src/Content/Text/Plaintext.php

   1 <?php
   2 /**
   3  * @file src/Content/Text/Plaintext.php
   4  */
   5 namespace Friendica\Content\Text;
   6
   7 use Friendica\App;
   8 use Friendica\Core\PConfig;
   9 use Friendica\Object\Image;
  10 use Friendica\Util\ParseUrl;
  11
  12 require_once "include/bbcode.php";
  13 require_once "include/html2plain.php";
  14 require_once "include/network.php";
  15
  16 class Plaintext
  17 {
  18         /**
  19          * @brief Fetches attachment data that were generated the old way
  20          *
  21          * @param string $body Message body
  22          * @return array
  23          * 'type' -> Message type ("link", "video", "photo")
  24          * 'text' -> Text before the shared message
  25          * 'after' -> Text after the shared message
  26          * 'image' -> Preview image of the message
  27          * 'url' -> Url to the attached message
  28          * 'title' -> Title of the attachment
  29          * 'description' -> Description of the attachment
  30          */
  31         private static function getOldAttachmentData($body)
  32         {
  33                 $post = [];
  34
  35                 // Simplify image codes
  36                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
  37
  38                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
  39                         foreach ($attached as $data) {
  40                                 if (!in_array($data[1], ["type-link", "type-video", "type-photo"])) {
  41                                         continue;
  42                                 }
  43
  44                                 $post["type"] = substr($data[1], 5);
  45
  46                                 $pos = strpos($body, $data[0]);
  47                                 if ($pos > 0) {
  48                                         $post["text"] = trim(substr($body, 0, $pos));
  49                                         $post["after"] = trim(substr($body, $pos + strlen($data[0])));
  50                                 } else {
  51                                         $post["text"] = trim(str_replace($data[0], "", $body));
  52                                 }
  53
  54                                 $attacheddata = $data[2];
  55
  56                                 $URLSearchString = "^\[\]";
  57
  58                                 if (preg_match("/\[img\]([$URLSearchString]*)\[\/img\]/ism", $attacheddata, $matches)) {
  59
  60                                         $picturedata = Image::getInfoFromURL($matches[1]);
  61
  62                                         if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
  63                                                 $post["image"] = $matches[1];
  64                                         } else {
  65                                                 $post["preview"] = $matches[1];
  66                                         }
  67                                 }
  68
  69                                 if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
  70                                         $post["url"] = $matches[1];
  71                                         $post["title"] = $matches[2];
  72                                 }
  73                                 if (($post["url"] == "") && (in_array($post["type"], ["link", "video"]))
  74                                         && preg_match("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
  75                                         $post["url"] = $matches[1];
  76                                 }
  77
  78                                 // Search for description
  79                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
  80                                         $post["description"] = $matches[1];
  81                                 }
  82                         }
  83                 }
  84                 return $post;
  85         }
  86
  87         /**
  88          * @brief Fetches attachment data that were generated with the "attachment" element
  89          *
  90          * @param string $body Message body
  91          * @return array
  92          * 'type' -> Message type ("link", "video", "photo")
  93          * 'text' -> Text before the shared message
  94          * 'after' -> Text after the shared message
  95          * 'image' -> Preview image of the message
  96          * 'url' -> Url to the attached message
  97          * 'title' -> Title of the attachment
  98          * 'description' -> Description of the attachment
  99          */
 100         public static function getAttachmentData($body)
 101         {
 102                 $data = [];
 103
 104                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
 105                         return self::getOldAttachmentData($body);
 106                 }
 107
 108                 $attributes = $match[2];
 109
 110                 $data["text"] = trim($match[1]);
 111
 112                 $type = "";
 113                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
 114                 if (x($matches, 1)) {
 115                         $type = strtolower($matches[1]);
 116                 }
 117
 118                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
 119                 if (x($matches, 1)) {
 120                         $type = strtolower($matches[1]);
 121                 }
 122
 123                 if ($type == "") {
 124                         return [];
 125                 }
 126
 127                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
 128                         return [];
 129                 }
 130
 131                 if ($type != "") {
 132                         $data["type"] = $type;
 133                 }
 134
 135                 $url = "";
 136                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
 137                 if (x($matches, 1)) {
 138                         $url = $matches[1];
 139                 }
 140
 141                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
 142                 if (x($matches, 1)) {
 143                         $url = $matches[1];
 144                 }
 145
 146                 if ($url != "") {
 147                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
 148                 }
 149
 150                 $title = "";
 151                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
 152                 if (x($matches, 1)) {
 153                         $title = $matches[1];
 154                 }
 155
 156                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
 157                 if (x($matches, 1)) {
 158                         $title = $matches[1];
 159                 }
 160
 161                 if ($title != "") {
 162                         $title = bbcode(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, false, true);
 163                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
 164                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
 165                         $data["title"] = $title;
 166                 }
 167
 168                 $image = "";
 169                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
 170                 if (x($matches, 1)) {
 171                         $image = $matches[1];
 172                 }
 173
 174                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
 175                 if (x($matches, 1)) {
 176                         $image = $matches[1];
 177                 }
 178
 179                 if ($image != "") {
 180                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
 181                 }
 182
 183                 $preview = "";
 184                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
 185                 if (x($matches, 1)) {
 186                         $preview = $matches[1];
 187                 }
 188
 189                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
 190                 if (x($matches, 1)) {
 191                         $preview = $matches[1];
 192                 }
 193
 194                 if ($preview != "") {
 195                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
 196                 }
 197
 198                 $data["description"] = trim($match[3]);
 199
 200                 $data["after"] = trim($match[4]);
 201
 202                 return $data;
 203         }
 204
 205         public static function getAttachedData($body, $item = [])
 206         {
 207                 /*
 208                 - text:
 209                 - type: link, video, photo
 210                 - title:
 211                 - url:
 212                 - image:
 213                 - description:
 214                 - (thumbnail)
 215                 */
 216
 217                 $has_title = !empty($item['title']);
 218                 $plink = (!empty($item['plink']) ? $item['plink'] : '');
 219                 $post = self::getAttachmentData($body);
 220
 221                 // if nothing is found, it maybe having an image.
 222                 if (!isset($post["type"])) {
 223                         // Simplify image codes
 224                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
 225
 226                         $URLSearchString = "^\[\]";
 227                         if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
 228                                 if ((count($pictures) == 1) && !$has_title) {
 229                                         // Checking, if the link goes to a picture
 230                                         $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
 231
 232                                         // Workaround:
 233                                         // Sometimes photo posts to the own album are not detected at the start.
 234                                         // So we seem to cannot use the cache for these cases. That's strange.
 235                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
 236                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
 237                                         }
 238
 239                                         if ($data["type"] == "photo") {
 240                                                 $post["type"] = "photo";
 241                                                 if (isset($data["images"][0])) {
 242                                                         $post["image"] = $data["images"][0]["src"];
 243                                                         $post["url"] = $data["url"];
 244                                                 } else {
 245                                                         $post["image"] = $data["url"];
 246                                                 }
 247
 248                                                 $post["preview"] = $pictures[0][2];
 249                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
 250                                         } else {
 251                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
 252                                                 if (substr($imgdata["mime"], 0, 6) == "image/") {
 253                                                         $post["type"] = "photo";
 254                                                         $post["image"] = $pictures[0][1];
 255                                                         $post["preview"] = $pictures[0][2];
 256                                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 257                                                 }
 258                                         }
 259                                 } elseif (count($pictures) > 0) {
 260                                         $post["type"] = "link";
 261                                         $post["url"] = $plink;
 262                                         $post["image"] = $pictures[0][2];
 263                                         $post["text"] = $body;
 264                                 }
 265                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
 266                                 if ((count($pictures) == 1) && !$has_title) {
 267                                         $post["type"] = "photo";
 268                                         $post["image"] = $pictures[0][1];
 269                                         $post["text"] = str_replace($pictures[0][0], "", $body);
 270                                 } elseif (count($pictures) > 0) {
 271                                         $post["type"] = "link";
 272                                         $post["url"] = $plink;
 273                                         $post["image"] = $pictures[0][1];
 274                                         $post["text"] = $body;
 275                                 }
 276                         }
 277
 278                         // Test for the external links
 279                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
 280                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
 281
 282                         $links = array_merge($links1, $links2);
 283
 284                         // If there is only a single one, then use it.
 285                         // This should cover link posts via API.
 286                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
 287                                 $post["type"] = "link";
 288                                 $post["text"] = trim($body);
 289                                 $post["url"] = $links[0][1];
 290                         }
 291
 292                         // Now count the number of external media links
 293                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
 294                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
 295                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
 296                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
 297
 298                         // Add them to the other external links
 299                         $links = array_merge($links, $links1, $links2, $links3, $links4);
 300
 301                         // Are there more than one?
 302                         if (count($links) > 1) {
 303                                 // The post will be the type "text", which means a blog post
 304                                 unset($post["type"]);
 305                                 $post["url"] = $plink;
 306                         }
 307
 308                         if (!isset($post["type"])) {
 309                                 $post["type"] = "text";
 310                                 $post["text"] = trim($body);
 311                         }
 312                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
 313                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
 314
 315                         if (isset($data["images"][0])) {
 316                                 $post["image"] = $data["images"][0]["src"];
 317                         }
 318                 }
 319
 320                 return $post;
 321         }
 322
 323         /**
 324          * Shortens message
 325          *
 326          * @param type $msg
 327          * @param type $limit
 328          * @return type
 329          *
 330          * @todo For Twitter URLs aren't shortened, but they have to be calculated as if.
 331          */
 332         public static function shortenMsg($msg, $limit)
 333         {
 334                 $lines = explode("\n", $msg);
 335                 $msg = "";
 336                 $recycle = html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8');
 337                 $ellipsis = html_entity_decode("&#x2026;", ENT_QUOTES, 'UTF-8');
 338                 foreach ($lines as $row => $line) {
 339                         if (iconv_strlen(trim($msg . "\n" . $line), "UTF-8") <= $limit) {
 340                                 $msg = trim($msg . "\n" . $line);
 341                         } elseif (($msg == "") || (($row == 1) && (substr($msg, 0, 4) == $recycle))) {
 342                                 // Is the new message empty by now or is it a reshared message?
 343                                 $msg = iconv_substr(iconv_substr(trim($msg . "\n" . $line), 0, $limit, "UTF-8"), 0, -3, "UTF-8") . $ellipsis;
 344                         } else {
 345                                 break;
 346                         }
 347                 }
 348
 349                 return $msg;
 350         }
 351
 352         /**
 353          * @brief Convert a message into plaintext for connectors to other networks
 354          *
 355          * @param array $b The message array that is about to be posted
 356          * @param int $limit The maximum number of characters when posting to that network
 357          * @param bool $includedlinks Has an attached link to be included into the message?
 358          * @param int $htmlmode This triggers the behaviour of the bbcode conversion
 359          * @param string $target_network Name of the network where the post should go to.
 360          *
 361          * @return string The converted message
 362          */
 363         public static function toPlaintext($b, $limit = 0, $includedlinks = false, $htmlmode = 2, $target_network = "")
 364         {
 365                 // Remove the hash tags
 366                 $URLSearchString = "^\[\]";
 367                 $body = preg_replace("/([#@])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $b["body"]);
 368
 369                 // Add an URL element if the text contains a raw link
 370                 $body = preg_replace("/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url]$2[/url]', $body);
 371
 372                 // Remove the abstract
 373                 $body = remove_abstract($body);
 374
 375                 // At first look at data that is attached via "type-..." stuff
 376                 // This will hopefully replaced with a dedicated bbcode later
 377                 //$post = self::getAttachedData($b["body"]);
 378                 $post = self::getAttachedData($body, $b);
 379
 380                 if (($b["title"] != "") && ($post["text"] != "")) {
 381                         $post["text"] = trim($b["title"]."\n\n".$post["text"]);
 382                 } elseif ($b["title"] != "") {
 383                         $post["text"] = trim($b["title"]);
 384                 }
 385
 386                 $abstract = "";
 387
 388                 // Fetch the abstract from the given target network
 389                 if ($target_network != "") {
 390                         $default_abstract = fetch_abstract($b["body"]);
 391                         $abstract = fetch_abstract($b["body"], $target_network);
 392
 393                         // If we post to a network with no limit we only fetch
 394                         // an abstract exactly for this network
 395                         if (($limit == 0) && ($abstract == $default_abstract)) {
 396                                 $abstract = "";
 397                         }
 398                 } else {// Try to guess the correct target network
 399                         switch ($htmlmode) {
 400                                 case 8:
 401                                         $abstract = fetch_abstract($b["body"], NETWORK_TWITTER);
 402                                         break;
 403                                 case 7:
 404                                         $abstract = fetch_abstract($b["body"], NETWORK_STATUSNET);
 405                                         break;
 406                                 case 6:
 407                                         $abstract = fetch_abstract($b["body"], NETWORK_APPNET);
 408                                         break;
 409                                 default: // We don't know the exact target.
 410                                         // We fetch an abstract since there is a posting limit.
 411                                         if ($limit > 0) {
 412                                                 $abstract = fetch_abstract($b["body"]);
 413                                         }
 414                         }
 415                 }
 416
 417                 if ($abstract != "") {
 418                         $post["text"] = $abstract;
 419
 420                         if ($post["type"] == "text") {
 421                                 $post["type"] = "link";
 422                                 $post["url"] = $b["plink"];
 423                         }
 424                 }
 425
 426                 $html = bbcode($post["text"].$post["after"], false, false, $htmlmode);
 427                 $msg = html2plain($html, 0, true);
 428                 $msg = trim(html_entity_decode($msg, ENT_QUOTES, 'UTF-8'));
 429
 430                 $link = "";
 431                 if ($includedlinks) {
 432                         if ($post["type"] == "link") {
 433                                 $link = $post["url"];
 434                         } elseif ($post["type"] == "text") {
 435                                 $link = $post["url"];
 436                         } elseif ($post["type"] == "video") {
 437                                 $link = $post["url"];
 438                         } elseif ($post["type"] == "photo") {
 439                                 $link = $post["image"];
 440                         }
 441
 442                         if (($msg == "") && isset($post["title"])) {
 443                                 $msg = trim($post["title"]);
 444                         }
 445
 446                         if (($msg == "") && isset($post["description"])) {
 447                                 $msg = trim($post["description"]);
 448                         }
 449
 450                         // If the link is already contained in the post, then it neeedn't to be added again
 451                         // But: if the link is beyond the limit, then it has to be added.
 452                         if (($link != "") && strstr($msg, $link)) {
 453                                 $pos = strpos($msg, $link);
 454
 455                                 // Will the text be shortened in the link?
 456                                 // Or is the link the last item in the post?
 457                                 if (($limit > 0) && ($pos < $limit) && (($pos + 23 > $limit) || ($pos + strlen($link) == strlen($msg)))) {
 458                                         $msg = trim(str_replace($link, "", $msg));
 459                                 } elseif (($limit == 0) || ($pos < $limit)) {
 460                                         // The limit has to be increased since it will be shortened - but not now
 461                                         // Only do it with Twitter (htmlmode = 8)
 462                                         if (($limit > 0) && (strlen($link) > 23) && ($htmlmode == 8)) {
 463                                                 $limit = $limit - 23 + strlen($link);
 464                                         }
 465
 466                                         $link = "";
 467
 468                                         if ($post["type"] == "text") {
 469                                                 unset($post["url"]);
 470                                         }
 471                                 }
 472                         }
 473                 }
 474
 475                 if ($limit > 0) {
 476                         // Reduce multiple spaces
 477                         // When posted to a network with limited space, we try to gain space where possible
 478                         while (strpos($msg, "  ") !== false) {
 479                                 $msg = str_replace("  ", " ", $msg);
 480                         }
 481
 482                         // Twitter is using its own limiter, so we always assume that shortened links will have this length
 483                         if (iconv_strlen($link, "UTF-8") > 0) {
 484                                 $limit = $limit - 23;
 485                         }
 486
 487                         if (iconv_strlen($msg, "UTF-8") > $limit) {
 488                                 if (($post["type"] == "text") && isset($post["url"])) {
 489                                         $post["url"] = $b["plink"];
 490                                 } elseif (!isset($post["url"])) {
 491                                         $limit = $limit - 23;
 492                                         $post["url"] = $b["plink"];
 493                                 // Which purpose has this line? It is now uncommented, but left as a reminder
 494                                 //} elseif (strpos($b["body"], "[share") !== false) {
 495                                 //      $post["url"] = $b["plink"];
 496                                 } elseif (PConfig::get($b["uid"], "system", "no_intelligent_shortening")) {
 497                                         $post["url"] = $b["plink"];
 498                                 }
 499                                 $msg = self::shortenMsg($msg, $limit);
 500                         }
 501                 }
 502
 503                 $post["text"] = trim($msg);
 504
 505                 return($post);
 506         }
 507 }