src/Content/Text/BBCode.php

   1 <?php
   2 /**
   3  * @copyright Copyright (C) 2020, Friendica
   4  *
   5  * @license GNU AGPL version 3 or any later version
   6  *
   7  * This program is free software: you can redistribute it and/or modify
   8  * it under the terms of the GNU Affero General Public License as
   9  * published by the Free Software Foundation, either version 3 of the
  10  * License, or (at your option) any later version.
  11  *
  12  * This program is distributed in the hope that it will be useful,
  13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15  * GNU Affero General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU Affero General Public License
  18  * along with this program.  If not, see <https://www.gnu.org/licenses/>.
  19  *
  20  */
  21
  22 namespace Friendica\Content\Text;
  23
  24 use DOMDocument;
  25 use DOMXPath;
  26 use Exception;
  27 use Friendica\Content\ContactSelector;
  28 use Friendica\Content\Item;
  29 use Friendica\Content\OEmbed;
  30 use Friendica\Content\Smilies;
  31 use Friendica\Core\Hook;
  32 use Friendica\Core\Logger;
  33 use Friendica\Core\Protocol;
  34 use Friendica\Core\Renderer;
  35 use Friendica\Core\System;
  36 use Friendica\DI;
  37 use Friendica\Model\Contact;
  38 use Friendica\Model\Event;
  39 use Friendica\Model\Photo;
  40 use Friendica\Model\Tag;
  41 use Friendica\Object\Image;
  42 use Friendica\Protocol\Activity;
  43 use Friendica\Util\Images;
  44 use Friendica\Util\Map;
  45 use Friendica\Util\ParseUrl;
  46 use Friendica\Util\Proxy as ProxyUtils;
  47 use Friendica\Util\Strings;
  48 use Friendica\Util\XML;
  49
  50 class BBCode
  51 {
  52         const INTERNAL = 0;
  53         const API = 2;
  54         const DIASPORA = 3;
  55         const CONNECTORS = 4;
  56         const OSTATUS = 7;
  57         const TWITTER = 8;
  58         const BACKLINK = 8;
  59         const ACTIVITYPUB = 9;
  60
  61         /**
  62          * Fetches attachment data that were generated the old way
  63          *
  64          * @param string $body Message body
  65          * @return array
  66          *                     'type' -> Message type ('link', 'video', 'photo')
  67          *                     'text' -> Text before the shared message
  68          *                     'after' -> Text after the shared message
  69          *                     'image' -> Preview image of the message
  70          *                     'url' -> Url to the attached message
  71          *                     'title' -> Title of the attachment
  72          *                     'description' -> Description of the attachment
  73          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  74          */
  75         private static function getOldAttachmentData($body)
  76         {
  77                 $post = [];
  78
  79                 // Simplify image codes
  80                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  81
  82                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  83                         foreach ($attached as $data) {
  84                                 if (!in_array($data[1], ['type-link', 'type-video', 'type-photo'])) {
  85                                         continue;
  86                                 }
  87
  88                                 $post['type'] = substr($data[1], 5);
  89
  90                                 $pos = strpos($body, $data[0]);
  91                                 if ($pos > 0) {
  92                                         $post['text'] = trim(substr($body, 0, $pos));
  93                                         $post['after'] = trim(substr($body, $pos + strlen($data[0])));
  94                                 } else {
  95                                         $post['text'] = trim(str_replace($data[0], '', $body));
  96                                         $post['after'] = '';
  97                                 }
  98
  99                                 $attacheddata = $data[2];
 100
 101                                 if (preg_match("/\[img\](.*?)\[\/img\]/ism", $attacheddata, $matches)) {
 102
 103                                         $picturedata = Images::getInfoFromURLCached($matches[1]);
 104
 105                                         if ($picturedata) {
 106                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
 107                                                         $post['image'] = $matches[1];
 108                                                 } else {
 109                                                         $post['preview'] = $matches[1];
 110                                                 }
 111                                         }
 112                                 }
 113
 114                                 if (preg_match("/\[bookmark\=(.*?)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
 115                                         $post['url'] = $matches[1];
 116                                         $post['title'] = $matches[2];
 117                                 }
 118                                 if (!empty($post['url']) && (in_array($post['type'], ['link', 'video']))
 119                                         && preg_match("/\[url\=(.*?)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
 120                                         $post['url'] = $matches[1];
 121                                 }
 122
 123                                 // Search for description
 124                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
 125                                         $post['description'] = $matches[1];
 126                                 }
 127                         }
 128                 }
 129                 return $post;
 130         }
 131
 132         /**
 133          * Fetches attachment data that were generated with the "attachment" element
 134          *
 135          * @param string $body Message body
 136          * @return array
 137          *                     'type' -> Message type ('link', 'video', 'photo')
 138          *                     'text' -> Text before the shared message
 139          *                     'after' -> Text after the shared message
 140          *                     'image' -> Preview image of the message
 141          *                     'url' -> Url to the attached message
 142          *                     'title' -> Title of the attachment
 143          *                     'description' -> Description of the attachment
 144          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 145          */
 146         public static function getAttachmentData($body)
 147         {
 148                 $data = [
 149                         'type'          => '',
 150                         'text'          => '',
 151                         'after'         => '',
 152                         'image'         => null,
 153                         'url'           => '',
 154                         'provider_name' => '',
 155                         'provider_url'  => '',
 156                         'title'         => '',
 157                         'description'   => '',
 158                 ];
 159
 160                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 161                         return self::getOldAttachmentData($body);
 162                 }
 163
 164                 $attributes = $match[2];
 165
 166                 $data['text'] = trim($match[1]);
 167
 168                 $type = '';
 169                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 170                 if (!empty($matches[1])) {
 171                         $type = strtolower($matches[1]);
 172                 }
 173
 174                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 175                 if (!empty($matches[1])) {
 176                         $type = strtolower($matches[1]);
 177                 }
 178
 179                 if ($type == '') {
 180                         return [];
 181                 }
 182
 183                 if (!in_array($type, ['link', 'audio', 'photo', 'video'])) {
 184                         return [];
 185                 }
 186
 187                 if ($type != '') {
 188                         $data['type'] = $type;
 189                 }
 190
 191                 $url = '';
 192                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 193                 if (!empty($matches[1])) {
 194                         $url = $matches[1];
 195                 }
 196
 197                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 198                 if (!empty($matches[1])) {
 199                         $url = $matches[1];
 200                 }
 201
 202                 if ($url != '') {
 203                         $data['url'] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 204                 }
 205
 206                 $title = '';
 207                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 208                 if (!empty($matches[1])) {
 209                         $title = $matches[1];
 210                 }
 211
 212                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 213                 if (!empty($matches[1])) {
 214                         $title = $matches[1];
 215                 }
 216
 217                 if ($title != '') {
 218                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 219                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 220                         $title = str_replace(['[', ']'], ['&#91;', '&#93;'], $title);
 221                         $data['title'] = $title;
 222                 }
 223
 224                 $image = '';
 225                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 226                 if (!empty($matches[1])) {
 227                         $image = $matches[1];
 228                 }
 229
 230                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 231                 if (!empty($matches[1])) {
 232                         $image = $matches[1];
 233                 }
 234
 235                 if ($image != '') {
 236                         $data['image'] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 237                 }
 238
 239                 $preview = '';
 240                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 241                 if (!empty($matches[1])) {
 242                         $preview = $matches[1];
 243                 }
 244
 245                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 246                 if (!empty($matches[1])) {
 247                         $preview = $matches[1];
 248                 }
 249
 250                 if ($preview != '') {
 251                         $data['preview'] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 252                 }
 253
 254                 $data['description'] = trim($match[3]);
 255
 256                 $data['after'] = trim($match[4]);
 257
 258                 $parts = parse_url($data['url']);
 259                 if (!empty($parts['scheme']) && !empty($parts['host'])) {
 260                         $data['provider_name'] = $parts['host'];
 261                         $data['provider_url'] = $parts['scheme'] . '://' . $parts['host'];
 262
 263                         if (!empty($parts['port'])) {
 264                                 $data['provider_url'] .= ':' . $parts['port'];
 265                         }
 266                 }
 267
 268                 return $data;
 269         }
 270
 271         public static function getAttachedData($body, $item = [])
 272         {
 273                 /*
 274                 - text:
 275                 - type: link, video, photo
 276                 - title:
 277                 - url:
 278                 - image:
 279                 - description:
 280                 - (thumbnail)
 281                 */
 282
 283                 $has_title = !empty($item['title']);
 284                 $plink = $item['plink'] ?? '';
 285                 $post = self::getAttachmentData($body);
 286
 287                 // Get all linked images with alternative image description
 288                 if (preg_match_all("/\[img=([^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 289                         foreach ($pictures as $picture) {
 290                                 if (Photo::isLocal($picture[1])) {
 291                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]];
 292                                 }
 293                         }
 294                         if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
 295                                 $post['image_description'] = $post['images'][0]['description'];
 296                         }
 297                 }
 298
 299                 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
 300                         foreach ($pictures as $picture) {
 301                                 if (Photo::isLocal($picture[1])) {
 302                                         $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
 303                                 }
 304                         }
 305                 }
 306
 307                 // if nothing is found, it maybe having an image.
 308                 if (!isset($post['type'])) {
 309                         // Simplify image codes
 310                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 311                         $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 312                         $post['text'] = $body;
 313
 314                         if (preg_match_all("#\[url=([^\]]+?)\]\s*\[img\]([^\[]+?)\[/img\]\s*\[/url\]#ism", $body, $pictures, PREG_SET_ORDER)) {
 315                                 if ((count($pictures) == 1) && !$has_title) {
 316                                         if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) {
 317                                                 // Replace the preview picture with the real picture
 318                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
 319                                                 $data = ['url' => $url, 'type' => 'photo'];
 320                                         } else {
 321                                                 // Checking, if the link goes to a picture
 322                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 323                                         }
 324
 325                                         // Workaround:
 326                                         // Sometimes photo posts to the own album are not detected at the start.
 327                                         // So we seem to cannot use the cache for these cases. That's strange.
 328                                         if (($data['type'] != 'photo') && strstr($pictures[0][1], "/photos/")) {
 329                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 330                                         }
 331
 332                                         if ($data['type'] == 'photo') {
 333                                                 $post['type'] = 'photo';
 334                                                 if (isset($data['images'][0])) {
 335                                                         $post['image'] = $data['images'][0]['src'];
 336                                                         $post['url'] = $data['url'];
 337                                                 } else {
 338                                                         $post['image'] = $data['url'];
 339                                                 }
 340
 341                                                 $post['preview'] = $pictures[0][2];
 342                                                 $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 343                                         } else {
 344                                                 $imgdata = Images::getInfoFromURLCached($pictures[0][1]);
 345                                                 if ($imgdata && substr($imgdata['mime'], 0, 6) == 'image/') {
 346                                                         $post['type'] = 'photo';
 347                                                         $post['image'] = $pictures[0][1];
 348                                                         $post['preview'] = $pictures[0][2];
 349                                                         $post['text'] = trim(str_replace($pictures[0][0], '', $body));
 350                                                 }
 351                                         }
 352                                 } elseif (count($pictures) > 0) {
 353                                         $post['type'] = 'link';
 354                                         $post['url'] = $plink;
 355                                         $post['image'] = $pictures[0][2];
 356                                         $post['text'] = $body;
 357
 358                                         foreach ($pictures as $picture) {
 359                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 360                                         }
 361                                 }
 362                         } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 363                                 if ((count($pictures) == 1) && !$has_title) {
 364                                         $post['type'] = 'photo';
 365                                         $post['image'] = $pictures[0][1];
 366                                         $post['text'] = str_replace($pictures[0][0], '', $body);
 367                                 } elseif (count($pictures) > 0) {
 368                                         $post['type'] = 'link';
 369                                         $post['url'] = $plink;
 370                                         $post['image'] = $pictures[0][1];
 371                                         $post['text'] = $body;
 372
 373                                         foreach ($pictures as $picture) {
 374                                                 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
 375                                         }
 376                                 }
 377                         }
 378
 379                         // Test for the external links
 380                         preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER);
 381                         preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER);
 382
 383                         $links = array_merge($links1, $links2);
 384
 385                         // If there is only a single one, then use it.
 386                         // This should cover link posts via API.
 387                         if ((count($links) == 1) && !isset($post['preview']) && !$has_title) {
 388                                 $post['type'] = 'link';
 389                                 $post['url'] = $links[0][1];
 390                         }
 391
 392                         // Simplify "video" element
 393                         $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']);
 394
 395                         // Now count the number of external media links
 396                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
 397                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
 398                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
 399                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
 400
 401                         // Add them to the other external links
 402                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 403
 404                         // Are there more than one?
 405                         if (count($links) > 1) {
 406                                 // The post will be the type "text", which means a blog post
 407                                 unset($post['type']);
 408                                 $post['url'] = $plink;
 409                         }
 410
 411                         if (!isset($post['type'])) {
 412                                 $post['type'] = "text";
 413                                 $post['text'] = trim($body);
 414                         }
 415                 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
 416                         $data = ParseUrl::getSiteinfoCached($post['url'], true);
 417
 418                         if (isset($data['images'][0])) {
 419                                 $post['image'] = $data['images'][0]['src'];
 420                         }
 421                 }
 422
 423                 return $post;
 424         }
 425
 426         /**
 427          * Remove [attachment] BBCode and replaces it with a regular [url]
 428          *
 429          * @param string  $body
 430          * @param boolean $no_link_desc No link description
 431          *
 432          * @return string with replaced body
 433          */
 434         public static function removeAttachment($body, $no_link_desc = false)
 435         {
 436                 return preg_replace_callback("/\s*\[attachment (.*)\](.*?)\[\/attachment\]\s*/ism",
 437                         function ($match) use ($no_link_desc) {
 438                                 $attach_data = self::getAttachmentData($match[0]);
 439                                 if (empty($attach_data['url'])) {
 440                                         return $match[0];
 441                                 } elseif (empty($attach_data['title']) || $no_link_desc) {
 442                                         return "\n[url]" . $attach_data['url'] . "[/url]\n";
 443                                 } else {
 444                                         return "\n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
 445                                 }
 446                 }, $body);
 447         }
 448
 449         /**
 450          * Converts a BBCode text into plaintext
 451          *
 452          * @param      $text
 453          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 454          *
 455          * @return string
 456          */
 457         public static function toPlaintext($text, $keep_urls = true)
 458         {
 459                 $naked_text = HTML::toPlaintext(self::convert($text, false, 0, true), 0, !$keep_urls);
 460
 461                 return $naked_text;
 462         }
 463
 464         private static function proxyUrl($image, $simplehtml = self::INTERNAL)
 465         {
 466                 // Only send proxied pictures to API and for internal display
 467                 if (in_array($simplehtml, [self::INTERNAL, self::API])) {
 468                         return ProxyUtils::proxifyUrl($image);
 469                 } else {
 470                         return $image;
 471                 }
 472         }
 473
 474         /**
 475          * This function changing the visual size (not the real size) of images.
 476          * The function does not work for pictures with an alternate text description.
 477          * This could only be changed by using some new "img" BBCode format.
 478          *
 479          * @param string $srctext The body with images
 480          * @return string The body with possibly scaled images
 481          */
 482         public static function scaleExternalImages(string $srctext)
 483         {
 484                 $s = $srctext;
 485
 486                 // Simplify image links
 487                 $s = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $s);
 488
 489                 $matches = null;
 490                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 491                 if ($c) {
 492                         foreach ($matches as $mtch) {
 493                                 Logger::log('scale_external_image: ' . $mtch[1]);
 494
 495                                 $hostname = str_replace('www.', '', substr(DI::baseUrl(), strpos(DI::baseUrl(), '://') + 3));
 496                                 if (stristr($mtch[1], $hostname)) {
 497                                         continue;
 498                                 }
 499
 500                                 $curlResult = DI::httpRequest()->get($mtch[1]);
 501                                 if (!$curlResult->isSuccess()) {
 502                                         continue;
 503                                 }
 504
 505                                 $i = $curlResult->getBody();
 506                                 $type = $curlResult->getContentType();
 507                                 $type = Images::getMimeTypeByData($i, $mtch[1], $type);
 508
 509                                 if ($i) {
 510                                         $Image = new Image($i, $type);
 511                                         if ($Image->isValid()) {
 512                                                 $orig_width = $Image->getWidth();
 513                                                 $orig_height = $Image->getHeight();
 514
 515                                                 if ($orig_width > 640 || $orig_height > 640) {
 516                                                         $Image->scaleDown(640);
 517                                                         $new_width = $Image->getWidth();
 518                                                         $new_height = $Image->getHeight();
 519                                                         Logger::info('External images scaled', ['orig_width' => $orig_width, 'new_width' => $new_width, 'orig_height' => $orig_height, 'new_height' => $new_height, 'match' => $mtch[0]]);
 520                                                         $s = str_replace(
 521                                                                 $mtch[0],
 522                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]'
 523                                                                 . "\n",
 524                                                                 $s
 525                                                         );
 526                                                         Logger::info('New string', ['image' => $s]);
 527                                                 }
 528                                         }
 529                                 }
 530                         }
 531                 }
 532
 533                 return $s;
 534         }
 535
 536         /**
 537          * Truncates imported message body string length to max_import_size
 538          *
 539          * The purpose of this function is to apply system message length limits to
 540          * imported messages without including any embedded photos in the length
 541          *
 542          * @param string $body
 543          * @return string
 544          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 545          */
 546         public static function limitBodySize($body)
 547         {
 548                 $maxlen = DI::config()->get('config', 'max_import_size', 0);
 549
 550                 // If the length of the body, including the embedded images, is smaller
 551                 // than the maximum, then don't waste time looking for the images
 552                 if ($maxlen && (strlen($body) > $maxlen)) {
 553
 554                         Logger::info('the total body length exceeds the limit', ['maxlen' => $maxlen, 'body_len' => strlen($body)]);
 555
 556                         $orig_body = $body;
 557                         $new_body = '';
 558                         $textlen = 0;
 559
 560                         $img_start = strpos($orig_body, '[img');
 561                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 562                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 563                         while (($img_st_close !== false) && ($img_end !== false)) {
 564
 565                                 $img_st_close++; // make it point to AFTER the closing bracket
 566                                 $img_end += $img_start;
 567                                 $img_end += strlen('[/img]');
 568
 569                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 570                                         // This is an embedded image
 571
 572                                         if (($textlen + $img_start) > $maxlen) {
 573                                                 if ($textlen < $maxlen) {
 574                                                         Logger::info('the limit happens before an embedded image');
 575                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 576                                                         $textlen = $maxlen;
 577                                                 }
 578                                         } else {
 579                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 580                                                 $textlen += $img_start;
 581                                         }
 582
 583                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 584                                 } else {
 585
 586                                         if (($textlen + $img_end) > $maxlen) {
 587                                                 if ($textlen < $maxlen) {
 588                                                         Logger::info('the limit happens before the end of a non-embedded image');
 589                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 590                                                         $textlen = $maxlen;
 591                                                 }
 592                                         } else {
 593                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 594                                                 $textlen += $img_end;
 595                                         }
 596                                 }
 597                                 $orig_body = substr($orig_body, $img_end);
 598
 599                                 if ($orig_body === false) {
 600                                         // in case the body ends on a closing image tag
 601                                         $orig_body = '';
 602                                 }
 603
 604                                 $img_start = strpos($orig_body, '[img');
 605                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 606                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 607                         }
 608
 609                         if (($textlen + strlen($orig_body)) > $maxlen) {
 610                                 if ($textlen < $maxlen) {
 611                                         Logger::info('the limit happens after the end of the last image');
 612                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 613                                 }
 614                         } else {
 615                                 Logger::info('the text size with embedded images extracted did not violate the limit');
 616                                 $new_body = $new_body . $orig_body;
 617                         }
 618
 619                         return $new_body;
 620                 } else {
 621                         return $body;
 622                 }
 623         }
 624
 625         /**
 626          * Processes [attachment] tags
 627          *
 628          * Note: Can produce a [bookmark] tag in the returned string
 629          *
 630          * @param string  $text
 631          * @param integer $simplehtml
 632          * @param bool    $tryoembed
 633          * @return string
 634          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 635          */
 636         private static function convertAttachment($text, $simplehtml = self::INTERNAL, $tryoembed = true)
 637         {
 638                 $data = self::getAttachmentData($text);
 639                 if (empty($data) || empty($data['url'])) {
 640                         return $text;
 641                 }
 642
 643                 if (isset($data['title'])) {
 644                         $data['title'] = strip_tags($data['title']);
 645                         $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
 646                 } else {
 647                         $data['title'] = null;
 648                 }
 649
 650                 if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) {
 651                         $data['preview'] = $data['image'];
 652                         $data['image'] = '';
 653                 }
 654
 655                 $return = '';
 656                 try {
 657                         if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 658                                 $return = OEmbed::getHTML($data['url'], $data['title']);
 659                         } else {
 660                                 throw new Exception('OEmbed is disabled for this attachment.');
 661                         }
 662                 } catch (Exception $e) {
 663                         $data['title'] = ($data['title'] ?? '') ?: $data['url'];
 664
 665                         if ($simplehtml != self::CONNECTORS) {
 666                                 $return = sprintf('<div class="type-%s">', $data['type']);
 667                         }
 668
 669                         if (!empty($data['title']) && !empty($data['url'])) {
 670                                 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
 671                                         $return .= sprintf('<a href="%s" target="_blank" rel="noopener noreferrer"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data['url'], self::proxyUrl($data['image'], $simplehtml), $data['title']);
 672                                 } else {
 673                                         if (!empty($data['image'])) {
 674                                                 $return .= sprintf('<a href="%s" target="_blank" rel="noopener noreferrer"><img src="%s" alt="" title="%s" class="attachment-image" /></a><br />', $data['url'], self::proxyUrl($data['image'], $simplehtml), $data['title']);
 675                                         } elseif (!empty($data['preview'])) {
 676                                                 $return .= sprintf('<a href="%s" target="_blank" rel="noopener noreferrer"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br />', $data['url'], self::proxyUrl($data['preview'], $simplehtml), $data['title']);
 677                                         }
 678                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 679                                 }
 680                         }
 681
 682                         if (!empty($data['description']) && $data['description'] != $data['title']) {
 683                                 // Sanitize the HTML by converting it to BBCode
 684                                 $bbcode = HTML::toBBCode($data['description']);
 685                                 $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 686                         }
 687
 688                         if (!empty($data['url'])) {
 689                                 $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 690                         }
 691
 692                         if ($simplehtml != self::CONNECTORS) {
 693                                 $return .= '</div>';
 694                         }
 695                 }
 696
 697                 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
 698         }
 699
 700         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 701         {
 702                 $data = self::getAttachmentData($Text);
 703
 704                 if (!$data) {
 705                         return $Text;
 706                 } elseif ($nolink) {
 707                         return $data['text'] . ($data['after'] ?? '');
 708                 }
 709
 710                 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
 711                 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
 712                 if ($plaintext || (($title != '') && strstr($text, $title))) {
 713                         $data['title'] = $data['url'];
 714                 } elseif (($text != '') && strstr($title, $text)) {
 715                         $data['text'] = $data['title'];
 716                         $data['title'] = $data['url'];
 717                 }
 718
 719                 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
 720                         return $data['title'] . $data['after'];
 721                 }
 722
 723                 // If the link already is included in the post, don't add it again
 724                 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
 725                         return $data['text'] . $data['after'];
 726                 }
 727
 728                 $text = $data['text'];
 729
 730                 if (!empty($data['url']) && !empty($data['title'])) {
 731                         $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
 732                 } elseif (!empty($data['url'])) {
 733                         $text .= "\n[url]" . $data['url'] . '[/url]';
 734                 }
 735
 736                 return $text . "\n" . $data['after'];
 737         }
 738
 739         /**
 740          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 741          *
 742          * @param array $match Array with the matching values
 743          * @return string reformatted link including HTML codes
 744          */
 745         private static function convertUrlForActivityPubCallback($match)
 746         {
 747                 $url = $match[1];
 748
 749                 if (isset($match[2]) && ($match[1] != $match[2])) {
 750                         return $match[0];
 751                 }
 752
 753                 $parts = parse_url($url);
 754                 if (!isset($parts['scheme'])) {
 755                         return $match[0];
 756                 }
 757
 758                 return self::convertUrlForActivityPub($url);
 759         }
 760
 761         /**
 762          * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
 763          *
 764          * @param string $url URL that is about to be reformatted
 765          * @return string reformatted link including HTML codes
 766          */
 767         private static function convertUrlForActivityPub($url)
 768         {
 769                 $html = '<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>';
 770                 return sprintf($html, $url, self::getStyledURL($url));
 771         }
 772
 773         /**
 774          * Converts an URL in a nicer format (without the scheme and possibly shortened)
 775          *
 776          * @param string $url URL that is about to be reformatted
 777          * @return string reformatted link
 778          */
 779         private static function getStyledURL($url)
 780         {
 781                 $parts = parse_url($url);
 782                 $scheme = $parts['scheme'] . '://';
 783                 $styled_url = str_replace($scheme, '', $url);
 784
 785                 if (strlen($styled_url) > 30) {
 786                         $styled_url = substr($styled_url, 0, 30) . "…";
 787                 }
 788
 789                 return $styled_url;
 790         }
 791
 792         /*
 793          * [noparse][i]italic[/i][/noparse] turns into
 794          * [noparse][ i ]italic[ /i ][/noparse],
 795          * to hide them from parser.
 796          */
 797         private static function escapeNoparseCallback($match)
 798         {
 799                 $whole_match = $match[0];
 800                 $captured = $match[1];
 801                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 802                 $new_str = str_replace($captured, $spacefied, $whole_match);
 803                 return $new_str;
 804         }
 805
 806         /*
 807          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 808          * now turns back and the [noparse] tags are trimed
 809          * returning [i]italic[/i]
 810          */
 811         private static function unescapeNoparseCallback($match)
 812         {
 813                 $captured = $match[1];
 814                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 815                 return $unspacefied;
 816         }
 817
 818         /**
 819          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 820          * occurrences
 821          *
 822          * @param string $text        Text to search
 823          * @param string $name        Tag name
 824          * @param int    $occurrences Number of first occurrences to skip
 825          * @return boolean|array
 826          */
 827         public static function getTagPosition($text, $name, $occurrences = 0)
 828         {
 829                 if ($occurrences < 0) {
 830                         $occurrences = 0;
 831                 }
 832
 833                 $start_open = -1;
 834                 for ($i = 0; $i <= $occurrences; $i++) {
 835                         if ($start_open !== false) {
 836                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 837                         }
 838                 }
 839
 840                 if ($start_open === false) {
 841                         return false;
 842                 }
 843
 844                 $start_equal = strpos($text, '=', $start_open);
 845                 $start_close = strpos($text, ']', $start_open);
 846
 847                 if ($start_close === false) {
 848                         return false;
 849                 }
 850
 851                 $start_close++;
 852
 853                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 854
 855                 if ($end_open === false) {
 856                         return false;
 857                 }
 858
 859                 $res = [
 860                         'start' => [
 861                                 'open' => $start_open,
 862                                 'close' => $start_close
 863                         ],
 864                         'end' => [
 865                                 'open' => $end_open,
 866                                 'close' => $end_open + strlen('[/' . $name . ']')
 867                         ],
 868                 ];
 869
 870                 if ($start_equal !== false) {
 871                         $res['start']['equal'] = $start_equal + 1;
 872                 }
 873
 874                 return $res;
 875         }
 876
 877         /**
 878          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 879          *
 880          * @param string $pattern Preg pattern string
 881          * @param string $replace Preg replace string
 882          * @param string $name    BBCode tag name
 883          * @param string $text    Text to search
 884          * @return string
 885          */
 886         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 887         {
 888                 $occurrences = 0;
 889                 $pos = self::getTagPosition($text, $name, $occurrences);
 890                 while ($pos !== false && $occurrences++ < 1000) {
 891                         $start = substr($text, 0, $pos['start']['open']);
 892                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 893                         $end = substr($text, $pos['end']['close']);
 894                         if ($end === false) {
 895                                 $end = '';
 896                         }
 897
 898                         $subject = preg_replace($pattern, $replace, $subject);
 899                         $text = $start . $subject . $end;
 900
 901                         $pos = self::getTagPosition($text, $name, $occurrences);
 902                 }
 903
 904                 return $text;
 905         }
 906
 907         private static function extractImagesFromItemBody($body)
 908         {
 909                 $saved_image = [];
 910                 $orig_body = $body;
 911                 $new_body = '';
 912
 913                 $cnt = 0;
 914                 $img_start = strpos($orig_body, '[img');
 915                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 916                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 917                 while (($img_st_close !== false) && ($img_end !== false)) {
 918                         $img_st_close++; // make it point to AFTER the closing bracket
 919                         $img_end += $img_start;
 920
 921                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 922                                 // This is an embedded image
 923                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 924                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 925
 926                                 $cnt++;
 927                         } else {
 928                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 929                         }
 930
 931                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 932
 933                         if ($orig_body === false) {
 934                                 // in case the body ends on a closing image tag
 935                                 $orig_body = '';
 936                         }
 937
 938                         $img_start = strpos($orig_body, '[img');
 939                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 940                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 941                 }
 942
 943                 $new_body = $new_body . $orig_body;
 944
 945                 return ['body' => $new_body, 'images' => $saved_image];
 946         }
 947
 948         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 949         {
 950                 $newbody = $body;
 951
 952                 $cnt = 0;
 953                 foreach ($images as $image) {
 954                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 955                         // it loops over the array starting from the first element and going sequentially
 956                         // to the last element
 957                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 958                                 '<img src="' . self::proxyUrl($image) . '" alt="' . DI::l10n()->t('Image/photo') . '" />', $newbody);
 959                         $cnt++;
 960                 }
 961
 962                 return $newbody;
 963         }
 964
 965         /**
 966          * This function converts a [share] block to text according to a provided callback function whose signature is:
 967          *
 968          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 969          *
 970          * Where:
 971          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 972          * data lookup
 973          * - $author_contact is a contact record array
 974          * - $content is the inner content of the [share] block
 975          * - $is_quote_share indicates whether there's any content before the [share] block
 976          * - Return value is the string that should replace the [share] block in the provided text
 977          *
 978          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 979          *
 980          * @param  string   $text     A BBCode string
 981          * @param  callable $callback
 982          * @return string The BBCode string with all [share] blocks replaced
 983          */
 984         public static function convertShare($text, callable $callback)
 985         {
 986                 $return = preg_replace_callback(
 987                         "/(.*?)\[share(.*?)\](.*)\[\/share\]/ism",
 988                         function ($match) use ($callback) {
 989                                 $attribute_string = $match[2];
 990                                 $attributes = [];
 991                                 foreach (['author', 'profile', 'avatar', 'link', 'posted', 'guid'] as $field) {
 992                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 993                                         $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
 994                                 }
 995
 996                                 $author_contact = Contact::getByURL($attributes['profile'], false, ['url', 'addr', 'name', 'micro']);
 997                                 $author_contact['url'] = ($author_contact['url'] ?? $attributes['profile']);
 998                                 $author_contact['addr'] = ($author_contact['addr'] ?? '') ?: Protocol::getAddrFromProfileUrl($attributes['profile']);
 999
1000                                 $attributes['author']   = ($author_contact['name']  ?? '') ?: $attributes['author'];
1001                                 $attributes['avatar']   = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
1002                                 $attributes['profile']  = ($author_contact['url']   ?? '') ?: $attributes['profile'];
1003
1004                                 if ($attributes['avatar']) {
1005                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
1006                                 }
1007
1008                                 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
1009                         },
1010                         $text
1011                 );
1012
1013                 return $return;
1014         }
1015
1016         /**
1017          * Default [share] tag conversion callback
1018          *
1019          * Note: Can produce a [bookmark] tag in the output
1020          *
1021          * @see BBCode::convertShare()
1022          * @param array   $attributes     [share] block attribute values
1023          * @param array   $author_contact Contact row of the shared author
1024          * @param string  $content        Inner content of the [share] block
1025          * @param boolean $is_quote_share Whether there is content before the [share] block
1026          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
1027          * @return string
1028          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1029          */
1030         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
1031         {
1032                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
1033
1034                 switch ($simplehtml) {
1035                         case self::API:
1036                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1037                                 break;
1038                         case self::DIASPORA:
1039                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1040                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1041                                 } else {
1042                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1043
1044                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1045                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1046                                         }
1047
1048                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1049
1050                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
1051                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1052                                         }
1053                                 }
1054
1055                                 break;
1056                         case self::CONNECTORS:
1057                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
1058                                 $headline .= DI::l10n()->t('<a href="%1$s" target="_blank" rel="noopener noreferrer">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1059                                 $headline .= ':</b></p>' . "\n";
1060
1061                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1062
1063                                 break;
1064                         case self::OSTATUS:
1065                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1066                                 break;
1067                         case self::ACTIVITYPUB:
1068                                 $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>';
1069                                 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1070                                 break;
1071                         default:
1072                                 $text = ($is_quote_share? "\n" : '');
1073
1074                                 $contact = Contact::getByURL($attributes['profile'], false, ['network']);
1075                                 $network = $contact['network'] ?? Protocol::PHANTOM;
1076
1077                                 $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1078                                 $text .= Renderer::replaceMacros($tpl, [
1079                                         '$profile'      => $attributes['profile'],
1080                                         '$avatar'       => $attributes['avatar'],
1081                                         '$author'       => $attributes['author'],
1082                                         '$link'         => $attributes['link'],
1083                                         '$link_title'   => DI::l10n()->t('link to source'),
1084                                         '$posted'       => $attributes['posted'],
1085                                         '$guid'         => $attributes['guid'],
1086                                         '$network_name' => ContactSelector::networkToName($network, $attributes['profile']),
1087                                         '$network_icon' => ContactSelector::networkToIcon($network, $attributes['profile']),
1088                                         '$content'      => self::setMentions(trim($content), 0, $network),
1089                                 ]);
1090                                 break;
1091                 }
1092
1093                 return $text;
1094         }
1095
1096         private static function removePictureLinksCallback($match)
1097         {
1098                 $cache_key = 'remove:' . $match[1];
1099                 $text = DI::cache()->get($cache_key);
1100
1101                 if (is_null($text)) {
1102                         $a = DI::app();
1103
1104                         $stamp1 = microtime(true);
1105
1106                         $ch = @curl_init($match[1]);
1107                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1108                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1109                         @curl_setopt($ch, CURLOPT_USERAGENT, DI::httpRequest()->getUserAgent());
1110                         @curl_exec($ch);
1111                         $curl_info = @curl_getinfo($ch);
1112
1113                         DI::profiler()->saveTimestamp($stamp1, "network");
1114
1115                         if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1116                                 $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]";
1117                         } else {
1118                                 $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]";
1119
1120                                 // if its not a picture then look if its a page that contains a picture link
1121                                 $body = DI::httpRequest()->fetch($match[1]);
1122
1123                                 $doc = new DOMDocument();
1124                                 @$doc->loadHTML($body);
1125                                 $xpath = new DOMXPath($doc);
1126                                 $list = $xpath->query("//meta[@name]");
1127                                 foreach ($list as $node) {
1128                                         $attr = [];
1129
1130                                         if ($node->attributes->length) {
1131                                                 foreach ($node->attributes as $attribute) {
1132                                                         $attr[$attribute->name] = $attribute->value;
1133                                                 }
1134                                         }
1135
1136                                         if (strtolower($attr['name']) == 'twitter:image') {
1137                                                 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1138                                         }
1139                                 }
1140                         }
1141                         DI::cache()->set($cache_key, $text);
1142                 }
1143
1144                 return $text;
1145         }
1146
1147         private static function expandLinksCallback($match)
1148         {
1149                 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1150                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1151                 } else {
1152                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1153                 }
1154         }
1155
1156         private static function cleanPictureLinksCallback($match)
1157         {
1158                 $a = DI::app();
1159
1160                 // When the picture link is the own photo path then we can avoid fetching the link
1161                 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()->get()) . '/photos/');
1162                 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1163                         if (!empty($match[3])) {
1164                                 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1165                         } else {
1166                                 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1167                         }
1168                         return $text;
1169                 }
1170
1171                 $cache_key = 'clean:' . $match[1];
1172                 $text = DI::cache()->get($cache_key);
1173                 if (!is_null($text)) {
1174                         return $text;
1175                 }
1176
1177                 // Only fetch the header, not the content
1178                 $stamp1 = microtime(true);
1179
1180                 $ch = @curl_init($match[1]);
1181                 @curl_setopt($ch, CURLOPT_NOBODY, true);
1182                 @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1183                 @curl_setopt($ch, CURLOPT_USERAGENT, DI::httpRequest()->getUserAgent());
1184                 @curl_exec($ch);
1185                 $curl_info = @curl_getinfo($ch);
1186
1187                 DI::profiler()->saveTimestamp($stamp1, "network");
1188
1189                 // if its a link to a picture then embed this picture
1190                 if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1191                         $text = '[img]' . $match[1] . '[/img]';
1192                 } else {
1193                         if (!empty($match[3])) {
1194                                 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1195                         } else {
1196                                 $text = '[img]' . $match[2] . '[/img]';
1197                         }
1198
1199                         // if its not a picture then look if its a page that contains a picture link
1200                         $body = DI::httpRequest()->fetch($match[1]);
1201
1202                         $doc = new DOMDocument();
1203                         @$doc->loadHTML($body);
1204                         $xpath = new DOMXPath($doc);
1205                         $list = $xpath->query("//meta[@name]");
1206                         foreach ($list as $node) {
1207                                 $attr = [];
1208                                 if ($node->attributes->length) {
1209                                         foreach ($node->attributes as $attribute) {
1210                                                 $attr[$attribute->name] = $attribute->value;
1211                                         }
1212                                 }
1213
1214                                 if (strtolower($attr['name']) == "twitter:image") {
1215                                         if (!empty($match[3])) {
1216                                                 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1217                                         } else {
1218                                                 $text = "[img]" . $attr['content'] . "[/img]";
1219                                         }
1220                                 }
1221                         }
1222                 }
1223                 DI::cache()->set($cache_key, $text);
1224
1225                 return $text;
1226         }
1227
1228         public static function cleanPictureLinks($text)
1229         {
1230                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1231                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1232                 return $return;
1233         }
1234
1235         public static function removeLinks(string $bbcode)
1236         {
1237                 $bbcode = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $1 ', $bbcode);
1238                 $bbcode = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $bbcode);
1239
1240                 $bbcode = preg_replace('/[@!#]\[url\=.*?\].*?\[\/url\]/ism', '', $bbcode);
1241                 $bbcode = preg_replace("/\[url=[^\[\]]*\](.*)\[\/url\]/Usi", ' $1 ', $bbcode);
1242                 $bbcode = preg_replace('/[@!#]?\[url.*?\[\/url\]/ism', '', $bbcode);
1243                 return $bbcode;
1244         }
1245
1246         /**
1247          * Converts a BBCode message to HTML message
1248          *
1249          * BBcode 2 HTML was written by WAY2WEB.net
1250          * extended to work with Mistpark/Friendica - Mike Macgirvin
1251          *
1252          * Simple HTML values meaning:
1253          * - 0: Friendica display
1254          * - 1: Unused
1255          * - 2: Used for Windows Phone push, Friendica API
1256          * - 3: Used before converting to Markdown in bb2diaspora.php
1257          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1258          * - 5: Unused
1259          * - 6: Unused
1260          * - 7: Used for dfrn, OStatus
1261          * - 8: Used for Twitter, WP backlink text setting
1262          * - 9: ActivityPub
1263          *
1264          * @param string $text
1265          * @param bool   $try_oembed
1266          * @param int    $simple_html
1267          * @param bool   $for_plaintext
1268          * @return string
1269          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1270          */
1271         public static function convert(string $text = null, $try_oembed = true, $simple_html = self::INTERNAL, $for_plaintext = false)
1272         {
1273                 // Accounting for null default column values
1274                 if (is_null($text) || $text === '') {
1275                         return '';
1276                 }
1277
1278                 $a = DI::app();
1279
1280                 $text = self::performWithEscapedTags($text, ['code'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) {
1281                         $text = self::performWithEscapedTags($text, ['noparse', 'nobb', 'pre'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) {
1282                                 /*
1283                                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1284                                  *
1285                                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1286                                  * $match[1] = $url
1287                                  * $match[2] = $title or absent
1288                                  */
1289                                 $try_oembed_callback = function ($match)
1290                                 {
1291                                         $url = $match[1];
1292                                         $title = $match[2] ?? null;
1293
1294                                         try {
1295                                                 $return = OEmbed::getHTML($url, $title);
1296                                         } catch (Exception $ex) {
1297                                                 $return = $match[0];
1298                                         }
1299
1300                                         return $return;
1301                                 };
1302
1303
1304
1305                                 // Remove the abstract element. It is a non visible element.
1306                                 $text = self::stripAbstract($text);
1307
1308                                 // Move new lines outside of tags
1309                                 $text = preg_replace("#\[(\w*)](\n*)#ism", '$2[$1]', $text);
1310                                 $text = preg_replace("#(\n*)\[/(\w*)]#ism", '[/$2]$1', $text);
1311
1312                                 // Extract the private images which use data urls since preg has issues with
1313                                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1314                                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1315
1316                                 $extracted = self::extractImagesFromItemBody($text);
1317                                 $text = $extracted['body'];
1318                                 $saved_image = $extracted['images'];
1319
1320                                 // If we find any event code, turn it into an event.
1321                                 // After we're finished processing the bbcode we'll
1322                                 // replace all of the event code with a reformatted version.
1323
1324                                 $ev = Event::fromBBCode($text);
1325
1326                                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1327                                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1328
1329                                 $text = str_replace("<", "&lt;", $text);
1330                                 $text = str_replace(">", "&gt;", $text);
1331
1332                                 // remove some newlines before the general conversion
1333                                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1334                                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1335
1336                                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1337                                 if (!$try_oembed) {
1338                                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1339                                 }
1340
1341                                 // Convert new line chars to html <br /> tags
1342
1343                                 // nlbr seems to be hopelessly messed up
1344                                 //      $Text = nl2br($Text);
1345
1346                                 // We'll emulate it.
1347
1348                                 $text = trim($text);
1349                                 $text = str_replace("\r\n", "\n", $text);
1350
1351                                 // Remove linefeeds inside of the table elements. See issue #6799
1352                                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1353                                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1354                                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1355                                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1356                                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1357                                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1358                                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1359                                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1360                                 do {
1361                                         $oldtext = $text;
1362                                         $text = str_replace($search, $replace, $text);
1363                                 } while ($oldtext != $text);
1364
1365                                 // Replace these here only once
1366                                 $search = ["\n[table]", "[/table]\n"];
1367                                 $replace = ["[table]", "[/table]"];
1368                                 $text = str_replace($search, $replace, $text);
1369
1370                                 // removing multiplicated newlines
1371                                 if (DI::config()->get('system', 'remove_multiplicated_lines')) {
1372                                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1373                                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1374                                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1375                                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1376                                         do {
1377                                                 $oldtext = $text;
1378                                                 $text = str_replace($search, $replace, $text);
1379                                         } while ($oldtext != $text);
1380                                 }
1381
1382                                 /// @todo Have a closer look at the different html modes
1383                                 // Handle attached links or videos
1384                                 if ($simple_html == self::ACTIVITYPUB) {
1385                                         $text = self::removeAttachment($text);
1386                                 } elseif (!in_array($simple_html, [self::INTERNAL, self::CONNECTORS])) {
1387                                         $text = self::removeAttachment($text, true);
1388                                 } else {
1389                                         $text = self::convertAttachment($text, $simple_html, $try_oembed);
1390                                 }
1391
1392                                 // leave open the posibility of [map=something]
1393                                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1394                                 if (strpos($text, '[/map]') !== false) {
1395                                         $text = preg_replace_callback(
1396                                                 "/\[map\](.*?)\[\/map\]/ism",
1397                                                 function ($match) use ($simple_html) {
1398                                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1399                                                 },
1400                                                 $text
1401                                         );
1402                                 }
1403
1404                                 if (strpos($text, '[map=') !== false) {
1405                                         $text = preg_replace_callback(
1406                                                 "/\[map=(.*?)\]/ism",
1407                                                 function ($match) use ($simple_html) {
1408                                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1409                                                 },
1410                                                 $text
1411                                         );
1412                                 }
1413
1414                                 if (strpos($text, '[map]') !== false) {
1415                                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1416                                 }
1417
1418                                 // Check for headers
1419                                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1420                                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1421                                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1422                                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1423                                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1424                                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1425
1426                                 // Check for paragraph
1427                                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1428
1429                                 // Check for bold text
1430                                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1431
1432                                 // Check for Italics text
1433                                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1434
1435                                 // Check for Underline text
1436                                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1437
1438                                 // Check for strike-through text
1439                                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1440
1441                                 // Check for over-line text
1442                                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1443
1444                                 // Check for colored text
1445                                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1446
1447                                 // Check for sized text
1448                                 // [size=50] --> font-size: 50px (with the unit).
1449                                 if ($simple_html != self::DIASPORA) {
1450                                         $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1451                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1452                                 } else {
1453                                         // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1454                                         $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1455                                 }
1456
1457
1458                                 // Check for centered text
1459                                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1460
1461                                 // Check for list text
1462                                 $text = str_replace("[*]", "<li>", $text);
1463
1464                                 // Check for style sheet commands
1465                                 $text = preg_replace_callback(
1466                                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1467                                         function ($match) {
1468                                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1469                                         },
1470                                         $text
1471                                 );
1472
1473                                 // Check for CSS classes
1474                                 $text = preg_replace_callback(
1475                                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1476                                         function ($match) {
1477                                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1478                                         },
1479                                         $text
1480                                 );
1481
1482                                 // handle nested lists
1483                                 $endlessloop = 0;
1484
1485                                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1486                                                 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1487                                                 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1488                                                 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1489                                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1490                                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1491                                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1492                                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1493                                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1494                                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1495                                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1496                                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1497                                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1498                                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1499                                 }
1500
1501                                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1502                                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1503                                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1504                                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1505
1506                                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1507                                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1508
1509                                 $text = str_replace('[hr]', '<hr />', $text);
1510
1511                                 if (!$for_plaintext) {
1512                                         $escaped = [];
1513
1514                                         // Escaping BBCodes susceptible to contain rogue URL we don'' want the autolinker to catch
1515                                         $text = preg_replace_callback('#\[(url|img|audio|video|youtube|vimeo|share|attachment|iframe|bookmark).+?\[/\1\]#ism',
1516                                                 function ($matches) use (&$escaped) {
1517                                                         $return = '{escaped-' . count($escaped) . '}';
1518                                                         $escaped[] = $matches[0];
1519
1520                                                         return $return;
1521                                                 },
1522                                                 $text
1523                                         );
1524
1525                                         // Autolinker for isolated URLs
1526                                         $text = preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1527
1528                                         // Restoring escaped blocks
1529                                         $text = preg_replace_callback('/{escaped-([0-9]+)}/iU',
1530                                                 function ($matches) use ($escaped) {
1531                                                         return $escaped[intval($matches[1])] ?? $matches[0];
1532                                                 },
1533                                                 $text
1534                                         );
1535                                 }
1536
1537                                 // This is actually executed in Item::prepareBody()
1538
1539                                 $nosmile = strpos($text, '[nosmile]') !== false;
1540                                 $text = str_replace('[nosmile]', '', $text);
1541
1542                                 // Check for font change text
1543                                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1544
1545                                 // Declare the format for [spoiler] layout
1546                                 $SpoilerLayout = '<details class="spoiler"><summary>' . DI::l10n()->t('Click to open/close') . '</summary>$1</details>';
1547
1548                                 // Check for [spoiler] text
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", $SpoilerLayout, $text);
1553                                 }
1554
1555                                 // Check for [spoiler=Title] text
1556
1557                                 // handle nested quotes
1558                                 $endlessloop = 0;
1559                                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1560                                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1561                                                 '<details class="spoiler"><summary>$1</summary>$2</details>',
1562                                                 $text);
1563                                 }
1564
1565                                 // Declare the format for [quote] layout
1566                                 $QuoteLayout = '<blockquote>$1</blockquote>';
1567
1568                                 // Check for [quote] text
1569                                 // handle nested quotes
1570                                 $endlessloop = 0;
1571                                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1572                                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1573                                 }
1574
1575                                 // Check for [quote=Author] text
1576
1577                                 $t_wrote = DI::l10n()->t('$1 wrote:');
1578
1579                                 // handle nested quotes
1580                                 $endlessloop = 0;
1581                                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1582                                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1583                                                 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1584                                                 $text);
1585                                 }
1586
1587
1588                                 // [img=widthxheight]image source[/img]
1589                                 $text = preg_replace_callback(
1590                                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1591                                         function ($matches) use ($simple_html) {
1592                                                 if (strpos($matches[3], "data:image/") === 0) {
1593                                                         return $matches[0];
1594                                                 }
1595
1596                                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1597                                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1598                                         },
1599                                         $text
1600                                 );
1601
1602                                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1603                                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1604
1605                                 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1606                                         function ($matches) use ($simple_html) {
1607                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1608                                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1609                                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1610                                         },
1611                                         $text);
1612
1613                                 // Images
1614                                 // [img]pathtoimage[/img]
1615                                 $text = preg_replace_callback(
1616                                         "/\[img\](.*?)\[\/img\]/ism",
1617                                         function ($matches) use ($simple_html) {
1618                                                 if (strpos($matches[1], "data:image/") === 0) {
1619                                                         return $matches[0];
1620                                                 }
1621
1622                                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1623                                                 return "[img]" . $matches[1] . "[/img]";
1624                                         },
1625                                         $text
1626                                 );
1627
1628                                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1629                                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1630
1631                                 $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);
1632                                 $text = preg_replace("/\[crypt(.*?)\](.*?)\[\/crypt\]/ism", '<br/><img src="' .DI::baseUrl() . '/images/lock_icon.gif" alt="' . DI::l10n()->t('Encrypted content') . '" title="' . '$1' . ' ' . DI::l10n()->t('Encrypted content') . '" /><br />', $text);
1633                                 //$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);
1634
1635                                 // Simplify "video" element
1636                                 $text = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1637
1638                                 // Try to Oembed
1639                                 if ($try_oembed) {
1640                                         $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);
1641                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '<audio src="$1" controls="controls"><a href="$1">$1</a></audio>', $text);
1642
1643                                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1644                                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1645                                 } else {
1646                                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1647                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1648                                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1649                                                 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1650                                 }
1651
1652                                 // html5 video and audio
1653
1654
1655                                 if ($try_oembed) {
1656                                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1657                                 } else {
1658                                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1659                                 }
1660
1661                                 // Youtube extensions
1662                                 if ($try_oembed) {
1663                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1664                                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1665                                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1666                                 }
1667
1668                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1669                                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1670                                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1671
1672                                 if ($try_oembed) {
1673                                         $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);
1674                                 } else {
1675                                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1676                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank" rel="noopener noreferrer">https://www.youtube.com/watch?v=$1</a>', $text);
1677                                 }
1678
1679                                 if ($try_oembed) {
1680                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1681                                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1682                                 }
1683
1684                                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1685                                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1686
1687                                 if ($try_oembed) {
1688                                         $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);
1689                                 } else {
1690                                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1691                                                 '<a href="https://vimeo.com/$1" target="_blank" rel="noopener noreferrer">https://vimeo.com/$1</a>', $text);
1692                                 }
1693
1694                                 // oembed tag
1695                                 $text = OEmbed::BBCode2HTML($text);
1696
1697                                 // Avoid triple linefeeds through oembed
1698                                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1699
1700                                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1701                                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1702                                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1703                                 // start which is always required). Allow desc with a missing summary for compatibility.
1704
1705                                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1706                                         $sub = Event::getHTML($ev, $simple_html);
1707
1708                                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1709                                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1710                                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1711                                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1712                                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1713                                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1714                                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1715                                 }
1716
1717                                 // Replace non graphical smilies for external posts
1718                                 if (!$nosmile && !$for_plaintext) {
1719                                         $text = Smilies::replace($text);
1720                                 }
1721
1722                                 if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA)) {
1723                                         $conv = html_entity_decode(str_replace([' ', "\n", "\r"], '', $text));
1724                                         // Emojis are always 4 byte Unicode characters
1725                                         if (!empty($conv) && (strlen($conv) / mb_strlen($conv) == 4)) {
1726                                                 $text = '<span style="font-size: xx-large; line-height: initial;">' . $text . '</span>';
1727                                         }
1728                                 }
1729
1730                                 if (!$for_plaintext) {
1731                                         if (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1732                                                 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1733                                                 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1734                                         }
1735                                 } else {
1736                                         $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1737                                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1738                                 }
1739
1740                                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1741
1742                                 // Remove all hashtag addresses
1743                                 if ($simple_html && !in_array($simple_html, [self::DIASPORA, self::OSTATUS, self::ACTIVITYPUB])) {
1744                                         $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1745                                 } elseif ($simple_html == self::DIASPORA) {
1746                                         // The ! is converted to @ since Diaspora only understands the @
1747                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1748                                                 '@<a href="$2">$3</a>',
1749                                                 $text);
1750                                 } elseif (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1751                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1752                                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1753                                                 $text);
1754                                 } elseif (!$simple_html) {
1755                                         $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1756                                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1757                                                 $text);
1758                                 }
1759
1760                                 // Bookmarks in red - will be converted to bookmarks in friendica
1761                                 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1762                                 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1763                                 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1764                                                         "[bookmark=$1]$2[/bookmark]", $text);
1765
1766                                 if (in_array($simple_html, [self::API, self::OSTATUS, self::TWITTER])) {
1767                                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1768                                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1769                                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1770                                 }
1771
1772                                 // Perform URL Search
1773                                 if ($try_oembed) {
1774                                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1775                                 }
1776
1777                                 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1778
1779                                 // Handle Diaspora posts
1780                                 $text = preg_replace_callback(
1781                                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1782                                         function ($match) {
1783                                                 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1784                                         }, $text
1785                                 );
1786
1787                                 $text = preg_replace_callback(
1788                                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1789                                         function ($match) {
1790                                                 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1791                                         }, $text
1792                                 );
1793
1794                                 // Server independent link to posts and comments
1795                                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1796                                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1797                                 $text = preg_replace($expression, DI::baseUrl()."/display/$1", $text);
1798
1799                                 /* Tag conversion
1800                                  * Supports:
1801                                  * - #[url=<anything>]<term>[/url]
1802                                  * - [url=<anything>]#<term>[/url]
1803                                  */
1804                                 $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) use ($simple_html) {
1805                                         if ($simple_html == BBCode::ACTIVITYPUB) {
1806                                                 return '<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1807                                                         . '" data-tag="' . XML::escape($matches[1]) . '" rel="tag ugc">#'
1808                                                         . XML::escape($matches[1]) . '</a>';
1809                                         } else {
1810                                                 return '#<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1811                                                         . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1812                                                         . XML::escape($matches[1]) . '</a>';
1813                                         }
1814                                 }, $text);
1815
1816                                 // We need no target="_blank" rel="noopener noreferrer" for local links
1817                                 // convert links start with DI::baseUrl() as local link without the target="_blank" rel="noopener noreferrer" attribute
1818                                 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
1819                                 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1820                                 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1821
1822                                 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1823                                 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1824
1825                                 // Red compatibility, though the link can't be authenticated on Friendica
1826                                 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1827
1828
1829                                 // we may need to restrict this further if it picks up too many strays
1830                                 // link acct:user@host to a webfinger profile redirector
1831
1832                                 $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);
1833
1834                                 // Perform MAIL Search
1835                                 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1836                                 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1837
1838                                 /// @todo What is the meaning of these lines?
1839                                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1840                                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1841
1842                                 // Currently deactivated, it made problems with " inside of alt texts.
1843                                 //$text = preg_replace('/\&quot\;/', '"', $text);
1844
1845                                 // fix any escaped ampersands that may have been converted into links
1846                                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1847
1848                                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1849                                 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1850
1851                                 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1852
1853                                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1854                                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">', $text);
1855
1856                                 // sanitize href attributes (only allowlisted protocols URLs)
1857                                 // default value for backward compatibility
1858                                 $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
1859
1860                                 // Always allowed protocol even if config isn't set or not including it
1861                                 $allowed_link_protocols[] = '//';
1862                                 $allowed_link_protocols[] = 'http://';
1863                                 $allowed_link_protocols[] = 'https://';
1864                                 $allowed_link_protocols[] = 'redir/';
1865
1866                                 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1867
1868                                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1869                                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text);
1870
1871                                 // Shared content
1872                                 $text = self::convertShare(
1873                                         $text,
1874                                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1875                                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1876                                         }
1877                                 );
1878
1879                                 $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1880
1881                                 return $text;
1882                         }); // Escaped noparse, nobb, pre
1883
1884                         // Remove escaping tags
1885                         $text = preg_replace("/\[noparse\](.*?)\[\/noparse\]/ism", '\1', $text);
1886                         $text = preg_replace("/\[nobb\](.*?)\[\/nobb\]/ism", '\1', $text);
1887
1888                         // Additionally, [pre] tags preserve spaces
1889                         $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", function ($match) {
1890                                 return str_replace(' ', '&nbsp;', $match[1]);
1891                         }, $text);
1892
1893                         return $text;
1894                 }); // Escaped code
1895
1896                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1897                         function ($matches) {
1898                                 if (strpos($matches[2], "\n") !== false) {
1899                                         $return = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlspecialchars(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
1900                                 } else {
1901                                         $return = '<code>' . htmlspecialchars($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
1902                                 }
1903
1904                                 return $return;
1905                         },
1906                         $text
1907                 );
1908
1909                 // Clean up the HTML by loading and saving the HTML with the DOM.
1910                 // Bad structured html can break a whole page.
1911                 // For performance reasons do it only with activated item cache or at export.
1912                 if (!$try_oembed || (get_itemcachepath() != '')) {
1913                         $doc = new DOMDocument();
1914                         $doc->preserveWhiteSpace = false;
1915
1916                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1917
1918                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1919                         $encoding = '<?xml encoding="UTF-8">';
1920                         @$doc->loadHTML($encoding . $doctype . '<html><body>' . $text . '</body></html>');
1921                         $doc->encoding = 'UTF-8';
1922                         $text = $doc->saveHTML();
1923                         $text = str_replace(['<html><body>', '</body></html>', $doctype, $encoding], ['', '', '', ''], $text);
1924
1925                         $text = str_replace('<br></li>', '</li>', $text);
1926
1927                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1928                 }
1929
1930                 // Clean up some useless linebreaks in lists
1931                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1932                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1933                 //$Text = str_replace('</li><br />', '</li>', $Text);
1934                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1935                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1936
1937                 Hook::callAll('bbcode', $text);
1938
1939                 return trim($text);
1940         }
1941
1942         /**
1943          * Strips the "abstract" tag from the provided text
1944          *
1945          * @param string $text The text with BBCode
1946          * @return string The same text - but without "abstract" element
1947          */
1948         public static function stripAbstract($text)
1949         {
1950                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1951                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1952
1953                 return $text;
1954         }
1955
1956         /**
1957          * Returns the value of the "abstract" element
1958          *
1959          * @param string $text The text that maybe contains the element
1960          * @param string $addon The addon for which the abstract is meant for
1961          * @return string The abstract
1962          */
1963         public static function getAbstract($text, $addon = '')
1964         {
1965                 $abstract = '';
1966                 $abstracts = [];
1967                 $addon = strtolower($addon);
1968
1969                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1970                         foreach ($results AS $result) {
1971                                 $abstracts[strtolower($result[1])] = $result[2];
1972                         }
1973                 }
1974
1975                 if (isset($abstracts[$addon])) {
1976                         $abstract = $abstracts[$addon];
1977                 }
1978
1979                 if ($abstract == '' && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1980                         $abstract = $result[1];
1981                 }
1982
1983                 return $abstract;
1984         }
1985
1986         /**
1987          * Callback function to replace a Friendica style mention in a mention for Diaspora
1988          *
1989          * @param array $match Matching values for the callback
1990          *                     [1] = Mention type (! or @)
1991          *                     [2] = Name
1992          *                     [3] = Address
1993          * @return string Replaced mention
1994          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1995          * @throws \ImagickException
1996          */
1997         private static function bbCodeMention2DiasporaCallback($match)
1998         {
1999                 $contact = Contact::getByURL($match[3], false, ['addr']);
2000                 if (empty($contact['addr'])) {
2001                         return $match[0];
2002                 }
2003
2004                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
2005                 return $mention;
2006         }
2007
2008         /**
2009          * Converts a BBCode text into Markdown
2010          *
2011          * This function converts a BBCode item body to be sent to Markdown-enabled
2012          * systems like Diaspora and Libertree
2013          *
2014          * @param string $text
2015          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
2016          * @return string
2017          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2018          */
2019         public static function toMarkdown($text, $for_diaspora = true)
2020         {
2021                 $original_text = $text;
2022
2023                 // Since Diaspora is creating a summary for links, this function removes them before posting
2024                 if ($for_diaspora) {
2025                         $text = self::removeShareInformation($text);
2026                 }
2027
2028                 /**
2029                  * Transform #tags, strip off the [url] and replace spaces with underscore
2030                  */
2031                 $url_search_string = "^\[\]";
2032                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2033                         function ($matches) {
2034                                 return '#' . str_replace(' ', '_', $matches[2]);
2035                         },
2036                         $text
2037                 );
2038
2039                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2040                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2041
2042                 // Convert it to HTML - don't try oembed
2043                 if ($for_diaspora) {
2044                         $text = self::convert($text, false, self::DIASPORA);
2045
2046                         // Add all tags that maybe were removed
2047                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2048                                 $tagline = '';
2049                                 foreach ($tags[2] as $tag) {
2050                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2051                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2052                                                 $tagline .= '#' . $tag . ' ';
2053                                         }
2054                                 }
2055                                 $text = $text . " " . $tagline;
2056                         }
2057                 } else {
2058                         $text = self::convert($text, false, self::CONNECTORS);
2059                 }
2060
2061                 // If a link is followed by a quote then there should be a newline before it
2062                 // Maybe we should make this newline at every time before a quote.
2063                 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2064
2065                 $stamp1 = microtime(true);
2066
2067                 // Now convert HTML to Markdown
2068                 $text = HTML::toMarkdown($text);
2069
2070                 DI::profiler()->saveTimestamp($stamp1, "parser");
2071
2072                 // Libertree has a problem with escaped hashtags.
2073                 $text = str_replace(['\#'], ['#'], $text);
2074
2075                 // Remove any leading or trailing whitespace, as this will mess up
2076                 // the Diaspora signature verification and cause the item to disappear
2077                 $text = trim($text);
2078
2079                 if ($for_diaspora) {
2080                         $url_search_string = "^\[\]";
2081                         $text = preg_replace_callback(
2082                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2083                                 ['self', 'bbCodeMention2DiasporaCallback'],
2084                                 $text
2085                         );
2086                 }
2087
2088                 Hook::callAll('bb2diaspora', $text);
2089
2090                 return $text;
2091         }
2092
2093         /**
2094          * Pull out all #hashtags and @person tags from $string.
2095          *
2096          * We also get @person@domain.com - which would make
2097          * the regex quite complicated as tags can also
2098          * end a sentence. So we'll run through our results
2099          * and strip the period from any tags which end with one.
2100          * Returns array of tags found, or empty array.
2101          *
2102          * @param string $string Post content
2103          *
2104          * @return array List of tag and person names
2105          */
2106         public static function getTags($string)
2107         {
2108                 $ret = [];
2109
2110                 BBCode::performWithEscapedTags($string, ['noparse', 'pre', 'code'], function ($string) use (&$ret) {
2111                         // Convert hashtag links to hashtags
2112                         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2 ', $string);
2113
2114                         // Force line feeds at bbtags
2115                         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2116
2117                         // ignore anything in a bbtag
2118                         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2119
2120                         // Match full names against @tags including the space between first and last
2121                         // We will look these up afterward to see if they are full names or not recognisable.
2122
2123                         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2124                                 foreach ($matches[1] as $match) {
2125                                         if (strstr($match, ']')) {
2126                                                 // we might be inside a bbcode color tag - leave it alone
2127                                                 continue;
2128                                         }
2129
2130                                         if (substr($match, -1, 1) === '.') {
2131                                                 $ret[] = substr($match, 0, -1);
2132                                         } else {
2133                                                 $ret[] = $match;
2134                                         }
2135                                 }
2136                         }
2137
2138                         // Otherwise pull out single word tags. These can be @nickname, @first_last
2139                         // and #hash tags.
2140
2141                         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?\']*[^\^ \x0D\x0A,;:?!\'.])/', $string, $matches)) {
2142                                 foreach ($matches[1] as $match) {
2143                                         if (strstr($match, ']')) {
2144                                                 // we might be inside a bbcode color tag - leave it alone
2145                                                 continue;
2146                                         }
2147
2148                                         // ignore strictly numeric tags like #1
2149                                         if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
2150                                                 continue;
2151                                         }
2152
2153                                         // try not to catch url fragments
2154                                         if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2155                                                 continue;
2156                                         }
2157
2158                                         $ret[] = $match;
2159                                 }
2160                         }
2161                 });
2162
2163                 return array_unique($ret);
2164         }
2165
2166         /**
2167          * Perform a custom function on a text after having escaped blocks enclosed in the provided tag list.
2168          *
2169          * @param string   $text
2170          * @param array    $tagList A list of tag names, e.g ['noparse', 'nobb', 'pre']
2171          * @param callable $callback
2172          * @return string
2173          * @throws Exception
2174          *@see Strings::performWithEscapedBlocks
2175          *
2176          */
2177         public static function performWithEscapedTags(string $text, array $tagList, callable $callback)
2178         {
2179                 $tagList = array_map('preg_quote', $tagList);
2180
2181                 return Strings::performWithEscapedBlocks($text, '#\[(?:' . implode('|', $tagList) . ').*?\[/(?:' . implode('|', $tagList) . ')]#ism', $callback);
2182         }
2183
2184         /**
2185          * Replaces mentions in the provided message body for the provided user and network if any
2186          *
2187          * @param $body
2188          * @param $profile_uid
2189          * @param $network
2190          * @return string
2191          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2192          * @throws \ImagickException
2193          */
2194         public static function setMentions($body, $profile_uid = 0, $network = '')
2195         {
2196                 BBCode::performWithEscapedTags($body, ['noparse', 'pre', 'code', 'img'], function ($body) use ($profile_uid, $network) {
2197                         $tags = BBCode::getTags($body);
2198
2199                         $tagged = [];
2200                         $inform = '';
2201
2202                         foreach ($tags as $tag) {
2203                                 $tag_type = substr($tag, 0, 1);
2204
2205                                 if ($tag_type == Tag::TAG_CHARACTER[Tag::HASHTAG]) {
2206                                         continue;
2207                                 }
2208
2209                                 /*
2210                                  * If we already tagged 'Robert Johnson', don't try and tag 'Robert'.
2211                                  * Robert Johnson should be first in the $tags array
2212                                  */
2213                                 foreach ($tagged as $nextTag) {
2214                                         if (stristr($nextTag, $tag . ' ')) {
2215                                                 continue 2;
2216                                         }
2217                                 }
2218
2219                                 $success = Item::replaceTag($body, $inform, $profile_uid, $tag, $network);
2220
2221                                 if ($success['replaced']) {
2222                                         $tagged[] = $tag;
2223                                 }
2224                         }
2225
2226                         return $body;
2227                 });
2228
2229                 return $body;
2230         }
2231
2232         /**
2233          * @param string      $author  Author display name
2234          * @param string      $profile Author profile URL
2235          * @param string      $avatar  Author profile picture URL
2236          * @param string      $link    Post source URL
2237          * @param string      $posted  Post created date
2238          * @param string|null $guid    Post guid (if any)
2239          * @return string
2240          * @TODO Rewrite to handle over whole record array
2241          */
2242         public static function getShareOpeningTag(string $author, string $profile, string $avatar, string $link, string $posted, string $guid = null)
2243         {
2244                 $header = "[share author='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $author) .
2245                         "' profile='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $profile) .
2246                         "' avatar='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $avatar) .
2247                         "' link='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $link) .
2248                         "' posted='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $posted);
2249
2250                 if ($guid) {
2251                         $header .= "' guid='" . str_replace(["'", "[", "]"], ["&#x27;", "&#x5B;", "&#x5D;"], $guid);
2252                 }
2253
2254                 $header  .= "']";
2255
2256                 return $header;
2257         }
2258 }