src/Content/Text/BBCode.php

   1 <?php
   2 /**
   3  * @file src/Content/Text/BBCode.php
   4  */
   5
   6 namespace Friendica\Content\Text;
   7
   8 use DOMDocument;
   9 use DOMXPath;
  10 use Exception;
  11 use Friendica\BaseObject;
  12 use Friendica\Content\OEmbed;
  13 use Friendica\Content\Smilies;
  14 use Friendica\Core\Cache;
  15 use Friendica\Core\Config;
  16 use Friendica\Core\Hook;
  17 use Friendica\Core\L10n;
  18 use Friendica\Core\Logger;
  19 use Friendica\Core\Protocol;
  20 use Friendica\Core\Renderer;
  21 use Friendica\Core\System;
  22 use Friendica\Model\Contact;
  23 use Friendica\Model\Event;
  24 use Friendica\Network\Probe;
  25 use Friendica\Object\Image;
  26 use Friendica\Util\Map;
  27 use Friendica\Util\Network;
  28 use Friendica\Util\ParseUrl;
  29 use Friendica\Util\Proxy as ProxyUtils;
  30 use Friendica\Util\Strings;
  31 use Friendica\Util\XML;
  32
  33 class BBCode extends BaseObject
  34 {
  35         /**
  36          * @brief Fetches attachment data that were generated the old way
  37          *
  38          * @param string $body Message body
  39          * @return array
  40          *                     'type' -> Message type ("link", "video", "photo")
  41          *                     'text' -> Text before the shared message
  42          *                     'after' -> Text after the shared message
  43          *                     'image' -> Preview image of the message
  44          *                     'url' -> Url to the attached message
  45          *                     'title' -> Title of the attachment
  46          *                     'description' -> Description of the attachment
  47          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
  48          */
  49         private static function getOldAttachmentData($body)
  50         {
  51                 $post = [];
  52
  53                 // Simplify image codes
  54                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  55
  56                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  57                         foreach ($attached as $data) {
  58                                 if (!in_array($data[1], ["type-link", "type-video", "type-photo"])) {
  59                                         continue;
  60                                 }
  61
  62                                 $post["type"] = substr($data[1], 5);
  63
  64                                 $pos = strpos($body, $data[0]);
  65                                 if ($pos > 0) {
  66                                         $post["text"] = trim(substr($body, 0, $pos));
  67                                         $post["after"] = trim(substr($body, $pos + strlen($data[0])));
  68                                 } else {
  69                                         $post["text"] = trim(str_replace($data[0], "", $body));
  70                                         $post["after"] = '';
  71                                 }
  72
  73                                 $attacheddata = $data[2];
  74
  75                                 $URLSearchString = "^\[\]";
  76
  77                                 if (preg_match("/\[img\]([$URLSearchString]*)\[\/img\]/ism", $attacheddata, $matches)) {
  78
  79                                         $picturedata = Image::getInfoFromURL($matches[1]);
  80
  81                                         if ($picturedata) {
  82                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
  83                                                         $post["image"] = $matches[1];
  84                                                 } else {
  85                                                         $post["preview"] = $matches[1];
  86                                                 }
  87                                         }
  88                                 }
  89
  90                                 if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  91                                         $post["url"] = $matches[1];
  92                                         $post["title"] = $matches[2];
  93                                 }
  94                                 if (!empty($post["url"]) && (in_array($post["type"], ["link", "video"]))
  95                                         && preg_match("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
  96                                         $post["url"] = $matches[1];
  97                                 }
  98
  99                                 // Search for description
 100                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
 101                                         $post["description"] = $matches[1];
 102                                 }
 103                         }
 104                 }
 105                 return $post;
 106         }
 107
 108         /**
 109          * @brief Fetches attachment data that were generated with the "attachment" element
 110          *
 111          * @param string $body Message body
 112          * @return array
 113          *                     'type' -> Message type ("link", "video", "photo")
 114          *                     'text' -> Text before the shared message
 115          *                     'after' -> Text after the shared message
 116          *                     'image' -> Preview image of the message
 117          *                     'url' -> Url to the attached message
 118          *                     'title' -> Title of the attachment
 119          *                     'description' -> Description of the attachment
 120          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 121          */
 122         public static function getAttachmentData($body)
 123         {
 124                 $data = [];
 125
 126                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 127                         return self::getOldAttachmentData($body);
 128                 }
 129
 130                 $attributes = $match[2];
 131
 132                 $data["text"] = trim($match[1]);
 133
 134                 $type = "";
 135                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 136                 if (!empty($matches[1])) {
 137                         $type = strtolower($matches[1]);
 138                 }
 139
 140                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 141                 if (!empty($matches[1])) {
 142                         $type = strtolower($matches[1]);
 143                 }
 144
 145                 if ($type == "") {
 146                         return [];
 147                 }
 148
 149                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
 150                         return [];
 151                 }
 152
 153                 if ($type != "") {
 154                         $data["type"] = $type;
 155                 }
 156
 157                 $url = "";
 158                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 159                 if (!empty($matches[1])) {
 160                         $url = $matches[1];
 161                 }
 162
 163                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 164                 if (!empty($matches[1])) {
 165                         $url = $matches[1];
 166                 }
 167
 168                 if ($url != "") {
 169                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 170                 }
 171
 172                 $title = "";
 173                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 174                 if (!empty($matches[1])) {
 175                         $title = $matches[1];
 176                 }
 177
 178                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 179                 if (!empty($matches[1])) {
 180                         $title = $matches[1];
 181                 }
 182
 183                 if ($title != "") {
 184                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
 185                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 186                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
 187                         $data["title"] = $title;
 188                 }
 189
 190                 $image = "";
 191                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 192                 if (!empty($matches[1])) {
 193                         $image = $matches[1];
 194                 }
 195
 196                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 197                 if (!empty($matches[1])) {
 198                         $image = $matches[1];
 199                 }
 200
 201                 if ($image != "") {
 202                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 203                 }
 204
 205                 $preview = "";
 206                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 207                 if (!empty($matches[1])) {
 208                         $preview = $matches[1];
 209                 }
 210
 211                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 212                 if (!empty($matches[1])) {
 213                         $preview = $matches[1];
 214                 }
 215
 216                 if ($preview != "") {
 217                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 218                 }
 219
 220                 $data["description"] = trim($match[3]);
 221
 222                 $data["after"] = trim($match[4]);
 223
 224                 return $data;
 225         }
 226
 227         public static function getAttachedData($body, $item = [])
 228         {
 229                 /*
 230                 - text:
 231                 - type: link, video, photo
 232                 - title:
 233                 - url:
 234                 - image:
 235                 - description:
 236                 - (thumbnail)
 237                 */
 238
 239                 $has_title = !empty($item['title']);
 240                 $plink = defaults($item, 'plink', '');
 241                 $post = self::getAttachmentData($body);
 242
 243                 // if nothing is found, it maybe having an image.
 244                 if (!isset($post["type"])) {
 245                         // Simplify image codes
 246                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 247
 248                         $URLSearchString = "^\[\]";
 249
 250                         $body = preg_replace("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
 251
 252                         if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 253                                 if ((count($pictures) == 1) && !$has_title) {
 254                                         if (!empty($item['object-type']) && ($item['object-type'] == ACTIVITY_OBJ_IMAGE)) {
 255                                                 $data = ['url' => $pictures[0][1], 'type' => 'photo'];
 256                                         } else {
 257                                                 // Checking, if the link goes to a picture
 258                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 259                                         }
 260
 261                                         // Workaround:
 262                                         // Sometimes photo posts to the own album are not detected at the start.
 263                                         // So we seem to cannot use the cache for these cases. That's strange.
 264                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
 265                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 266                                         }
 267
 268                                         if ($data["type"] == "photo") {
 269                                                 $post["type"] = "photo";
 270                                                 if (isset($data["images"][0])) {
 271                                                         $post["image"] = $data["images"][0]["src"];
 272                                                         $post["url"] = $data["url"];
 273                                                 } else {
 274                                                         $post["image"] = $data["url"];
 275                                                 }
 276
 277                                                 $post["preview"] = $pictures[0][2];
 278                                                 $post["text"] = trim(str_replace($pictures[0][0], "", $body));
 279                                         } else {
 280                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 281                                                 if ($imgdata && substr($imgdata["mime"], 0, 6) == "image/") {
 282                                                         $post["type"] = "photo";
 283                                                         $post["image"] = $pictures[0][1];
 284                                                         $post["preview"] = $pictures[0][2];
 285                                                         $post["text"] = trim(str_replace($pictures[0][0], "", $body));
 286                                                 }
 287                                         }
 288                                 } elseif (count($pictures) > 0) {
 289                                         $post["type"] = "link";
 290                                         $post["url"] = $plink;
 291                                         $post["image"] = $pictures[0][2];
 292                                         $post["text"] = $body;
 293                                 }
 294                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 295                                 if ((count($pictures) == 1) && !$has_title) {
 296                                         $post["type"] = "photo";
 297                                         $post["image"] = $pictures[0][1];
 298                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 299                                 } elseif (count($pictures) > 0) {
 300                                         $post["type"] = "link";
 301                                         $post["url"] = $plink;
 302                                         $post["image"] = $pictures[0][1];
 303                                         $post["text"] = $body;
 304                                 }
 305                         }
 306
 307                         // Test for the external links
 308                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
 309                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
 310
 311                         $links = array_merge($links1, $links2);
 312
 313                         // If there is only a single one, then use it.
 314                         // This should cover link posts via API.
 315                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 316                                 $post["type"] = "link";
 317                                 $post["text"] = trim($body);
 318                                 $post["url"] = $links[0][1];
 319                         }
 320
 321                         // Now count the number of external media links
 322                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
 323                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
 324                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
 325                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
 326
 327                         // Add them to the other external links
 328                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 329
 330                         // Are there more than one?
 331                         if (count($links) > 1) {
 332                                 // The post will be the type "text", which means a blog post
 333                                 unset($post["type"]);
 334                                 $post["url"] = $plink;
 335                         }
 336
 337                         if (!isset($post["type"])) {
 338                                 $post["type"] = "text";
 339                                 $post["text"] = trim($body);
 340                         }
 341                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 342                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
 343
 344                         if (isset($data["images"][0])) {
 345                                 $post["image"] = $data["images"][0]["src"];
 346                         }
 347                 }
 348
 349                 return $post;
 350         }
 351
 352         /**
 353          * @brief Converts a BBCode text into plaintext
 354          *
 355          * @param      $text
 356          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 357          *
 358          * @return string
 359          */
 360         public static function toPlaintext($text, $keep_urls = true)
 361         {
 362                 $naked_text = preg_replace('/\[(.+?)\]\s*/','', $text);
 363                 if (!$keep_urls) {
 364                         $naked_text = preg_replace('#https?\://[^\s<]+[^\s\.\)]#i', '', $naked_text);
 365                 }
 366
 367                 return $naked_text;
 368         }
 369
 370         private static function proxyUrl($image, $simplehtml = false)
 371         {
 372                 // Only send proxied pictures to API and for internal display
 373                 if (in_array($simplehtml, [false, 2])) {
 374                         return ProxyUtils::proxifyUrl($image);
 375                 } else {
 376                         return $image;
 377                 }
 378         }
 379
 380         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 381         {
 382                 // Suppress "view full size"
 383                 if (intval(Config::get('system', 'no_view_full_size'))) {
 384                         $include_link = false;
 385                 }
 386
 387                 // Picture addresses can contain special characters
 388                 $s = htmlspecialchars_decode($srctext);
 389
 390                 $matches = null;
 391                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 392                 if ($c) {
 393                         foreach ($matches as $mtch) {
 394                                 Logger::log('scale_external_image: ' . $mtch[1]);
 395
 396                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 397                                 if (stristr($mtch[1], $hostname)) {
 398                                         continue;
 399                                 }
 400
 401                                 // $scale_replace, if passed, is an array of two elements. The
 402                                 // first is the name of the full-size image. The second is the
 403                                 // name of a remote, scaled-down version of the full size image.
 404                                 // This allows Friendica to display the smaller remote image if
 405                                 // one exists, while still linking to the full-size image
 406                                 if ($scale_replace) {
 407                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 408                                 } else {
 409                                         $scaled = $mtch[1];
 410                                 }
 411                                 $i = Network::fetchUrl($scaled);
 412                                 if (!$i) {
 413                                         return $srctext;
 414                                 }
 415
 416                                 // guess mimetype from headers or filename
 417                                 $type = Image::guessType($mtch[1], true);
 418
 419                                 if ($i) {
 420                                         $Image = new Image($i, $type);
 421                                         if ($Image->isValid()) {
 422                                                 $orig_width = $Image->getWidth();
 423                                                 $orig_height = $Image->getHeight();
 424
 425                                                 if ($orig_width > 640 || $orig_height > 640) {
 426                                                         $Image->scaleDown(640);
 427                                                         $new_width = $Image->getWidth();
 428                                                         $new_height = $Image->getHeight();
 429                                                         Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG);
 430                                                         $s = str_replace(
 431                                                                 $mtch[0],
 432                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 433                                                                 . "\n" . (($include_link)
 434                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 435                                                                         : ''),
 436                                                                 $s
 437                                                         );
 438                                                         Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG);
 439                                                 }
 440                                         }
 441                                 }
 442                         }
 443                 }
 444
 445                 // replace the special char encoding
 446                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 447                 return $s;
 448         }
 449
 450         /**
 451          * The purpose of this function is to apply system message length limits to
 452          * imported messages without including any embedded photos in the length
 453          *
 454          * @brief Truncates imported message body string length to max_import_size
 455          * @param string $body
 456          * @return string
 457          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 458          */
 459         public static function limitBodySize($body)
 460         {
 461                 $maxlen = Config::get('config', 'max_import_size', 0);
 462
 463                 // If the length of the body, including the embedded images, is smaller
 464                 // than the maximum, then don't waste time looking for the images
 465                 if ($maxlen && (strlen($body) > $maxlen)) {
 466
 467                         Logger::log('the total body length exceeds the limit', Logger::DEBUG);
 468
 469                         $orig_body = $body;
 470                         $new_body = '';
 471                         $textlen = 0;
 472
 473                         $img_start = strpos($orig_body, '[img');
 474                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 475                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 476                         while (($img_st_close !== false) && ($img_end !== false)) {
 477
 478                                 $img_st_close++; // make it point to AFTER the closing bracket
 479                                 $img_end += $img_start;
 480                                 $img_end += strlen('[/img]');
 481
 482                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 483                                         // This is an embedded image
 484
 485                                         if (($textlen + $img_start) > $maxlen) {
 486                                                 if ($textlen < $maxlen) {
 487                                                         Logger::log('the limit happens before an embedded image', Logger::DEBUG);
 488                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 489                                                         $textlen = $maxlen;
 490                                                 }
 491                                         } else {
 492                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 493                                                 $textlen += $img_start;
 494                                         }
 495
 496                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 497                                 } else {
 498
 499                                         if (($textlen + $img_end) > $maxlen) {
 500                                                 if ($textlen < $maxlen) {
 501                                                         Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG);
 502                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 503                                                         $textlen = $maxlen;
 504                                                 }
 505                                         } else {
 506                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 507                                                 $textlen += $img_end;
 508                                         }
 509                                 }
 510                                 $orig_body = substr($orig_body, $img_end);
 511
 512                                 if ($orig_body === false) {
 513                                         // in case the body ends on a closing image tag
 514                                         $orig_body = '';
 515                                 }
 516
 517                                 $img_start = strpos($orig_body, '[img');
 518                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 519                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 520                         }
 521
 522                         if (($textlen + strlen($orig_body)) > $maxlen) {
 523                                 if ($textlen < $maxlen) {
 524                                         Logger::log('the limit happens after the end of the last image', Logger::DEBUG);
 525                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 526                                 }
 527                         } else {
 528                                 Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG);
 529                                 $new_body = $new_body . $orig_body;
 530                         }
 531
 532                         return $new_body;
 533                 } else {
 534                         return $body;
 535                 }
 536         }
 537
 538         /**
 539          * Processes [attachment] tags
 540          *
 541          * Note: Can produce a [bookmark] tag in the returned string
 542          *
 543          * @brief Processes [attachment] tags
 544          * @param string   $return
 545          * @param bool|int $simplehtml
 546          * @param bool     $tryoembed
 547          * @return string
 548          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 549          */
 550         private static function convertAttachment($return, $simplehtml = false, $tryoembed = true)
 551         {
 552                 $data = self::getAttachmentData($return);
 553                 if (empty($data) || empty($data["url"])) {
 554                         return $return;
 555                 }
 556
 557                 if (isset($data["title"])) {
 558                         $data["title"] = strip_tags($data["title"]);
 559                         $data["title"] = str_replace(["http://", "https://"], "", $data["title"]);
 560                 } else {
 561                         $data["title"] = null;
 562                 }
 563
 564                 if (((strpos($data["text"], "[img=") !== false) || (strpos($data["text"], "[img]") !== false) || Config::get('system', 'always_show_preview')) && !empty($data["image"])) {
 565                         $data["preview"] = $data["image"];
 566                         $data["image"] = "";
 567                 }
 568
 569                 $return = '';
 570                 if ($simplehtml == 7) {
 571                         $return = self::convertUrlForOStatus($data["url"]);
 572                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
 573                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data["url"], $data["title"]);
 574                 } else {
 575                         try {
 576                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 577                                         $return = OEmbed::getHTML($data['url'], $data['title']);
 578                                 } else {
 579                                         throw new Exception('OEmbed is disabled for this attachment.');
 580                                 }
 581                         } catch (Exception $e) {
 582                                 $data["title"] = defaults($data, 'title', $data['url']);
 583
 584                                 if ($simplehtml != 4) {
 585                                         $return = sprintf('<div class="type-%s">', $data["type"]);
 586                                 }
 587
 588                                 if (!empty($data['title']) && !empty($data['url'])) {
 589                                         if (!empty($data["image"]) && empty($data["text"]) && ($data["type"] == "photo")) {
 590                                                 $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data["url"], self::proxyUrl($data["image"], $simplehtml), $data["title"]);
 591                                         } else {
 592                                                 if (!empty($data["image"])) {
 593                                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a><br />', $data["url"], self::proxyUrl($data["image"], $simplehtml), $data["title"]);
 594                                                 } elseif (!empty($data["preview"])) {
 595                                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br />', $data["url"], self::proxyUrl($data["preview"], $simplehtml), $data["title"]);
 596                                                 }
 597                                                 $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 598                                         }
 599                                 }
 600
 601                                 if (!empty($data["description"]) && $data["description"] != $data["title"]) {
 602                                         // Sanitize the HTML by converting it to BBCode
 603                                         $bbcode = HTML::toBBCode($data["description"]);
 604                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 605                                 }
 606
 607                                 if (!empty($data['url'])) {
 608                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 609                                 }
 610
 611                                 if ($simplehtml != 4) {
 612                                         $return .= '</div>';
 613                                 }
 614                         }
 615                 }
 616
 617                 return trim(defaults($data, 'text', '') . ' ' . $return . ' ' . defaults($data, 'after', ''));
 618         }
 619
 620         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 621         {
 622                 $data = self::getAttachmentData($Text);
 623
 624                 if (!$data) {
 625                         return $Text;
 626                 } elseif ($nolink) {
 627                         return $data["text"] . defaults($data, 'after', '');
 628                 }
 629
 630                 $title = htmlentities(defaults($data, 'title', ''), ENT_QUOTES, 'UTF-8', false);
 631                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
 632                 if ($plaintext || (($title != "") && strstr($text, $title))) {
 633                         $data["title"] = $data["url"];
 634                 } elseif (($text != "") && strstr($title, $text)) {
 635                         $data["text"] = $data["title"];
 636                         $data["title"] = $data["url"];
 637                 }
 638
 639                 if (empty($data["text"]) && !empty($data["title"]) && empty($data["url"])) {
 640                         return $data["title"] . $data["after"];
 641                 }
 642
 643                 // If the link already is included in the post, don't add it again
 644                 if (!empty($data["url"]) && strpos($data["text"], $data["url"])) {
 645                         return $data["text"] . $data["after"];
 646                 }
 647
 648                 $text = $data["text"];
 649
 650                 if (!empty($data["url"]) && !empty($data["title"])) {
 651                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
 652                 } elseif (!empty($data["url"])) {
 653                         $text .= "\n[url]" . $data["url"] . "[/url]";
 654                 }
 655
 656                 return $text . "\n" . $data["after"];
 657         }
 658
 659         /**
 660          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 661          *
 662          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 663          * @param array $match Array with the matching values
 664          * @return string reformatted link including HTML codes
 665          */
 666         private static function convertUrlForOStatusCallback($match)
 667         {
 668                 $url = $match[1];
 669
 670                 if (isset($match[2]) && ($match[1] != $match[2])) {
 671                         return $match[0];
 672                 }
 673
 674                 $parts = parse_url($url);
 675                 if (!isset($parts['scheme'])) {
 676                         return $match[0];
 677                 }
 678
 679                 return self::convertUrlForOStatus($url);
 680         }
 681
 682         /**
 683          * @brief Converts [url] BBCodes in a format that looks fine on OStatus systems.
 684          * @param string $url URL that is about to be reformatted
 685          * @return string reformatted link including HTML codes
 686          */
 687         private static function convertUrlForOStatus($url)
 688         {
 689                 $parts = parse_url($url);
 690                 $scheme = $parts['scheme'] . '://';
 691                 $styled_url = str_replace($scheme, '', $url);
 692
 693                 if (strlen($styled_url) > 30) {
 694                         $styled_url = substr($styled_url, 0, 30) . "…";
 695                 }
 696
 697                 $html = '<a href="%s" target="_blank">%s</a>';
 698
 699                 return sprintf($html, $url, $styled_url);
 700         }
 701
 702         /*
 703          * [noparse][i]italic[/i][/noparse] turns into
 704          * [noparse][ i ]italic[ /i ][/noparse],
 705          * to hide them from parser.
 706          */
 707         private static function escapeNoparseCallback($match)
 708         {
 709                 $whole_match = $match[0];
 710                 $captured = $match[1];
 711                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 712                 $new_str = str_replace($captured, $spacefied, $whole_match);
 713                 return $new_str;
 714         }
 715
 716         /*
 717          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 718          * now turns back and the [noparse] tags are trimed
 719          * returning [i]italic[/i]
 720          */
 721         private static function unescapeNoparseCallback($match)
 722         {
 723                 $captured = $match[1];
 724                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 725                 return $unspacefied;
 726         }
 727
 728         /**
 729          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 730          * occurrences
 731          *
 732          * @param string $text        Text to search
 733          * @param string $name        Tag name
 734          * @param int    $occurrences Number of first occurrences to skip
 735          * @return boolean|array
 736          */
 737         public static function getTagPosition($text, $name, $occurrences = 0)
 738         {
 739                 if ($occurrences < 0) {
 740                         $occurrences = 0;
 741                 }
 742
 743                 $start_open = -1;
 744                 for ($i = 0; $i <= $occurrences; $i++) {
 745                         if ($start_open !== false) {
 746                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 747                         }
 748                 }
 749
 750                 if ($start_open === false) {
 751                         return false;
 752                 }
 753
 754                 $start_equal = strpos($text, '=', $start_open);
 755                 $start_close = strpos($text, ']', $start_open);
 756
 757                 if ($start_close === false) {
 758                         return false;
 759                 }
 760
 761                 $start_close++;
 762
 763                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 764
 765                 if ($end_open === false) {
 766                         return false;
 767                 }
 768
 769                 $res = [
 770                         'start' => [
 771                                 'open' => $start_open,
 772                                 'close' => $start_close
 773                         ],
 774                         'end' => [
 775                                 'open' => $end_open,
 776                                 'close' => $end_open + strlen('[/' . $name . ']')
 777                         ],
 778                 ];
 779
 780                 if ($start_equal !== false) {
 781                         $res['start']['equal'] = $start_equal + 1;
 782                 }
 783
 784                 return $res;
 785         }
 786
 787         /**
 788          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 789          *
 790          * @param string $pattern Preg pattern string
 791          * @param string $replace Preg replace string
 792          * @param string $name    BBCode tag name
 793          * @param string $text    Text to search
 794          * @return string
 795          */
 796         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 797         {
 798                 $occurrences = 0;
 799                 $pos = self::getTagPosition($text, $name, $occurrences);
 800                 while ($pos !== false && $occurrences++ < 1000) {
 801                         $start = substr($text, 0, $pos['start']['open']);
 802                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 803                         $end = substr($text, $pos['end']['close']);
 804                         if ($end === false) {
 805                                 $end = '';
 806                         }
 807
 808                         $subject = preg_replace($pattern, $replace, $subject);
 809                         $text = $start . $subject . $end;
 810
 811                         $pos = self::getTagPosition($text, $name, $occurrences);
 812                 }
 813
 814                 return $text;
 815         }
 816
 817         private static function extractImagesFromItemBody($body)
 818         {
 819                 $saved_image = [];
 820                 $orig_body = $body;
 821                 $new_body = '';
 822
 823                 $cnt = 0;
 824                 $img_start = strpos($orig_body, '[img');
 825                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 826                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 827                 while (($img_st_close !== false) && ($img_end !== false)) {
 828                         $img_st_close++; // make it point to AFTER the closing bracket
 829                         $img_end += $img_start;
 830
 831                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 832                                 // This is an embedded image
 833                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 834                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 835
 836                                 $cnt++;
 837                         } else {
 838                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 839                         }
 840
 841                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 842
 843                         if ($orig_body === false) {
 844                                 // in case the body ends on a closing image tag
 845                                 $orig_body = '';
 846                         }
 847
 848                         $img_start = strpos($orig_body, '[img');
 849                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 850                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 851                 }
 852
 853                 $new_body = $new_body . $orig_body;
 854
 855                 return ['body' => $new_body, 'images' => $saved_image];
 856         }
 857
 858         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 859         {
 860                 $newbody = $body;
 861
 862                 $cnt = 0;
 863                 foreach ($images as $image) {
 864                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 865                         // it loops over the array starting from the first element and going sequentially
 866                         // to the last element
 867                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 868                                 '<img src="' . self::proxyUrl($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
 869                         $cnt++;
 870                 }
 871
 872                 return $newbody;
 873         }
 874
 875         /**
 876          * This function converts a [share] block to text according to a provided callback function whose signature is:
 877          *
 878          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 879          *
 880          * Where:
 881          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 882          * data lookup
 883          * - $author_contact is a contact record array
 884          * - $content is the inner content of the [share] block
 885          * - $is_quote_share indicates whether there's any content before the [share] block
 886          * - Return value is the string that should replace the [share] block in the provided text
 887          *
 888          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 889          *
 890          * @param  string   $text     A BBCode string
 891          * @param  callable $callback
 892          * @return string The BBCode string with all [share] blocks replaced
 893          */
 894         public static function convertShare($text, callable $callback)
 895         {
 896                 $return = preg_replace_callback(
 897                         "/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
 898                         function ($match) use ($callback) {
 899                                 $attribute_string = $match[2];
 900
 901                                 $attributes = [];
 902                                 foreach(['author', 'profile', 'avatar', 'link', 'posted'] as $field) {
 903                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 904                                         $attributes[$field] = html_entity_decode(defaults($matches, 2, ''), ENT_QUOTES, 'UTF-8');
 905                                 }
 906
 907                                 // We only call this so that a previously unknown contact can be added.
 908                                 // This is important for the function "Model\Contact::getDetailsByURL()".
 909                                 // This function then can fetch an entry from the contact table.
 910                                 Contact::getIdForURL($attributes['profile'], 0, true);
 911
 912                                 $author_contact = Contact::getDetailsByURL($attributes['profile']);
 913                                 $author_contact['addr'] = defaults($author_contact, 'addr' , Protocol::getAddrFromProfileUrl($attributes['profile']));
 914
 915                                 $attributes['author']   = defaults($author_contact, 'name' , $attributes['author']);
 916                                 $attributes['avatar']   = defaults($author_contact, 'micro', $attributes['avatar']);
 917                                 $attributes['profile']  = defaults($author_contact, 'url'  , $attributes['profile']);
 918
 919                                 if ($attributes['avatar']) {
 920                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
 921                                 }
 922
 923                                 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
 924                         },
 925                         $text
 926                 );
 927
 928                 return $return;
 929         }
 930
 931         /**
 932          * Default [share] tag conversion callback
 933          *
 934          * Note: Can produce a [bookmark] tag in the output
 935          *
 936          * @see BBCode::convertShare()
 937          * @param array   $attributes     [share] block attribute values
 938          * @param array   $author_contact Contact row of the shared author
 939          * @param string  $content        Inner content of the [share] block
 940          * @param boolean $is_quote_share Whether there is content before the [share] block
 941          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
 942          * @return string
 943          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 944          */
 945         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
 946         {
 947                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
 948
 949                 switch ($simplehtml) {
 950                         case 1:
 951                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' <a href="' . $attributes['profile'] . '">' . $mention . '</a>: </p>' . "\n" . '«' . $content . '»';
 952                                 break;
 953                         case 2:
 954                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
 955                                 break;
 956                         case 3: // Diaspora
 957                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
 958                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
 959                                 } else {
 960                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
 961
 962                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
 963                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
 964                                         }
 965
 966                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
 967
 968                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
 969                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
 970                                         }
 971                                 }
 972
 973                                 break;
 974                         case 4:
 975                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
 976                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
 977                                 $headline .= ':</b></p>' . "\n";
 978
 979                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
 980
 981                                 break;
 982                         case 5:
 983                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
 984                                 break;
 985                         case 7: // statusnet/GNU Social
 986                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
 987                                 break;
 988                         case 9: // Google+
 989                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n";
 990                                 $text .= '<p>' . $content . '</p>' . "\n";
 991
 992                                 if ($attributes['link'] != '') {
 993                                         $text .= '<p>' . $attributes['link'] . '</p>';
 994                                 }
 995                                 break;
 996                         default:
 997                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
 998                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($attributes['link'])) {
 999                                         try {
1000                                                 $text = ($is_quote_share? '<br />' : '') . OEmbed::getHTML($attributes['link']);
1001                                         } catch (Exception $e) {
1002                                                 $text = ($is_quote_share? '<br />' : '') . sprintf('[bookmark=%s]%s[/bookmark]', $attributes['link'], $content);
1003                                         }
1004                                 } else {
1005                                         $text = ($is_quote_share? "\n" : '');
1006
1007                                         $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1008                                         $text .= Renderer::replaceMacros($tpl, [
1009                                                 '$profile' => $attributes['profile'],
1010                                                 '$avatar'  => $attributes['avatar'],
1011                                                 '$author'  => $attributes['author'],
1012                                                 '$link'    => $attributes['link'],
1013                                                 '$posted'  => $attributes['posted'],
1014                                                 '$content' => trim($content)
1015                                         ]);
1016                                 }
1017                                 break;
1018                 }
1019
1020                 return $text;
1021         }
1022
1023         private static function removePictureLinksCallback($match)
1024         {
1025                 $text = Cache::get($match[1]);
1026
1027                 if (is_null($text)) {
1028                         $a = self::getApp();
1029
1030                         $stamp1 = microtime(true);
1031
1032                         $ch = @curl_init($match[1]);
1033                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1034                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1035                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1036                         @curl_exec($ch);
1037                         $curl_info = @curl_getinfo($ch);
1038
1039                         $a->getProfiler()->saveTimestamp($stamp1, "network", System::callstack());
1040
1041                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1042                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1043                         } else {
1044                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1045
1046                                 // if its not a picture then look if its a page that contains a picture link
1047                                 $body = Network::fetchUrl($match[1]);
1048
1049                                 $doc = new DOMDocument();
1050                                 @$doc->loadHTML($body);
1051                                 $xpath = new DOMXPath($doc);
1052                                 $list = $xpath->query("//meta[@name]");
1053                                 foreach ($list as $node) {
1054                                         $attr = [];
1055
1056                                         if ($node->attributes->length) {
1057                                                 foreach ($node->attributes as $attribute) {
1058                                                         $attr[$attribute->name] = $attribute->value;
1059                                                 }
1060                                         }
1061
1062                                         if (strtolower($attr["name"]) == "twitter:image") {
1063                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1064                                         }
1065                                 }
1066                         }
1067                         Cache::set($match[1], $text);
1068                 }
1069
1070                 return $text;
1071         }
1072
1073         private static function expandLinksCallback($match)
1074         {
1075                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1076                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1077                 } else {
1078                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1079                 }
1080         }
1081
1082         private static function cleanPictureLinksCallback($match)
1083         {
1084                 $text = Cache::get($match[1]);
1085
1086                 if (is_null($text)) {
1087                         $a = self::getApp();
1088
1089                         $stamp1 = microtime(true);
1090
1091                         $ch = @curl_init($match[1]);
1092                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1093                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1094                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1095                         @curl_exec($ch);
1096                         $curl_info = @curl_getinfo($ch);
1097
1098                         $a->getProfiler()->saveTimestamp($stamp1, "network", System::callstack());
1099
1100                         // if its a link to a picture then embed this picture
1101                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1102                                 $text = "[img]" . $match[1] . "[/img]";
1103                         } else {
1104                                 $text = "[img]" . $match[2] . "[/img]";
1105
1106                                 // if its not a picture then look if its a page that contains a picture link
1107                                 $body = Network::fetchUrl($match[1]);
1108
1109                                 $doc = new DOMDocument();
1110                                 @$doc->loadHTML($body);
1111                                 $xpath = new DOMXPath($doc);
1112                                 $list = $xpath->query("//meta[@name]");
1113                                 foreach ($list as $node) {
1114                                         $attr = [];
1115                                         if ($node->attributes->length) {
1116                                                 foreach ($node->attributes as $attribute) {
1117                                                         $attr[$attribute->name] = $attribute->value;
1118                                                 }
1119                                         }
1120
1121                                         if (strtolower($attr["name"]) == "twitter:image") {
1122                                                 $text = "[img]" . $attr["content"] . "[/img]";
1123                                         }
1124                                 }
1125                         }
1126                         Cache::set($match[1], $text);
1127                 }
1128
1129                 return $text;
1130         }
1131
1132         public static function cleanPictureLinks($text)
1133         {
1134                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1135                 return $return;
1136         }
1137
1138         /**
1139          * @brief Converts a BBCode message to HTML message
1140          *
1141          * BBcode 2 HTML was written by WAY2WEB.net
1142          * extended to work with Mistpark/Friendica - Mike Macgirvin
1143          *
1144          * Simple HTML values meaning:
1145          * - 0: Friendica display
1146          * - 1: Unused
1147          * - 2: Used for Google+, Windows Phone push, Friendica API
1148          * - 3: Used before converting to Markdown in bb2diaspora.php
1149          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1150          * - 5: Unused
1151          * - 6: Unused
1152          * - 7: Used for dfrn, OStatus
1153          * - 8: Used for WP backlink text setting
1154          *
1155          * @param string $text
1156          * @param bool   $try_oembed
1157          * @param int    $simple_html
1158          * @param bool   $for_plaintext
1159          * @return string
1160          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1161          */
1162         public static function convert($text, $try_oembed = true, $simple_html = 0, $for_plaintext = false)
1163         {
1164                 $a = self::getApp();
1165
1166                 /*
1167                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1168                  *
1169                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1170                  * $match[1] = $url
1171                  * $match[2] = $title or absent
1172                  */
1173                 $try_oembed_callback = function ($match)
1174                 {
1175                         $url = $match[1];
1176                         $title = defaults($match, 2, null);
1177
1178                         try {
1179                                 $return = OEmbed::getHTML($url, $title);
1180                         } catch (Exception $ex) {
1181                                 $return = $match[0];
1182                         }
1183
1184                         return $return;
1185                 };
1186
1187                 // Extracting code blocks before the whitespace processing and the autolinker
1188                 $codeblocks = [];
1189
1190                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1191                         function ($matches) use (&$codeblocks) {
1192                                 $return = '#codeblock-' . count($codeblocks) . '#';
1193                                 if (strpos($matches[2], "\n") !== false) {
1194                                         $codeblocks[] = '<pre><code class="language-' . trim($matches[1]) . '">' . trim($matches[2], "\n\r") . '</code></pre>';
1195                                 } else {
1196                                         $codeblocks[] = '<code>' . $matches[2] . '</code>';
1197                                 }
1198
1199                                 return $return;
1200                         },
1201                         $text
1202                 );
1203
1204                 // Hide all [noparse] contained bbtags by spacefying them
1205                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1206
1207                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1208                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1209                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1210
1211                 // Remove the abstract element. It is a non visible element.
1212                 $text = self::stripAbstract($text);
1213
1214                 // Move all spaces out of the tags
1215                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1216                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1217
1218                 // Extract the private images which use data urls since preg has issues with
1219                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1220                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1221
1222                 $extracted = self::extractImagesFromItemBody($text);
1223                 $text = $extracted['body'];
1224                 $saved_image = $extracted['images'];
1225
1226                 // If we find any event code, turn it into an event.
1227                 // After we're finished processing the bbcode we'll
1228                 // replace all of the event code with a reformatted version.
1229
1230                 $ev = Event::fromBBCode($text);
1231
1232                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1233                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1234
1235                 $text = str_replace("<", "&lt;", $text);
1236                 $text = str_replace(">", "&gt;", $text);
1237
1238                 // remove some newlines before the general conversion
1239                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1240                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1241
1242                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1243                 if (!$try_oembed) {
1244                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1245                 }
1246
1247                 // Convert new line chars to html <br /> tags
1248
1249                 // nlbr seems to be hopelessly messed up
1250                 //      $Text = nl2br($Text);
1251
1252                 // We'll emulate it.
1253
1254                 $text = trim($text);
1255                 $text = str_replace("\r\n", "\n", $text);
1256
1257                 // Remove linefeeds inside of the table elements. See issue #6799
1258                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1259                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1260                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1261                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1262                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1263                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1264                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1265                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1266                 do {
1267                         $oldtext = $text;
1268                         $text = str_replace($search, $replace, $text);
1269                 } while ($oldtext != $text);
1270
1271                 // Replace these here only once
1272                 $search = ["\n[table]", "[/table]\n"];
1273                 $replace = ["[table]", "[/table]"];
1274                 $text = str_replace($search, $replace, $text);
1275
1276                 // removing multiplicated newlines
1277                 if (Config::get("system", "remove_multiplicated_lines")) {
1278                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1279                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1280                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1281                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1282                         do {
1283                                 $oldtext = $text;
1284                                 $text = str_replace($search, $replace, $text);
1285                         } while ($oldtext != $text);
1286                 }
1287
1288                 // Set up the parameters for a URL search string
1289                 $URLSearchString = "^\[\]";
1290                 // Set up the parameters for a MAIL search string
1291                 $MAILSearchString = $URLSearchString;
1292
1293                 // Handle attached links or videos
1294                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1295
1296                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1297                 if (!$for_plaintext) {
1298                         $text = preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1299                         if ($simple_html == 7) {
1300                                 $text = preg_replace_callback("/\[url\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1301                                 $text = preg_replace_callback("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1302                         }
1303                 } else {
1304                         $text = preg_replace("(\[url\]([$URLSearchString]*)\[\/url\])ism", " $1 ", $text);
1305                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1306                 }
1307
1308                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1309
1310                 // Remove all hashtag addresses
1311                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7])) {
1312                         $text = preg_replace("/([#@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $text);
1313                 } elseif ($simple_html == 3) {
1314                         // The ! is converted to @ since Diaspora only understands the @
1315                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1316                                 '@<a href="$2">$3</a>',
1317                                 $text);
1318                 } elseif ($simple_html == 7) {
1319                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1320                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1321                                 $text);
1322                 } elseif (!$simple_html) {
1323                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1324                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1325                                 $text);
1326                 }
1327
1328                 // Bookmarks in red - will be converted to bookmarks in friendica
1329                 $text = preg_replace("/#\^\[url\]([$URLSearchString]*)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1330                 $text = preg_replace("/#\^\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1331                 $text = preg_replace("/#\[url\=[$URLSearchString]*\]\^\[\/url\]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i",
1332                                         "[bookmark=$1]$2[/bookmark]", $text);
1333
1334                 if (in_array($simple_html, [2, 6, 7, 8, 9])) {
1335                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1336                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1337                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1338                 }
1339
1340                 if ($simple_html == 5) {
1341                         $text = preg_replace("/[^#@!]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1342                 }
1343
1344                 // Perform URL Search
1345                 if ($try_oembed) {
1346                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1347                 }
1348
1349                 if ($simple_html == 5) {
1350                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1351                 } else {
1352                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1353                 }
1354
1355                 // Handle Diaspora posts
1356                 $text = preg_replace_callback(
1357                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1358                         function ($match) {
1359                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1360                         }, $text
1361                 );
1362
1363                 $text = preg_replace_callback(
1364                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1365                         function ($match) {
1366                                 return "[url=" . System::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1367                         }, $text
1368                 );
1369
1370                 // Server independent link to posts and comments
1371                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1372                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1373                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1374
1375                 /* Tag conversion
1376                  * Supports:
1377                  * - #[url=<anything>]<term>[/url]
1378                  * - [url=<anything>]#<term>[/url]
1379                  */
1380                 $text = preg_replace_callback("/(?:#\[url\=[$URLSearchString]*\]|\[url\=[$URLSearchString]*\]#)(.*?)\[\/url\]/ism", function($matches) {
1381                         return '#<a href="'
1382                                 . System::baseUrl()     . '/search?tag=' . rawurlencode($matches[1])
1383                                 . '" class="tag" title="' . XML::escape($matches[1]) . '">'
1384                                 . XML::escape($matches[1])
1385                                 . '</a>';
1386                 }, $text);
1387
1388                 // We need no target="_blank" for local links
1389                 // convert links start with System::baseUrl() as local link without the target="_blank" attribute
1390                 $escapedBaseUrl = preg_quote(System::baseUrl(), '/');
1391                 $text = preg_replace("/\[url\](".$escapedBaseUrl."[$URLSearchString]*)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1392                 $text = preg_replace("/\[url\=(".$escapedBaseUrl."[$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1393
1394                 $text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1395                 $text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1396
1397                 // Red compatibility, though the link can't be authenticated on Friendica
1398                 $text = preg_replace("/\[zrl\=([$URLSearchString]*)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1399
1400
1401                 // we may need to restrict this further if it picks up too many strays
1402                 // link acct:user@host to a webfinger profile redirector
1403
1404                 $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="' . System::baseUrl() . '/acctlink?addr=$1@$2" target="extlink">acct:$1@$2</a>', $text);
1405
1406                 // Perform MAIL Search
1407                 $text = preg_replace("/\[mail\]([$MAILSearchString]*)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1408                 $text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1409
1410                 // leave open the posibility of [map=something]
1411                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1412
1413                 if (strpos($text, '[/map]') !== false) {
1414                         $text = preg_replace_callback(
1415                                 "/\[map\](.*?)\[\/map\]/ism",
1416                                 function ($match) use ($simple_html) {
1417                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1418                                 },
1419                                 $text
1420                         );
1421                 }
1422                 if (strpos($text, '[map=') !== false) {
1423                         $text = preg_replace_callback(
1424                                 "/\[map=(.*?)\]/ism",
1425                                 function ($match) use ($simple_html) {
1426                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1427                                 },
1428                                 $text
1429                         );
1430                 }
1431                 if (strpos($text, '[map]') !== false) {
1432                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1433                 }
1434
1435                 // Check for headers
1436                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1437                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1438                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1439                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1440                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1441                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1442
1443                 // Check for paragraph
1444                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1445
1446                 // Check for bold text
1447                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1448
1449                 // Check for Italics text
1450                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1451
1452                 // Check for Underline text
1453                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1454
1455                 // Check for strike-through text
1456                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1457
1458                 // Check for over-line text
1459                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1460
1461                 // Check for colored text
1462                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1463
1464                 // Check for sized text
1465                 // [size=50] --> font-size: 50px (with the unit).
1466                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1467                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1468
1469                 // Check for centered text
1470                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1471
1472                 // Check for list text
1473                 $text = str_replace("[*]", "<li>", $text);
1474
1475                 // Check for style sheet commands
1476                 $text = preg_replace_callback(
1477                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1478                         function ($match) {
1479                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1480                         },
1481                         $text
1482                 );
1483
1484                 // Check for CSS classes
1485                 $text = preg_replace_callback(
1486                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1487                         function ($match) {
1488                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1489                         },
1490                         $text
1491                 );
1492
1493                 // handle nested lists
1494                 $endlessloop = 0;
1495
1496                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1497                            ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1498                            ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1499                            ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1500                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1501                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1502                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1503                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1504                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1505                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1506                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1507                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1508                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1509                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1510                 }
1511
1512                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1513                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1514                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1515                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1516
1517                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1518                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1519
1520                 $text = str_replace('[hr]', '<hr />', $text);
1521
1522                 // This is actually executed in Item::prepareBody()
1523
1524                 $text = str_replace('[nosmile]', '', $text);
1525
1526                 // Check for font change text
1527                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1528
1529                 // Declare the format for [spoiler] layout
1530                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
1531
1532                 // Check for [spoiler] text
1533                 // handle nested quotes
1534                 $endlessloop = 0;
1535                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1536                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", "$SpoilerLayout", $text);
1537                 }
1538
1539                 // Check for [spoiler=Author] text
1540
1541                 $t_wrote = L10n::t('$1 wrote:');
1542
1543                 // handle nested quotes
1544                 $endlessloop = 0;
1545                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1546                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1547                                                  "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1548                                                  $text);
1549                 }
1550
1551                 // Declare the format for [quote] layout
1552                 $QuoteLayout = '<blockquote>$1</blockquote>';
1553
1554                 // Check for [quote] text
1555                 // handle nested quotes
1556                 $endlessloop = 0;
1557                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1558                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1559                 }
1560
1561                 // Check for [quote=Author] text
1562
1563                 $t_wrote = L10n::t('$1 wrote:');
1564
1565                 // handle nested quotes
1566                 $endlessloop = 0;
1567                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1568                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1569                                                  "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1570                                                  $text);
1571                 }
1572
1573
1574                 // [img=widthxheight]image source[/img]
1575                 $text = preg_replace_callback(
1576                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1577                         function ($matches) use ($simple_html) {
1578                                 if (strpos($matches[3], "data:image/") === 0) {
1579                                         return $matches[0];
1580                                 }
1581
1582                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1583                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1584                         },
1585                         $text
1586                 );
1587
1588                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1589                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1590
1591                 $text = preg_replace_callback("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism",
1592                         function ($matches) use ($simple_html) {
1593                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1594                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1595                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '">';
1596                         },
1597                         $text);
1598
1599                 // Images
1600                 // [img]pathtoimage[/img]
1601                 $text = preg_replace_callback(
1602                         "/\[img\](.*?)\[\/img\]/ism",
1603                         function ($matches) use ($simple_html) {
1604                                 if (strpos($matches[1], "data:image/") === 0) {
1605                                         return $matches[0];
1606                                 }
1607
1608                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1609                                 return "[img]" . $matches[1] . "[/img]";
1610                         },
1611                         $text
1612                 );
1613
1614                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1615                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1616
1617                 // Shared content
1618                 $text = self::convertShare(
1619                         $text,
1620                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1621                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1622                         }
1623                 );
1624
1625                 $text = preg_replace("/\[crypt\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . L10n::t('Encrypted content') . '" /><br />', $text);
1626                 $text = preg_replace("/\[crypt(.*?)\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . '$1' . ' ' . L10n::t('Encrypted content') . '" /><br />', $text);
1627                 //$Text = preg_replace("/\[crypt=(.*?)\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . '$1' . ' ' . L10n::t('Encrypted content') . '" /><br />', $Text);
1628
1629                 // Try to Oembed
1630                 if ($try_oembed) {
1631                         $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);
1632                         $text = preg_replace("/\[audio\](.*?\.(ogg|ogv|oga|ogm|webm|mp4|mp3).*?)\[\/audio\]/ism", '<audio src="$1" controls="controls"><a href="$1">$1</a></audio>', $text);
1633
1634                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1635                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1636                 } else {
1637                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1638                                                 '<a href="$1" target="_blank">$1</a>', $text);
1639                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1640                                                 '<a href="$1" target="_blank">$1</a>', $text);
1641                 }
1642
1643                 // html5 video and audio
1644
1645
1646                 if ($try_oembed) {
1647                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1648                 } else {
1649                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1650                 }
1651
1652                 // Youtube extensions
1653                 if ($try_oembed) {
1654                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1655                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1656                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1657                 }
1658
1659                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1660                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1661                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1662
1663                 if ($try_oembed) {
1664                         $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);
1665                 } else {
1666                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1667                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1668                 }
1669
1670                 if ($try_oembed) {
1671                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1672                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1673                 }
1674
1675                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1676                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1677
1678                 if ($try_oembed) {
1679                         $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);
1680                 } else {
1681                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1682                                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1683                 }
1684
1685                 // oembed tag
1686                 $text = OEmbed::BBCode2HTML($text);
1687
1688                 // Avoid triple linefeeds through oembed
1689                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1690
1691                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1692                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1693                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1694                 // start which is always required). Allow desc with a missing summary for compatibility.
1695
1696                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1697                         $sub = Event::getHTML($ev, $simple_html);
1698
1699                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1700                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1701                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1702                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1703                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1704                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1705                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1706                 }
1707
1708                 // Replace non graphical smilies for external posts
1709                 if ($simple_html) {
1710                         $text = Smilies::replace($text);
1711                 }
1712
1713                 // Unhide all [noparse] contained bbtags unspacefying them
1714                 // and triming the [noparse] tag.
1715
1716                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1717                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1718                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1719
1720                 /// @todo What is the meaning of these lines?
1721                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1722                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1723
1724                 // Currently deactivated, it made problems with " inside of alt texts.
1725                 //$text = preg_replace('/\&quot\;/', '"', $text);
1726
1727                 // fix any escaped ampersands that may have been converted into links
1728                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1729
1730                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1731                 $allowed_src_protocols = ['http', 'redir', 'cid'];
1732                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1733                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1734
1735                 // sanitize href attributes (only whitelisted protocols URLs)
1736                 // default value for backward compatibility
1737                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', ['ftp', 'mailto', 'gopher', 'cid']);
1738
1739                 // Always allowed protocol even if config isn't set or not including it
1740                 $allowed_link_protocols[] = 'http';
1741                 $allowed_link_protocols[] = 'redir/';
1742
1743                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1744                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1745
1746                 if ($saved_image) {
1747                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1748                 }
1749
1750                 // Restore code blocks
1751                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1752                         function ($matches) use ($codeblocks) {
1753                                 $return = $matches[0];
1754                                 if (isset($codeblocks[intval($matches[1])])) {
1755                                         $return = $codeblocks[$matches[1]];
1756                                 }
1757                                 return $return;
1758                         },
1759                         $text
1760                 );
1761
1762                 // Clean up the HTML by loading and saving the HTML with the DOM.
1763                 // Bad structured html can break a whole page.
1764                 // For performance reasons do it only with activated item cache or at export.
1765                 if (!$try_oembed || (get_itemcachepath() != "")) {
1766                         $doc = new DOMDocument();
1767                         $doc->preserveWhiteSpace = false;
1768
1769                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1770
1771                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1772                         $encoding = '<?xml encoding="UTF-8">';
1773                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1774                         $doc->encoding = 'UTF-8';
1775                         $text = $doc->saveHTML();
1776                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1777
1778                         $text = str_replace('<br></li>', '</li>', $text);
1779
1780                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1781                 }
1782
1783                 // Clean up some useless linebreaks in lists
1784                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1785                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1786                 //$Text = str_replace('</li><br />', '</li>', $Text);
1787                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1788                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1789
1790                 Hook::callAll('bbcode', $text);
1791
1792                 return trim($text);
1793         }
1794
1795         /**
1796          * @brief Strips the "abstract" tag from the provided text
1797          *
1798          * @param string $text The text with BBCode
1799          * @return string The same text - but without "abstract" element
1800          */
1801         public static function stripAbstract($text)
1802         {
1803                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1804                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1805
1806                 return $text;
1807         }
1808
1809         /**
1810          * @brief Returns the value of the "abstract" element
1811          *
1812          * @param string $text The text that maybe contains the element
1813          * @param string $addon The addon for which the abstract is meant for
1814          * @return string The abstract
1815          */
1816         public static function getAbstract($text, $addon = "")
1817         {
1818                 $abstract = "";
1819                 $abstracts = [];
1820                 $addon = strtolower($addon);
1821
1822                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1823                         foreach ($results AS $result) {
1824                                 $abstracts[strtolower($result[1])] = $result[2];
1825                         }
1826                 }
1827
1828                 if (isset($abstracts[$addon])) {
1829                         $abstract = $abstracts[$addon];
1830                 }
1831
1832                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1833                         $abstract = $result[1];
1834                 }
1835
1836                 return $abstract;
1837         }
1838
1839         /**
1840          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1841          *
1842          * @param array $match Matching values for the callback
1843          *                     [1] = Mention type (! or @)
1844          *                     [2] = Name
1845          *                     [3] = Address
1846          * @return string Replaced mention
1847          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1848          * @throws \ImagickException
1849          */
1850         private static function bbCodeMention2DiasporaCallback($match)
1851         {
1852                 $contact = Contact::getDetailsByURL($match[3]);
1853
1854                 if (empty($contact['addr'])) {
1855                         $contact = Probe::uri($match[3]);
1856                 }
1857
1858                 if (empty($contact['addr'])) {
1859                         return $match[0];
1860                 }
1861
1862                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
1863                 return $mention;
1864         }
1865
1866         /**
1867          * @brief Converts a BBCode text into Markdown
1868          *
1869          * This function converts a BBCode item body to be sent to Markdown-enabled
1870          * systems like Diaspora and Libertree
1871          *
1872          * @param string $text
1873          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1874          * @return string
1875          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1876          */
1877         public static function toMarkdown($text, $for_diaspora = true)
1878         {
1879                 $a = self::getApp();
1880
1881                 $original_text = $text;
1882
1883                 // Since Diaspora is creating a summary for links, this function removes them before posting
1884                 if ($for_diaspora) {
1885                         $text = self::removeShareInformation($text);
1886                 }
1887
1888                 /**
1889                  * Transform #tags, strip off the [url] and replace spaces with underscore
1890                  */
1891                 $url_search_string = "^\[\]";
1892                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
1893                         function ($matches) {
1894                                 return '#' . str_replace(' ', '_', $matches[2]);
1895                         },
1896                         $text
1897                 );
1898
1899                 // Converting images with size parameters to simple images. Markdown doesn't know it.
1900                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
1901
1902                 // Convert it to HTML - don't try oembed
1903                 if ($for_diaspora) {
1904                         $text = self::convert($text, false, 3);
1905
1906                         // Add all tags that maybe were removed
1907                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
1908                                 $tagline = "";
1909                                 foreach ($tags[2] as $tag) {
1910                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
1911                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
1912                                                 $tagline .= '#' . $tag . ' ';
1913                                         }
1914                                 }
1915                                 $text = $text . " " . $tagline;
1916                         }
1917                 } else {
1918                         $text = self::convert($text, false, 4);
1919                 }
1920
1921                 // mask some special HTML chars from conversation to markdown
1922                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
1923
1924                 // If a link is followed by a quote then there should be a newline before it
1925                 // Maybe we should make this newline at every time before a quote.
1926                 $text = str_replace(["</a><blockquote>"], ["</a><br><blockquote>"], $text);
1927
1928                 $stamp1 = microtime(true);
1929
1930                 // Now convert HTML to Markdown
1931                 $text = HTML::toMarkdown($text);
1932
1933                 // unmask the special chars back to HTML
1934                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
1935
1936                 $a->getProfiler()->saveTimestamp($stamp1, "parser", System::callstack());
1937
1938                 // Libertree has a problem with escaped hashtags.
1939                 $text = str_replace(['\#'], ['#'], $text);
1940
1941                 // Remove any leading or trailing whitespace, as this will mess up
1942                 // the Diaspora signature verification and cause the item to disappear
1943                 $text = trim($text);
1944
1945                 if ($for_diaspora) {
1946                         $url_search_string = "^\[\]";
1947                         $text = preg_replace_callback(
1948                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
1949                                 ['self', 'bbCodeMention2DiasporaCallback'],
1950                                 $text
1951                         );
1952                 }
1953
1954                 Hook::callAll('bb2diaspora', $text);
1955
1956                 return $text;
1957         }
1958
1959         /**
1960      * @brief Pull out all #hashtags and @person tags from $string.
1961      *
1962      * We also get @person@domain.com - which would make
1963      * the regex quite complicated as tags can also
1964      * end a sentence. So we'll run through our results
1965      * and strip the period from any tags which end with one.
1966      * Returns array of tags found, or empty array.
1967      *
1968      * @param string $string Post content
1969      *
1970      * @return array List of tag and person names
1971      */
1972     public static function getTags($string)
1973     {
1974         $ret = [];
1975
1976         // Convert hashtag links to hashtags
1977         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2', $string);
1978
1979         // ignore anything in a code block
1980         $string = preg_replace('/\[code.*?\].*?\[\/code\]/sm', '', $string);
1981
1982         // Force line feeds at bbtags
1983         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
1984
1985         // ignore anything in a bbtag
1986         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
1987
1988         // Match full names against @tags including the space between first and last
1989         // We will look these up afterward to see if they are full names or not recognisable.
1990
1991         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
1992             foreach ($matches[1] as $match) {
1993                 if (strstr($match, ']')) {
1994                     // we might be inside a bbcode color tag - leave it alone
1995                     continue;
1996                 }
1997
1998                 if (substr($match, -1, 1) === '.') {
1999                     $ret[] = substr($match, 0, -1);
2000                 } else {
2001                     $ret[] = $match;
2002                 }
2003             }
2004         }
2005
2006         // Otherwise pull out single word tags. These can be @nickname, @first_last
2007         // and #hash tags.
2008
2009         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?]+)([ \x0D\x0A,;:?]|$)/', $string, $matches)) {
2010             foreach ($matches[1] as $match) {
2011                 if (strstr($match, ']')) {
2012                     // we might be inside a bbcode color tag - leave it alone
2013                     continue;
2014                 }
2015                 if (substr($match, -1, 1) === '.') {
2016                     $match = substr($match,0,-1);
2017                 }
2018                 // ignore strictly numeric tags like #1
2019                 if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
2020                     continue;
2021                 }
2022                 // try not to catch url fragments
2023                 if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2024                     continue;
2025                 }
2026                 $ret[] = $match;
2027             }
2028         }
2029
2030         return $ret;
2031     }
2032 }