src/Content/Text/BBCode.php

   1 <?php
   2 /**
   3  * @file src/Content/Text/BBCode.php
   4  */
   5 namespace Friendica\Content\Text;
   6
   7 use Friendica\App;
   8 use Friendica\Content\Text\Plaintext;
   9 use Friencica\Core\Config;
  10 use Friendica\Core\L10n;
  11 use Friendica\Core\PConfig;
  12 use Friendica\Core\System;
  13 use Friendica\Object\Image;
  14 use Friendica\Util\ParseUrl;
  15
  16 require_once "include/bbcode.php";
  17 require_once "include/html2plain.php";
  18
  19 class BBCode
  20 {
  21         /**
  22          * @brief Fetches attachment data that were generated the old way
  23          *
  24          * @param string $body Message body
  25          * @return array
  26          * 'type' -> Message type ("link", "video", "photo")
  27          * 'text' -> Text before the shared message
  28          * 'after' -> Text after the shared message
  29          * 'image' -> Preview image of the message
  30          * 'url' -> Url to the attached message
  31          * 'title' -> Title of the attachment
  32          * 'description' -> Description of the attachment
  33          */
  34         private static function getOldAttachmentData($body)
  35         {
  36                 $post = [];
  37
  38                 // Simplify image codes
  39                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  40
  41                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  42                         foreach ($attached as $data) {
  43                                 if (!in_array($data[1], ["type-link", "type-video", "type-photo"])) {
  44                                         continue;
  45                                 }
  46
  47                                 $post["type"] = substr($data[1], 5);
  48
  49                                 $pos = strpos($body, $data[0]);
  50                                 if ($pos > 0) {
  51                                         $post["text"] = trim(substr($body, 0, $pos));
  52                                         $post["after"] = trim(substr($body, $pos + strlen($data[0])));
  53                                 } else {
  54                                         $post["text"] = trim(str_replace($data[0], "", $body));
  55                                 }
  56
  57                                 $attacheddata = $data[2];
  58
  59                                 $URLSearchString = "^\[\]";
  60
  61                                 if (preg_match("/\[img\]([$URLSearchString]*)\[\/img\]/ism", $attacheddata, $matches)) {
  62
  63                                         $picturedata = Image::getInfoFromURL($matches[1]);
  64
  65                                         if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
  66                                                 $post["image"] = $matches[1];
  67                                         } else {
  68                                                 $post["preview"] = $matches[1];
  69                                         }
  70                                 }
  71
  72                                 if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  73                                         $post["url"] = $matches[1];
  74                                         $post["title"] = $matches[2];
  75                                 }
  76                                 if (($post["url"] == "") && (in_array($post["type"], ["link", "video"]))
  77                                         && preg_match("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
  78                                         $post["url"] = $matches[1];
  79                                 }
  80
  81                                 // Search for description
  82                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
  83                                         $post["description"] = $matches[1];
  84                                 }
  85                         }
  86                 }
  87                 return $post;
  88         }
  89
  90         /**
  91          * @brief Fetches attachment data that were generated with the "attachment" element
  92          *
  93          * @param string $body Message body
  94          * @return array
  95          * 'type' -> Message type ("link", "video", "photo")
  96          * 'text' -> Text before the shared message
  97          * 'after' -> Text after the shared message
  98          * 'image' -> Preview image of the message
  99          * 'url' -> Url to the attached message
 100          * 'title' -> Title of the attachment
 101          * 'description' -> Description of the attachment
 102          */
 103         public static function getAttachmentData($body)
 104         {
 105                 $data = [];
 106
 107                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 108                         return self::getOldAttachmentData($body);
 109                 }
 110
 111                 $attributes = $match[2];
 112
 113                 $data["text"] = trim($match[1]);
 114
 115                 $type = "";
 116                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 117                 if (x($matches, 1)) {
 118                         $type = strtolower($matches[1]);
 119                 }
 120
 121                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 122                 if (x($matches, 1)) {
 123                         $type = strtolower($matches[1]);
 124                 }
 125
 126                 if ($type == "") {
 127                         return [];
 128                 }
 129
 130                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
 131                         return [];
 132                 }
 133
 134                 if ($type != "") {
 135                         $data["type"] = $type;
 136                 }
 137
 138                 $url = "";
 139                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 140                 if (x($matches, 1)) {
 141                         $url = $matches[1];
 142                 }
 143
 144                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 145                 if (x($matches, 1)) {
 146                         $url = $matches[1];
 147                 }
 148
 149                 if ($url != "") {
 150                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 151                 }
 152
 153                 $title = "";
 154                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 155                 if (x($matches, 1)) {
 156                         $title = $matches[1];
 157                 }
 158
 159                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 160                 if (x($matches, 1)) {
 161                         $title = $matches[1];
 162                 }
 163
 164                 if ($title != "") {
 165                         $title = bbcode(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, false, true);
 166                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 167                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
 168                         $data["title"] = $title;
 169                 }
 170
 171                 $image = "";
 172                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 173                 if (x($matches, 1)) {
 174                         $image = $matches[1];
 175                 }
 176
 177                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 178                 if (x($matches, 1)) {
 179                         $image = $matches[1];
 180                 }
 181
 182                 if ($image != "") {
 183                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 184                 }
 185
 186                 $preview = "";
 187                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 188                 if (x($matches, 1)) {
 189                         $preview = $matches[1];
 190                 }
 191
 192                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 193                 if (x($matches, 1)) {
 194                         $preview = $matches[1];
 195                 }
 196
 197                 if ($preview != "") {
 198                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 199                 }
 200
 201                 $data["description"] = trim($match[3]);
 202
 203                 $data["after"] = trim($match[4]);
 204
 205                 return $data;
 206         }
 207
 208         public static function getAttachedData($body, $item = [])
 209         {
 210                 /*
 211                 - text:
 212                 - type: link, video, photo
 213                 - title:
 214                 - url:
 215                 - image:
 216                 - description:
 217                 - (thumbnail)
 218                 */
 219
 220                 $has_title = !empty($item['title']);
 221                 $plink = (!empty($item['plink']) ? $item['plink'] : '');
 222                 $post = self::getAttachmentData($body);
 223
 224                 // if nothing is found, it maybe having an image.
 225                 if (!isset($post["type"])) {
 226                         // Simplify image codes
 227                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 228
 229                         $URLSearchString = "^\[\]";
 230                         if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 231                                 if ((count($pictures) == 1) && !$has_title) {
 232                                         // Checking, if the link goes to a picture
 233                                         $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 234
 235                                         // Workaround:
 236                                         // Sometimes photo posts to the own album are not detected at the start.
 237                                         // So we seem to cannot use the cache for these cases. That's strange.
 238                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
 239                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 240                                         }
 241
 242                                         if ($data["type"] == "photo") {
 243                                                 $post["type"] = "photo";
 244                                                 if (isset($data["images"][0])) {
 245                                                         $post["image"] = $data["images"][0]["src"];
 246                                                         $post["url"] = $data["url"];
 247                                                 } else {
 248                                                         $post["image"] = $data["url"];
 249                                                 }
 250
 251                                                 $post["preview"] = $pictures[0][2];
 252                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
 253                                         } else {
 254                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 255                                                 if (substr($imgdata["mime"], 0, 6) == "image/") {
 256                                                         $post["type"] = "photo";
 257                                                         $post["image"] = $pictures[0][1];
 258                                                         $post["preview"] = $pictures[0][2];
 259                                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 260                                                 }
 261                                         }
 262                                 } elseif (count($pictures) > 0) {
 263                                         $post["type"] = "link";
 264                                         $post["url"] = $plink;
 265                                         $post["image"] = $pictures[0][2];
 266                                         $post["text"] = $body;
 267                                 }
 268                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 269                                 if ((count($pictures) == 1) && !$has_title) {
 270                                         $post["type"] = "photo";
 271                                         $post["image"] = $pictures[0][1];
 272                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 273                                 } elseif (count($pictures) > 0) {
 274                                         $post["type"] = "link";
 275                                         $post["url"] = $plink;
 276                                         $post["image"] = $pictures[0][1];
 277                                         $post["text"] = $body;
 278                                 }
 279                         }
 280
 281                         // Test for the external links
 282                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
 283                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
 284
 285                         $links = array_merge($links1, $links2);
 286
 287                         // If there is only a single one, then use it.
 288                         // This should cover link posts via API.
 289                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 290                                 $post["type"] = "link";
 291                                 $post["text"] = trim($body);
 292                                 $post["url"] = $links[0][1];
 293                         }
 294
 295                         // Now count the number of external media links
 296                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
 297                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
 298                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
 299                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
 300
 301                         // Add them to the other external links
 302                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 303
 304                         // Are there more than one?
 305                         if (count($links) > 1) {
 306                                 // The post will be the type "text", which means a blog post
 307                                 unset($post["type"]);
 308                                 $post["url"] = $plink;
 309                         }
 310
 311                         if (!isset($post["type"])) {
 312                                 $post["type"] = "text";
 313                                 $post["text"] = trim($body);
 314                         }
 315                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 316                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
 317
 318                         if (isset($data["images"][0])) {
 319                                 $post["image"] = $data["images"][0]["src"];
 320                         }
 321                 }
 322
 323                 return $post;
 324         }
 325
 326         /**
 327          * @brief Convert a message into plaintext for connectors to other networks
 328          *
 329          * @param array $b The message array that is about to be posted
 330          * @param int $limit The maximum number of characters when posting to that network
 331          * @param bool $includedlinks Has an attached link to be included into the message?
 332          * @param int $htmlmode This triggers the behaviour of the bbcode conversion
 333          * @param string $target_network Name of the network where the post should go to.
 334          *
 335          * @return string The converted message
 336          */
 337         public static function toPlaintext($b, $limit = 0, $includedlinks = false, $htmlmode = 2, $target_network = "")
 338         {
 339                 // Remove the hash tags
 340                 $URLSearchString = "^\[\]";
 341                 $body = preg_replace("/([#@])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $b["body"]);
 342
 343                 // Add an URL element if the text contains a raw link
 344                 $body = preg_replace("/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url]$2[/url]', $body);
 345
 346                 // Remove the abstract
 347                 $body = remove_abstract($body);
 348
 349                 // At first look at data that is attached via "type-..." stuff
 350                 // This will hopefully replaced with a dedicated bbcode later
 351                 //$post = self::getAttachedData($b["body"]);
 352                 $post = self::getAttachedData($body, $b);
 353
 354                 if (($b["title"] != "") && ($post["text"] != "")) {
 355                         $post["text"] = trim($b["title"]."\n\n".$post["text"]);
 356                 } elseif ($b["title"] != "") {
 357                         $post["text"] = trim($b["title"]);
 358                 }
 359
 360                 $abstract = "";
 361
 362                 // Fetch the abstract from the given target network
 363                 if ($target_network != "") {
 364                         $default_abstract = fetch_abstract($b["body"]);
 365                         $abstract = fetch_abstract($b["body"], $target_network);
 366
 367                         // If we post to a network with no limit we only fetch
 368                         // an abstract exactly for this network
 369                         if (($limit == 0) && ($abstract == $default_abstract)) {
 370                                 $abstract = "";
 371                         }
 372                 } else {// Try to guess the correct target network
 373                         switch ($htmlmode) {
 374                                 case 8:
 375                                         $abstract = fetch_abstract($b["body"], NETWORK_TWITTER);
 376                                         break;
 377                                 case 7:
 378                                         $abstract = fetch_abstract($b["body"], NETWORK_STATUSNET);
 379                                         break;
 380                                 case 6:
 381                                         $abstract = fetch_abstract($b["body"], NETWORK_APPNET);
 382                                         break;
 383                                 default: // We don't know the exact target.
 384                                         // We fetch an abstract since there is a posting limit.
 385                                         if ($limit > 0) {
 386                                                 $abstract = fetch_abstract($b["body"]);
 387                                         }
 388                         }
 389                 }
 390
 391                 if ($abstract != "") {
 392                         $post["text"] = $abstract;
 393
 394                         if ($post["type"] == "text") {
 395                                 $post["type"] = "link";
 396                                 $post["url"] = $b["plink"];
 397                         }
 398                 }
 399
 400                 $html = bbcode($post["text"].$post["after"], false, false, $htmlmode);
 401                 $msg = html2plain($html, 0, true);
 402                 $msg = trim(html_entity_decode($msg, ENT_QUOTES, 'UTF-8'));
 403
 404                 $link = "";
 405                 if ($includedlinks) {
 406                         if ($post["type"] == "link") {
 407                                 $link = $post["url"];
 408                         } elseif ($post["type"] == "text") {
 409                                 $link = $post["url"];
 410                         } elseif ($post["type"] == "video") {
 411                                 $link = $post["url"];
 412                         } elseif ($post["type"] == "photo") {
 413                                 $link = $post["image"];
 414                         }
 415
 416                         if (($msg == "") && isset($post["title"])) {
 417                                 $msg = trim($post["title"]);
 418                         }
 419
 420                         if (($msg == "") && isset($post["description"])) {
 421                                 $msg = trim($post["description"]);
 422                         }
 423
 424                         // If the link is already contained in the post, then it neeedn't to be added again
 425                         // But: if the link is beyond the limit, then it has to be added.
 426                         if (($link != "") && strstr($msg, $link)) {
 427                                 $pos = strpos($msg, $link);
 428
 429                                 // Will the text be shortened in the link?
 430                                 // Or is the link the last item in the post?
 431                                 if (($limit > 0) && ($pos < $limit) && (($pos + 23 > $limit) || ($pos + strlen($link) == strlen($msg)))) {
 432                                         $msg = trim(str_replace($link, "", $msg));
 433                                 } elseif (($limit == 0) || ($pos < $limit)) {
 434                                         // The limit has to be increased since it will be shortened - but not now
 435                                         // Only do it with Twitter (htmlmode = 8)
 436                                         if (($limit > 0) && (strlen($link) > 23) && ($htmlmode == 8)) {
 437                                                 $limit = $limit - 23 + strlen($link);
 438                                         }
 439
 440                                         $link = "";
 441
 442                                         if ($post["type"] == "text") {
 443                                                 unset($post["url"]);
 444                                         }
 445                                 }
 446                         }
 447                 }
 448
 449                 if ($limit > 0) {
 450                         // Reduce multiple spaces
 451                         // When posted to a network with limited space, we try to gain space where possible
 452                         while (strpos($msg, "  ") !== false) {
 453                                 $msg = str_replace("  ", " ", $msg);
 454                         }
 455
 456                         // Twitter is using its own limiter, so we always assume that shortened links will have this length
 457                         if (iconv_strlen($link, "UTF-8") > 0) {
 458                                 $limit = $limit - 23;
 459                         }
 460
 461                         if (iconv_strlen($msg, "UTF-8") > $limit) {
 462                                 if (($post["type"] == "text") && isset($post["url"])) {
 463                                         $post["url"] = $b["plink"];
 464                                 } elseif (!isset($post["url"])) {
 465                                         $limit = $limit - 23;
 466                                         $post["url"] = $b["plink"];
 467                                 // Which purpose has this line? It is now uncommented, but left as a reminder
 468                                 //} elseif (strpos($b["body"], "[share") !== false) {
 469                                 //      $post["url"] = $b["plink"];
 470                                 } elseif (PConfig::get($b["uid"], "system", "no_intelligent_shortening")) {
 471                                         $post["url"] = $b["plink"];
 472                                 }
 473                                 $msg = Plaintext::shorten($msg, $limit);
 474                         }
 475                 }
 476
 477                 $post["text"] = trim($msg);
 478
 479                 return($post);
 480         }
 481
 482         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
 483         {
 484                 // Suppress "view full size"
 485                 if (intval(Config::get('system', 'no_view_full_size'))) {
 486                         $include_link = false;
 487                 }
 488
 489                 // Picture addresses can contain special characters
 490                 $s = htmlspecialchars_decode($srctext);
 491
 492                 $matches = null;
 493                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
 494                 if ($c) {
 495                         foreach ($matches as $mtch) {
 496                                 logger('scale_external_image: ' . $mtch[1]);
 497
 498                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
 499                                 if (stristr($mtch[1], $hostname)) {
 500                                         continue;
 501                                 }
 502
 503                                 // $scale_replace, if passed, is an array of two elements. The
 504                                 // first is the name of the full-size image. The second is the
 505                                 // name of a remote, scaled-down version of the full size image.
 506                                 // This allows Friendica to display the smaller remote image if
 507                                 // one exists, while still linking to the full-size image
 508                                 if ($scale_replace) {
 509                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
 510                                 } else {
 511                                         $scaled = $mtch[1];
 512                                 }
 513                                 $i = self::fetchURL($scaled);
 514                                 if (! $i) {
 515                                         return $srctext;
 516                                 }
 517
 518                                 // guess mimetype from headers or filename
 519                                 $type = Image::guessType($mtch[1], true);
 520
 521                                 if ($i) {
 522                                         $Image = new Image($i, $type);
 523                                         if ($Image->isValid()) {
 524                                                 $orig_width = $Image->getWidth();
 525                                                 $orig_height = $Image->getHeight();
 526
 527                                                 if ($orig_width > 640 || $orig_height > 640) {
 528                                                         $Image->scaleDown(640);
 529                                                         $new_width = $Image->getWidth();
 530                                                         $new_height = $Image->getHeight();
 531                                                         logger('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], LOGGER_DEBUG);
 532                                                         $s = str_replace(
 533                                                                 $mtch[0],
 534                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
 535                                                                 . "\n" . (($include_link)
 536                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
 537                                                                         : ''),
 538                                                                 $s
 539                                                         );
 540                                                         logger('scale_external_images: new string: ' . $s, LOGGER_DEBUG);
 541                                                 }
 542                                         }
 543                                 }
 544                         }
 545                 }
 546
 547                 // replace the special char encoding
 548                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
 549                 return $s;
 550         }
 551 }