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                                                 // Replace the preview picture with the real picture
 256                                                 $url = str_replace('-1.', '-0.', $pictures[0][2]);
 257                                                 $data = ['url' => $url, 'type' => 'photo'];
 258                                         } else {
 259                                                 // Checking, if the link goes to a picture
 260                                                 $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 261                                         }
 262
 263                                         // Workaround:
 264                                         // Sometimes photo posts to the own album are not detected at the start.
 265                                         // So we seem to cannot use the cache for these cases. That's strange.
 266                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
 267                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 268                                         }
 269
 270                                         if ($data["type"] == "photo") {
 271                                                 $post["type"] = "photo";
 272                                                 if (isset($data["images"][0])) {
 273                                                         $post["image"] = $data["images"][0]["src"];
 274                                                         $post["url"] = $data["url"];
 275                                                 } else {
 276                                                         $post["image"] = $data["url"];
 277                                                 }
 278
 279                                                 $post["preview"] = $pictures[0][2];
 280                                                 $post["text"] = trim(str_replace($pictures[0][0], "", $body));
 281                                         } else {
 282                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 283                                                 if ($imgdata && substr($imgdata["mime"], 0, 6) == "image/") {
 284                                                         $post["type"] = "photo";
 285                                                         $post["image"] = $pictures[0][1];
 286                                                         $post["preview"] = $pictures[0][2];
 287                                                         $post["text"] = trim(str_replace($pictures[0][0], "", $body));
 288                                                 }
 289                                         }
 290                                 } elseif (count($pictures) > 0) {
 291                                         $post["type"] = "link";
 292                                         $post["url"] = $plink;
 293                                         $post["image"] = $pictures[0][2];
 294                                         $post["text"] = $body;
 295                                 }
 296                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 297                                 if ((count($pictures) == 1) && !$has_title) {
 298                                         $post["type"] = "photo";
 299                                         $post["image"] = $pictures[0][1];
 300                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 301                                 } elseif (count($pictures) > 0) {
 302                                         $post["type"] = "link";
 303                                         $post["url"] = $plink;
 304                                         $post["image"] = $pictures[0][1];
 305                                         $post["text"] = $body;
 306                                 }
 307                         }
 308
 309                         // Test for the external links
 310                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
 311                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
 312
 313                         $links = array_merge($links1, $links2);
 314
 315                         // If there is only a single one, then use it.
 316                         // This should cover link posts via API.
 317                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 318                                 $post["type"] = "link";
 319                                 $post["text"] = trim($body);
 320                                 $post["url"] = $links[0][1];
 321                         }
 322
 323                         // Now count the number of external media links
 324                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
 325                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
 326                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
 327                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
 328
 329                         // Add them to the other external links
 330                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 331
 332                         // Are there more than one?
 333                         if (count($links) > 1) {
 334                                 // The post will be the type "text", which means a blog post
 335                                 unset($post["type"]);
 336                                 $post["url"] = $plink;
 337                         }
 338
 339                         if (!isset($post["type"])) {
 340                                 $post["type"] = "text";
 341                                 $post["text"] = trim($body);
 342                         }
 343                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 344                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
 345
 346                         if (isset($data["images"][0])) {
 347                                 $post["image"] = $data["images"][0]["src"];
 348                         }
 349                 }
 350
 351                 return $post;
 352         }
 353
 354         /**
 355          * @brief Converts a BBCode text into plaintext
 356          *
 357          * @param      $text
 358          * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
 359          *
 360          * @return string
 361          */
 362         public static function toPlaintext($text, $keep_urls = true)
 363         {
 364                 $naked_text = preg_replace('/\[(.+?)\]\s*/','', $text);
 365                 if (!$keep_urls) {
 366                         $naked_text = preg_replace('#https?\://[^\s<]+[^\s\.\)]#i', '', $naked_text);
 367                 }
 368
 369                 return $naked_text;
 370         }
 371
 372         private static function proxyUrl($image, $simplehtml = false)
 373         {
 374                 // Only send proxied pictures to API and for internal display
 375                 if (in_array($simplehtml, [false, 2])) {
 376                         return ProxyUtils::proxifyUrl($image);
 377                 } else {
 378                         return $image;
 379                 }
 380         }
 381
 382         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 383         {
 384                 // Suppress "view full size"
 385                 if (intval(Config::get('system', 'no_view_full_size'))) {
 386                         $include_link = false;
 387                 }
 388
 389                 // Picture addresses can contain special characters
 390                 $s = htmlspecialchars_decode($srctext);
 391
 392                 $matches = null;
 393                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 394                 if ($c) {
 395                         foreach ($matches as $mtch) {
 396                                 Logger::log('scale_external_image: ' . $mtch[1]);
 397
 398                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 399                                 if (stristr($mtch[1], $hostname)) {
 400                                         continue;
 401                                 }
 402
 403                                 // $scale_replace, if passed, is an array of two elements. The
 404                                 // first is the name of the full-size image. The second is the
 405                                 // name of a remote, scaled-down version of the full size image.
 406                                 // This allows Friendica to display the smaller remote image if
 407                                 // one exists, while still linking to the full-size image
 408                                 if ($scale_replace) {
 409                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 410                                 } else {
 411                                         $scaled = $mtch[1];
 412                                 }
 413                                 $i = Network::fetchUrl($scaled);
 414                                 if (!$i) {
 415                                         return $srctext;
 416                                 }
 417
 418                                 // guess mimetype from headers or filename
 419                                 $type = Image::guessType($mtch[1], true);
 420
 421                                 if ($i) {
 422                                         $Image = new Image($i, $type);
 423                                         if ($Image->isValid()) {
 424                                                 $orig_width = $Image->getWidth();
 425                                                 $orig_height = $Image->getHeight();
 426
 427                                                 if ($orig_width > 640 || $orig_height > 640) {
 428                                                         $Image->scaleDown(640);
 429                                                         $new_width = $Image->getWidth();
 430                                                         $new_height = $Image->getHeight();
 431                                                         Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG);
 432                                                         $s = str_replace(
 433                                                                 $mtch[0],
 434                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 435                                                                 . "\n" . (($include_link)
 436                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 437                                                                         : ''),
 438                                                                 $s
 439                                                         );
 440                                                         Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG);
 441                                                 }
 442                                         }
 443                                 }
 444                         }
 445                 }
 446
 447                 // replace the special char encoding
 448                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 449                 return $s;
 450         }
 451
 452         /**
 453          * The purpose of this function is to apply system message length limits to
 454          * imported messages without including any embedded photos in the length
 455          *
 456          * @brief Truncates imported message body string length to max_import_size
 457          * @param string $body
 458          * @return string
 459          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 460          */
 461         public static function limitBodySize($body)
 462         {
 463                 $maxlen = Config::get('config', 'max_import_size', 0);
 464
 465                 // If the length of the body, including the embedded images, is smaller
 466                 // than the maximum, then don't waste time looking for the images
 467                 if ($maxlen && (strlen($body) > $maxlen)) {
 468
 469                         Logger::log('the total body length exceeds the limit', Logger::DEBUG);
 470
 471                         $orig_body = $body;
 472                         $new_body = '';
 473                         $textlen = 0;
 474
 475                         $img_start = strpos($orig_body, '[img');
 476                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 477                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 478                         while (($img_st_close !== false) && ($img_end !== false)) {
 479
 480                                 $img_st_close++; // make it point to AFTER the closing bracket
 481                                 $img_end += $img_start;
 482                                 $img_end += strlen('[/img]');
 483
 484                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 485                                         // This is an embedded image
 486
 487                                         if (($textlen + $img_start) > $maxlen) {
 488                                                 if ($textlen < $maxlen) {
 489                                                         Logger::log('the limit happens before an embedded image', Logger::DEBUG);
 490                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 491                                                         $textlen = $maxlen;
 492                                                 }
 493                                         } else {
 494                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
 495                                                 $textlen += $img_start;
 496                                         }
 497
 498                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
 499                                 } else {
 500
 501                                         if (($textlen + $img_end) > $maxlen) {
 502                                                 if ($textlen < $maxlen) {
 503                                                         Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG);
 504                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 505                                                         $textlen = $maxlen;
 506                                                 }
 507                                         } else {
 508                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
 509                                                 $textlen += $img_end;
 510                                         }
 511                                 }
 512                                 $orig_body = substr($orig_body, $img_end);
 513
 514                                 if ($orig_body === false) {
 515                                         // in case the body ends on a closing image tag
 516                                         $orig_body = '';
 517                                 }
 518
 519                                 $img_start = strpos($orig_body, '[img');
 520                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 521                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 522                         }
 523
 524                         if (($textlen + strlen($orig_body)) > $maxlen) {
 525                                 if ($textlen < $maxlen) {
 526                                         Logger::log('the limit happens after the end of the last image', Logger::DEBUG);
 527                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
 528                                 }
 529                         } else {
 530                                 Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG);
 531                                 $new_body = $new_body . $orig_body;
 532                         }
 533
 534                         return $new_body;
 535                 } else {
 536                         return $body;
 537                 }
 538         }
 539
 540         /**
 541          * Processes [attachment] tags
 542          *
 543          * Note: Can produce a [bookmark] tag in the returned string
 544          *
 545          * @brief Processes [attachment] tags
 546          * @param string   $return
 547          * @param bool|int $simplehtml
 548          * @param bool     $tryoembed
 549          * @return string
 550          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 551          */
 552         private static function convertAttachment($return, $simplehtml = false, $tryoembed = true)
 553         {
 554                 $data = self::getAttachmentData($return);
 555                 if (empty($data) || empty($data["url"])) {
 556                         return $return;
 557                 }
 558
 559                 if (isset($data["title"])) {
 560                         $data["title"] = strip_tags($data["title"]);
 561                         $data["title"] = str_replace(["http://", "https://"], "", $data["title"]);
 562                 } else {
 563                         $data["title"] = null;
 564                 }
 565
 566                 if (((strpos($data["text"], "[img=") !== false) || (strpos($data["text"], "[img]") !== false) || Config::get('system', 'always_show_preview')) && !empty($data["image"])) {
 567                         $data["preview"] = $data["image"];
 568                         $data["image"] = "";
 569                 }
 570
 571                 $return = '';
 572                 if ($simplehtml == 7) {
 573                         $return = self::convertUrlForOStatus($data["url"]);
 574                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
 575                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data["url"], $data["title"]);
 576                 } else {
 577                         try {
 578                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
 579                                         $return = OEmbed::getHTML($data['url'], $data['title']);
 580                                 } else {
 581                                         throw new Exception('OEmbed is disabled for this attachment.');
 582                                 }
 583                         } catch (Exception $e) {
 584                                 $data["title"] = defaults($data, 'title', $data['url']);
 585
 586                                 if ($simplehtml != 4) {
 587                                         $return = sprintf('<div class="type-%s">', $data["type"]);
 588                                 }
 589
 590                                 if (!empty($data['title']) && !empty($data['url'])) {
 591                                         if (!empty($data["image"]) && empty($data["text"]) && ($data["type"] == "photo")) {
 592                                                 $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data["url"], self::proxyUrl($data["image"], $simplehtml), $data["title"]);
 593                                         } else {
 594                                                 if (!empty($data["image"])) {
 595                                                         $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), $data["title"]);
 596                                                 } elseif (!empty($data["preview"])) {
 597                                                         $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"]);
 598                                                 }
 599                                                 $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
 600                                         }
 601                                 }
 602
 603                                 if (!empty($data["description"]) && $data["description"] != $data["title"]) {
 604                                         // Sanitize the HTML by converting it to BBCode
 605                                         $bbcode = HTML::toBBCode($data["description"]);
 606                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
 607                                 }
 608
 609                                 if (!empty($data['url'])) {
 610                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
 611                                 }
 612
 613                                 if ($simplehtml != 4) {
 614                                         $return .= '</div>';
 615                                 }
 616                         }
 617                 }
 618
 619                 return trim(defaults($data, 'text', '') . ' ' . $return . ' ' . defaults($data, 'after', ''));
 620         }
 621
 622         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
 623         {
 624                 $data = self::getAttachmentData($Text);
 625
 626                 if (!$data) {
 627                         return $Text;
 628                 } elseif ($nolink) {
 629                         return $data["text"] . defaults($data, 'after', '');
 630                 }
 631
 632                 $title = htmlentities(defaults($data, 'title', ''), ENT_QUOTES, 'UTF-8', false);
 633                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
 634                 if ($plaintext || (($title != "") && strstr($text, $title))) {
 635                         $data["title"] = $data["url"];
 636                 } elseif (($text != "") && strstr($title, $text)) {
 637                         $data["text"] = $data["title"];
 638                         $data["title"] = $data["url"];
 639                 }
 640
 641                 if (empty($data["text"]) && !empty($data["title"]) && empty($data["url"])) {
 642                         return $data["title"] . $data["after"];
 643                 }
 644
 645                 // If the link already is included in the post, don't add it again
 646                 if (!empty($data["url"]) && strpos($data["text"], $data["url"])) {
 647                         return $data["text"] . $data["after"];
 648                 }
 649
 650                 $text = $data["text"];
 651
 652                 if (!empty($data["url"]) && !empty($data["title"])) {
 653                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
 654                 } elseif (!empty($data["url"])) {
 655                         $text .= "\n[url]" . $data["url"] . "[/url]";
 656                 }
 657
 658                 return $text . "\n" . $data["after"];
 659         }
 660
 661         /**
 662          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 663          *
 664          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
 665          * @param array $match Array with the matching values
 666          * @return string reformatted link including HTML codes
 667          */
 668         private static function convertUrlForOStatusCallback($match)
 669         {
 670                 $url = $match[1];
 671
 672                 if (isset($match[2]) && ($match[1] != $match[2])) {
 673                         return $match[0];
 674                 }
 675
 676                 $parts = parse_url($url);
 677                 if (!isset($parts['scheme'])) {
 678                         return $match[0];
 679                 }
 680
 681                 return self::convertUrlForOStatus($url);
 682         }
 683
 684         /**
 685          * @brief Converts [url] BBCodes in a format that looks fine on OStatus systems.
 686          * @param string $url URL that is about to be reformatted
 687          * @return string reformatted link including HTML codes
 688          */
 689         private static function convertUrlForOStatus($url)
 690         {
 691                 $parts = parse_url($url);
 692                 $scheme = $parts['scheme'] . '://';
 693                 $styled_url = str_replace($scheme, '', $url);
 694
 695                 if (strlen($styled_url) > 30) {
 696                         $styled_url = substr($styled_url, 0, 30) . "…";
 697                 }
 698
 699                 $html = '<a href="%s" target="_blank">%s</a>';
 700
 701                 return sprintf($html, $url, $styled_url);
 702         }
 703
 704         /*
 705          * [noparse][i]italic[/i][/noparse] turns into
 706          * [noparse][ i ]italic[ /i ][/noparse],
 707          * to hide them from parser.
 708          */
 709         private static function escapeNoparseCallback($match)
 710         {
 711                 $whole_match = $match[0];
 712                 $captured = $match[1];
 713                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
 714                 $new_str = str_replace($captured, $spacefied, $whole_match);
 715                 return $new_str;
 716         }
 717
 718         /*
 719          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
 720          * now turns back and the [noparse] tags are trimed
 721          * returning [i]italic[/i]
 722          */
 723         private static function unescapeNoparseCallback($match)
 724         {
 725                 $captured = $match[1];
 726                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
 727                 return $unspacefied;
 728         }
 729
 730         /**
 731          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
 732          * occurrences
 733          *
 734          * @param string $text        Text to search
 735          * @param string $name        Tag name
 736          * @param int    $occurrences Number of first occurrences to skip
 737          * @return boolean|array
 738          */
 739         public static function getTagPosition($text, $name, $occurrences = 0)
 740         {
 741                 if ($occurrences < 0) {
 742                         $occurrences = 0;
 743                 }
 744
 745                 $start_open = -1;
 746                 for ($i = 0; $i <= $occurrences; $i++) {
 747                         if ($start_open !== false) {
 748                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
 749                         }
 750                 }
 751
 752                 if ($start_open === false) {
 753                         return false;
 754                 }
 755
 756                 $start_equal = strpos($text, '=', $start_open);
 757                 $start_close = strpos($text, ']', $start_open);
 758
 759                 if ($start_close === false) {
 760                         return false;
 761                 }
 762
 763                 $start_close++;
 764
 765                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
 766
 767                 if ($end_open === false) {
 768                         return false;
 769                 }
 770
 771                 $res = [
 772                         'start' => [
 773                                 'open' => $start_open,
 774                                 'close' => $start_close
 775                         ],
 776                         'end' => [
 777                                 'open' => $end_open,
 778                                 'close' => $end_open + strlen('[/' . $name . ']')
 779                         ],
 780                 ];
 781
 782                 if ($start_equal !== false) {
 783                         $res['start']['equal'] = $start_equal + 1;
 784                 }
 785
 786                 return $res;
 787         }
 788
 789         /**
 790          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
 791          *
 792          * @param string $pattern Preg pattern string
 793          * @param string $replace Preg replace string
 794          * @param string $name    BBCode tag name
 795          * @param string $text    Text to search
 796          * @return string
 797          */
 798         public static function pregReplaceInTag($pattern, $replace, $name, $text)
 799         {
 800                 $occurrences = 0;
 801                 $pos = self::getTagPosition($text, $name, $occurrences);
 802                 while ($pos !== false && $occurrences++ < 1000) {
 803                         $start = substr($text, 0, $pos['start']['open']);
 804                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
 805                         $end = substr($text, $pos['end']['close']);
 806                         if ($end === false) {
 807                                 $end = '';
 808                         }
 809
 810                         $subject = preg_replace($pattern, $replace, $subject);
 811                         $text = $start . $subject . $end;
 812
 813                         $pos = self::getTagPosition($text, $name, $occurrences);
 814                 }
 815
 816                 return $text;
 817         }
 818
 819         private static function extractImagesFromItemBody($body)
 820         {
 821                 $saved_image = [];
 822                 $orig_body = $body;
 823                 $new_body = '';
 824
 825                 $cnt = 0;
 826                 $img_start = strpos($orig_body, '[img');
 827                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 828                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 829                 while (($img_st_close !== false) && ($img_end !== false)) {
 830                         $img_st_close++; // make it point to AFTER the closing bracket
 831                         $img_end += $img_start;
 832
 833                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
 834                                 // This is an embedded image
 835                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
 836                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
 837
 838                                 $cnt++;
 839                         } else {
 840                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
 841                         }
 842
 843                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
 844
 845                         if ($orig_body === false) {
 846                                 // in case the body ends on a closing image tag
 847                                 $orig_body = '';
 848                         }
 849
 850                         $img_start = strpos($orig_body, '[img');
 851                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
 852                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
 853                 }
 854
 855                 $new_body = $new_body . $orig_body;
 856
 857                 return ['body' => $new_body, 'images' => $saved_image];
 858         }
 859
 860         private static function interpolateSavedImagesIntoItemBody($body, array $images)
 861         {
 862                 $newbody = $body;
 863
 864                 $cnt = 0;
 865                 foreach ($images as $image) {
 866                         // We're depending on the property of 'foreach' (specified on the PHP website) that
 867                         // it loops over the array starting from the first element and going sequentially
 868                         // to the last element
 869                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
 870                                 '<img src="' . self::proxyUrl($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
 871                         $cnt++;
 872                 }
 873
 874                 return $newbody;
 875         }
 876
 877         /**
 878          * This function converts a [share] block to text according to a provided callback function whose signature is:
 879          *
 880          * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
 881          *
 882          * Where:
 883          * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
 884          * data lookup
 885          * - $author_contact is a contact record array
 886          * - $content is the inner content of the [share] block
 887          * - $is_quote_share indicates whether there's any content before the [share] block
 888          * - Return value is the string that should replace the [share] block in the provided text
 889          *
 890          * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
 891          *
 892          * @param  string   $text     A BBCode string
 893          * @param  callable $callback
 894          * @return string The BBCode string with all [share] blocks replaced
 895          */
 896         public static function convertShare($text, callable $callback)
 897         {
 898                 $return = preg_replace_callback(
 899                         "/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
 900                         function ($match) use ($callback) {
 901                                 $attribute_string = $match[2];
 902
 903                                 $attributes = [];
 904                                 foreach(['author', 'profile', 'avatar', 'link', 'posted'] as $field) {
 905                                         preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
 906                                         $attributes[$field] = html_entity_decode(defaults($matches, 2, ''), ENT_QUOTES, 'UTF-8');
 907                                 }
 908
 909                                 // We only call this so that a previously unknown contact can be added.
 910                                 // This is important for the function "Model\Contact::getDetailsByURL()".
 911                                 // This function then can fetch an entry from the contact table.
 912                                 Contact::getIdForURL($attributes['profile'], 0, true);
 913
 914                                 $author_contact = Contact::getDetailsByURL($attributes['profile']);
 915                                 $author_contact['addr'] = defaults($author_contact, 'addr' , Protocol::getAddrFromProfileUrl($attributes['profile']));
 916
 917                                 $attributes['author']   = defaults($author_contact, 'name' , $attributes['author']);
 918                                 $attributes['avatar']   = defaults($author_contact, 'micro', $attributes['avatar']);
 919                                 $attributes['profile']  = defaults($author_contact, 'url'  , $attributes['profile']);
 920
 921                                 if ($attributes['avatar']) {
 922                                         $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
 923                                 }
 924
 925                                 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
 926                         },
 927                         $text
 928                 );
 929
 930                 return $return;
 931         }
 932
 933         /**
 934          * Default [share] tag conversion callback
 935          *
 936          * Note: Can produce a [bookmark] tag in the output
 937          *
 938          * @see BBCode::convertShare()
 939          * @param array   $attributes     [share] block attribute values
 940          * @param array   $author_contact Contact row of the shared author
 941          * @param string  $content        Inner content of the [share] block
 942          * @param boolean $is_quote_share Whether there is content before the [share] block
 943          * @param integer $simplehtml     Mysterious integer value depending on the target network/formatting style
 944          * @return string
 945          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
 946          */
 947         private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
 948         {
 949                 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
 950
 951                 switch ($simplehtml) {
 952                         case 1:
 953                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' <a href="' . $attributes['profile'] . '">' . $mention . '</a>: </p>' . "\n" . '«' . $content . '»';
 954                                 break;
 955                         case 2:
 956                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
 957                                 break;
 958                         case 3: // Diaspora
 959                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
 960                                         $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
 961                                 } else {
 962                                         $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
 963
 964                                         if (!empty($attributes['posted']) && !empty($attributes['link'])) {
 965                                                 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
 966                                         }
 967
 968                                         $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
 969
 970                                         if (empty($attributes['posted']) && !empty($attributes['link'])) {
 971                                                 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
 972                                         }
 973                                 }
 974
 975                                 break;
 976                         case 4:
 977                                 $headline = '<p><b>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8');
 978                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
 979                                 $headline .= ':</b></p>' . "\n";
 980
 981                                 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
 982
 983                                 break;
 984                         case 5:
 985                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
 986                                 break;
 987                         case 7: // statusnet/GNU Social
 988                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
 989                                 break;
 990                         case 9: // Google+
 991                                 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('&#x2672; ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n";
 992                                 $text .= '<p>' . $content . '</p>' . "\n";
 993
 994                                 if ($attributes['link'] != '') {
 995                                         $text .= '<p>' . $attributes['link'] . '</p>';
 996                                 }
 997                                 break;
 998                         default:
 999                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1000                                 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($attributes['link'])) {
1001                                         try {
1002                                                 $text = ($is_quote_share? '<br />' : '') . OEmbed::getHTML($attributes['link']);
1003                                         } catch (Exception $e) {
1004                                                 $text = ($is_quote_share? '<br />' : '') . sprintf('[bookmark=%s]%s[/bookmark]', $attributes['link'], $content);
1005                                         }
1006                                 } else {
1007                                         $text = ($is_quote_share? "\n" : '');
1008
1009                                         $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1010                                         $text .= Renderer::replaceMacros($tpl, [
1011                                                 '$profile' => $attributes['profile'],
1012                                                 '$avatar'  => $attributes['avatar'],
1013                                                 '$author'  => $attributes['author'],
1014                                                 '$link'    => $attributes['link'],
1015                                                 '$posted'  => $attributes['posted'],
1016                                                 '$content' => trim($content)
1017                                         ]);
1018                                 }
1019                                 break;
1020                 }
1021
1022                 return $text;
1023         }
1024
1025         private static function removePictureLinksCallback($match)
1026         {
1027                 $text = Cache::get($match[1]);
1028
1029                 if (is_null($text)) {
1030                         $a = self::getApp();
1031
1032                         $stamp1 = microtime(true);
1033
1034                         $ch = @curl_init($match[1]);
1035                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1036                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1037                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1038                         @curl_exec($ch);
1039                         $curl_info = @curl_getinfo($ch);
1040
1041                         $a->getProfiler()->saveTimestamp($stamp1, "network", System::callstack());
1042
1043                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1044                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1045                         } else {
1046                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1047
1048                                 // if its not a picture then look if its a page that contains a picture link
1049                                 $body = Network::fetchUrl($match[1]);
1050
1051                                 $doc = new DOMDocument();
1052                                 @$doc->loadHTML($body);
1053                                 $xpath = new DOMXPath($doc);
1054                                 $list = $xpath->query("//meta[@name]");
1055                                 foreach ($list as $node) {
1056                                         $attr = [];
1057
1058                                         if ($node->attributes->length) {
1059                                                 foreach ($node->attributes as $attribute) {
1060                                                         $attr[$attribute->name] = $attribute->value;
1061                                                 }
1062                                         }
1063
1064                                         if (strtolower($attr["name"]) == "twitter:image") {
1065                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1066                                         }
1067                                 }
1068                         }
1069                         Cache::set($match[1], $text);
1070                 }
1071
1072                 return $text;
1073         }
1074
1075         private static function expandLinksCallback($match)
1076         {
1077                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1078                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1079                 } else {
1080                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1081                 }
1082         }
1083
1084         private static function cleanPictureLinksCallback($match)
1085         {
1086                 $text = Cache::get($match[1]);
1087
1088                 if (is_null($text)) {
1089                         $a = self::getApp();
1090
1091                         $stamp1 = microtime(true);
1092
1093                         $ch = @curl_init($match[1]);
1094                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1095                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1096                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1097                         @curl_exec($ch);
1098                         $curl_info = @curl_getinfo($ch);
1099
1100                         $a->getProfiler()->saveTimestamp($stamp1, "network", System::callstack());
1101
1102                         // if its a link to a picture then embed this picture
1103                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1104                                 $text = "[img]" . $match[1] . "[/img]";
1105                         } else {
1106                                 $text = "[img]" . $match[2] . "[/img]";
1107
1108                                 // if its not a picture then look if its a page that contains a picture link
1109                                 $body = Network::fetchUrl($match[1]);
1110
1111                                 $doc = new DOMDocument();
1112                                 @$doc->loadHTML($body);
1113                                 $xpath = new DOMXPath($doc);
1114                                 $list = $xpath->query("//meta[@name]");
1115                                 foreach ($list as $node) {
1116                                         $attr = [];
1117                                         if ($node->attributes->length) {
1118                                                 foreach ($node->attributes as $attribute) {
1119                                                         $attr[$attribute->name] = $attribute->value;
1120                                                 }
1121                                         }
1122
1123                                         if (strtolower($attr["name"]) == "twitter:image") {
1124                                                 $text = "[img]" . $attr["content"] . "[/img]";
1125                                         }
1126                                 }
1127                         }
1128                         Cache::set($match[1], $text);
1129                 }
1130
1131                 return $text;
1132         }
1133
1134         public static function cleanPictureLinks($text)
1135         {
1136                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1137                 return $return;
1138         }
1139
1140         /**
1141          * @brief Converts a BBCode message to HTML message
1142          *
1143          * BBcode 2 HTML was written by WAY2WEB.net
1144          * extended to work with Mistpark/Friendica - Mike Macgirvin
1145          *
1146          * Simple HTML values meaning:
1147          * - 0: Friendica display
1148          * - 1: Unused
1149          * - 2: Used for Google+, Windows Phone push, Friendica API
1150          * - 3: Used before converting to Markdown in bb2diaspora.php
1151          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1152          * - 5: Unused
1153          * - 6: Unused
1154          * - 7: Used for dfrn, OStatus
1155          * - 8: Used for WP backlink text setting
1156          *
1157          * @param string $text
1158          * @param bool   $try_oembed
1159          * @param int    $simple_html
1160          * @param bool   $for_plaintext
1161          * @return string
1162          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1163          */
1164         public static function convert($text, $try_oembed = true, $simple_html = 0, $for_plaintext = false)
1165         {
1166                 $a = self::getApp();
1167
1168                 /*
1169                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1170                  *
1171                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1172                  * $match[1] = $url
1173                  * $match[2] = $title or absent
1174                  */
1175                 $try_oembed_callback = function ($match)
1176                 {
1177                         $url = $match[1];
1178                         $title = defaults($match, 2, null);
1179
1180                         try {
1181                                 $return = OEmbed::getHTML($url, $title);
1182                         } catch (Exception $ex) {
1183                                 $return = $match[0];
1184                         }
1185
1186                         return $return;
1187                 };
1188
1189                 // Extracting code blocks before the whitespace processing and the autolinker
1190                 $codeblocks = [];
1191
1192                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1193                         function ($matches) use (&$codeblocks) {
1194                                 $return = '#codeblock-' . count($codeblocks) . '#';
1195                                 if (strpos($matches[2], "\n") !== false) {
1196                                         $codeblocks[] = '<pre><code class="language-' . trim($matches[1]) . '">' . trim($matches[2], "\n\r") . '</code></pre>';
1197                                 } else {
1198                                         $codeblocks[] = '<code>' . $matches[2] . '</code>';
1199                                 }
1200
1201                                 return $return;
1202                         },
1203                         $text
1204                 );
1205
1206                 // Hide all [noparse] contained bbtags by spacefying them
1207                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1208
1209                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1210                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1211                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1212
1213                 // Remove the abstract element. It is a non visible element.
1214                 $text = self::stripAbstract($text);
1215
1216                 // Move all spaces out of the tags
1217                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1218                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1219
1220                 // Extract the private images which use data urls since preg has issues with
1221                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1222                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1223
1224                 $extracted = self::extractImagesFromItemBody($text);
1225                 $text = $extracted['body'];
1226                 $saved_image = $extracted['images'];
1227
1228                 // If we find any event code, turn it into an event.
1229                 // After we're finished processing the bbcode we'll
1230                 // replace all of the event code with a reformatted version.
1231
1232                 $ev = Event::fromBBCode($text);
1233
1234                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1235                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1236
1237                 $text = str_replace("<", "&lt;", $text);
1238                 $text = str_replace(">", "&gt;", $text);
1239
1240                 // remove some newlines before the general conversion
1241                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1242                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1243
1244                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1245                 if (!$try_oembed) {
1246                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1247                 }
1248
1249                 // Convert new line chars to html <br /> tags
1250
1251                 // nlbr seems to be hopelessly messed up
1252                 //      $Text = nl2br($Text);
1253
1254                 // We'll emulate it.
1255
1256                 $text = trim($text);
1257                 $text = str_replace("\r\n", "\n", $text);
1258
1259                 // Remove linefeeds inside of the table elements. See issue #6799
1260                 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1261                         "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1262                         "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1263                         "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1264                 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1265                         "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1266                         "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1267                         "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1268                 do {
1269                         $oldtext = $text;
1270                         $text = str_replace($search, $replace, $text);
1271                 } while ($oldtext != $text);
1272
1273                 // Replace these here only once
1274                 $search = ["\n[table]", "[/table]\n"];
1275                 $replace = ["[table]", "[/table]"];
1276                 $text = str_replace($search, $replace, $text);
1277
1278                 // removing multiplicated newlines
1279                 if (Config::get("system", "remove_multiplicated_lines")) {
1280                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1281                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1282                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1283                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1284                         do {
1285                                 $oldtext = $text;
1286                                 $text = str_replace($search, $replace, $text);
1287                         } while ($oldtext != $text);
1288                 }
1289
1290                 // Set up the parameters for a URL search string
1291                 $URLSearchString = "^\[\]";
1292                 // Set up the parameters for a MAIL search string
1293                 $MAILSearchString = $URLSearchString;
1294
1295                 // Handle attached links or videos
1296                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1297
1298                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1299                 if (!$for_plaintext) {
1300                         $text = preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1301                         if ($simple_html == 7) {
1302                                 $text = preg_replace_callback("/\[url\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1303                                 $text = preg_replace_callback("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1304                         }
1305                 } else {
1306                         $text = preg_replace("(\[url\]([$URLSearchString]*)\[\/url\])ism", " $1 ", $text);
1307                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1308                 }
1309
1310                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1311
1312                 // Remove all hashtag addresses
1313                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7])) {
1314                         $text = preg_replace("/([#@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $text);
1315                 } elseif ($simple_html == 3) {
1316                         // The ! is converted to @ since Diaspora only understands the @
1317                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1318                                 '@<a href="$2">$3</a>',
1319                                 $text);
1320                 } elseif ($simple_html == 7) {
1321                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1322                                 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1323                                 $text);
1324                 } elseif (!$simple_html) {
1325                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1326                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1327                                 $text);
1328                 }
1329
1330                 // Bookmarks in red - will be converted to bookmarks in friendica
1331                 $text = preg_replace("/#\^\[url\]([$URLSearchString]*)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1332                 $text = preg_replace("/#\^\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1333                 $text = preg_replace("/#\[url\=[$URLSearchString]*\]\^\[\/url\]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i",
1334                                         "[bookmark=$1]$2[/bookmark]", $text);
1335
1336                 if (in_array($simple_html, [2, 6, 7, 8, 9])) {
1337                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1338                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1339                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1340                 }
1341
1342                 if ($simple_html == 5) {
1343                         $text = preg_replace("/[^#@!]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1344                 }
1345
1346                 // Perform URL Search
1347                 if ($try_oembed) {
1348                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1349                 }
1350
1351                 if ($simple_html == 5) {
1352                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1353                 } else {
1354                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1355                 }
1356
1357                 // Handle Diaspora posts
1358                 $text = preg_replace_callback(
1359                         "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1360                         function ($match) {
1361                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1362                         }, $text
1363                 );
1364
1365                 $text = preg_replace_callback(
1366                         "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1367                         function ($match) {
1368                                 return "[url=" . System::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1369                         }, $text
1370                 );
1371
1372                 // Server independent link to posts and comments
1373                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1374                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1375                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1376
1377                 /* Tag conversion
1378                  * Supports:
1379                  * - #[url=<anything>]<term>[/url]
1380                  * - [url=<anything>]#<term>[/url]
1381                  */
1382                 $text = preg_replace_callback("/(?:#\[url\=[$URLSearchString]*\]|\[url\=[$URLSearchString]*\]#)(.*?)\[\/url\]/ism", function($matches) {
1383                         return '#<a href="'
1384                                 . System::baseUrl()     . '/search?tag=' . rawurlencode($matches[1])
1385                                 . '" class="tag" title="' . XML::escape($matches[1]) . '">'
1386                                 . XML::escape($matches[1])
1387                                 . '</a>';
1388                 }, $text);
1389
1390                 // We need no target="_blank" for local links
1391                 // convert links start with System::baseUrl() as local link without the target="_blank" attribute
1392                 $escapedBaseUrl = preg_quote(System::baseUrl(), '/');
1393                 $text = preg_replace("/\[url\](".$escapedBaseUrl."[$URLSearchString]*)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1394                 $text = preg_replace("/\[url\=(".$escapedBaseUrl."[$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1395
1396                 $text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1397                 $text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1398
1399                 // Red compatibility, though the link can't be authenticated on Friendica
1400                 $text = preg_replace("/\[zrl\=([$URLSearchString]*)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1401
1402
1403                 // we may need to restrict this further if it picks up too many strays
1404                 // link acct:user@host to a webfinger profile redirector
1405
1406                 $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);
1407
1408                 // Perform MAIL Search
1409                 $text = preg_replace("/\[mail\]([$MAILSearchString]*)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1410                 $text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1411
1412                 // leave open the posibility of [map=something]
1413                 // this is replaced in Item::prepareBody() which has knowledge of the item location
1414
1415                 if (strpos($text, '[/map]') !== false) {
1416                         $text = preg_replace_callback(
1417                                 "/\[map\](.*?)\[\/map\]/ism",
1418                                 function ($match) use ($simple_html) {
1419                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1420                                 },
1421                                 $text
1422                         );
1423                 }
1424                 if (strpos($text, '[map=') !== false) {
1425                         $text = preg_replace_callback(
1426                                 "/\[map=(.*?)\]/ism",
1427                                 function ($match) use ($simple_html) {
1428                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1429                                 },
1430                                 $text
1431                         );
1432                 }
1433                 if (strpos($text, '[map]') !== false) {
1434                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1435                 }
1436
1437                 // Check for headers
1438                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1439                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1440                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1441                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1442                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1443                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1444
1445                 // Check for paragraph
1446                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1447
1448                 // Check for bold text
1449                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1450
1451                 // Check for Italics text
1452                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1453
1454                 // Check for Underline text
1455                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1456
1457                 // Check for strike-through text
1458                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1459
1460                 // Check for over-line text
1461                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1462
1463                 // Check for colored text
1464                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1465
1466                 // Check for sized text
1467                 // [size=50] --> font-size: 50px (with the unit).
1468                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1469                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1470
1471                 // Check for centered text
1472                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1473
1474                 // Check for list text
1475                 $text = str_replace("[*]", "<li>", $text);
1476
1477                 // Check for style sheet commands
1478                 $text = preg_replace_callback(
1479                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1480                         function ($match) {
1481                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1482                         },
1483                         $text
1484                 );
1485
1486                 // Check for CSS classes
1487                 $text = preg_replace_callback(
1488                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1489                         function ($match) {
1490                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1491                         },
1492                         $text
1493                 );
1494
1495                 // handle nested lists
1496                 $endlessloop = 0;
1497
1498                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1499                            ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1500                            ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1501                            ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1502                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1503                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1504                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1505                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1506                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1507                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1508                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1509                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1510                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1511                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1512                 }
1513
1514                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1515                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1516                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1517                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1518
1519                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1520                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1521
1522                 $text = str_replace('[hr]', '<hr />', $text);
1523
1524                 // This is actually executed in Item::prepareBody()
1525
1526                 $text = str_replace('[nosmile]', '', $text);
1527
1528                 // Check for font change text
1529                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1530
1531                 // Declare the format for [spoiler] layout
1532                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
1533
1534                 // Check for [spoiler] text
1535                 // handle nested quotes
1536                 $endlessloop = 0;
1537                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1538                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", "$SpoilerLayout", $text);
1539                 }
1540
1541                 // Check for [spoiler=Author] text
1542
1543                 $t_wrote = L10n::t('$1 wrote:');
1544
1545                 // handle nested quotes
1546                 $endlessloop = 0;
1547                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1548                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1549                                                  "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1550                                                  $text);
1551                 }
1552
1553                 // Declare the format for [quote] layout
1554                 $QuoteLayout = '<blockquote>$1</blockquote>';
1555
1556                 // Check for [quote] text
1557                 // handle nested quotes
1558                 $endlessloop = 0;
1559                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1560                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1561                 }
1562
1563                 // Check for [quote=Author] text
1564
1565                 $t_wrote = L10n::t('$1 wrote:');
1566
1567                 // handle nested quotes
1568                 $endlessloop = 0;
1569                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1570                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1571                                                  "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1572                                                  $text);
1573                 }
1574
1575
1576                 // [img=widthxheight]image source[/img]
1577                 $text = preg_replace_callback(
1578                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1579                         function ($matches) use ($simple_html) {
1580                                 if (strpos($matches[3], "data:image/") === 0) {
1581                                         return $matches[0];
1582                                 }
1583
1584                                 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1585                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1586                         },
1587                         $text
1588                 );
1589
1590                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1591                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1592
1593                 $text = preg_replace_callback("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism",
1594                         function ($matches) use ($simple_html) {
1595                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1596                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1597                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '">';
1598                         },
1599                         $text);
1600
1601                 // Images
1602                 // [img]pathtoimage[/img]
1603                 $text = preg_replace_callback(
1604                         "/\[img\](.*?)\[\/img\]/ism",
1605                         function ($matches) use ($simple_html) {
1606                                 if (strpos($matches[1], "data:image/") === 0) {
1607                                         return $matches[0];
1608                                 }
1609
1610                                 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1611                                 return "[img]" . $matches[1] . "[/img]";
1612                         },
1613                         $text
1614                 );
1615
1616                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1617                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1618
1619                 // Shared content
1620                 $text = self::convertShare(
1621                         $text,
1622                         function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1623                                 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1624                         }
1625                 );
1626
1627                 $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);
1628                 $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 />', $text);
1629                 //$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 />', $Text);
1630
1631                 // Try to Oembed
1632                 if ($try_oembed) {
1633                         $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);
1634                         $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);
1635
1636                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1637                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1638                 } else {
1639                         $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1640                                                 '<a href="$1" target="_blank">$1</a>', $text);
1641                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1642                                                 '<a href="$1" target="_blank">$1</a>', $text);
1643                 }
1644
1645                 // html5 video and audio
1646
1647
1648                 if ($try_oembed) {
1649                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1650                 } else {
1651                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1652                 }
1653
1654                 // Youtube extensions
1655                 if ($try_oembed) {
1656                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1657                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1658                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1659                 }
1660
1661                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1662                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1663                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1664
1665                 if ($try_oembed) {
1666                         $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);
1667                 } else {
1668                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1669                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1670                 }
1671
1672                 if ($try_oembed) {
1673                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1674                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1675                 }
1676
1677                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1678                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1679
1680                 if ($try_oembed) {
1681                         $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);
1682                 } else {
1683                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1684                                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1685                 }
1686
1687                 // oembed tag
1688                 $text = OEmbed::BBCode2HTML($text);
1689
1690                 // Avoid triple linefeeds through oembed
1691                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1692
1693                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1694                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1695                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1696                 // start which is always required). Allow desc with a missing summary for compatibility.
1697
1698                 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1699                         $sub = Event::getHTML($ev, $simple_html);
1700
1701                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1702                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1703                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1704                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1705                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1706                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1707                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1708                 }
1709
1710                 // Replace non graphical smilies for external posts
1711                 if ($simple_html) {
1712                         $text = Smilies::replace($text);
1713                 }
1714
1715                 // Unhide all [noparse] contained bbtags unspacefying them
1716                 // and triming the [noparse] tag.
1717
1718                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1719                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1720                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1721
1722                 /// @todo What is the meaning of these lines?
1723                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1724                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1725
1726                 // Currently deactivated, it made problems with " inside of alt texts.
1727                 //$text = preg_replace('/\&quot\;/', '"', $text);
1728
1729                 // fix any escaped ampersands that may have been converted into links
1730                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1731
1732                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1733                 $allowed_src_protocols = ['http', 'redir', 'cid'];
1734                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1735                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1736
1737                 // sanitize href attributes (only whitelisted protocols URLs)
1738                 // default value for backward compatibility
1739                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', ['ftp', 'mailto', 'gopher', 'cid']);
1740
1741                 // Always allowed protocol even if config isn't set or not including it
1742                 $allowed_link_protocols[] = 'http';
1743                 $allowed_link_protocols[] = 'redir/';
1744
1745                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1746                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1747
1748                 if ($saved_image) {
1749                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1750                 }
1751
1752                 // Restore code blocks
1753                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1754                         function ($matches) use ($codeblocks) {
1755                                 $return = $matches[0];
1756                                 if (isset($codeblocks[intval($matches[1])])) {
1757                                         $return = $codeblocks[$matches[1]];
1758                                 }
1759                                 return $return;
1760                         },
1761                         $text
1762                 );
1763
1764                 // Clean up the HTML by loading and saving the HTML with the DOM.
1765                 // Bad structured html can break a whole page.
1766                 // For performance reasons do it only with activated item cache or at export.
1767                 if (!$try_oembed || (get_itemcachepath() != "")) {
1768                         $doc = new DOMDocument();
1769                         $doc->preserveWhiteSpace = false;
1770
1771                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1772
1773                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1774                         $encoding = '<?xml encoding="UTF-8">';
1775                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1776                         $doc->encoding = 'UTF-8';
1777                         $text = $doc->saveHTML();
1778                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1779
1780                         $text = str_replace('<br></li>', '</li>', $text);
1781
1782                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1783                 }
1784
1785                 // Clean up some useless linebreaks in lists
1786                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1787                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1788                 //$Text = str_replace('</li><br />', '</li>', $Text);
1789                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1790                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1791
1792                 Hook::callAll('bbcode', $text);
1793
1794                 return trim($text);
1795         }
1796
1797         /**
1798          * @brief Strips the "abstract" tag from the provided text
1799          *
1800          * @param string $text The text with BBCode
1801          * @return string The same text - but without "abstract" element
1802          */
1803         public static function stripAbstract($text)
1804         {
1805                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1806                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1807
1808                 return $text;
1809         }
1810
1811         /**
1812          * @brief Returns the value of the "abstract" element
1813          *
1814          * @param string $text The text that maybe contains the element
1815          * @param string $addon The addon for which the abstract is meant for
1816          * @return string The abstract
1817          */
1818         public static function getAbstract($text, $addon = "")
1819         {
1820                 $abstract = "";
1821                 $abstracts = [];
1822                 $addon = strtolower($addon);
1823
1824                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1825                         foreach ($results AS $result) {
1826                                 $abstracts[strtolower($result[1])] = $result[2];
1827                         }
1828                 }
1829
1830                 if (isset($abstracts[$addon])) {
1831                         $abstract = $abstracts[$addon];
1832                 }
1833
1834                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1835                         $abstract = $result[1];
1836                 }
1837
1838                 return $abstract;
1839         }
1840
1841         /**
1842          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1843          *
1844          * @param array $match Matching values for the callback
1845          *                     [1] = Mention type (! or @)
1846          *                     [2] = Name
1847          *                     [3] = Address
1848          * @return string Replaced mention
1849          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1850          * @throws \ImagickException
1851          */
1852         private static function bbCodeMention2DiasporaCallback($match)
1853         {
1854                 $contact = Contact::getDetailsByURL($match[3]);
1855
1856                 if (empty($contact['addr'])) {
1857                         $contact = Probe::uri($match[3]);
1858                 }
1859
1860                 if (empty($contact['addr'])) {
1861                         return $match[0];
1862                 }
1863
1864                 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
1865                 return $mention;
1866         }
1867
1868         /**
1869          * @brief Converts a BBCode text into Markdown
1870          *
1871          * This function converts a BBCode item body to be sent to Markdown-enabled
1872          * systems like Diaspora and Libertree
1873          *
1874          * @param string $text
1875          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1876          * @return string
1877          * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1878          */
1879         public static function toMarkdown($text, $for_diaspora = true)
1880         {
1881                 $a = self::getApp();
1882
1883                 $original_text = $text;
1884
1885                 // Since Diaspora is creating a summary for links, this function removes them before posting
1886                 if ($for_diaspora) {
1887                         $text = self::removeShareInformation($text);
1888                 }
1889
1890                 /**
1891                  * Transform #tags, strip off the [url] and replace spaces with underscore
1892                  */
1893                 $url_search_string = "^\[\]";
1894                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
1895                         function ($matches) {
1896                                 return '#' . str_replace(' ', '_', $matches[2]);
1897                         },
1898                         $text
1899                 );
1900
1901                 // Converting images with size parameters to simple images. Markdown doesn't know it.
1902                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
1903
1904                 // Convert it to HTML - don't try oembed
1905                 if ($for_diaspora) {
1906                         $text = self::convert($text, false, 3);
1907
1908                         // Add all tags that maybe were removed
1909                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
1910                                 $tagline = "";
1911                                 foreach ($tags[2] as $tag) {
1912                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
1913                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
1914                                                 $tagline .= '#' . $tag . ' ';
1915                                         }
1916                                 }
1917                                 $text = $text . " " . $tagline;
1918                         }
1919                 } else {
1920                         $text = self::convert($text, false, 4);
1921                 }
1922
1923                 // mask some special HTML chars from conversation to markdown
1924                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
1925
1926                 // If a link is followed by a quote then there should be a newline before it
1927                 // Maybe we should make this newline at every time before a quote.
1928                 $text = str_replace(["</a><blockquote>"], ["</a><br><blockquote>"], $text);
1929
1930                 $stamp1 = microtime(true);
1931
1932                 // Now convert HTML to Markdown
1933                 $text = HTML::toMarkdown($text);
1934
1935                 // unmask the special chars back to HTML
1936                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
1937
1938                 $a->getProfiler()->saveTimestamp($stamp1, "parser", System::callstack());
1939
1940                 // Libertree has a problem with escaped hashtags.
1941                 $text = str_replace(['\#'], ['#'], $text);
1942
1943                 // Remove any leading or trailing whitespace, as this will mess up
1944                 // the Diaspora signature verification and cause the item to disappear
1945                 $text = trim($text);
1946
1947                 if ($for_diaspora) {
1948                         $url_search_string = "^\[\]";
1949                         $text = preg_replace_callback(
1950                                 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
1951                                 ['self', 'bbCodeMention2DiasporaCallback'],
1952                                 $text
1953                         );
1954                 }
1955
1956                 Hook::callAll('bb2diaspora', $text);
1957
1958                 return $text;
1959         }
1960
1961         /**
1962      * @brief Pull out all #hashtags and @person tags from $string.
1963      *
1964      * We also get @person@domain.com - which would make
1965      * the regex quite complicated as tags can also
1966      * end a sentence. So we'll run through our results
1967      * and strip the period from any tags which end with one.
1968      * Returns array of tags found, or empty array.
1969      *
1970      * @param string $string Post content
1971      *
1972      * @return array List of tag and person names
1973      */
1974     public static function getTags($string)
1975     {
1976         $ret = [];
1977
1978         // Convert hashtag links to hashtags
1979         $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2', $string);
1980
1981         // ignore anything in a code block
1982         $string = preg_replace('/\[code.*?\].*?\[\/code\]/sm', '', $string);
1983
1984         // Force line feeds at bbtags
1985         $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
1986
1987         // ignore anything in a bbtag
1988         $string = preg_replace('/\[(.*?)\]/sm', '', $string);
1989
1990         // Match full names against @tags including the space between first and last
1991         // We will look these up afterward to see if they are full names or not recognisable.
1992
1993         if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
1994             foreach ($matches[1] as $match) {
1995                 if (strstr($match, ']')) {
1996                     // we might be inside a bbcode color tag - leave it alone
1997                     continue;
1998                 }
1999
2000                 if (substr($match, -1, 1) === '.') {
2001                     $ret[] = substr($match, 0, -1);
2002                 } else {
2003                     $ret[] = $match;
2004                 }
2005             }
2006         }
2007
2008         // Otherwise pull out single word tags. These can be @nickname, @first_last
2009         // and #hash tags.
2010
2011         if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?]+)([ \x0D\x0A,;:?]|$)/', $string, $matches)) {
2012             foreach ($matches[1] as $match) {
2013                 if (strstr($match, ']')) {
2014                     // we might be inside a bbcode color tag - leave it alone
2015                     continue;
2016                 }
2017                 if (substr($match, -1, 1) === '.') {
2018                     $match = substr($match,0,-1);
2019                 }
2020                 // ignore strictly numeric tags like #1
2021                 if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
2022                     continue;
2023                 }
2024                 // try not to catch url fragments
2025                 if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2026                     continue;
2027                 }
2028                 $ret[] = $match;
2029             }
2030         }
2031
2032         return $ret;
2033     }
2034 }