127                                 }
1128                         }
1129                         DI::cache()->set($cache_key, $text);
1130                 }
1131
1132                 return $text;
1133         }
1134
1135         private static function expandLinksCallback($match)
1136         {
1137                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1138                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1139                 } else {
1140                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1141                 }
1142         }
1143
1144         private static function cleanPictureLinksCallback($match)
1145         {
1146                 $a = DI::app();
1147
1148                 // When the picture link is the own photo path then we can avoid fetching the link
1149                 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()->get()) . '/photos/');
1150                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1151                         if (!empty($match[3])) {
1152                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1153                         } else {
1154                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1155                         }
1156                         return $text;
1157                 }
1158
1159                 $cache_key = 'clean:' . $match[1];
1160                 $text = DI::cache()->get($cache_key);
1161                 if (!is_null($text)) {
1162                         return $text;
1163                 }
1164
1165                 // Only fetch the header, not the content
1166                 $stamp1 = microtime(true);
1167
1168                 $ch = @curl_init($match[1]);
1169                 @curl_setopt($ch, CURLOPT_NOBODY, true);
1170                 @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1171                 @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1172                 @curl_exec($ch);
1173                 $curl_info = @curl_getinfo($ch);
1174
1175                 DI::profiler()->saveTimestamp($stamp1, "network", System::callstack());
1176
1177                 // if its a link to a picture then embed this picture
1178                 if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1179                         $text = '[img]' . $match[1] . '[/img]';
1180                 } else {
1181                         if (!empty($match[3])) {
1182                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1183                         } else {
1184                                 $text = '[img]' . $match[2] . '[/img]';
1185                         }
1186
1187                         // if its not a picture then look if its a page that contains a picture link
1188                         $body = Network::fetchUrl($match[1]);
1189
1190                         $doc = new DOMDocument();
1191                         @$doc->loadHTML($body);
1192                         $xpath = new DOMXPath($doc);
1193                         $list = $xpath->query("//meta[@name]");
1194                         foreach ($list as $node) {
1195                                 $attr = [];
1196                                 if ($node->attributes->length) {
1197                                         foreach ($node->attributes as $attribute) {
1198                                                 $attr[$attribute->name] = $attribute->value;
1199                                         }
1200                                 }
1201
1202                                 if (strtolower($attr['name']) == "twitter:image") {
1203                                         if (!empty($match[3])) {
1204                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1205                                         } else {
1206                                                 $text = "[img]" . $attr['content'] . "[/img]";
1207                                         }
1208                                 }
1209                         }
1210                 }
1211                 DI::cache()->set($cache_key, $text);
1212
1213                 return $text;
1214         }
1215
1216         public static function cleanPictureLinks($text)
1217         {
1218                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1219                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1220                 return $return;
1221         }
1222
1223         /**
1224          * Converts a BBCode message to HTML message
1225          *
1226          * BBcode 2 HTML was written by WAY2WEB.net
1227          * extended to work with Mistpark/Friendica - Mike Macgirvin
1228          *
1229          * Simple HTML values meaning:
1230          * - 0: Friendica display
1231          * - 1: Unused
1232          * - 2: Used for Windows Phone push, Friendica API
1233          * - 3: Used before converting to Markdown in bb2diaspora.php
1234          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1235          * - 5: Unused
1236          * - 6: Unused
1237          * - 7: Used for dfrn, OStatus
1238          * - 8: Used for Twitter, WP backlink text setting
1239          * - 9: ActivityPub
1240          *
1241          * @param string $text
1242          * @param bool   $try_oembed
1243          * @param int    $simple_html
1244          * @param bool   $for_plaintext
1245          * @return string
1246          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1247          */
1248         public static function convert($text, $try_oembed = true, $simple_html = 0, $for_plaintext = false)
1249         {
1250                 $a = DI::app();
1251
1252                 /*
1253                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1254                  *
1255                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1256                  * $match[1] = $url
1257                  * $match[2] = $title or absent
1258                  */
1259                 $try_oembed_callback = function ($match)
1260                 {
1261                         $url = $match[1];
1262                         $title = $match[2] ?? null;
1263
1264                         try {
1265                                 $return = OEmbed::getHTML($url, $title);
1266                         } catch (Exception $ex) {
1267                                 $return = $match[0];
1268                         }
1269
1270                         return $return;
1271                 };
1272
1273                 // Extracting code blocks before the whitespace processing and the autolinker
1274                 $codeblocks = [];
1275
1276                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1277                         function ($matches) use (&$codeblocks) {
1278                                 $return = '#codeblock-' . count($codeblocks) . '#';
1279                                 if (strpos($matches[2], "\n") !== false) {
1280                                         $codeblocks[] = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlspecialchars(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
1281                                 } else {
1282                                         $codeblocks[] = '<code>' . htmlspecialchars($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
1283                                 }
1284
1285                                 return $return;
1286                         },
1287                         $text
1288                 );
1289
1290                 // Hide all [noparse] contained bbtags by spacefying them
1291                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1292
1293                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1294                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1295                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1296
1297                 // Remove the abstract element. It is a non visible element.
1298                 $text = self::stripAbstract($text);
1299
1300                 // Move all spaces out of the tags
1301                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1302                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1303
1304                 // Extract the private images which use data urls since preg has issues with
1305                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1306                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1307
1308                 $extracted = self::extractImagesFromItemBody($text);
1309                 $text = $extracted['body'];
1310                 $saved_image = $extracted['images'];
1311
1312                 // If we find any event code, turn it into an event.
1313                 // After we're finished processing the bbcode we'll
1314                 // replace all of the event code with a reformatted version.
1315
1316                 $ev = Event::fromBBCode($text);
1317
1318                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1319                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1320
1321                 $text = str_replace("<", "&lt;", $text);
1322                 $text = str_replace(">", "&gt;", $text);
1323
1324                 // remove some newlines before the general conversion
1325                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1326                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1327
1328                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1329                 if (!$try_oembed) {
1330                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1331                 }
1332
1333                 // Convert new line chars to html <br /> tags
1334
1335                 // nlbr seems to be hopelessly messed up
1336                 //      $Text = nl2br($Text);
1337
1338                 // We'll emulate it.
1339
1340                 $text = trim($text);
1341                 $text = str_replace("\r\n", "\n", $text);
1342
1343                 // Remove linefeeds inside of the table elements. See issue #6799
1344                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1345                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1346                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1347                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1348                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1349                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1350                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1351                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1352                 do {
1353                         $oldtext = $text;
1354                         $text = str_replace($search, $replace, $text);
1355                 } while ($oldtext != $text);
1356
1357                 // Replace these here only once
1358                 $search = ["\n[table]", "[/table]\n"];
1359                 $replace = ["[table]", "[/table]"];
1360                 $text = str_replace($search, $replace, $text);
1361
1362                 // removing multiplicated newlines
1363                 if (DI::config()->get('system', 'remove_multiplicated_lines')) {
1364                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1365                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1366                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1367                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1368                         do {
1369                                 $oldtext = $text;
1370                                 $text = str_replace($search, $replace, $text);
1371                         } while ($oldtext != $text);
1372                 }
1373
1374                 /// @todo Have a closer look at the different html modes
1375                 // Handle attached links or videos
1376                 if (in_array($simple_html, [9])) {
1377                         $text = self::removeAttachment($text);
1378                 } elseif (!in_array($simple_html, [0, 4])) {
1379                         $text = self::removeAttachment($text, true);
1380                 } else {
1381                         $text = self::convertAttachment($text, $simple_html, $try_oembed);
1382                 }
1383
1384                 // leave open the posibility of [map=something]
1385                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1386                 if (strpos($text, '[/map]') !== false) {
1387                         $text = preg_replace_callback(
1388                                 "/\[map\](.*?)\[\/map\]/ism",
1389                                 function ($match) use ($simple_html) {
1390                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1391                                 },
1392                                 $text
1393                         );
1394                 }
1395
1396                 if (strpos($text, '[map=') !== false) {
1397                         $text = preg_replace_callback(
1398                                 "/\[map=(.*?)\]/ism",
1399                                 function ($match) use ($simple_html) {
1400                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1401                                 },
1402                                 $text
1403                         );
1404                 }
1405
1406                 if (strpos($text, '[map]') !== false) {
1407                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1408                 }
1409
1410                 // Check for headers
1411                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1412                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1413                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1414                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1415                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1416                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1417
1418                 // Check for paragraph
1419                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1420
1421                 // Check for bold text
1422                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1423
1424                 // Check for Italics text
1425                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1426
1427                 // Check for Underline text
1428                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1429
1430                 // Check for strike-through text
1431                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1432
1433                 // Check for over-line text
1434                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1435
1436                 // Check for colored text
1437                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1438
1439                 // Check for sized text
1440                 // [size=50] --> font-size: 50px (with the unit).
1441                 if ($simple_html != 3) {
1442                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1443                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1444                 } else {
1445                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1446                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1447                 }
1448
1449
1450                 // Check for centered text
1451                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1452
1453                 // Check for list text
1454                 $text = str_replace("[*]", "<li>", $text);
1455
1456                 // Check for style sheet commands
1457                 $text = preg_replace_callback(
1458                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1459                         function ($match) {
1460                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1461                         },
1462                         $text
1463                 );
1464
1465                 // Check for CSS classes
1466                 $text = preg_replace_callback(
1467                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1468                         function ($match) {
1469                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1470                         },
1471                         $text
1472                 );
1473
1474                 // handle nested lists
1475                 $endlessloop = 0;
1476
1477                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1478                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1479                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1480                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1481                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1482                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1483                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1484                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1485                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1486                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1487                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1488                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1489                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1490                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1491                 }
1492
1493                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1494                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1495                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1496                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1497
1498                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1499                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1500
1501                 $text = str_replace('[hr]', '<hr />', $text);
1502
1503                 if (!$for_plaintext) {
1504                         $escaped = [];
1505
1506                         // Escaping BBCodes susceptible to contain rogue URL we don'' want the autolinker to catch
1507                         $text = preg_replace_callback('#\[(url|img|audio|video|youtube|vimeo|share|attachment|iframe|bookmark).+?\[/\1\]#ism',
1508                                 function ($matches) use (&$escaped) {
1509                                         $return = '{escaped-' . count($escaped) . '}';
1510                                         $escaped[] = $matches[0];
1511
1512                                         return $return;
1513                                 },
1514                                 $text
1515                         );
1516
1517                         // Autolinker for isolated URLs
1518                         $text = preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1519
1520                         // Restoring escaped blocks
1521                         $text = preg_replace_callback('/{escaped-([0-9]+)}/iU',
1522                                 function ($matches) use ($escaped) {
1523                                         return $escaped[intval($matches[1])] ?? $matches[0];
1524                                 },
1525                                 $text
1526                         );
1527                 }
1528
1529                 // This is actually executed in Item::prepareBody()
1530
1531                 $nosmile = strpos($text, '[nosmile]') !== false;
1532                 $text = str_replace('[nosmile]', '', $text);
1533
1534                 // Check for font change text
1535                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1536
1537                 // Declare the format for [spoiler] layout
1538                 $SpoilerLayout = '<details class="spoiler"><summary>' . DI::l10n()->t('Click to open/close') . '</summary>$1</details>';
1539
1540                 // Check for [spoiler] text
1541                 // handle nested quotes
1542                 $endlessloop = 0;
1543                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1544                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1545                 }
1546
1547                 // Check for [spoiler=Title] text
1548
1549                 // handle nested quotes
1550                 $endlessloop = 0;
1551                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1552                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1553                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1554                                 $text);
1555                 }
1556
1557                 // Declare the format for [quote] layout
1558                 $QuoteLayout = '<blockquote>$1</blockquote>';
1559
1560                 // Check for [quote] text
1561                 // handle nested quotes
1562                 $endlessloop = 0;
1563                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1564                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1565                 }
1566
1567                 // Check for [quote=Author] text
1568
1569                 $t_wrote = DI::l10n()->t('$1 wrote:');
1570
1571                 // handle nested quotes
1572                 $endlessloop = 0;
1573                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1574                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1575                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1576                                 $text);
1577                 }
1578
1579
1580                 // [img=widthxheight]image source[/img]
1581                 $text = preg_replace_callback(
1582                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1583                         function ($matches) use ($simple_html) {
1584                                 if (strpos($matches[3], "data:image/") === 0) {
1585                                         return $matches[0];
1586                                 }
1587
1588                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1589                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1590                         },
1591                         $text
1592                 );
1593
1594                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1595                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1596
1597                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1598                         function ($matches) use ($simple_html) {
1599                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1600                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1601                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1602                         },
1603                         $text);
1604
1605                 // Images
1606                 // [img]pathtoimage[/img]
1607                 $text = preg_replace_callback(
1608                         "/\[img\](.*?)\[\/img\]/ism",
1609                         function ($matches) use ($simple_html) {
1610                                 if (strpos($matches[1], "data:image/") === 0) {
1611                                         return $matches[0];
1612                                 }
1613
1614                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1615                                 return "[img]" . $matches[1] . "[/img]";
1616                         },
1617                         $text
1618                 );
1619
1620                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1621                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1622
1623                 $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);
1624                 $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 />', $text);
1625                 //$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);
1626
1627                 // Simplify "video" element
1628                 $text = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1629
1630                 // Try to Oembed
1631                 if ($try_oembed) {
1632                         $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);
1633                         $text = preg_replace("/\[audio\](.*?\.(ogg|ogv|oga|ogm|webm|mp4|mp3).*?)\[\/audio\]/ism", '<audio src="$1" controls="controls"><a href="$1">$1</a></audio>', $text);
1634
1635                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1636                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1637                 } else {
1638                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1639                                 '<a href="$1" target="_blank">$1</a>', $text);
1640                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1641                                 '<a href="$1" target="_blank">$1</a>', $text);
1642                 }
1643
1644                 // html5 video and audio
1645
1646
1647                 if ($try_oembed) {
1648                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1649                 } else {
1650                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1651                 }
1652
1653                 // Youtube extensions
1654                 if ($try_oembed) {
1655                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1656                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1657                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1658                 }
1659
1660                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1661                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1662                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1663
1664                 if ($try_oembed) {
1665                         $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);
1666                 } else {
1667                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1668                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1669                 }
1670
1671                 if ($try_oembed) {
1672                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1673                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1674                 }
1675
1676                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1677                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1678
1679                 if ($try_oembed) {
1680                         $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);
1681                 } else {
1682                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1683                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1684                 }
1685
1686                 // oembed tag
1687                 $text = OEmbed::BBCode2HTML($text);
1688
1689                 // Avoid triple linefeeds through oembed
1690                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1691
1692                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1693                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1694                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1695                 // start which is always required). Allow desc with a missing summary for compatibility.
1696
1697                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1698                         $sub = Event::getHTML($ev, $simple_html);
1699
1700                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1701                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1702                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1703                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1704                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1705                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1706                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1707                 }
1708
1709                 // Replace non graphical smilies for external posts
1710                 if (!$nosmile && !$for_plaintext) {
1711                         $text = Smilies::replace($text);
1712                 }
1713
1714                 if (!$for_plaintext) {
1715                         if (in_array($simple_html, [7, 9])) {
1716                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1717                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1718                         }
1719                 } else {
1720                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1721                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1722                 }
1723
1724                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1725
1726                 // Remove all hashtag addresses
1727                 if ($simple_html && !in_array($simple_html, [3, 7, 9])) {
1728                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1729                 } elseif ($simple_html == 3) {
1730                         // The ! is converted to @ since Diaspora only understands the @
1731                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1732                                 '@<a href="$2">$3</a>',
1733                                 $text);
1734                 } elseif (in_array($simple_html, [7, 9])) {
1735                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1736                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1737                                 $text);
1738                 } elseif (!$simple_html) {
1739                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1740                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1741                                 $text);
1742                 }
1743
1744                 // Bookmarks in red - will be converted to bookmarks in friendica
1745                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1746                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1747                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1748                                         "[bookmark=$1]$2[/bookmark]", $text);
1749
1750                 if (in_array($simple_html, [2, 6, 7, 8])) {
1751                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1752                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1753                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1754                 }
1755
1756                 if ($simple_html == 5) {
1757                         $text = preg_replace("/[^#@!]\[url\=(.*?)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1758                 }
1759
1760                 // Perform URL Search
1761                 if ($try_oembed) {
1762                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1763                 }
1764
1765                 if ($simple_html == 5) {
1766                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1767                 } else {
1768                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1769                 }
1770
1771                 // Handle Diaspora posts
1772                 $text = preg_replace_callback(
1773                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1774                         function ($match) {
1775                                 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1776                         }, $text
1777                 );
1778
1779                 $text = preg_replace_callback(
1780                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1781                         function ($match) {
1782                                 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1783                         }, $text
1784                 );
1785
1786                 // Server independent link to posts and comments
1787                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1788                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1789                 $text = preg_replace($expression, DI::baseUrl()."/display/$1", $text);
1790
1791                 /* Tag conversion
1792                  * Supports:
1793                  * - #[url=<anything>]<term>[/url]
1794                  * - [url=<anything>]#<term>[/url]
1795                  */
1796                 $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) {
1797                         return '#<a href="'
1798                                 . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1799                                 . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1800                                 . XML::escape($matches[1])
1801                                 . '</a>';
1802                 }, $text);
1803
1804                 // We need no target="_blank" for local links
1805                 // convert links start with DI::baseUrl() as local link without the target="_blank" attribute
1806                 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
1807                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1808                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1809
1810                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1811                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1812
1813                 // Red compatibility, though the link can't be authenticated on Friendica
1814                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1815
1816
1817                 // we may need to restrict this further if it picks up too many strays
1818                 // link acct:user@host to a webfinger profile redirector
1819
1820                 $text = preg_replace('/acct:([^@]+)@((?!\-)(?:[a-zA-Z\d\-]{0,62}[a-zA-Z\d]\.){1,126}(?!\d+)[a-zA-Z\d]{1,63})/', '<a href="' . DI::baseUrl() . '/acctlink?addr=$1@$2" target="extlink">acct:$1@$2</a>', $text);
1821
1822                 // Perform MAIL Search
1823                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1824                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1825
1826                 // Unhide all [noparse] contained bbtags unspacefying them
1827                 // and triming the [noparse] tag.
1828
1829                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1830                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1831                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1832
1833                 /// @todo What is the meaning of these lines?
1834                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1835                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1836
1837                 // Currently deactivated, it made problems with " inside of alt texts.
1838                 //$text = preg_replace('/\&quot\;/', '"', $text);
1839
1840                 // fix any escaped ampersands that may have been converted into links
1841                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1842
1843                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1844                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1845
1846                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1847
1848                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1849                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">', $text);
1850
1851                 // sanitize href attributes (only whitelisted protocols URLs)
1852                 // default value for backward compatibility
1853                 $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
1854
1855                 // Always allowed protocol even if config isn't set or not including it
1856                 $allowed_link_protocols[] = '//';
1857                 $allowed_link_protocols[] = 'http://';
1858                 $allowed_link_protocols[] = 'https://';
1859                 $allowed_link_protocols[] = 'redir/';
1860
1861                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1862
1863                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1864                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text);
1865
1866                 // Shared content
1867                 $text = self::convertShare(
1868                         $text,
1869                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1870                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1871                         }
1872                 );
1873
1874                 if ($saved_image) {
1875                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1876                 }
1877
1878                 // Restore code blocks
1879                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1880                         function ($matches) use ($codeblocks) {
1881                                 $return = $matches[0];
1882                                 if (isset($codeblocks[intval($matches[1])])) {
1883                                         $return = $codeblocks[$matches[1]];
1884                                 }
1885                                 return $return;
1886                         },
1887                         $text
1888                 );
1889
1890                 // Clean up the HTML by loading and saving the HTML with the DOM.
1891                 // Bad structured html can break a whole page.
1892                 // For performance reasons do it only with activated item cache or at export.
1893                 if (!$try_oembed || (get_itemcachepath() != '')) {
1894                         $doc = new DOMDocument();
1895                         $doc->preserveWhiteSpace = false;
1896
1897                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1898
1899                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1900                         $encoding = '<?xml encoding="UTF-8">';
1901                         @$doc->loadHTML($encoding . $doctype . '<html><body>' . $text . '</body></html>');
1902                         $doc->encoding = 'UTF-8';
1903                         $text = $doc->saveHTML();
1904                         $text = str_replace(['<html><body>', '</body></html>', $doctype, $encoding], ['', '', '', ''], $text);
1905
1906                         $text = str_replace('<br></li>', '</li>', $text);
1907
1908                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1909                 }
1910
1911                 // Clean up some useless linebreaks in lists
1912                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1913                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1914                 //$Text = str_replace('</li><br />', '</li>', $Text);
1915                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1916                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1917
1918                 Hook::callAll('bbcode', $text);
1919
1920                 return trim($text);
1921         }
1922
1923         /**
1924          * Strips the "abstract" tag from the provided text
1925          *
1926          * @param string $text The text with BBCode
1927          * @return string The same text - but without "abstract" element
1928          */
1929         public static function stripAbstract($text)
1930         {
1931                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1932                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1933
1934                 return $text;
1935         }
1936
1937         /**
1938          * Returns the value of the "abstract" element
1939          *
1940          * @param string $text The text that maybe contains the element
1941          * @param string $addon The addon for which the abstract is meant for
1942          * @return string The abstract
1943          */
1944         public static function getAbstract($text, $addon = '')
1945         {
1946                 $abstract = '';
1947                 $abstracts = [];
1948                 $addon = strtolower($addon);
1949
1950                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1951                         foreach ($results AS $result) {
1952                                 $abstracts[strtolower($result[1])] = $result[2];
1953                         }
1954                 }
1955
1956                 if (isset($abstracts[$addon])) {
1957                         $abstract = $abstracts[$addon];
1958                 }
1959
1960                 if ($abstract == '' && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1961                         $abstract = $result[1];
1962                 }
1963
1964                 return $abstract;
1965         }
1966
1967         /**
1968          * Callback function to replace a Friendica style mention in a mention for Diaspora
1969          *
1970          * @param array $match Matching values for the callback
1971          *                     [1] = Mention type (! or @)
1972          *                     [2] = Name
1973          *                     [3] = Address
1974          * @return string Replaced mention
1975          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1976          * @throws \ImagickException
1977          */
1978         private static function bbCodeMention2DiasporaCallback($match)
1979         {
1980                 $contact = Contact::getDetailsByURL($match[3]);
1981
1982                 if (empty($contact['addr'])) {
1983                         $contact = Probe::uri($match[3]);
1984                 }
1985
1986                 if (empty($contact['addr'])) {
1987                         return $match[0];
1988                 }
1989
1990                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
1991                 return $mention;
1992         }
1993
1994         /**
1995          * Converts a BBCode text into Markdown
1996          *
1997          * This function converts a BBCode item body to be sent to Markdown-enabled
1998          * systems like Diaspora and Libertree
1999          *
2000          * @param string $text
2001          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
2002          * @return string
2003          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2004          */
2005         public static function toMarkdown($text, $for_diaspora = true)
2006         {
2007                 $a = DI::app();
2008
2009                 $original_text = $text;
2010
2011                 // Since Diaspora is creating a summary for links, this function removes them before posting
2012                 if ($for_diaspora) {
2013                         $text = self::removeShareInformation($text);
2014                 }
2015
2016                 /**
2017                  * Transform #tags, strip off the [url] and replace spaces with underscore
2018                  */
2019                 $url_search_string = "^\[\]";
2020                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2021                         function ($matches) {
2022                                 return '#' . str_replace(' ', '_', $matches[2]);
2023                         },
2024                         $text
2025                 );
2026
2027                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2028                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2029
2030                 // Convert it to HTML - don't try oembed
2031                 if ($for_diaspora) {
2032                         $text = self::convert($text, false, 3);
2033
2034                         // Add all tags that maybe were removed
2035                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2036                                 $tagline = '';
2037                                 foreach ($tags[2] as $tag) {
2038                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2039                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2040                                                 $tagline .= '#' . $tag . ' ';
2041                                         }
2042                                 }
2043                                 $text = $text . " " . $tagline;
2044                         }
2045                 } else {
2046                         $text = self::convert($text, false, 4);
2047                 }
2048
2049                 // If a link is followed by a quote then there should be a newline before it
2050                 // Maybe we should make this newline at every time before a quote.
2051                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2052
2053                 $stamp1 = microtime(true);
2054
2055                 // Now convert HTML to Markdown
2056                 $text = HTML::toMarkdown($text);
2057
2058                 DI::profiler()->saveTimestamp($stamp1, "parser", System::callstack());
2059
2060                 // Libertree has a problem with escaped hashtags.
2061                 $text = str_replace(['\#'], ['#'], $text);
2062
2063                 // Remove any leading or trailing whitespace, as this will mess up
2064                 // the Diaspora signature verification and cause the item to disappear
2065                 $text = trim($text);
2066
2067                 if ($for_diaspora) {
2068                         $url_search_string = "^\[\]";
2069                         $text = preg_replace_callback(
2070                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2071                                 ['self', 'bbCodeMention2DiasporaCallback'],
2072                                 $text
2073                         );
2074                 }
2075
2076                 Hook::callAll('bb2diaspora', $text);
2077
2078                 return $text;
2079         }
2080
2081         /**
2082          * Pull out all #hashtags and @person tags from $string.
2083          *
2084          * We also get @person@domain.com - which would make
2085          * the regex quite complicated as tags can also
2086          * end a sentence. So we'll run through our results
2087          * and strip the period from any tags which end with one.
2088          * Returns array of tags found, or empty array.
2089          *
2090          * @param string $string Post content
2091          *
2092          * @return array List of tag and person names
2093          */
2094         public static function getTags($string)
2095         {
2096                 $ret = [];
2097
2098                 // Convert hashtag links to hashtags
2099                 $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2', $string);
2100
2101                 // ignore anything in a code block
2102                 $string = preg_replace('/\[code.*?\].*?\[\/code\]/sm', '', $string);
2103
2104                 // Force line feeds at bbtags
2105                 $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2106
2107                 // ignore anything in a bbtag
2108                 $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2109
2110                 // Match full names against @tags including the space between first and last
2111                 // We will look these up afterward to see if they are full names or not recognisable.
2112
2113                 if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2114                         foreach ($matches[1] as $match) {
2115                                 if (strstr($match, ']')) {
2116                                         // we might be inside a bbcode color tag - leave it alone
2117                                         continue;
2118                                 }
2119
2120                                 if (substr($match, -1, 1) === '.') {
2121                                         $ret[] = substr($match, 0, -1);
2122                                 } else {
2123                                         $ret[] = $match;
2124                                 }
2125                         }
2126                 }
2127
2128                 // Otherwise pull out single word tags. These can be @nickname, @first_last
2129                 // and #hash tags.
2130
2131                 if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?]+)([ \x0D\x0A,;:?]|$)/', $string, $matches)) {
2132                         foreach ($matches[1] as $match) {
2133                                 if (strstr($match, ']')) {
2134                                         // we might be inside a bbcode color tag - leave it alone
2135                                         continue;
2136                                 }
2137
2138                                 if (substr($match, -1, 1) === '.') {
2139                                         $match = substr($match,0,-1);
2140                                 }
2141
2142                                 // ignore strictly numeric tags like #1
2143                                 if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
2144                                         continue;
2145                                 }
2146
2147                                 // try not to catch url fragments
2148                                 if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2149                                         continue;
2150                                 }
2151                                 $ret[] = $match;
2152                         }
2153                 }
2154
2155                 return $ret;
2156         }
2157 }