include/plaintext.php

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