Add bash to text_highlight language list
[friendica.git/.git] / src / Content / Text / BBCode.php
1 <?php
2
3 /**
4  * @file src/Content/Text/BBCode.php
5  */
6
7 namespace Friendica\Content\Text;
8
9 use DOMDocument;
10 use DOMXPath;
11 use Exception;
12 use Friendica\BaseObject;
13 use Friendica\Content\OEmbed;
14 use Friendica\Content\Smilies;
15 use Friendica\Core\Addon;
16 use Friendica\Core\Cache;
17 use Friendica\Core\Config;
18 use Friendica\Core\L10n;
19 use Friendica\Core\PConfig;
20 use Friendica\Core\Protocol;
21 use Friendica\Core\System;
22 use Friendica\Model\Contact;
23 use Friendica\Model\Event;
24 use Friendica\Network\Probe;
25 use Friendica\Object\Image;
26 use Friendica\Util\Map;
27 use Friendica\Util\Network;
28 use Friendica\Util\ParseUrl;
29 use League\HTMLToMarkdown\HtmlConverter;
30
31 require_once "mod/proxy.php";
32
33 class BBCode extends BaseObject
34 {
35         /**
36          * @brief Fetches attachment data that were generated the old way
37          *
38          * @param string $body Message body
39          * @return array
40          * 'type' -> Message type ("link", "video", "photo")
41          * 'text' -> Text before the shared message
42          * 'after' -> Text after the shared message
43          * 'image' -> Preview image of the message
44          * 'url' -> Url to the attached message
45          * 'title' -> Title of the attachment
46          * 'description' -> Description of the attachment
47          */
48         private static function getOldAttachmentData($body)
49         {
50                 $post = [];
51
52                 // Simplify image codes
53                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
54
55                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
56                         foreach ($attached as $data) {
57                                 if (!in_array($data[1], ["type-link", "type-video", "type-photo"])) {
58                                         continue;
59                                 }
60
61                                 $post["type"] = substr($data[1], 5);
62
63                                 $pos = strpos($body, $data[0]);
64                                 if ($pos > 0) {
65                                         $post["text"] = trim(substr($body, 0, $pos));
66                                         $post["after"] = trim(substr($body, $pos + strlen($data[0])));
67                                 } else {
68                                         $post["text"] = trim(str_replace($data[0], "", $body));
69                                 }
70
71                                 $attacheddata = $data[2];
72
73                                 $URLSearchString = "^\[\]";
74
75                                 if (preg_match("/\[img\]([$URLSearchString]*)\[\/img\]/ism", $attacheddata, $matches)) {
76
77                                         $picturedata = Image::getInfoFromURL($matches[1]);
78
79                                         if ($picturedata) {
80                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
81                                                         $post["image"] = $matches[1];
82                                                 } else {
83                                                         $post["preview"] = $matches[1];
84                                                 }
85                                         }
86                                 }
87
88                                 if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
89                                         $post["url"] = $matches[1];
90                                         $post["title"] = $matches[2];
91                                 }
92                                 if (($post["url"] == "") && (in_array($post["type"], ["link", "video"]))
93                                         && preg_match("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
94                                         $post["url"] = $matches[1];
95                                 }
96
97                                 // Search for description
98                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
99                                         $post["description"] = $matches[1];
100                                 }
101                         }
102                 }
103                 return $post;
104         }
105
106         /**
107          * @brief Fetches attachment data that were generated with the "attachment" element
108          *
109          * @param string $body Message body
110          * @return array
111          * 'type' -> Message type ("link", "video", "photo")
112          * 'text' -> Text before the shared message
113          * 'after' -> Text after the shared message
114          * 'image' -> Preview image of the message
115          * 'url' -> Url to the attached message
116          * 'title' -> Title of the attachment
117          * 'description' -> Description of the attachment
118          */
119         public static function getAttachmentData($body)
120         {
121                 $data = [];
122
123                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
124                         return self::getOldAttachmentData($body);
125                 }
126
127                 $attributes = $match[2];
128
129                 $data["text"] = trim($match[1]);
130
131                 $type = "";
132                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
133                 if (x($matches, 1)) {
134                         $type = strtolower($matches[1]);
135                 }
136
137                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
138                 if (x($matches, 1)) {
139                         $type = strtolower($matches[1]);
140                 }
141
142                 if ($type == "") {
143                         return [];
144                 }
145
146                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
147                         return [];
148                 }
149
150                 if ($type != "") {
151                         $data["type"] = $type;
152                 }
153
154                 $url = "";
155                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
156                 if (x($matches, 1)) {
157                         $url = $matches[1];
158                 }
159
160                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
161                 if (x($matches, 1)) {
162                         $url = $matches[1];
163                 }
164
165                 if ($url != "") {
166                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
167                 }
168
169                 $title = "";
170                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
171                 if (x($matches, 1)) {
172                         $title = $matches[1];
173                 }
174
175                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
176                 if (x($matches, 1)) {
177                         $title = $matches[1];
178                 }
179
180                 if ($title != "") {
181                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
182                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
183                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
184                         $data["title"] = $title;
185                 }
186
187                 $image = "";
188                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
189                 if (x($matches, 1)) {
190                         $image = $matches[1];
191                 }
192
193                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
194                 if (x($matches, 1)) {
195                         $image = $matches[1];
196                 }
197
198                 if ($image != "") {
199                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
200                 }
201
202                 $preview = "";
203                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
204                 if (x($matches, 1)) {
205                         $preview = $matches[1];
206                 }
207
208                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
209                 if (x($matches, 1)) {
210                         $preview = $matches[1];
211                 }
212
213                 if ($preview != "") {
214                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
215                 }
216
217                 $data["description"] = trim($match[3]);
218
219                 $data["after"] = trim($match[4]);
220
221                 return $data;
222         }
223
224         public static function getAttachedData($body, $item = [])
225         {
226                 /*
227                 - text:
228                 - type: link, video, photo
229                 - title:
230                 - url:
231                 - image:
232                 - description:
233                 - (thumbnail)
234                 */
235
236                 $has_title = !empty($item['title']);
237                 $plink = (!empty($item['plink']) ? $item['plink'] : '');
238                 $post = self::getAttachmentData($body);
239
240                 // if nothing is found, it maybe having an image.
241                 if (!isset($post["type"])) {
242                         // Simplify image codes
243                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
244
245                         $URLSearchString = "^\[\]";
246                         if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
247                                 if ((count($pictures) == 1) && !$has_title) {
248                                         // Checking, if the link goes to a picture
249                                         $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
250
251                                         // Workaround:
252                                         // Sometimes photo posts to the own album are not detected at the start.
253                                         // So we seem to cannot use the cache for these cases. That's strange.
254                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
255                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
256                                         }
257
258                                         if ($data["type"] == "photo") {
259                                                 $post["type"] = "photo";
260                                                 if (isset($data["images"][0])) {
261                                                         $post["image"] = $data["images"][0]["src"];
262                                                         $post["url"] = $data["url"];
263                                                 } else {
264                                                         $post["image"] = $data["url"];
265                                                 }
266
267                                                 $post["preview"] = $pictures[0][2];
268                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
269                                         } else {
270                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
271                                                 if ($imgdata && substr($imgdata["mime"], 0, 6) == "image/") {
272                                                         $post["type"] = "photo";
273                                                         $post["image"] = $pictures[0][1];
274                                                         $post["preview"] = $pictures[0][2];
275                                                         $post["text"] = str_replace($pictures[0][0], "", $body);
276                                                 }
277                                         }
278                                 } elseif (count($pictures) > 0) {
279                                         $post["type"] = "link";
280                                         $post["url"] = $plink;
281                                         $post["image"] = $pictures[0][2];
282                                         $post["text"] = $body;
283                                 }
284                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
285                                 if ((count($pictures) == 1) && !$has_title) {
286                                         $post["type"] = "photo";
287                                         $post["image"] = $pictures[0][1];
288                                         $post["text"] = str_replace($pictures[0][0], "", $body);
289                                 } elseif (count($pictures) > 0) {
290                                         $post["type"] = "link";
291                                         $post["url"] = $plink;
292                                         $post["image"] = $pictures[0][1];
293                                         $post["text"] = $body;
294                                 }
295                         }
296
297                         // Test for the external links
298                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
299                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
300
301                         $links = array_merge($links1, $links2);
302
303                         // If there is only a single one, then use it.
304                         // This should cover link posts via API.
305                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
306                                 $post["type"] = "link";
307                                 $post["text"] = trim($body);
308                                 $post["url"] = $links[0][1];
309                         }
310
311                         // Now count the number of external media links
312                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
313                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
314                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
315                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
316
317                         // Add them to the other external links
318                         $links = array_merge($links, $links1, $links2, $links3, $links4);
319
320                         // Are there more than one?
321                         if (count($links) > 1) {
322                                 // The post will be the type "text", which means a blog post
323                                 unset($post["type"]);
324                                 $post["url"] = $plink;
325                         }
326
327                         if (!isset($post["type"])) {
328                                 $post["type"] = "text";
329                                 $post["text"] = trim($body);
330                         }
331                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
332                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
333
334                         if (isset($data["images"][0])) {
335                                 $post["image"] = $data["images"][0]["src"];
336                         }
337                 }
338
339                 return $post;
340         }
341
342         /**
343          * @brief Convert a message into plaintext for connectors to other networks
344          *
345          * @param array $b The message array that is about to be posted
346          * @param int $limit The maximum number of characters when posting to that network
347          * @param bool $includedlinks Has an attached link to be included into the message?
348          * @param int $htmlmode This triggers the behaviour of the bbcode conversion
349          * @param string $target_network Name of the network where the post should go to.
350          *
351          * @return string The converted message
352          */
353         public static function toPlaintext($b, $limit = 0, $includedlinks = false, $htmlmode = 2, $target_network = "")
354         {
355                 // Remove the hash tags
356                 $URLSearchString = "^\[\]";
357                 $body = preg_replace("/([#@])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $b["body"]);
358
359                 // Add an URL element if the text contains a raw link
360                 $body = preg_replace("/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url]$2[/url]', $body);
361
362                 // Remove the abstract
363                 $body = self::stripAbstract($body);
364
365                 // At first look at data that is attached via "type-..." stuff
366                 // This will hopefully replaced with a dedicated bbcode later
367                 //$post = self::getAttachedData($b["body"]);
368                 $post = self::getAttachedData($body, $b);
369
370                 if (($b["title"] != "") && ($post["text"] != "")) {
371                         $post["text"] = trim($b["title"]."\n\n".$post["text"]);
372                 } elseif ($b["title"] != "") {
373                         $post["text"] = trim($b["title"]);
374                 }
375
376                 $abstract = "";
377
378                 // Fetch the abstract from the given target network
379                 if ($target_network != "") {
380                         $default_abstract = self::getAbstract($b["body"]);
381                         $abstract = self::getAbstract($b["body"], $target_network);
382
383                         // If we post to a network with no limit we only fetch
384                         // an abstract exactly for this network
385                         if (($limit == 0) && ($abstract == $default_abstract)) {
386                                 $abstract = "";
387                         }
388                 } else {// Try to guess the correct target network
389                         switch ($htmlmode) {
390                                 case 8:
391                                         $abstract = self::getAbstract($b["body"], NETWORK_TWITTER);
392                                         break;
393                                 case 7:
394                                         $abstract = self::getAbstract($b["body"], NETWORK_STATUSNET);
395                                         break;
396                                 case 6:
397                                         $abstract = self::getAbstract($b["body"], NETWORK_APPNET);
398                                         break;
399                                 default: // We don't know the exact target.
400                                         // We fetch an abstract since there is a posting limit.
401                                         if ($limit > 0) {
402                                                 $abstract = self::getAbstract($b["body"]);
403                                         }
404                         }
405                 }
406
407                 if ($abstract != "") {
408                         $post["text"] = $abstract;
409
410                         if ($post["type"] == "text") {
411                                 $post["type"] = "link";
412                                 $post["url"] = $b["plink"];
413                         }
414                 }
415
416                 $html = self::convert($post["text"].$post["after"], false, $htmlmode);
417                 $msg = HTML::toPlaintext($html, 0, true);
418                 $msg = trim(html_entity_decode($msg, ENT_QUOTES, 'UTF-8'));
419
420                 $link = "";
421                 if ($includedlinks) {
422                         if ($post["type"] == "link") {
423                                 $link = $post["url"];
424                         } elseif ($post["type"] == "text") {
425                                 $link = $post["url"];
426                         } elseif ($post["type"] == "video") {
427                                 $link = $post["url"];
428                         } elseif ($post["type"] == "photo") {
429                                 $link = $post["image"];
430                         }
431
432                         if (($msg == "") && isset($post["title"])) {
433                                 $msg = trim($post["title"]);
434                         }
435
436                         if (($msg == "") && isset($post["description"])) {
437                                 $msg = trim($post["description"]);
438                         }
439
440                         // If the link is already contained in the post, then it neeedn't to be added again
441                         // But: if the link is beyond the limit, then it has to be added.
442                         if (($link != "") && strstr($msg, $link)) {
443                                 $pos = strpos($msg, $link);
444
445                                 // Will the text be shortened in the link?
446                                 // Or is the link the last item in the post?
447                                 if (($limit > 0) && ($pos < $limit) && (($pos + 23 > $limit) || ($pos + strlen($link) == strlen($msg)))) {
448                                         $msg = trim(str_replace($link, "", $msg));
449                                 } elseif (($limit == 0) || ($pos < $limit)) {
450                                         // The limit has to be increased since it will be shortened - but not now
451                                         // Only do it with Twitter (htmlmode = 8)
452                                         if (($limit > 0) && (strlen($link) > 23) && ($htmlmode == 8)) {
453                                                 $limit = $limit - 23 + strlen($link);
454                                         }
455
456                                         $link = "";
457
458                                         if ($post["type"] == "text") {
459                                                 unset($post["url"]);
460                                         }
461                                 }
462                         }
463                 }
464
465                 if ($limit > 0) {
466                         // Reduce multiple spaces
467                         // When posted to a network with limited space, we try to gain space where possible
468                         while (strpos($msg, "  ") !== false) {
469                                 $msg = str_replace("  ", " ", $msg);
470                         }
471
472                         // Twitter is using its own limiter, so we always assume that shortened links will have this length
473                         if (iconv_strlen($link, "UTF-8") > 0) {
474                                 $limit = $limit - 23;
475                         }
476
477                         if (iconv_strlen($msg, "UTF-8") > $limit) {
478                                 if (($post["type"] == "text") && isset($post["url"])) {
479                                         $post["url"] = $b["plink"];
480                                 } elseif (!isset($post["url"])) {
481                                         $limit = $limit - 23;
482                                         $post["url"] = $b["plink"];
483                                 // Which purpose has this line? It is now uncommented, but left as a reminder
484                                 //} elseif (strpos($b["body"], "[share") !== false) {
485                                 //      $post["url"] = $b["plink"];
486                                 } elseif (PConfig::get($b["uid"], "system", "no_intelligent_shortening")) {
487                                         $post["url"] = $b["plink"];
488                                 }
489                                 $msg = Plaintext::shorten($msg, $limit);
490                         }
491                 }
492
493                 $post["text"] = trim($msg);
494
495                 return($post);
496         }
497
498         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
499         {
500                 // Suppress "view full size"
501                 if (intval(Config::get('system', 'no_view_full_size'))) {
502                         $include_link = false;
503                 }
504
505                 // Picture addresses can contain special characters
506                 $s = htmlspecialchars_decode($srctext);
507
508                 $matches = null;
509                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
510                 if ($c) {
511                         foreach ($matches as $mtch) {
512                                 logger('scale_external_image: ' . $mtch[1]);
513
514                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
515                                 if (stristr($mtch[1], $hostname)) {
516                                         continue;
517                                 }
518
519                                 // $scale_replace, if passed, is an array of two elements. The
520                                 // first is the name of the full-size image. The second is the
521                                 // name of a remote, scaled-down version of the full size image.
522                                 // This allows Friendica to display the smaller remote image if
523                                 // one exists, while still linking to the full-size image
524                                 if ($scale_replace) {
525                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
526                                 } else {
527                                         $scaled = $mtch[1];
528                                 }
529                                 $i = Network::fetchUrl($scaled);
530                                 if (!$i) {
531                                         return $srctext;
532                                 }
533
534                                 // guess mimetype from headers or filename
535                                 $type = Image::guessType($mtch[1], true);
536
537                                 if ($i) {
538                                         $Image = new Image($i, $type);
539                                         if ($Image->isValid()) {
540                                                 $orig_width = $Image->getWidth();
541                                                 $orig_height = $Image->getHeight();
542
543                                                 if ($orig_width > 640 || $orig_height > 640) {
544                                                         $Image->scaleDown(640);
545                                                         $new_width = $Image->getWidth();
546                                                         $new_height = $Image->getHeight();
547                                                         logger('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], LOGGER_DEBUG);
548                                                         $s = str_replace(
549                                                                 $mtch[0],
550                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
551                                                                 . "\n" . (($include_link)
552                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
553                                                                         : ''),
554                                                                 $s
555                                                         );
556                                                         logger('scale_external_images: new string: ' . $s, LOGGER_DEBUG);
557                                                 }
558                                         }
559                                 }
560                         }
561                 }
562
563                 // replace the special char encoding
564                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
565                 return $s;
566         }
567
568         /**
569          * The purpose of this function is to apply system message length limits to
570          * imported messages without including any embedded photos in the length
571          *
572          * @brief Truncates imported message body string length to max_import_size
573          * @param string $body
574          * @return string
575          */
576         public static function limitBodySize($body)
577         {
578                 $maxlen = get_max_import_size();
579
580                 // If the length of the body, including the embedded images, is smaller
581                 // than the maximum, then don't waste time looking for the images
582                 if ($maxlen && (strlen($body) > $maxlen)) {
583
584                         logger('the total body length exceeds the limit', LOGGER_DEBUG);
585
586                         $orig_body = $body;
587                         $new_body = '';
588                         $textlen = 0;
589
590                         $img_start = strpos($orig_body, '[img');
591                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
592                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
593                         while (($img_st_close !== false) && ($img_end !== false)) {
594
595                                 $img_st_close++; // make it point to AFTER the closing bracket
596                                 $img_end += $img_start;
597                                 $img_end += strlen('[/img]');
598
599                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
600                                         // This is an embedded image
601
602                                         if (($textlen + $img_start) > $maxlen) {
603                                                 if ($textlen < $maxlen) {
604                                                         logger('the limit happens before an embedded image', LOGGER_DEBUG);
605                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
606                                                         $textlen = $maxlen;
607                                                 }
608                                         } else {
609                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
610                                                 $textlen += $img_start;
611                                         }
612
613                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
614                                 } else {
615
616                                         if (($textlen + $img_end) > $maxlen) {
617                                                 if ($textlen < $maxlen) {
618                                                         logger('the limit happens before the end of a non-embedded image', LOGGER_DEBUG);
619                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
620                                                         $textlen = $maxlen;
621                                                 }
622                                         } else {
623                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
624                                                 $textlen += $img_end;
625                                         }
626                                 }
627                                 $orig_body = substr($orig_body, $img_end);
628
629                                 if ($orig_body === false) {
630                                         // in case the body ends on a closing image tag
631                                         $orig_body = '';
632                                 }
633
634                                 $img_start = strpos($orig_body, '[img');
635                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
636                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
637                         }
638
639                         if (($textlen + strlen($orig_body)) > $maxlen) {
640                                 if ($textlen < $maxlen) {
641                                         logger('the limit happens after the end of the last image', LOGGER_DEBUG);
642                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
643                                 }
644                         } else {
645                                 logger('the text size with embedded images extracted did not violate the limit', LOGGER_DEBUG);
646                                 $new_body = $new_body . $orig_body;
647                         }
648
649                         return $new_body;
650                 } else {
651                         return $body;
652                 }
653         }
654
655         /**
656          * Processes [attachment] tags
657          *
658          * Note: Can produce a [bookmark] tag in the returned string
659          *
660          * @brief Processes [attachment] tags
661          * @param string $return
662          * @param bool|int $simplehtml
663          * @param bool $tryoembed
664          * @return string
665          */
666         private static function convertAttachment($return, $simplehtml = false, $tryoembed = true)
667         {
668                 $data = self::getAttachmentData($return);
669                 if (!$data) {
670                         return $return;
671                 }
672
673                 if (isset($data["title"])) {
674                         $data["title"] = strip_tags($data["title"]);
675                         $data["title"] = str_replace(["http://", "https://"], "", $data["title"]);
676                 }
677
678                 if (((strpos($data["text"], "[img=") !== false) || (strpos($data["text"], "[img]") !== false) || Config::get('system', 'always_show_preview')) && ($data["image"] != "")) {
679                         $data["preview"] = $data["image"];
680                         $data["image"] = "";
681                 }
682
683                 $return = '';
684                 if ($simplehtml == 7) {
685                         $return = self::convertUrlForOStatus($data["url"]);
686                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
687                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data["url"], $data["title"]);
688                 } else {
689                         try {
690                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
691                                         $return = OEmbed::getHTML($data['url'], $data['title']);
692                                 } else {
693                                         throw new Exception('OEmbed is disabled for this attachment.');
694                                 }
695                         } catch (Exception $e) {
696                                 if ($simplehtml != 4) {
697                                         $return = sprintf('<div class="type-%s">', $data["type"]);
698                                 }
699
700                                 if ($data["image"] != "") {
701                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a><br />', $data["url"], proxy_url($data["image"]), $data["title"]);
702                                 } elseif ($data["preview"] != "") {
703                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br />', $data["url"], proxy_url($data["preview"]), $data["title"]);
704                                 }
705
706                                 if (($data["type"] == "photo") && ($data["url"] != "") && ($data["image"] != "")) {
707                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data["url"], proxy_url($data["image"]), $data["title"]);
708                                 } else {
709                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
710                                 }
711
712                                 if ($data["description"] != "" && $data["description"] != $data["title"]) {
713                                         // Sanitize the HTML by converting it to BBCode
714                                         $bbcode = HTML::toBBCode($data["description"]);
715                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
716                                 }
717                                 if ($data["type"] == "link") {
718                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
719                                 }
720
721                                 if ($simplehtml != 4) {
722                                         $return .= '</div>';
723                                 }
724                         }
725                 }
726
727                 return trim($data["text"] . ' ' . $return . ' ' . $data["after"]);
728         }
729
730         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
731         {
732                 $data = self::getAttachmentData($Text);
733
734                 if (!$data) {
735                         return $Text;
736                 } elseif ($nolink) {
737                         return $data["text"] . $data["after"];
738                 }
739
740                 $title = htmlentities($data["title"], ENT_QUOTES, 'UTF-8', false);
741                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
742                 if ($plaintext || (($title != "") && strstr($text, $title))) {
743                         $data["title"] = $data["url"];
744                 } elseif (($text != "") && strstr($title, $text)) {
745                         $data["text"] = $data["title"];
746                         $data["title"] = $data["url"];
747                 }
748
749                 if (($data["text"] == "") && ($data["title"] != "") && ($data["url"] == "")) {
750                         return $data["title"] . $data["after"];
751                 }
752
753                 // If the link already is included in the post, don't add it again
754                 if (($data["url"] != "") && strpos($data["text"], $data["url"])) {
755                         return $data["text"] . $data["after"];
756                 }
757
758                 $text = $data["text"];
759
760                 if (($data["url"] != "") && ($data["title"] != "")) {
761                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
762                 } elseif (($data["url"] != "")) {
763                         $text .= "\n[url]" . $data["url"] . "[/url]";
764                 }
765
766                 return $text . "\n" . $data["after"];
767         }
768
769         /**
770          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
771          *
772          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
773          * @param array $match Array with the matching values
774          * @return string reformatted link including HTML codes
775          */
776         private static function convertUrlForOStatusCallback($match)
777         {
778                 $url = $match[1];
779
780                 if (isset($match[2]) && ($match[1] != $match[2])) {
781                         return $match[0];
782                 }
783
784                 $parts = parse_url($url);
785                 if (!isset($parts['scheme'])) {
786                         return $match[0];
787                 }
788
789                 return self::convertUrlForOStatus($url);
790         }
791
792         /**
793          * @brief Converts [url] BBCodes in a format that looks fine on OStatus systems.
794          * @param string $url URL that is about to be reformatted
795          * @return string reformatted link including HTML codes
796          */
797         private static function convertUrlForOStatus($url)
798         {
799                 $parts = parse_url($url);
800                 $scheme = $parts['scheme'] . '://';
801                 $styled_url = str_replace($scheme, '', $url);
802
803                 if (strlen($styled_url) > 30) {
804                         $styled_url = substr($styled_url, 0, 30) . "…";
805                 }
806
807                 $html = '<a href="%s" target="_blank">%s</a>';
808
809                 return sprintf($html, $url, $styled_url);
810         }
811
812         /*
813          * [noparse][i]italic[/i][/noparse] turns into
814          * [noparse][ i ]italic[ /i ][/noparse],
815          * to hide them from parser.
816          */
817         private static function escapeNoparseCallback($match)
818         {
819                 $whole_match = $match[0];
820                 $captured = $match[1];
821                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
822                 $new_str = str_replace($captured, $spacefied, $whole_match);
823                 return $new_str;
824         }
825
826         /*
827          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
828          * now turns back and the [noparse] tags are trimed
829          * returning [i]italic[/i]
830          */
831         private static function unescapeNoparseCallback($match)
832         {
833                 $captured = $match[1];
834                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
835                 return $unspacefied;
836         }
837
838         /**
839          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
840          * occurrences
841          *
842          * @param string $text        Text to search
843          * @param string $name        Tag name
844          * @param int    $occurrences Number of first occurrences to skip
845          * @return boolean|array
846          */
847         public static function getTagPosition($text, $name, $occurrences = 0)
848         {
849                 if ($occurrences < 0) {
850                         $occurrences = 0;
851                 }
852
853                 $start_open = -1;
854                 for ($i = 0; $i <= $occurrences; $i++) {
855                         if ($start_open !== false) {
856                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
857                         }
858                 }
859
860                 if ($start_open === false) {
861                         return false;
862                 }
863
864                 $start_equal = strpos($text, '=', $start_open);
865                 $start_close = strpos($text, ']', $start_open);
866
867                 if ($start_close === false) {
868                         return false;
869                 }
870
871                 $start_close++;
872
873                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
874
875                 if ($end_open === false) {
876                         return false;
877                 }
878
879                 $res = [
880                         'start' => [
881                                 'open' => $start_open,
882                                 'close' => $start_close
883                         ],
884                         'end' => [
885                                 'open' => $end_open,
886                                 'close' => $end_open + strlen('[/' . $name . ']')
887                         ],
888                 ];
889
890                 if ($start_equal !== false) {
891                         $res['start']['equal'] = $start_equal + 1;
892                 }
893
894                 return $res;
895         }
896
897         /**
898          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
899          *
900          * @param type $pattern Preg pattern string
901          * @param type $replace Preg replace string
902          * @param type $name    BBCode tag name
903          * @param type $text    Text to search
904          * @return string
905          */
906         public static function pregReplaceInTag($pattern, $replace, $name, $text)
907         {
908                 $occurrences = 0;
909                 $pos = self::getTagPosition($text, $name, $occurrences);
910                 while ($pos !== false && $occurrences++ < 1000) {
911                         $start = substr($text, 0, $pos['start']['open']);
912                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
913                         $end = substr($text, $pos['end']['close']);
914                         if ($end === false) {
915                                 $end = '';
916                         }
917
918                         $subject = preg_replace($pattern, $replace, $subject);
919                         $text = $start . $subject . $end;
920
921                         $pos = self::getTagPosition($text, $name, $occurrences);
922                 }
923
924                 return $text;
925         }
926
927         private static function extractImagesFromItemBody($body)
928         {
929                 $saved_image = [];
930                 $orig_body = $body;
931                 $new_body = '';
932
933                 $cnt = 0;
934                 $img_start = strpos($orig_body, '[img');
935                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
936                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
937                 while (($img_st_close !== false) && ($img_end !== false)) {
938                         $img_st_close++; // make it point to AFTER the closing bracket
939                         $img_end += $img_start;
940
941                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
942                                 // This is an embedded image
943                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
944                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
945
946                                 $cnt++;
947                         } else {
948                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
949                         }
950
951                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
952
953                         if ($orig_body === false) {
954                                 // in case the body ends on a closing image tag
955                                 $orig_body = '';
956                         }
957
958                         $img_start = strpos($orig_body, '[img');
959                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
960                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
961                 }
962
963                 $new_body = $new_body . $orig_body;
964
965                 return ['body' => $new_body, 'images' => $saved_image];
966         }
967
968         private static function interpolateSavedImagesIntoItemBody($body, array $images)
969         {
970                 $newbody = $body;
971
972                 $cnt = 0;
973                 foreach ($images as $image) {
974                         // We're depending on the property of 'foreach' (specified on the PHP website) that
975                         // it loops over the array starting from the first element and going sequentially
976                         // to the last element
977                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
978                                 '<img src="' . proxy_url($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
979                         $cnt++;
980                 }
981
982                 return $newbody;
983         }
984
985         /**
986          * Processes [share] tags
987          *
988          * Note: Can produce a [bookmark] tag in the output
989          *
990          * @brief Processes [share] tags
991          * @param array    $share      preg_match_callback result array
992          * @param bool|int $simplehtml
993          * @return string
994          */
995         private static function convertShare($share, $simplehtml)
996         {
997                 $attributes = $share[2];
998
999                 $author = "";
1000                 preg_match("/author='(.*?)'/ism", $attributes, $matches);
1001                 if (x($matches, 1)) {
1002                         $author = html_entity_decode($matches[1], ENT_QUOTES, 'UTF-8');
1003                 }
1004
1005                 preg_match('/author="(.*?)"/ism', $attributes, $matches);
1006                 if (x($matches, 1)) {
1007                         $author = $matches[1];
1008                 }
1009
1010                 $profile = "";
1011                 preg_match("/profile='(.*?)'/ism", $attributes, $matches);
1012                 if (x($matches, 1)) {
1013                         $profile = $matches[1];
1014                 }
1015
1016                 preg_match('/profile="(.*?)"/ism', $attributes, $matches);
1017                 if (x($matches, 1)) {
1018                         $profile = $matches[1];
1019                 }
1020
1021                 $avatar = "";
1022                 preg_match("/avatar='(.*?)'/ism", $attributes, $matches);
1023                 if (x($matches, 1)) {
1024                         $avatar = $matches[1];
1025                 }
1026
1027                 preg_match('/avatar="(.*?)"/ism', $attributes, $matches);
1028                 if (x($matches, 1)) {
1029                         $avatar = $matches[1];
1030                 }
1031
1032                 $link = "";
1033                 preg_match("/link='(.*?)'/ism", $attributes, $matches);
1034                 if (x($matches, 1)) {
1035                         $link = $matches[1];
1036                 }
1037
1038                 preg_match('/link="(.*?)"/ism', $attributes, $matches);
1039                 if (x($matches, 1)) {
1040                         $link = $matches[1];
1041                 }
1042
1043                 $posted = "";
1044
1045                 preg_match("/posted='(.*?)'/ism", $attributes, $matches);
1046                 if (x($matches, 1)) {
1047                         $posted = $matches[1];
1048                 }
1049
1050                 preg_match('/posted="(.*?)"/ism', $attributes, $matches);
1051                 if (x($matches, 1)) {
1052                         $posted = $matches[1];
1053                 }
1054
1055                 // We only call this so that a previously unknown contact can be added.
1056                 // This is important for the function "Model\Contact::getDetailsByURL()".
1057                 // This function then can fetch an entry from the contact table.
1058                 Contact::getIdForURL($profile, 0, true);
1059
1060                 $data = Contact::getDetailsByURL($profile);
1061
1062                 if (x($data, "name") && x($data, "addr")) {
1063                         $userid_compact = $data["name"] . " (" . $data["addr"] . ")";
1064                 } else {
1065                         $userid_compact = Protocol::getAddrFromProfileUrl($profile, $author);
1066                 }
1067
1068                 if (x($data, "addr")) {
1069                         $userid = $data["addr"];
1070                 } else {
1071                         $userid = Protocol::formatMention($profile, $author);
1072                 }
1073
1074                 if (x($data, "name")) {
1075                         $author = $data["name"];
1076                 }
1077
1078                 if (x($data, "micro")) {
1079                         $avatar = $data["micro"];
1080                 }
1081
1082                 $preshare = trim($share[1]);
1083                 if ($preshare != "") {
1084                         $preshare .= "<br />";
1085                 }
1086
1087                 switch ($simplehtml) {
1088                         case 1:
1089                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' <a href="' . $profile . '">' . $userid . "</a>: <br />»" . $share[3] . "«";
1090                                 break;
1091                         case 2:
1092                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1093                                 break;
1094                         case 3: // Diaspora
1095                                 $headline = '<b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . $userid . ':</b><br />';
1096
1097                                 $text = trim($share[1]);
1098
1099                                 if ($text != "") {
1100                                         $text .= "<hr />";
1101                                 }
1102
1103                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0) {
1104                                         $text .= '<br /><a href="' . $link . '">' . $link . '</a>';
1105                                 } else {
1106                                         $text .= $headline . '<blockquote>' . trim($share[3]) . "</blockquote><br />";
1107
1108                                         if ($link != "") {
1109                                                 $text .= '<br /><a href="' . $link . '">[l]</a>';
1110                                         }
1111                                 }
1112
1113                                 break;
1114                         case 4:
1115                                 $headline = '<br /><b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8');
1116                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $link, $userid, $posted);
1117                                 $headline .= ":</b><br />";
1118
1119                                 $text = trim($share[1]);
1120
1121                                 if ($text != "") {
1122                                         $text .= "<hr />";
1123                                 }
1124
1125                                 $text .= $headline . '<blockquote class="shared_content">' . trim($share[3]) . "</blockquote><br />";
1126
1127                                 break;
1128                         case 5:
1129                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1130                                 break;
1131                         case 6: // app.net
1132                                 $text = $preshare . "&gt;&gt; @" . $userid_compact . ": <br />" . $share[3];
1133                                 break;
1134                         case 7: // statusnet/GNU Social
1135                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . " @" . $userid_compact . ": " . $share[3];
1136                                 break;
1137                         case 8: // twitter
1138                                 $text = $preshare . "RT @" . $userid_compact . ": " . $share[3];
1139                                 break;
1140                         case 9: // Google+/Facebook
1141                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1142
1143                                 if ($link != "") {
1144                                         $text .= "<br /><br />" . $link;
1145                                 }
1146                                 break;
1147                         default:
1148                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1149                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($link)) {
1150                                         try {
1151                                                 $oembed = OEmbed::getHTML($link, $preshare);
1152                                         } catch (Exception $e) {
1153                                                 $oembed = sprintf('[bookmark=%s]%s[/bookmark]', $link, $preshare);
1154                                         }
1155
1156                                         $text = $preshare . $oembed;
1157                                 } else {
1158                                         $text = trim($share[1]) . "\n";
1159
1160                                         $avatar = proxy_url($avatar, false, PROXY_SIZE_THUMB);
1161
1162                                         $tpl = get_markup_template('shared_content.tpl');
1163                                         $text .= replace_macros($tpl, [
1164                                                 '$profile' => $profile,
1165                                                 '$avatar' => $avatar,
1166                                                 '$author' => $author,
1167                                                 '$link' => $link,
1168                                                 '$posted' => $posted,
1169                                                 '$content' => trim($share[3])
1170                                         ]);
1171                                 }
1172                                 break;
1173                 }
1174
1175                 return $text;
1176         }
1177
1178         private static function removePictureLinksCallback($match)
1179         {
1180                 $text = Cache::get($match[1]);
1181
1182                 if (is_null($text)) {
1183                         $a = self::getApp();
1184
1185                         $stamp1 = microtime(true);
1186
1187                         $ch = @curl_init($match[1]);
1188                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1189                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1190                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1191                         @curl_exec($ch);
1192                         $curl_info = @curl_getinfo($ch);
1193
1194                         $a->save_timestamp($stamp1, "network");
1195
1196                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1197                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1198                         } else {
1199                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1200
1201                                 // if its not a picture then look if its a page that contains a picture link
1202                                 $body = Network::fetchUrl($match[1]);
1203
1204                                 $doc = new DOMDocument();
1205                                 @$doc->loadHTML($body);
1206                                 $xpath = new DOMXPath($doc);
1207                                 $list = $xpath->query("//meta[@name]");
1208                                 foreach ($list as $node) {
1209                                         $attr = [];
1210
1211                                         if ($node->attributes->length) {
1212                                                 foreach ($node->attributes as $attribute) {
1213                                                         $attr[$attribute->name] = $attribute->value;
1214                                                 }
1215                                         }
1216
1217                                         if (strtolower($attr["name"]) == "twitter:image") {
1218                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1219                                         }
1220                                 }
1221                         }
1222                         Cache::set($match[1], $text);
1223                 }
1224
1225                 return $text;
1226         }
1227
1228         private static function expandLinksCallback($match)
1229         {
1230                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1231                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1232                 } else {
1233                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1234                 }
1235         }
1236
1237         private static function cleanPictureLinksCallback($match)
1238         {
1239                 $text = Cache::get($match[1]);
1240
1241                 if (is_null($text)) {
1242                         $a = self::getApp();
1243
1244                         $stamp1 = microtime(true);
1245
1246                         $ch = @curl_init($match[1]);
1247                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1248                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1249                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1250                         @curl_exec($ch);
1251                         $curl_info = @curl_getinfo($ch);
1252
1253                         $a->save_timestamp($stamp1, "network");
1254
1255                         // if its a link to a picture then embed this picture
1256                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1257                                 $text = "[img]" . $match[1] . "[/img]";
1258                         } else {
1259                                 $text = "[img]" . $match[2] . "[/img]";
1260
1261                                 // if its not a picture then look if its a page that contains a picture link
1262                                 $body = Network::fetchUrl($match[1]);
1263
1264                                 $doc = new DOMDocument();
1265                                 @$doc->loadHTML($body);
1266                                 $xpath = new DOMXPath($doc);
1267                                 $list = $xpath->query("//meta[@name]");
1268                                 foreach ($list as $node) {
1269                                         $attr = [];
1270                                         if ($node->attributes->length) {
1271                                                 foreach ($node->attributes as $attribute) {
1272                                                         $attr[$attribute->name] = $attribute->value;
1273                                                 }
1274                                         }
1275
1276                                         if (strtolower($attr["name"]) == "twitter:image") {
1277                                                 $text = "[img]" . $attr["content"] . "[/img]";
1278                                         }
1279                                 }
1280                         }
1281                         Cache::set($match[1], $text);
1282                 }
1283
1284                 return $text;
1285         }
1286
1287         public static function cleanPictureLinks($text)
1288         {
1289                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1290                 return $return;
1291         }
1292
1293         private static function textHighlightCallback($match)
1294         {
1295                 if (in_array(strtolower($match[1]),
1296                                 ['php', 'css', 'mysql', 'sql', 'abap', 'diff', 'html', 'perl', 'ruby',
1297                                 'vbscript', 'avrc', 'dtd', 'java', 'xml', 'cpp', 'python', 'javascript', 'js', 'sh', 'bash'])
1298                 ) {
1299                         return text_highlight($match[2], strtolower($match[1]));
1300                 }
1301                 return $match[0];
1302         }
1303
1304         /**
1305          * @brief Converts a BBCode message to HTML message
1306          *
1307          * BBcode 2 HTML was written by WAY2WEB.net
1308          * extended to work with Mistpark/Friendica - Mike Macgirvin
1309          *
1310          * Simple HTML values meaning:
1311          * - 0: Friendica display
1312          * - 1: Unused
1313          * - 2: Used for Facebook, Google+, Windows Phone push, Friendica API
1314          * - 3: Used before converting to Markdown in bb2diaspora.php
1315          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1316          * - 5: Unused
1317          * - 6: Used for Appnet
1318          * - 7: Used for dfrn, OStatus
1319          * - 8: Used for WP backlink text setting
1320          *
1321          * @param string $text
1322          * @param bool   $try_oembed
1323          * @param int    $simple_html
1324          * @param bool   $for_plaintext
1325          * @return string
1326          */
1327         public static function convert($text, $try_oembed = true, $simple_html = false, $for_plaintext = false)
1328         {
1329                 $a = self::getApp();
1330
1331                 /*
1332                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1333                  *
1334                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1335                  * $match[1] = $url
1336                  * $match[2] = $title or absent
1337                  */
1338                 $try_oembed_callback = function ($match)
1339                 {
1340                         $url = $match[1];
1341                         $title = defaults($match, 2, null);
1342
1343                         try {
1344                                 $return = OEmbed::getHTML($url, $title);
1345                         } catch (Exception $ex) {
1346                                 $return = $match[0];
1347                         }
1348
1349                         return $return;
1350                 };
1351
1352                 // Hide all [noparse] contained bbtags by spacefying them
1353                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1354
1355                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1356                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1357                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1358
1359                 // Remove the abstract element. It is a non visible element.
1360                 $text = self::stripAbstract($text);
1361
1362                 // Move all spaces out of the tags
1363                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1364                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1365
1366                 // Extract the private images which use data urls since preg has issues with
1367                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1368                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1369
1370                 $extracted = self::extractImagesFromItemBody($text);
1371                 $text = $extracted['body'];
1372                 $saved_image = $extracted['images'];
1373
1374                 // If we find any event code, turn it into an event.
1375                 // After we're finished processing the bbcode we'll
1376                 // replace all of the event code with a reformatted version.
1377
1378                 $ev = Event::fromBBCode($text);
1379
1380                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1381                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1382
1383                 $text = str_replace("<", "&lt;", $text);
1384                 $text = str_replace(">", "&gt;", $text);
1385
1386                 // remove some newlines before the general conversion
1387                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1388                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1389
1390                 $text = preg_replace("/\n\[code\]/ism", "[code]", $text);
1391                 $text = preg_replace("/\[\/code\]\n/ism", "[/code]", $text);
1392
1393                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1394                 if (!$try_oembed) {
1395                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1396                 }
1397
1398                 // Check for [code] text here, before the linefeeds are messed with.
1399                 // The highlighter will unescape and re-escape the content.
1400                 if (strpos($text, '[code=') !== false) {
1401                         $text = preg_replace_callback("/\[code=(.*?)\](.*?)\[\/code\]/ism", 'self::textHighlightCallback', $text);
1402                 }
1403                 // Convert new line chars to html <br /> tags
1404
1405                 // nlbr seems to be hopelessly messed up
1406                 //      $Text = nl2br($Text);
1407
1408                 // We'll emulate it.
1409
1410                 $text = trim($text);
1411                 $text = str_replace("\r\n", "\n", $text);
1412
1413                 // removing multiplicated newlines
1414                 if (Config::get("system", "remove_multiplicated_lines")) {
1415                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1416                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1417                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1418                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1419                         do {
1420                                 $oldtext = $text;
1421                                 $text = str_replace($search, $replace, $text);
1422                         } while ($oldtext != $text);
1423                 }
1424
1425                 // Set up the parameters for a URL search string
1426                 $URLSearchString = "^\[\]";
1427                 // Set up the parameters for a MAIL search string
1428                 $MAILSearchString = $URLSearchString;
1429
1430                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1431                 if (!$for_plaintext) {
1432                         // Autolink feature (thanks to http://code.seebz.net/p/autolink-php/)
1433                         // Currently disabled, since the function is too greedy
1434                         // $autolink_regex = "`([^\]\=\"']|^)(https?\://[^\s<]+[^\s<\.\)])`ism";
1435                         $autolink_regex = "/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism";
1436                         $text = preg_replace($autolink_regex, '$1[url]$2[/url]', $text);
1437                         if ($simple_html == 7) {
1438                                 $text = preg_replace_callback("/\[url\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1439                                 $text = preg_replace_callback("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1440                         }
1441                 } else {
1442                         $text = preg_replace("(\[url\]([$URLSearchString]*)\[\/url\])ism", " $1 ", $text);
1443                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1444                 }
1445
1446
1447                 // Handle attached links or videos
1448                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1449
1450                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1451
1452                 // Remove all hashtag addresses
1453                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7])) {
1454                         $text = preg_replace("/([#@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $text);
1455                 } elseif ($simple_html == 3) {
1456                         // The ! is converted to @ since Diaspora only understands the @
1457                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1458                                 '@<a href="$2">$3</a>',
1459                                 $text);
1460                 } elseif ($simple_html == 7) {
1461                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1462                                 '$1<span class="vcard"><a href="$2" class="url" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1463                                 $text);
1464                 } elseif (!$simple_html) {
1465                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1466                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1467                                 $text);
1468                 }
1469
1470                 // Bookmarks in red - will be converted to bookmarks in friendica
1471                 $text = preg_replace("/#\^\[url\]([$URLSearchString]*)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1472                 $text = preg_replace("/#\^\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1473                 $text = preg_replace("/#\[url\=[$URLSearchString]*\]\^\[\/url\]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i",
1474                                         "[bookmark=$1]$2[/bookmark]", $text);
1475
1476                 if (in_array($simple_html, [2, 6, 7, 8, 9])) {
1477                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1478                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1479                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1480                 }
1481
1482                 if ($simple_html == 5) {
1483                         $text = preg_replace("/[^#@!]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1484                 }
1485
1486                 // Perform URL Search
1487                 if ($try_oembed) {
1488                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1489                 }
1490
1491                 if ($simple_html == 5) {
1492                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1493                 } else {
1494                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1495                 }
1496
1497                 // Handle Diaspora posts
1498                 $text = preg_replace_callback(
1499                         "&\[url=/posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1500                         function ($match) {
1501                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1502                         }, $text
1503                 );
1504
1505                 // Server independent link to posts and comments
1506                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1507                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1508                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1509
1510                 $text = preg_replace("/([#])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1511                                         '$1<a href="' . System::baseUrl() . '/search?tag=$3" class="tag" title="$3">$3</a>', $text);
1512
1513                 $text = preg_replace("/\[url\=([$URLSearchString]*)\]#(.*?)\[\/url\]/ism",
1514                                         '#<a href="' . System::baseUrl() . '/search?tag=$2" class="tag" title="$2">$2</a>', $text);
1515
1516                 $text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1517                 $text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1518                 //$Text = preg_replace("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $Text);
1519
1520                 // Red compatibility, though the link can't be authenticated on Friendica
1521                 $text = preg_replace("/\[zrl\=([$URLSearchString]*)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1522
1523
1524                 // we may need to restrict this further if it picks up too many strays
1525                 // link acct:user@host to a webfinger profile redirector
1526
1527                 $text = preg_replace('/acct:([^@]+)@((?!\-)(?:[a-zA-Z\d\-]{0,62}[a-zA-Z\d]\.){1,126}(?!\d+)[a-zA-Z\d]{1,63})/', '<a href="' . System::baseUrl() . '/acctlink?addr=$1@$2" target="extlink">acct:$1@$2</a>', $text);
1528
1529                 // Perform MAIL Search
1530                 $text = preg_replace("/\[mail\]([$MAILSearchString]*)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1531                 $text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1532
1533                 // leave open the posibility of [map=something]
1534                 // this is replaced in prepare_body() which has knowledge of the item location
1535
1536                 if (strpos($text, '[/map]') !== false) {
1537                         $text = preg_replace_callback(
1538                                 "/\[map\](.*?)\[\/map\]/ism",
1539                                 function ($match) use ($simple_html) {
1540                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1541                                 },
1542                                 $text
1543                         );
1544                 }
1545                 if (strpos($text, '[map=') !== false) {
1546                         $text = preg_replace_callback(
1547                                 "/\[map=(.*?)\]/ism",
1548                                 function ($match) use ($simple_html) {
1549                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1550                                 },
1551                                 $text
1552                         );
1553                 }
1554                 if (strpos($text, '[map]') !== false) {
1555                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1556                 }
1557
1558                 // Check for headers
1559                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1560                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1561                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1562                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1563                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1564                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1565
1566                 // Check for paragraph
1567                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1568
1569                 // Check for bold text
1570                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1571
1572                 // Check for Italics text
1573                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1574
1575                 // Check for Underline text
1576                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1577
1578                 // Check for strike-through text
1579                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<strike>$1</strike>', $text);
1580
1581                 // Check for over-line text
1582                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1583
1584                 // Check for colored text
1585                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1586
1587                 // Check for sized text
1588                 // [size=50] --> font-size: 50px (with the unit).
1589                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1590                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1591
1592                 // Check for centered text
1593                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1594
1595                 // Check for list text
1596                 $text = str_replace("[*]", "<li>", $text);
1597
1598                 // Check for style sheet commands
1599                 $text = preg_replace_callback(
1600                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1601                         function ($match) {
1602                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1603                         },
1604                         $text
1605                 );
1606
1607                 // Check for CSS classes
1608                 $text = preg_replace_callback(
1609                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1610                         function ($match) {
1611                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1612                         },
1613                         $text
1614                 );
1615
1616                 // handle nested lists
1617                 $endlessloop = 0;
1618
1619                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1620                            ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1621                            ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1622                            ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1623                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1624                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1625                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1626                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1627                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1628                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1629                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1630                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1631                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1632                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1633                 }
1634
1635                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1636                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1637                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1638                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1639
1640                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1641                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1642
1643                 $text = str_replace('[hr]', '<hr />', $text);
1644
1645                 // This is actually executed in prepare_body()
1646
1647                 $text = str_replace('[nosmile]', '', $text);
1648
1649                 // Check for font change text
1650                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1651
1652                 // Declare the format for [code] layout
1653
1654                 $CodeLayout = '<code>$1</code>';
1655                 // Check for [code] text
1656                 $text = preg_replace("/\[code\](.*?)\[\/code\]/ism", "$CodeLayout", $text);
1657
1658                 // Declare the format for [spoiler] layout
1659                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
1660
1661                 // Check for [spoiler] text
1662                 // handle nested quotes
1663                 $endlessloop = 0;
1664                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1665                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", "$SpoilerLayout", $text);
1666                 }
1667
1668                 // Check for [spoiler=Author] text
1669
1670                 $t_wrote = L10n::t('$1 wrote:');
1671
1672                 // handle nested quotes
1673                 $endlessloop = 0;
1674                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1675                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1676                                                  "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1677                                                  $text);
1678                 }
1679
1680                 // Declare the format for [quote] layout
1681                 $QuoteLayout = '<blockquote>$1</blockquote>';
1682
1683                 // Check for [quote] text
1684                 // handle nested quotes
1685                 $endlessloop = 0;
1686                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1687                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1688                 }
1689
1690                 // Check for [quote=Author] text
1691
1692                 $t_wrote = L10n::t('$1 wrote:');
1693
1694                 // handle nested quotes
1695                 $endlessloop = 0;
1696                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1697                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1698                                                  "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1699                                                  $text);
1700                 }
1701
1702
1703                 // [img=widthxheight]image source[/img]
1704                 $text = preg_replace_callback(
1705                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1706                         function ($matches) {
1707                                 if (strpos($matches[3], "data:image/") === 0) {
1708                                         return $matches[0];
1709                                 }
1710
1711                                 $matches[3] = proxy_url($matches[3]);
1712                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1713                         },
1714                         $text
1715                 );
1716
1717                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1718                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1719
1720                 // Images
1721                 // [img]pathtoimage[/img]
1722                 $text = preg_replace_callback(
1723                         "/\[img\](.*?)\[\/img\]/ism",
1724                         function ($matches) {
1725                                 if (strpos($matches[1], "data:image/") === 0) {
1726                                         return $matches[0];
1727                                 }
1728
1729                                 $matches[1] = proxy_url($matches[1]);
1730                                 return "[img]" . $matches[1] . "[/img]";
1731                         },
1732                         $text
1733                 );
1734
1735                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1736                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1737
1738                 // Shared content
1739                 $text = preg_replace_callback("/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
1740                         function ($match) use ($simple_html) {
1741                                 return self::convertShare($match, $simple_html);
1742                         }, $text);
1743
1744                 $text = preg_replace("/\[crypt\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . L10n::t('Encrypted content') . '" /><br />', $text);
1745                 $text = preg_replace("/\[crypt(.*?)\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . '$1' . ' ' . L10n::t('Encrypted content') . '" /><br />', $text);
1746                 //$Text = preg_replace("/\[crypt=(.*?)\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . '$1' . ' ' . L10n::t('Encrypted content') . '" /><br />', $Text);
1747
1748                 // Try to Oembed
1749                 if ($try_oembed) {
1750                         $text = preg_replace("/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4))\[\/video\]/ism", '<video src="$1" controls="controls" width="' . $a->videowidth . '" height="' . $a->videoheight . '" loop="true"><a href="$1">$1</a></video>', $text);
1751                         $text = preg_replace("/\[audio\](.*?\.(ogg|ogv|oga|ogm|webm|mp4|mp3))\[\/audio\]/ism", '<audio src="$1" controls="controls"><a href="$1">$1</a></audio>', $text);
1752
1753                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1754                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1755                 } else {
1756                         $text = preg_replace("/\[video\](.*?)\[\/video\]/",
1757                                                 '<a href="$1" target="_blank">$1</a>', $text);
1758                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/",
1759                                                 '<a href="$1" target="_blank">$1</a>', $text);
1760                 }
1761
1762                 // html5 video and audio
1763
1764
1765                 if ($try_oembed) {
1766                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1767                 } else {
1768                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1769                 }
1770
1771                 // Youtube extensions
1772                 if ($try_oembed) {
1773                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1774                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1775                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1776                 }
1777
1778                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1779                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1780                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1781
1782                 if ($try_oembed) {
1783                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism", '<iframe width="' . $a->videowidth . '" height="' . $a->videoheight . '" src="https://www.youtube.com/embed/$1" frameborder="0" ></iframe>', $text);
1784                 } else {
1785                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1786                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1787                 }
1788
1789                 if ($try_oembed) {
1790                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1791                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1792                 }
1793
1794                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1795                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1796
1797                 if ($try_oembed) {
1798                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism", '<iframe width="' . $a->videowidth . '" height="' . $a->videoheight . '" src="https://player.vimeo.com/video/$1" frameborder="0" ></iframe>', $text);
1799                 } else {
1800                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1801                                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1802                 }
1803
1804                 // oembed tag
1805                 $text = OEmbed::BBCode2HTML($text);
1806
1807                 // Avoid triple linefeeds through oembed
1808                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1809
1810                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1811                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1812                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1813                 // start which is always required). Allow desc with a missing summary for compatibility.
1814
1815                 if ((x($ev, 'desc') || x($ev, 'summary')) && x($ev, 'start')) {
1816                         $sub = Event::getHTML($ev, $simple_html);
1817
1818                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1819                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1820                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1821                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1822                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1823                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1824                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1825                 }
1826
1827                 // Replace non graphical smilies for external posts
1828                 if ($simple_html) {
1829                         $text = Smilies::replace($text, false, true);
1830                 }
1831
1832                 // Replace inline code blocks
1833                 $text = preg_replace_callback("|(?!<br[^>]*>)<code>([^<]*)</code>(?!<br[^>]*>)|ism",
1834                         function ($match) use ($simple_html) {
1835                                 $return = '<key>' . $match[1] . '</key>';
1836                                 // Use <code> for Diaspora inline code blocks
1837                                 if ($simple_html === 3) {
1838                                         $return = '<code>' . $match[1] . '</code>';
1839                                 }
1840                                 return $return;
1841                         }
1842                 , $text);
1843
1844                 // Unhide all [noparse] contained bbtags unspacefying them
1845                 // and triming the [noparse] tag.
1846
1847                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1848                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1849                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1850
1851
1852                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1853                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1854                 $text = preg_replace('/\&quot\;/', '"', $text);
1855
1856                 // fix any escaped ampersands that may have been converted into links
1857                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1858
1859                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1860                 $allowed_src_protocols = ['http', 'redir', 'cid'];
1861                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1862                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1863
1864                 // sanitize href attributes (only whitelisted protocols URLs)
1865                 // default value for backward compatibility
1866                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', ['ftp', 'mailto', 'gopher', 'cid']);
1867
1868                 // Always allowed protocol even if config isn't set or not including it
1869                 $allowed_link_protocols[] = 'http';
1870                 $allowed_link_protocols[] = 'redir/';
1871
1872                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1873                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1874
1875                 if ($saved_image) {
1876                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1877                 }
1878
1879                 // Clean up the HTML by loading and saving the HTML with the DOM.
1880                 // Bad structured html can break a whole page.
1881                 // For performance reasons do it only with ativated item cache or at export.
1882                 if (!$try_oembed || (get_itemcachepath() != "")) {
1883                         $doc = new DOMDocument();
1884                         $doc->preserveWhiteSpace = false;
1885
1886                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1887
1888                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1889                         $encoding = '<?xml encoding="UTF-8">';
1890                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1891                         $doc->encoding = 'UTF-8';
1892                         $text = $doc->saveHTML();
1893                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1894
1895                         $text = str_replace('<br></li>', '</li>', $text);
1896
1897                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1898                 }
1899
1900                 // Clean up some useless linebreaks in lists
1901                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1902                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1903                 //$Text = str_replace('</li><br />', '</li>', $Text);
1904                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1905                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1906
1907                 Addon::callHooks('bbcode', $text);
1908
1909                 return trim($text);
1910         }
1911
1912         /**
1913          * @brief Strips the "abstract" tag from the provided text
1914          *
1915          * @param string $text The text with BBCode
1916          * @return string The same text - but without "abstract" element
1917          */
1918         public static function stripAbstract($text)
1919         {
1920                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1921                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1922
1923                 return $text;
1924         }
1925
1926         /**
1927          * @brief Returns the value of the "abstract" element
1928          *
1929          * @param string $text The text that maybe contains the element
1930          * @param string $addon The addon for which the abstract is meant for
1931          * @return string The abstract
1932          */
1933         private static function getAbstract($text, $addon = "")
1934         {
1935                 $abstract = "";
1936                 $abstracts = [];
1937                 $addon = strtolower($addon);
1938
1939                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1940                         foreach ($results AS $result) {
1941                                 $abstracts[strtolower($result[1])] = $result[2];
1942                         }
1943                 }
1944
1945                 if (isset($abstracts[$addon])) {
1946                         $abstract = $abstracts[$addon];
1947                 }
1948
1949                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1950                         $abstract = $result[1];
1951                 }
1952
1953                 return $abstract;
1954         }
1955
1956         /**
1957          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1958          *
1959          * @param array $match Matching values for the callback
1960          * @return string Replaced mention
1961          */
1962         private static function bbCodeMention2DiasporaCallback($match)
1963         {
1964                 $contact = Contact::getDetailsByURL($match[3]);
1965
1966                 if (empty($contact['addr'])) {
1967                         $contact = Probe::uri($match[3]);
1968                 }
1969
1970                 if (empty($contact['addr'])) {
1971                         return $match[0];
1972                 }
1973
1974                 $mention = '@{' . $match[2] . '; ' . $contact['addr'] . '}';
1975                 return $mention;
1976         }
1977
1978         /**
1979          * @brief Converts a BBCode text into Markdown
1980          *
1981          * This function converts a BBCode item body to be sent to Markdown-enabled
1982          * systems like Diaspora and Libertree
1983          *
1984          * @param string $text
1985          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1986          * @return string
1987          */
1988         public static function toMarkdown($text, $for_diaspora = true)
1989         {
1990                 $a = self::getApp();
1991
1992                 $original_text = $text;
1993
1994                 // Since Diaspora is creating a summary for links, this function removes them before posting
1995                 if ($for_diaspora) {
1996                         $text = self::removeShareInformation($text);
1997                 }
1998
1999                 /**
2000                  * Transform #tags, strip off the [url] and replace spaces with underscore
2001                  */
2002                 $url_search_string = "^\[\]";
2003                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2004                         function ($matches) {
2005                                 return '#' . str_replace(' ', '_', $matches[2]);
2006                         },
2007                         $text
2008                 );
2009
2010                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2011                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2012
2013                 // Extracting multi-line code blocks before the whitespace processing/code highlighter in self::convert()
2014                 $codeblocks = [];
2015
2016                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#is",
2017                         function ($matches) use (&$codeblocks) {
2018                                 $return = $matches[0];
2019                                 if (strpos($matches[2], "\n") !== false) {
2020                                         $return = '#codeblock-' . count($codeblocks) . '#';
2021
2022                                         $prefix = '````' . $matches[1] . PHP_EOL;
2023                                         $codeblocks[] = $prefix . trim($matches[2]) . PHP_EOL . '````';
2024                                 }
2025                                 return $return;
2026                         },
2027                         $text
2028                 );
2029
2030                 // Convert it to HTML - don't try oembed
2031                 if ($for_diaspora) {
2032                         $text = self::convert($text, false, 3);
2033
2034                         // Add all tags that maybe were removed
2035                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2036                                 $tagline = "";
2037                                 foreach ($tags[2] as $tag) {
2038                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2039                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2040                                                 $tagline .= '#' . $tag . ' ';
2041                                         }
2042                                 }
2043                                 $text = $text . " " . $tagline;
2044                         }
2045                 } else {
2046                         $text = self::convert($text, false, 4);
2047                 }
2048
2049                 // mask some special HTML chars from conversation to markdown
2050                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
2051
2052                 // If a link is followed by a quote then there should be a newline before it
2053                 // Maybe we should make this newline at every time before a quote.
2054                 $text = str_replace(["</a><blockquote>"], ["</a><br><blockquote>"], $text);
2055
2056                 $stamp1 = microtime(true);
2057
2058                 // Now convert HTML to Markdown
2059                 $converter = new HtmlConverter();
2060                 $text = $converter->convert($text);
2061
2062                 // unmask the special chars back to HTML
2063                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
2064
2065                 $a->save_timestamp($stamp1, "parser");
2066
2067                 // Libertree has a problem with escaped hashtags.
2068                 $text = str_replace(['\#'], ['#'], $text);
2069
2070                 // Remove any leading or trailing whitespace, as this will mess up
2071                 // the Diaspora signature verification and cause the item to disappear
2072                 $text = trim($text);
2073
2074                 if ($for_diaspora) {
2075                         $url_search_string = "^\[\]";
2076                         $text = preg_replace_callback(
2077                                 "/([@]\[(.*?)\])\(([$url_search_string]*?)\)/ism",
2078                                 ['self', 'bbCodeMention2DiasporaCallback'],
2079                                 $text
2080                         );
2081                 }
2082
2083                 // Restore code blocks
2084                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
2085                         function ($matches) use ($codeblocks) {
2086                                 $return = '';
2087                                 if (isset($codeblocks[intval($matches[1])])) {
2088                                         $return = $codeblocks[$matches[1]];
2089                                 }
2090                                 return $return;
2091                         },
2092                         $text
2093                 );
2094
2095                 Addon::callHooks('bb2diaspora', $text);
2096
2097                 return $text;
2098         }
2099 }