include/plaintext.php

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