]> git.mxchange.org Git - friendica.git/blob - src/Content/Text/BBCode.php
c59d2433042a1d5780011a642205936f5ce349d3
[friendica.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\Network\Probe;
24 use Friendica\Object\Image;
25 use Friendica\Util\Map;
26 use Friendica\Util\Network;
27 use Friendica\Util\ParseUrl;
28 use League\HTMLToMarkdown\HtmlConverter;
29
30 require_once "include/event.php";
31 require_once "include/html2plain.php";
32 require_once "mod/proxy.php";
33
34 class BBCode extends BaseObject
35 {
36         /**
37          * @brief Fetches attachment data that were generated the old way
38          *
39          * @param string $body Message body
40          * @return array
41          * 'type' -> Message type ("link", "video", "photo")
42          * 'text' -> Text before the shared message
43          * 'after' -> Text after the shared message
44          * 'image' -> Preview image of the message
45          * 'url' -> Url to the attached message
46          * 'title' -> Title of the attachment
47          * 'description' -> Description of the attachment
48          */
49         private static function getOldAttachmentData($body)
50         {
51                 $post = [];
52
53                 // Simplify image codes
54                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
55
56                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
57                         foreach ($attached as $data) {
58                                 if (!in_array($data[1], ["type-link", "type-video", "type-photo"])) {
59                                         continue;
60                                 }
61
62                                 $post["type"] = substr($data[1], 5);
63
64                                 $pos = strpos($body, $data[0]);
65                                 if ($pos > 0) {
66                                         $post["text"] = trim(substr($body, 0, $pos));
67                                         $post["after"] = trim(substr($body, $pos + strlen($data[0])));
68                                 } else {
69                                         $post["text"] = trim(str_replace($data[0], "", $body));
70                                 }
71
72                                 $attacheddata = $data[2];
73
74                                 $URLSearchString = "^\[\]";
75
76                                 if (preg_match("/\[img\]([$URLSearchString]*)\[\/img\]/ism", $attacheddata, $matches)) {
77
78                                         $picturedata = Image::getInfoFromURL($matches[1]);
79
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                                 if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
88                                         $post["url"] = $matches[1];
89                                         $post["title"] = $matches[2];
90                                 }
91                                 if (($post["url"] == "") && (in_array($post["type"], ["link", "video"]))
92                                         && preg_match("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
93                                         $post["url"] = $matches[1];
94                                 }
95
96                                 // Search for description
97                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
98                                         $post["description"] = $matches[1];
99                                 }
100                         }
101                 }
102                 return $post;
103         }
104
105         /**
106          * @brief Fetches attachment data that were generated with the "attachment" element
107          *
108          * @param string $body Message body
109          * @return array
110          * 'type' -> Message type ("link", "video", "photo")
111          * 'text' -> Text before the shared message
112          * 'after' -> Text after the shared message
113          * 'image' -> Preview image of the message
114          * 'url' -> Url to the attached message
115          * 'title' -> Title of the attachment
116          * 'description' -> Description of the attachment
117          */
118         public static function getAttachmentData($body)
119         {
120                 $data = [];
121
122                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
123                         return self::getOldAttachmentData($body);
124                 }
125
126                 $attributes = $match[2];
127
128                 $data["text"] = trim($match[1]);
129
130                 $type = "";
131                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
132                 if (x($matches, 1)) {
133                         $type = strtolower($matches[1]);
134                 }
135
136                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
137                 if (x($matches, 1)) {
138                         $type = strtolower($matches[1]);
139                 }
140
141                 if ($type == "") {
142                         return [];
143                 }
144
145                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
146                         return [];
147                 }
148
149                 if ($type != "") {
150                         $data["type"] = $type;
151                 }
152
153                 $url = "";
154                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
155                 if (x($matches, 1)) {
156                         $url = $matches[1];
157                 }
158
159                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
160                 if (x($matches, 1)) {
161                         $url = $matches[1];
162                 }
163
164                 if ($url != "") {
165                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
166                 }
167
168                 $title = "";
169                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
170                 if (x($matches, 1)) {
171                         $title = $matches[1];
172                 }
173
174                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
175                 if (x($matches, 1)) {
176                         $title = $matches[1];
177                 }
178
179                 if ($title != "") {
180                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
181                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
182                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
183                         $data["title"] = $title;
184                 }
185
186                 $image = "";
187                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
188                 if (x($matches, 1)) {
189                         $image = $matches[1];
190                 }
191
192                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
193                 if (x($matches, 1)) {
194                         $image = $matches[1];
195                 }
196
197                 if ($image != "") {
198                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
199                 }
200
201                 $preview = "";
202                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
203                 if (x($matches, 1)) {
204                         $preview = $matches[1];
205                 }
206
207                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
208                 if (x($matches, 1)) {
209                         $preview = $matches[1];
210                 }
211
212                 if ($preview != "") {
213                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
214                 }
215
216                 $data["description"] = trim($match[3]);
217
218                 $data["after"] = trim($match[4]);
219
220                 return $data;
221         }
222
223         public static function getAttachedData($body, $item = [])
224         {
225                 /*
226                 - text:
227                 - type: link, video, photo
228                 - title:
229                 - url:
230                 - image:
231                 - description:
232                 - (thumbnail)
233                 */
234
235                 $has_title = !empty($item['title']);
236                 $plink = (!empty($item['plink']) ? $item['plink'] : '');
237                 $post = self::getAttachmentData($body);
238
239                 // if nothing is found, it maybe having an image.
240                 if (!isset($post["type"])) {
241                         // Simplify image codes
242                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
243
244                         $URLSearchString = "^\[\]";
245                         if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
246                                 if ((count($pictures) == 1) && !$has_title) {
247                                         // Checking, if the link goes to a picture
248                                         $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
249
250                                         // Workaround:
251                                         // Sometimes photo posts to the own album are not detected at the start.
252                                         // So we seem to cannot use the cache for these cases. That's strange.
253                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
254                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
255                                         }
256
257                                         if ($data["type"] == "photo") {
258                                                 $post["type"] = "photo";
259                                                 if (isset($data["images"][0])) {
260                                                         $post["image"] = $data["images"][0]["src"];
261                                                         $post["url"] = $data["url"];
262                                                 } else {
263                                                         $post["image"] = $data["url"];
264                                                 }
265
266                                                 $post["preview"] = $pictures[0][2];
267                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
268                                         } else {
269                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
270                                                 if (substr($imgdata["mime"], 0, 6) == "image/") {
271                                                         $post["type"] = "photo";
272                                                         $post["image"] = $pictures[0][1];
273                                                         $post["preview"] = $pictures[0][2];
274                                                         $post["text"] = str_replace($pictures[0][0], "", $body);
275                                                 }
276                                         }
277                                 } elseif (count($pictures) > 0) {
278                                         $post["type"] = "link";
279                                         $post["url"] = $plink;
280                                         $post["image"] = $pictures[0][2];
281                                         $post["text"] = $body;
282                                 }
283                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
284                                 if ((count($pictures) == 1) && !$has_title) {
285                                         $post["type"] = "photo";
286                                         $post["image"] = $pictures[0][1];
287                                         $post["text"] = str_replace($pictures[0][0], "", $body);
288                                 } elseif (count($pictures) > 0) {
289                                         $post["type"] = "link";
290                                         $post["url"] = $plink;
291                                         $post["image"] = $pictures[0][1];
292                                         $post["text"] = $body;
293                                 }
294                         }
295
296                         // Test for the external links
297                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
298                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
299
300                         $links = array_merge($links1, $links2);
301
302                         // If there is only a single one, then use it.
303                         // This should cover link posts via API.
304                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
305                                 $post["type"] = "link";
306                                 $post["text"] = trim($body);
307                                 $post["url"] = $links[0][1];
308                         }
309
310                         // Now count the number of external media links
311                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
312                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
313                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
314                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
315
316                         // Add them to the other external links
317                         $links = array_merge($links, $links1, $links2, $links3, $links4);
318
319                         // Are there more than one?
320                         if (count($links) > 1) {
321                                 // The post will be the type "text", which means a blog post
322                                 unset($post["type"]);
323                                 $post["url"] = $plink;
324                         }
325
326                         if (!isset($post["type"])) {
327                                 $post["type"] = "text";
328                                 $post["text"] = trim($body);
329                         }
330                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
331                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
332
333                         if (isset($data["images"][0])) {
334                                 $post["image"] = $data["images"][0]["src"];
335                         }
336                 }
337
338                 return $post;
339         }
340
341         /**
342          * @brief Convert a message into plaintext for connectors to other networks
343          *
344          * @param array $b The message array that is about to be posted
345          * @param int $limit The maximum number of characters when posting to that network
346          * @param bool $includedlinks Has an attached link to be included into the message?
347          * @param int $htmlmode This triggers the behaviour of the bbcode conversion
348          * @param string $target_network Name of the network where the post should go to.
349          *
350          * @return string The converted message
351          */
352         public static function toPlaintext($b, $limit = 0, $includedlinks = false, $htmlmode = 2, $target_network = "")
353         {
354                 // Remove the hash tags
355                 $URLSearchString = "^\[\]";
356                 $body = preg_replace("/([#@])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $b["body"]);
357
358                 // Add an URL element if the text contains a raw link
359                 $body = preg_replace("/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url]$2[/url]', $body);
360
361                 // Remove the abstract
362                 $body = self::stripAbstract($body);
363
364                 // At first look at data that is attached via "type-..." stuff
365                 // This will hopefully replaced with a dedicated bbcode later
366                 //$post = self::getAttachedData($b["body"]);
367                 $post = self::getAttachedData($body, $b);
368
369                 if (($b["title"] != "") && ($post["text"] != "")) {
370                         $post["text"] = trim($b["title"]."\n\n".$post["text"]);
371                 } elseif ($b["title"] != "") {
372                         $post["text"] = trim($b["title"]);
373                 }
374
375                 $abstract = "";
376
377                 // Fetch the abstract from the given target network
378                 if ($target_network != "") {
379                         $default_abstract = self::getAbstract($b["body"]);
380                         $abstract = self::getAbstract($b["body"], $target_network);
381
382                         // If we post to a network with no limit we only fetch
383                         // an abstract exactly for this network
384                         if (($limit == 0) && ($abstract == $default_abstract)) {
385                                 $abstract = "";
386                         }
387                 } else {// Try to guess the correct target network
388                         switch ($htmlmode) {
389                                 case 8:
390                                         $abstract = self::getAbstract($b["body"], NETWORK_TWITTER);
391                                         break;
392                                 case 7:
393                                         $abstract = self::getAbstract($b["body"], NETWORK_STATUSNET);
394                                         break;
395                                 case 6:
396                                         $abstract = self::getAbstract($b["body"], NETWORK_APPNET);
397                                         break;
398                                 default: // We don't know the exact target.
399                                         // We fetch an abstract since there is a posting limit.
400                                         if ($limit > 0) {
401                                                 $abstract = self::getAbstract($b["body"]);
402                                         }
403                         }
404                 }
405
406                 if ($abstract != "") {
407                         $post["text"] = $abstract;
408
409                         if ($post["type"] == "text") {
410                                 $post["type"] = "link";
411                                 $post["url"] = $b["plink"];
412                         }
413                 }
414
415                 $html = self::convert($post["text"].$post["after"], false, $htmlmode);
416                 $msg = HTML::toPlaintext($html, 0, true);
417                 $msg = trim(html_entity_decode($msg, ENT_QUOTES, 'UTF-8'));
418
419                 $link = "";
420                 if ($includedlinks) {
421                         if ($post["type"] == "link") {
422                                 $link = $post["url"];
423                         } elseif ($post["type"] == "text") {
424                                 $link = $post["url"];
425                         } elseif ($post["type"] == "video") {
426                                 $link = $post["url"];
427                         } elseif ($post["type"] == "photo") {
428                                 $link = $post["image"];
429                         }
430
431                         if (($msg == "") && isset($post["title"])) {
432                                 $msg = trim($post["title"]);
433                         }
434
435                         if (($msg == "") && isset($post["description"])) {
436                                 $msg = trim($post["description"]);
437                         }
438
439                         // If the link is already contained in the post, then it neeedn't to be added again
440                         // But: if the link is beyond the limit, then it has to be added.
441                         if (($link != "") && strstr($msg, $link)) {
442                                 $pos = strpos($msg, $link);
443
444                                 // Will the text be shortened in the link?
445                                 // Or is the link the last item in the post?
446                                 if (($limit > 0) && ($pos < $limit) && (($pos + 23 > $limit) || ($pos + strlen($link) == strlen($msg)))) {
447                                         $msg = trim(str_replace($link, "", $msg));
448                                 } elseif (($limit == 0) || ($pos < $limit)) {
449                                         // The limit has to be increased since it will be shortened - but not now
450                                         // Only do it with Twitter (htmlmode = 8)
451                                         if (($limit > 0) && (strlen($link) > 23) && ($htmlmode == 8)) {
452                                                 $limit = $limit - 23 + strlen($link);
453                                         }
454
455                                         $link = "";
456
457                                         if ($post["type"] == "text") {
458                                                 unset($post["url"]);
459                                         }
460                                 }
461                         }
462                 }
463
464                 if ($limit > 0) {
465                         // Reduce multiple spaces
466                         // When posted to a network with limited space, we try to gain space where possible
467                         while (strpos($msg, "  ") !== false) {
468                                 $msg = str_replace("  ", " ", $msg);
469                         }
470
471                         // Twitter is using its own limiter, so we always assume that shortened links will have this length
472                         if (iconv_strlen($link, "UTF-8") > 0) {
473                                 $limit = $limit - 23;
474                         }
475
476                         if (iconv_strlen($msg, "UTF-8") > $limit) {
477                                 if (($post["type"] == "text") && isset($post["url"])) {
478                                         $post["url"] = $b["plink"];
479                                 } elseif (!isset($post["url"])) {
480                                         $limit = $limit - 23;
481                                         $post["url"] = $b["plink"];
482                                 // Which purpose has this line? It is now uncommented, but left as a reminder
483                                 //} elseif (strpos($b["body"], "[share") !== false) {
484                                 //      $post["url"] = $b["plink"];
485                                 } elseif (PConfig::get($b["uid"], "system", "no_intelligent_shortening")) {
486                                         $post["url"] = $b["plink"];
487                                 }
488                                 $msg = Plaintext::shorten($msg, $limit);
489                         }
490                 }
491
492                 $post["text"] = trim($msg);
493
494                 return($post);
495         }
496
497         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
498         {
499                 // Suppress "view full size"
500                 if (intval(Config::get('system', 'no_view_full_size'))) {
501                         $include_link = false;
502                 }
503
504                 // Picture addresses can contain special characters
505                 $s = htmlspecialchars_decode($srctext);
506
507                 $matches = null;
508                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
509                 if ($c) {
510                         foreach ($matches as $mtch) {
511                                 logger('scale_external_image: ' . $mtch[1]);
512
513                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
514                                 if (stristr($mtch[1], $hostname)) {
515                                         continue;
516                                 }
517
518                                 // $scale_replace, if passed, is an array of two elements. The
519                                 // first is the name of the full-size image. The second is the
520                                 // name of a remote, scaled-down version of the full size image.
521                                 // This allows Friendica to display the smaller remote image if
522                                 // one exists, while still linking to the full-size image
523                                 if ($scale_replace) {
524                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
525                                 } else {
526                                         $scaled = $mtch[1];
527                                 }
528                                 $i = Network::fetchUrl($scaled);
529                                 if (!$i) {
530                                         return $srctext;
531                                 }
532
533                                 // guess mimetype from headers or filename
534                                 $type = Image::guessType($mtch[1], true);
535
536                                 if ($i) {
537                                         $Image = new Image($i, $type);
538                                         if ($Image->isValid()) {
539                                                 $orig_width = $Image->getWidth();
540                                                 $orig_height = $Image->getHeight();
541
542                                                 if ($orig_width > 640 || $orig_height > 640) {
543                                                         $Image->scaleDown(640);
544                                                         $new_width = $Image->getWidth();
545                                                         $new_height = $Image->getHeight();
546                                                         logger('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], LOGGER_DEBUG);
547                                                         $s = str_replace(
548                                                                 $mtch[0],
549                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
550                                                                 . "\n" . (($include_link)
551                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
552                                                                         : ''),
553                                                                 $s
554                                                         );
555                                                         logger('scale_external_images: new string: ' . $s, LOGGER_DEBUG);
556                                                 }
557                                         }
558                                 }
559                         }
560                 }
561
562                 // replace the special char encoding
563                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
564                 return $s;
565         }
566
567         /**
568          * The purpose of this function is to apply system message length limits to
569          * imported messages without including any embedded photos in the length
570          *
571          * @brief Truncates imported message body string length to max_import_size
572          * @param string $body
573          * @return string
574          */
575         public static function limitBodySize($body)
576         {
577                 $maxlen = get_max_import_size();
578
579                 // If the length of the body, including the embedded images, is smaller
580                 // than the maximum, then don't waste time looking for the images
581                 if ($maxlen && (strlen($body) > $maxlen)) {
582
583                         logger('the total body length exceeds the limit', LOGGER_DEBUG);
584
585                         $orig_body = $body;
586                         $new_body = '';
587                         $textlen = 0;
588
589                         $img_start = strpos($orig_body, '[img');
590                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
591                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
592                         while (($img_st_close !== false) && ($img_end !== false)) {
593
594                                 $img_st_close++; // make it point to AFTER the closing bracket
595                                 $img_end += $img_start;
596                                 $img_end += strlen('[/img]');
597
598                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
599                                         // This is an embedded image
600
601                                         if (($textlen + $img_start) > $maxlen) {
602                                                 if ($textlen < $maxlen) {
603                                                         logger('the limit happens before an embedded image', LOGGER_DEBUG);
604                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
605                                                         $textlen = $maxlen;
606                                                 }
607                                         } else {
608                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
609                                                 $textlen += $img_start;
610                                         }
611
612                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
613                                 } else {
614
615                                         if (($textlen + $img_end) > $maxlen) {
616                                                 if ($textlen < $maxlen) {
617                                                         logger('the limit happens before the end of a non-embedded image', LOGGER_DEBUG);
618                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
619                                                         $textlen = $maxlen;
620                                                 }
621                                         } else {
622                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
623                                                 $textlen += $img_end;
624                                         }
625                                 }
626                                 $orig_body = substr($orig_body, $img_end);
627
628                                 if ($orig_body === false) {
629                                         // in case the body ends on a closing image tag
630                                         $orig_body = '';
631                                 }
632
633                                 $img_start = strpos($orig_body, '[img');
634                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
635                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
636                         }
637
638                         if (($textlen + strlen($orig_body)) > $maxlen) {
639                                 if ($textlen < $maxlen) {
640                                         logger('the limit happens after the end of the last image', LOGGER_DEBUG);
641                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
642                                 }
643                         } else {
644                                 logger('the text size with embedded images extracted did not violate the limit', LOGGER_DEBUG);
645                                 $new_body = $new_body . $orig_body;
646                         }
647
648                         return $new_body;
649                 } else {
650                         return $body;
651                 }
652         }
653
654         /**
655          * Processes [attachment] tags
656          *
657          * Note: Can produce a [bookmark] tag in the returned string
658          *
659          * @brief Processes [attachment] tags
660          * @param string $return
661          * @param bool|int $simplehtml
662          * @param bool $tryoembed
663          * @return string
664          */
665         private static function convertAttachment($return, $simplehtml = false, $tryoembed = true)
666         {
667                 $data = self::getAttachmentData($return);
668                 if (!$data) {
669                         return $return;
670                 }
671
672                 if (isset($data["title"])) {
673                         $data["title"] = strip_tags($data["title"]);
674                         $data["title"] = str_replace(["http://", "https://"], "", $data["title"]);
675                 }
676
677                 if (((strpos($data["text"], "[img=") !== false) || (strpos($data["text"], "[img]") !== false) || Config::get('system', 'always_show_preview')) && ($data["image"] != "")) {
678                         $data["preview"] = $data["image"];
679                         $data["image"] = "";
680                 }
681
682                 $return = '';
683                 if ($simplehtml == 7) {
684                         $return = self::convertUrlForMastodon($data["url"]);
685                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
686                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data["url"], $data["title"]);
687                 } else {
688                         try {
689                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
690                                         $return = OEmbed::getHTML($data['url'], $data['title']);
691                                 } else {
692                                         throw new Exception('OEmbed is disabled for this attachment.');
693                                 }
694                         } catch (Exception $e) {
695                                 if ($simplehtml != 4) {
696                                         $return = sprintf('<div class="type-%s">', $data["type"]);
697                                 }
698
699                                 if ($data["image"] != "") {
700                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a><br />', $data["url"], proxy_url($data["image"]), $data["title"]);
701                                 } elseif ($data["preview"] != "") {
702                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br />', $data["url"], proxy_url($data["preview"]), $data["title"]);
703                                 }
704
705                                 if (($data["type"] == "photo") && ($data["url"] != "") && ($data["image"] != "")) {
706                                         $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"]);
707                                 } else {
708                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
709                                 }
710
711                                 if ($data["description"] != "" && $data["description"] != $data["title"]) {
712                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($data["description"])));
713                                 }
714
715                                 if ($data["type"] == "link") {
716                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
717                                 }
718
719                                 if ($simplehtml != 4) {
720                                         $return .= '</div>';
721                                 }
722                         }
723                 }
724
725                 return trim($data["text"] . ' ' . $return . ' ' . $data["after"]);
726         }
727
728         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
729         {
730                 $data = self::getAttachmentData($Text);
731
732                 if (!$data) {
733                         return $Text;
734                 } elseif ($nolink) {
735                         return $data["text"] . $data["after"];
736                 }
737
738                 $title = htmlentities($data["title"], ENT_QUOTES, 'UTF-8', false);
739                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
740                 if ($plaintext || (($title != "") && strstr($text, $title))) {
741                         $data["title"] = $data["url"];
742                 } elseif (($text != "") && strstr($title, $text)) {
743                         $data["text"] = $data["title"];
744                         $data["title"] = $data["url"];
745                 }
746
747                 if (($data["text"] == "") && ($data["title"] != "") && ($data["url"] == "")) {
748                         return $data["title"] . $data["after"];
749                 }
750
751                 // If the link already is included in the post, don't add it again
752                 if (($data["url"] != "") && strpos($data["text"], $data["url"])) {
753                         return $data["text"] . $data["after"];
754                 }
755
756                 $text = $data["text"];
757
758                 if (($data["url"] != "") && ($data["title"] != "")) {
759                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
760                 } elseif (($data["url"] != "")) {
761                         $text .= "\n" . $data["url"];
762                 }
763
764                 return $text . "\n" . $data["after"];
765         }
766
767         /**
768          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
769          *
770          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
771          * @param array $match Array with the matching values
772          * @return string reformatted link including HTML codes
773          */
774         private static function convertUrlForMastodonCallback($match)
775         {
776                 $url = $match[1];
777
778                 if (isset($match[2]) && ($match[1] != $match[2])) {
779                         return $match[0];
780                 }
781
782                 $parts = parse_url($url);
783                 if (!isset($parts['scheme'])) {
784                         return $match[0];
785                 }
786
787                 return self::convertUrlForMastodon($url);
788         }
789
790         /**
791          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon and GNU Social.
792          * @param string $url URL that is about to be reformatted
793          * @return string reformatted link including HTML codes
794          */
795         private static function convertUrlForMastodon($url)
796         {
797                 $parts = parse_url($url);
798                 $scheme = $parts['scheme'] . '://';
799                 $styled_url = str_replace($scheme, '', $url);
800
801                 $html = '<a href="%s" class="attachment" rel="nofollow noopener" target="_blank">' .
802                         '<span class="invisible">%s</span>';
803
804                 if (strlen($styled_url) > 30) {
805                         $html .= '<span class="ellipsis">%s</span>' .
806                                 '<span class="invisible">%s</span></a>';
807
808                         $ellipsis = substr($styled_url, 0, 30);
809                         $rest = substr($styled_url, 30);
810                         return sprintf($html, $url, $scheme, $ellipsis, $rest);
811                 } else {
812                         $html .= '%s</a>';
813                         return sprintf($html, $url, $scheme, $styled_url);
814                 }
815         }
816
817         /*
818          * [noparse][i]italic[/i][/noparse] turns into
819          * [noparse][ i ]italic[ /i ][/noparse],
820          * to hide them from parser.
821          */
822         private static function escapeNoparseCallback($match)
823         {
824                 $whole_match = $match[0];
825                 $captured = $match[1];
826                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
827                 $new_str = str_replace($captured, $spacefied, $whole_match);
828                 return $new_str;
829         }
830
831         /*
832          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
833          * now turns back and the [noparse] tags are trimed
834          * returning [i]italic[/i]
835          */
836         private static function unescapeNoparseCallback($match)
837         {
838                 $captured = $match[1];
839                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
840                 return $unspacefied;
841         }
842
843         /**
844          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
845          * occurrences
846          *
847          * @param string $text        Text to search
848          * @param string $name        Tag name
849          * @param int    $occurrences Number of first occurrences to skip
850          * @return boolean|array
851          */
852         public static function getTagPosition($text, $name, $occurrences = 0)
853         {
854                 if ($occurrences < 0) {
855                         $occurrences = 0;
856                 }
857
858                 $start_open = -1;
859                 for ($i = 0; $i <= $occurrences; $i++) {
860                         if ($start_open !== false) {
861                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
862                         }
863                 }
864
865                 if ($start_open === false) {
866                         return false;
867                 }
868
869                 $start_equal = strpos($text, '=', $start_open);
870                 $start_close = strpos($text, ']', $start_open);
871
872                 if ($start_close === false) {
873                         return false;
874                 }
875
876                 $start_close++;
877
878                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
879
880                 if ($end_open === false) {
881                         return false;
882                 }
883
884                 $res = [
885                         'start' => [
886                                 'open' => $start_open,
887                                 'close' => $start_close
888                         ],
889                         'end' => [
890                                 'open' => $end_open,
891                                 'close' => $end_open + strlen('[/' . $name . ']')
892                         ],
893                 ];
894
895                 if ($start_equal !== false) {
896                         $res['start']['equal'] = $start_equal + 1;
897                 }
898
899                 return $res;
900         }
901
902         /**
903          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
904          *
905          * @param type $pattern Preg pattern string
906          * @param type $replace Preg replace string
907          * @param type $name    BBCode tag name
908          * @param type $text    Text to search
909          * @return string
910          */
911         public static function pregReplaceInTag($pattern, $replace, $name, $text)
912         {
913                 $occurrences = 0;
914                 $pos = self::getTagPosition($text, $name, $occurrences);
915                 while ($pos !== false && $occurrences++ < 1000) {
916                         $start = substr($text, 0, $pos['start']['open']);
917                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
918                         $end = substr($text, $pos['end']['close']);
919                         if ($end === false) {
920                                 $end = '';
921                         }
922
923                         $subject = preg_replace($pattern, $replace, $subject);
924                         $text = $start . $subject . $end;
925
926                         $pos = self::getTagPosition($text, $name, $occurrences);
927                 }
928
929                 return $text;
930         }
931
932         private static function extractImagesFromItemBody($body)
933         {
934                 $saved_image = [];
935                 $orig_body = $body;
936                 $new_body = '';
937
938                 $cnt = 0;
939                 $img_start = strpos($orig_body, '[img');
940                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
941                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
942                 while (($img_st_close !== false) && ($img_end !== false)) {
943                         $img_st_close++; // make it point to AFTER the closing bracket
944                         $img_end += $img_start;
945
946                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
947                                 // This is an embedded image
948                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
949                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
950
951                                 $cnt++;
952                         } else {
953                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
954                         }
955
956                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
957
958                         if ($orig_body === false) {
959                                 // in case the body ends on a closing image tag
960                                 $orig_body = '';
961                         }
962
963                         $img_start = strpos($orig_body, '[img');
964                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
965                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
966                 }
967
968                 $new_body = $new_body . $orig_body;
969
970                 return ['body' => $new_body, 'images' => $saved_image];
971         }
972
973         private static function interpolateSavedImagesIntoItemBody($body, array $images)
974         {
975                 $newbody = $body;
976
977                 $cnt = 0;
978                 foreach ($images as $image) {
979                         // We're depending on the property of 'foreach' (specified on the PHP website) that
980                         // it loops over the array starting from the first element and going sequentially
981                         // to the last element
982                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
983                                 '<img src="' . proxy_url($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
984                         $cnt++;
985                 }
986
987                 return $newbody;
988         }
989
990         /**
991          * Processes [share] tags
992          *
993          * Note: Can produce a [bookmark] tag in the output
994          *
995          * @brief Processes [share] tags
996          * @param array    $share      preg_match_callback result array
997          * @param bool|int $simplehtml
998          * @return string
999          */
1000         private static function convertShare($share, $simplehtml)
1001         {
1002                 $attributes = $share[2];
1003
1004                 $author = "";
1005                 preg_match("/author='(.*?)'/ism", $attributes, $matches);
1006                 if (x($matches, 1)) {
1007                         $author = html_entity_decode($matches[1], ENT_QUOTES, 'UTF-8');
1008                 }
1009
1010                 preg_match('/author="(.*?)"/ism', $attributes, $matches);
1011                 if (x($matches, 1)) {
1012                         $author = $matches[1];
1013                 }
1014
1015                 $profile = "";
1016                 preg_match("/profile='(.*?)'/ism", $attributes, $matches);
1017                 if (x($matches, 1)) {
1018                         $profile = $matches[1];
1019                 }
1020
1021                 preg_match('/profile="(.*?)"/ism', $attributes, $matches);
1022                 if (x($matches, 1)) {
1023                         $profile = $matches[1];
1024                 }
1025
1026                 $avatar = "";
1027                 preg_match("/avatar='(.*?)'/ism", $attributes, $matches);
1028                 if (x($matches, 1)) {
1029                         $avatar = $matches[1];
1030                 }
1031
1032                 preg_match('/avatar="(.*?)"/ism', $attributes, $matches);
1033                 if (x($matches, 1)) {
1034                         $avatar = $matches[1];
1035                 }
1036
1037                 $link = "";
1038                 preg_match("/link='(.*?)'/ism", $attributes, $matches);
1039                 if (x($matches, 1)) {
1040                         $link = $matches[1];
1041                 }
1042
1043                 preg_match('/link="(.*?)"/ism', $attributes, $matches);
1044                 if (x($matches, 1)) {
1045                         $link = $matches[1];
1046                 }
1047
1048                 $posted = "";
1049
1050                 preg_match("/posted='(.*?)'/ism", $attributes, $matches);
1051                 if (x($matches, 1)) {
1052                         $posted = $matches[1];
1053                 }
1054
1055                 preg_match('/posted="(.*?)"/ism', $attributes, $matches);
1056                 if (x($matches, 1)) {
1057                         $posted = $matches[1];
1058                 }
1059
1060                 // We only call this so that a previously unknown contact can be added.
1061                 // This is important for the function "Model\Contact::getDetailsByURL()".
1062                 // This function then can fetch an entry from the contact table.
1063                 Contact::getIdForURL($profile, 0, true);
1064
1065                 $data = Contact::getDetailsByURL($profile);
1066
1067                 if (x($data, "name") && x($data, "addr")) {
1068                         $userid_compact = $data["name"] . " (" . $data["addr"] . ")";
1069                 } else {
1070                         $userid_compact = Protocol::getAddrFromProfileUrl($profile, $author);
1071                 }
1072
1073                 if (x($data, "addr")) {
1074                         $userid = $data["addr"];
1075                 } else {
1076                         $userid = Protocol::formatMention($profile, $author);
1077                 }
1078
1079                 if (x($data, "name")) {
1080                         $author = $data["name"];
1081                 }
1082
1083                 if (x($data, "micro")) {
1084                         $avatar = $data["micro"];
1085                 }
1086
1087                 $preshare = trim($share[1]);
1088                 if ($preshare != "") {
1089                         $preshare .= "<br />";
1090                 }
1091
1092                 switch ($simplehtml) {
1093                         case 1:
1094                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' <a href="' . $profile . '">' . $userid . "</a>: <br />»" . $share[3] . "«";
1095                                 break;
1096                         case 2:
1097                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1098                                 break;
1099                         case 3: // Diaspora
1100                                 $headline = '<b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . $userid . ':</b><br />';
1101
1102                                 $text = trim($share[1]);
1103
1104                                 if ($text != "") {
1105                                         $text .= "<hr />";
1106                                 }
1107
1108                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0) {
1109                                         $text .= $headline . '<blockquote>' . trim($share[3]) . "</blockquote><br />";
1110
1111                                         if ($link != "") {
1112                                                 $text .= '<br /><a href="' . $link . '">[l]</a>';
1113                                         }
1114                                 } else {
1115                                         $text .= '<br /><a href="' . $link . '">' . $link . '</a>';
1116                                 }
1117
1118                                 break;
1119                         case 4:
1120                                 $headline = '<br /><b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8');
1121                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $link, $userid, $posted);
1122                                 $headline .= ":</b><br />";
1123
1124                                 $text = trim($share[1]);
1125
1126                                 if ($text != "") {
1127                                         $text .= "<hr />";
1128                                 }
1129
1130                                 $text .= $headline . '<blockquote class="shared_content">' . trim($share[3]) . "</blockquote><br />";
1131
1132                                 break;
1133                         case 5:
1134                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1135                                 break;
1136                         case 6: // app.net
1137                                 $text = $preshare . "&gt;&gt; @" . $userid_compact . ": <br />" . $share[3];
1138                                 break;
1139                         case 7: // statusnet/GNU Social
1140                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . " @" . $userid_compact . ": " . $share[3];
1141                                 break;
1142                         case 8: // twitter
1143                                 $text = $preshare . "RT @" . $userid_compact . ": " . $share[3];
1144                                 break;
1145                         case 9: // Google+/Facebook
1146                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1147
1148                                 if ($link != "") {
1149                                         $text .= "<br /><br />" . $link;
1150                                 }
1151                                 break;
1152                         default:
1153                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1154                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($link)) {
1155                                         try {
1156                                                 $oembed = OEmbed::getHTML($link, $preshare);
1157                                         } catch (Exception $e) {
1158                                                 $oembed = sprintf('[bookmark=%s]%s[/bookmark]', $link, $preshare);
1159                                         }
1160
1161                                         $text = $preshare . $oembed;
1162                                 } else {
1163                                         $text = trim($share[1]) . "\n";
1164
1165                                         $avatar = proxy_url($avatar, false, PROXY_SIZE_THUMB);
1166
1167                                         $tpl = get_markup_template('shared_content.tpl');
1168                                         $text .= replace_macros($tpl, [
1169                                                 '$profile' => $profile,
1170                                                 '$avatar' => $avatar,
1171                                                 '$author' => $author,
1172                                                 '$link' => $link,
1173                                                 '$posted' => $posted,
1174                                                 '$content' => trim($share[3])
1175                                         ]);
1176                                 }
1177                                 break;
1178                 }
1179
1180                 return $text;
1181         }
1182
1183         private static function removePictureLinksCallback($match)
1184         {
1185                 $text = Cache::get($match[1]);
1186
1187                 if (is_null($text)) {
1188                         $a = self::getApp();
1189
1190                         $stamp1 = microtime(true);
1191
1192                         $ch = @curl_init($match[1]);
1193                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1194                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1195                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1196                         @curl_exec($ch);
1197                         $curl_info = @curl_getinfo($ch);
1198
1199                         $a->save_timestamp($stamp1, "network");
1200
1201                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1202                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1203                         } else {
1204                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1205
1206                                 // if its not a picture then look if its a page that contains a picture link
1207                                 $body = Network::fetchUrl($match[1]);
1208
1209                                 $doc = new DOMDocument();
1210                                 @$doc->loadHTML($body);
1211                                 $xpath = new DomXPath($doc);
1212                                 $list = $xpath->query("//meta[@name]");
1213                                 foreach ($list as $node) {
1214                                         $attr = [];
1215
1216                                         if ($node->attributes->length) {
1217                                                 foreach ($node->attributes as $attribute) {
1218                                                         $attr[$attribute->name] = $attribute->value;
1219                                                 }
1220                                         }
1221
1222                                         if (strtolower($attr["name"]) == "twitter:image") {
1223                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1224                                         }
1225                                 }
1226                         }
1227                         Cache::set($match[1], $text);
1228                 }
1229
1230                 return $text;
1231         }
1232
1233         private static function expandLinksCallback($match)
1234         {
1235                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1236                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1237                 } else {
1238                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1239                 }
1240         }
1241
1242         private static function cleanPictureLinksCallback($match)
1243         {
1244                 $text = Cache::get($match[1]);
1245
1246                 if (is_null($text)) {
1247                         $a = self::getApp();
1248
1249                         $stamp1 = microtime(true);
1250
1251                         $ch = @curl_init($match[1]);
1252                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1253                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1254                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1255                         @curl_exec($ch);
1256                         $curl_info = @curl_getinfo($ch);
1257
1258                         $a->save_timestamp($stamp1, "network");
1259
1260                         // if its a link to a picture then embed this picture
1261                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1262                                 $text = "[img]" . $match[1] . "[/img]";
1263                         } else {
1264                                 $text = "[img]" . $match[2] . "[/img]";
1265
1266                                 // if its not a picture then look if its a page that contains a picture link
1267                                 $body = Network::fetchUrl($match[1]);
1268
1269                                 $doc = new DOMDocument();
1270                                 @$doc->loadHTML($body);
1271                                 $xpath = new DOMXPath($doc);
1272                                 $list = $xpath->query("//meta[@name]");
1273                                 foreach ($list as $node) {
1274                                         $attr = [];
1275                                         if ($node->attributes->length) {
1276                                                 foreach ($node->attributes as $attribute) {
1277                                                         $attr[$attribute->name] = $attribute->value;
1278                                                 }
1279                                         }
1280
1281                                         if (strtolower($attr["name"]) == "twitter:image") {
1282                                                 $text = "[img]" . $attr["content"] . "[/img]";
1283                                         }
1284                                 }
1285                         }
1286                         Cache::set($match[1], $text);
1287                 }
1288
1289                 return $text;
1290         }
1291
1292         public static function cleanPictureLinks($text)
1293         {
1294                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1295                 return $return;
1296         }
1297
1298         private static function textHighlightCallback($match)
1299         {
1300                 if (in_array(strtolower($match[1]),
1301                                 ['php', 'css', 'mysql', 'sql', 'abap', 'diff', 'html', 'perl', 'ruby',
1302                                 'vbscript', 'avrc', 'dtd', 'java', 'xml', 'cpp', 'python', 'javascript', 'js', 'sh'])
1303                 ) {
1304                         return text_highlight($match[2], strtolower($match[1]));
1305                 }
1306                 return $match[0];
1307         }
1308
1309         /**
1310          * @brief Converts a BBCode message to HTML message
1311          *
1312          * BBcode 2 HTML was written by WAY2WEB.net
1313          * extended to work with Mistpark/Friendica - Mike Macgirvin
1314          *
1315          * Simple HTML values meaning:
1316          * - 0: Friendica display
1317          * - 1: Unused
1318          * - 2: Used for Facebook, Google+, Windows Phone push, Friendica API
1319          * - 3: Used before converting to Markdown in bb2diaspora.php
1320          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1321          * - 5: Unused
1322          * - 6: Used for Appnet
1323          * - 7: Used for dfrn, OStatus
1324          * - 8: Used for WP backlink text setting
1325          *
1326          * @param string $text
1327          * @param bool   $try_oembed
1328          * @param int    $simple_html
1329          * @param bool   $for_plaintext
1330          * @return string
1331          */
1332         public static function convert($text, $try_oembed = true, $simple_html = false, $for_plaintext = false)
1333         {
1334                 $a = self::getApp();
1335
1336                 /*
1337                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1338                  *
1339                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1340                  * $match[1] = $url
1341                  * $match[2] = $title or absent
1342                  */
1343                 $try_oembed_callback = function ($match)
1344                 {
1345                         $url = $match[1];
1346                         $title = defaults($match, 2, null);
1347
1348                         try {
1349                                 $return = OEmbed::getHTML($url, $title);
1350                         } catch (Exception $ex) {
1351                                 $return = $match[0];
1352                         }
1353
1354                         return $return;
1355                 };
1356
1357                 // Hide all [noparse] contained bbtags by spacefying them
1358                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1359
1360                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1361                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1362                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1363
1364                 // Remove the abstract element. It is a non visible element.
1365                 $text = self::stripAbstract($text);
1366
1367                 // Move all spaces out of the tags
1368                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1369                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1370
1371                 // Extract the private images which use data urls since preg has issues with
1372                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1373                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1374
1375                 $extracted = self::extractImagesFromItemBody($text);
1376                 $text = $extracted['body'];
1377                 $saved_image = $extracted['images'];
1378
1379                 // If we find any event code, turn it into an event.
1380                 // After we're finished processing the bbcode we'll
1381                 // replace all of the event code with a reformatted version.
1382
1383                 $ev = bbtoevent($text);
1384
1385                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1386                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1387
1388                 $text = str_replace("<", "&lt;", $text);
1389                 $text = str_replace(">", "&gt;", $text);
1390
1391                 // remove some newlines before the general conversion
1392                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1393                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1394
1395                 $text = preg_replace("/\n\[code\]/ism", "[code]", $text);
1396                 $text = preg_replace("/\[\/code\]\n/ism", "[/code]", $text);
1397
1398                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1399                 if (!$try_oembed) {
1400                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1401                 }
1402
1403                 // Check for [code] text here, before the linefeeds are messed with.
1404                 // The highlighter will unescape and re-escape the content.
1405                 if (strpos($text, '[code=') !== false) {
1406                         $text = preg_replace_callback("/\[code=(.*?)\](.*?)\[\/code\]/ism", 'self::textHighlightCallback', $text);
1407                 }
1408                 // Convert new line chars to html <br /> tags
1409
1410                 // nlbr seems to be hopelessly messed up
1411                 //      $Text = nl2br($Text);
1412
1413                 // We'll emulate it.
1414
1415                 $text = trim($text);
1416                 $text = str_replace("\r\n", "\n", $text);
1417
1418                 // removing multiplicated newlines
1419                 if (Config::get("system", "remove_multiplicated_lines")) {
1420                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1421                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1422                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1423                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1424                         do {
1425                                 $oldtext = $text;
1426                                 $text = str_replace($search, $replace, $text);
1427                         } while ($oldtext != $text);
1428                 }
1429
1430                 // Set up the parameters for a URL search string
1431                 $URLSearchString = "^\[\]";
1432                 // Set up the parameters for a MAIL search string
1433                 $MAILSearchString = $URLSearchString;
1434
1435                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1436                 if (!$for_plaintext) {
1437                         // Autolink feature (thanks to http://code.seebz.net/p/autolink-php/)
1438                         // Currently disabled, since the function is too greedy
1439                         // $autolink_regex = "`([^\]\=\"']|^)(https?\://[^\s<]+[^\s<\.\)])`ism";
1440                         $autolink_regex = "/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism";
1441                         $text = preg_replace($autolink_regex, '$1[url]$2[/url]', $text);
1442                         if ($simple_html == 7) {
1443                                 $text = preg_replace_callback("/\[url\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForMastodonCallback', $text);
1444                                 $text = preg_replace_callback("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForMastodonCallback', $text);
1445                         }
1446                 } else {
1447                         $text = preg_replace("(\[url\]([$URLSearchString]*)\[\/url\])ism", " $1 ", $text);
1448                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1449                 }
1450
1451
1452                 // Handle attached links or videos
1453                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1454
1455                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1456
1457                 // Remove all hashtag addresses
1458                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7])) {
1459                         $text = preg_replace("/([#@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $text);
1460                 } elseif ($simple_html == 3) {
1461                         // The ! is converted to @ since Diaspora only understands the @
1462                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1463                                 '@<a href="$2">$3</a>',
1464                                 $text);
1465                 } elseif ($simple_html == 7) {
1466                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1467                                 '$1<span class="vcard"><a href="$2" class="url" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1468                                 $text);
1469                 } elseif (!$simple_html) {
1470                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1471                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1472                                 $text);
1473                 }
1474
1475                 // Bookmarks in red - will be converted to bookmarks in friendica
1476                 $text = preg_replace("/#\^\[url\]([$URLSearchString]*)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1477                 $text = preg_replace("/#\^\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1478                 $text = preg_replace("/#\[url\=[$URLSearchString]*\]\^\[\/url\]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i",
1479                                         "[bookmark=$1]$2[/bookmark]", $text);
1480
1481                 if (in_array($simple_html, [2, 6, 7, 8, 9])) {
1482                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1483                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1484                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1485                 }
1486
1487                 if ($simple_html == 5) {
1488                         $text = preg_replace("/[^#@!]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1489                 }
1490
1491                 // Perform URL Search
1492                 if ($try_oembed) {
1493                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1494                 }
1495
1496                 if ($simple_html == 5) {
1497                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1498                 } else {
1499                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1500                 }
1501
1502                 // Handle Diaspora posts
1503                 $text = preg_replace_callback(
1504                         "&\[url=/posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1505                         function ($match) {
1506                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1507                         }, $text
1508                 );
1509
1510                 // Server independent link to posts and comments
1511                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1512                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1513                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1514
1515                 $text = preg_replace("/([#])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1516                                         '$1<a href="' . System::baseUrl() . '/search?tag=$3" class="tag" title="$3">$3</a>', $text);
1517
1518                 $text = preg_replace("/\[url\=([$URLSearchString]*)\]#(.*?)\[\/url\]/ism",
1519                                         '#<a href="' . System::baseUrl() . '/search?tag=$2" class="tag" title="$2">$2</a>', $text);
1520
1521                 $text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1522                 $text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1523                 //$Text = preg_replace("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $Text);
1524
1525                 // Red compatibility, though the link can't be authenticated on Friendica
1526                 $text = preg_replace("/\[zrl\=([$URLSearchString]*)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1527
1528
1529                 // we may need to restrict this further if it picks up too many strays
1530                 // link acct:user@host to a webfinger profile redirector
1531
1532                 $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);
1533
1534                 // Perform MAIL Search
1535                 $text = preg_replace("/\[mail\]([$MAILSearchString]*)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1536                 $text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1537
1538                 // leave open the posibility of [map=something]
1539                 // this is replaced in prepare_body() which has knowledge of the item location
1540
1541                 if (strpos($text, '[/map]') !== false) {
1542                         $text = preg_replace_callback(
1543                                 "/\[map\](.*?)\[\/map\]/ism",
1544                                 function ($match) {
1545                                         // the extra space in the following line is intentional
1546                                         // Whyyy? - @MrPetovan
1547                                         return str_replace($match[0], '<div class="map"  >' . Map::byLocation($match[1]) . '</div>', $match[0]);
1548                                 },
1549                                 $text
1550                         );
1551                 }
1552                 if (strpos($text, '[map=') !== false) {
1553                         $text = preg_replace_callback(
1554                                 "/\[map=(.*?)\]/ism",
1555                                 function ($match) {
1556                                         // the extra space in the following line is intentional
1557                                         // Whyyy? - @MrPetovan
1558                                         return str_replace($match[0], '<div class="map"  >' . Map::byCoordinates(str_replace('/', ' ', $match[1])) . '</div>', $match[0]);
1559                                 },
1560                                 $text
1561                         );
1562                 }
1563                 if (strpos($text, '[map]') !== false) {
1564                         $text = preg_replace("/\[map\]/", '<div class="map"></div>', $text);
1565                 }
1566
1567                 // Check for headers
1568                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1569                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1570                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1571                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1572                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1573                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1574
1575                 // Check for paragraph
1576                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1577
1578                 // Check for bold text
1579                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1580
1581                 // Check for Italics text
1582                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1583
1584                 // Check for Underline text
1585                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1586
1587                 // Check for strike-through text
1588                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<strike>$1</strike>', $text);
1589
1590                 // Check for over-line text
1591                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1592
1593                 // Check for colored text
1594                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1595
1596                 // Check for sized text
1597                 // [size=50] --> font-size: 50px (with the unit).
1598                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1599                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1600
1601                 // Check for centered text
1602                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1603
1604                 // Check for list text
1605                 $text = str_replace("[*]", "<li>", $text);
1606
1607                 // Check for style sheet commands
1608                 $text = preg_replace_callback(
1609                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1610                         function ($match) {
1611                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1612                         },
1613                         $text
1614                 );
1615
1616                 // Check for CSS classes
1617                 $text = preg_replace_callback(
1618                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1619                         function ($match) {
1620                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1621                         },
1622                         $text
1623                 );
1624
1625                 // handle nested lists
1626                 $endlessloop = 0;
1627
1628                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1629                            ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1630                            ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1631                            ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1632                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1633                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1634                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1635                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1636                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1637                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1638                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1639                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1640                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1641                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1642                 }
1643
1644                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1645                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1646                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1647                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1648
1649                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1650                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1651
1652                 $text = str_replace('[hr]', '<hr />', $text);
1653
1654                 // This is actually executed in prepare_body()
1655
1656                 $text = str_replace('[nosmile]', '', $text);
1657
1658                 // Check for font change text
1659                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1660
1661                 // Declare the format for [code] layout
1662
1663                 $CodeLayout = '<code>$1</code>';
1664                 // Check for [code] text
1665                 $text = preg_replace("/\[code\](.*?)\[\/code\]/ism", "$CodeLayout", $text);
1666
1667                 // Declare the format for [spoiler] layout
1668                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
1669
1670                 // Check for [spoiler] text
1671                 // handle nested quotes
1672                 $endlessloop = 0;
1673                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1674                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", "$SpoilerLayout", $text);
1675                 }
1676
1677                 // Check for [spoiler=Author] text
1678
1679                 $t_wrote = L10n::t('$1 wrote:');
1680
1681                 // handle nested quotes
1682                 $endlessloop = 0;
1683                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1684                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1685                                                  "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1686                                                  $text);
1687                 }
1688
1689                 // Declare the format for [quote] layout
1690                 $QuoteLayout = '<blockquote>$1</blockquote>';
1691
1692                 // Check for [quote] text
1693                 // handle nested quotes
1694                 $endlessloop = 0;
1695                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1696                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1697                 }
1698
1699                 // Check for [quote=Author] text
1700
1701                 $t_wrote = L10n::t('$1 wrote:');
1702
1703                 // handle nested quotes
1704                 $endlessloop = 0;
1705                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1706                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1707                                                  "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1708                                                  $text);
1709                 }
1710
1711
1712                 // [img=widthxheight]image source[/img]
1713                 $text = preg_replace_callback(
1714                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1715                         function ($matches) {
1716                                 if (strpos($matches[3], "data:image/") === 0) {
1717                                         return $matches[0];
1718                                 }
1719
1720                                 $matches[3] = proxy_url($matches[3]);
1721                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1722                         },
1723                         $text
1724                 );
1725
1726                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1727                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1728
1729                 // Images
1730                 // [img]pathtoimage[/img]
1731                 $text = preg_replace_callback(
1732                         "/\[img\](.*?)\[\/img\]/ism",
1733                         function ($matches) {
1734                                 if (strpos($matches[1], "data:image/") === 0) {
1735                                         return $matches[0];
1736                                 }
1737
1738                                 $matches[1] = proxy_url($matches[1]);
1739                                 return "[img]" . $matches[1] . "[/img]";
1740                         },
1741                         $text
1742                 );
1743
1744                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1745                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1746
1747                 // Shared content
1748                 $text = preg_replace_callback("/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
1749                         function ($match) use ($simple_html) {
1750                                 return self::convertShare($match, $simple_html);
1751                         }, $text);
1752
1753                 $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);
1754                 $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 />', $text);
1755                 //$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 />', $Text);
1756
1757                 // Try to Oembed
1758                 if ($try_oembed) {
1759                         $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);
1760                         $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);
1761
1762                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1763                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1764                 } else {
1765                         $text = preg_replace("/\[video\](.*?)\[\/video\]/",
1766                                                 '<a href="$1" target="_blank">$1</a>', $text);
1767                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/",
1768                                                 '<a href="$1" target="_blank">$1</a>', $text);
1769                 }
1770
1771                 // html5 video and audio
1772
1773
1774                 if ($try_oembed) {
1775                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1776                 } else {
1777                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1778                 }
1779
1780                 // Youtube extensions
1781                 if ($try_oembed) {
1782                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1783                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1784                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1785                 }
1786
1787                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1788                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1789                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1790
1791                 if ($try_oembed) {
1792                         $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);
1793                 } else {
1794                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1795                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1796                 }
1797
1798                 if ($try_oembed) {
1799                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1800                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1801                 }
1802
1803                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1804                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1805
1806                 if ($try_oembed) {
1807                         $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);
1808                 } else {
1809                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1810                                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1811                 }
1812
1813                 // oembed tag
1814                 $text = OEmbed::BBCode2HTML($text);
1815
1816                 // Avoid triple linefeeds through oembed
1817                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1818
1819                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1820                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1821                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1822                 // start which is always required). Allow desc with a missing summary for compatibility.
1823
1824                 if ((x($ev, 'desc') || x($ev, 'summary')) && x($ev, 'start')) {
1825                         $sub = format_event_html($ev, $simple_html);
1826
1827                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1828                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1829                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1830                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1831                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1832                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1833                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1834                 }
1835
1836                 // Replace non graphical smilies for external posts
1837                 if ($simple_html) {
1838                         $text = Smilies::replace($text, false, true);
1839                 }
1840
1841                 // Replace inline code blocks
1842                 $text = preg_replace_callback("|(?!<br[^>]*>)<code>([^<]*)</code>(?!<br[^>]*>)|ism",
1843                         function ($match) use ($simple_html) {
1844                                 $return = '<key>' . $match[1] . '</key>';
1845                                 // Use <code> for Diaspora inline code blocks
1846                                 if ($simple_html === 3) {
1847                                         $return = '<code>' . $match[1] . '</code>';
1848                                 }
1849                                 return $return;
1850                         }
1851                 , $text);
1852
1853                 // Unhide all [noparse] contained bbtags unspacefying them
1854                 // and triming the [noparse] tag.
1855
1856                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1857                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1858                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1859
1860
1861                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1862                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1863                 $text = preg_replace('/\&quot\;/', '"', $text);
1864
1865                 // fix any escaped ampersands that may have been converted into links
1866                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1867
1868                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1869                 $allowed_src_protocols = ['http', 'redir', 'cid'];
1870                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1871                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1872
1873                 // sanitize href attributes (only whitelisted protocols URLs)
1874                 // default value for backward compatibility
1875                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', ['ftp', 'mailto', 'gopher', 'cid']);
1876
1877                 // Always allowed protocol even if config isn't set or not including it
1878                 $allowed_link_protocols[] = 'http';
1879                 $allowed_link_protocols[] = 'redir/';
1880
1881                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1882                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1883
1884                 if ($saved_image) {
1885                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1886                 }
1887
1888                 // Clean up the HTML by loading and saving the HTML with the DOM.
1889                 // Bad structured html can break a whole page.
1890                 // For performance reasons do it only with ativated item cache or at export.
1891                 if (!$try_oembed || (get_itemcachepath() != "")) {
1892                         $doc = new DOMDocument();
1893                         $doc->preserveWhiteSpace = false;
1894
1895                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1896
1897                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1898                         $encoding = '<?xml encoding="UTF-8">';
1899                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1900                         $doc->encoding = 'UTF-8';
1901                         $text = $doc->saveHTML();
1902                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1903
1904                         $text = str_replace('<br></li>', '</li>', $text);
1905
1906                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1907                 }
1908
1909                 // Clean up some useless linebreaks in lists
1910                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1911                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1912                 //$Text = str_replace('</li><br />', '</li>', $Text);
1913                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1914                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1915
1916                 Addon::callHooks('bbcode', $text);
1917
1918                 return trim($text);
1919         }
1920
1921         /**
1922          * @brief Strips the "abstract" tag from the provided text
1923          *
1924          * @param string $text The text with BBCode
1925          * @return string The same text - but without "abstract" element
1926          */
1927         public static function stripAbstract($text)
1928         {
1929                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1930                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1931
1932                 return $text;
1933         }
1934
1935         /**
1936          * @brief Returns the value of the "abstract" element
1937          *
1938          * @param string $text The text that maybe contains the element
1939          * @param string $addon The addon for which the abstract is meant for
1940          * @return string The abstract
1941          */
1942         private static function getAbstract($text, $addon = "")
1943         {
1944                 $abstract = "";
1945                 $abstracts = [];
1946                 $addon = strtolower($addon);
1947
1948                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1949                         foreach ($results AS $result) {
1950                                 $abstracts[strtolower($result[1])] = $result[2];
1951                         }
1952                 }
1953
1954                 if (isset($abstracts[$addon])) {
1955                         $abstract = $abstracts[$addon];
1956                 }
1957
1958                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1959                         $abstract = $result[1];
1960                 }
1961
1962                 return $abstract;
1963         }
1964
1965         /**
1966          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1967          *
1968          * @param array $match Matching values for the callback
1969          * @return string Replaced mention
1970          */
1971         private static function bbCodeMention2DiasporaCallback($match)
1972         {
1973                 $contact = Contact::getDetailsByURL($match[3]);
1974
1975                 if (empty($contact['addr'])) {
1976                         $contact = Probe::uri($match[3]);
1977                 }
1978
1979                 if (empty($contact['addr'])) {
1980                         return $match[0];
1981                 }
1982
1983                 $mention = '@{' . $match[2] . '; ' . $contact['addr'] . '}';
1984                 return $mention;
1985         }
1986
1987         /**
1988          * @brief Converts a BBCode text into Markdown
1989          *
1990          * This function converts a BBCode item body to be sent to Markdown-enabled
1991          * systems like Diaspora and Libertree
1992          *
1993          * @param string $text
1994          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1995          * @return string
1996          */
1997         public static function toMarkdown($text, $for_diaspora = true)
1998         {
1999                 $a = self::getApp();
2000
2001                 $original_text = $text;
2002
2003                 // Since Diaspora is creating a summary for links, this function removes them before posting
2004                 if ($for_diaspora) {
2005                         $text = self::removeShareInformation($text);
2006                 }
2007
2008                 /**
2009                  * Transform #tags, strip off the [url] and replace spaces with underscore
2010                  */
2011                 $url_search_string = "^\[\]";
2012                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2013                         function ($matches) {
2014                                 return '#' . str_replace(' ', '_', $matches[2]);
2015                         },
2016                         $text
2017                 );
2018
2019                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2020                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2021
2022                 // Extracting multi-line code blocks before the whitespace processing/code highlighter in self::convert()
2023                 $codeblocks = [];
2024
2025                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#is",
2026                         function ($matches) use (&$codeblocks) {
2027                                 $return = $matches[0];
2028                                 if (strpos($matches[2], "\n") !== false) {
2029                                         $return = '#codeblock-' . count($codeblocks) . '#';
2030
2031                                         $prefix = '````' . $matches[1] . PHP_EOL;
2032                                         $codeblocks[] = $prefix . trim($matches[2]) . PHP_EOL . '````';
2033                                 }
2034                                 return $return;
2035                         },
2036                         $text
2037                 );
2038
2039                 // Convert it to HTML - don't try oembed
2040                 if ($for_diaspora) {
2041                         $text = self::convert($text, false, 3);
2042
2043                         // Add all tags that maybe were removed
2044                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2045                                 $tagline = "";
2046                                 foreach ($tags[2] as $tag) {
2047                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2048                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2049                                                 $tagline .= '#' . $tag . ' ';
2050                                         }
2051                                 }
2052                                 $text = $text . " " . $tagline;
2053                         }
2054                 } else {
2055                         $text = self::convert($text, false, 4);
2056                 }
2057
2058                 // mask some special HTML chars from conversation to markdown
2059                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
2060
2061                 // If a link is followed by a quote then there should be a newline before it
2062                 // Maybe we should make this newline at every time before a quote.
2063                 $text = str_replace(["</a><blockquote>"], ["</a><br><blockquote>"], $text);
2064
2065                 $stamp1 = microtime(true);
2066
2067                 // Now convert HTML to Markdown
2068                 $converter = new HtmlConverter();
2069                 $text = $converter->convert($text);
2070
2071                 // unmask the special chars back to HTML
2072                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
2073
2074                 $a->save_timestamp($stamp1, "parser");
2075
2076                 // Libertree has a problem with escaped hashtags.
2077                 $text = str_replace(['\#'], ['#'], $text);
2078
2079                 // Remove any leading or trailing whitespace, as this will mess up
2080                 // the Diaspora signature verification and cause the item to disappear
2081                 $text = trim($text);
2082
2083                 if ($for_diaspora) {
2084                         $url_search_string = "^\[\]";
2085                         $text = preg_replace_callback(
2086                                 "/([@]\[(.*?)\])\(([$url_search_string]*?)\)/ism",
2087                                 ['self', 'bbCodeMention2DiasporaCallback'],
2088                                 $text
2089                         );
2090                 }
2091
2092                 // Restore code blocks
2093                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
2094                         function ($matches) use ($codeblocks) {
2095                                 $return = '';
2096                                 if (isset($codeblocks[intval($matches[1])])) {
2097                                         $return = $codeblocks[$matches[1]];
2098                                 }
2099                                 return $return;
2100                         },
2101                         $text
2102                 );
2103
2104                 Addon::callHooks('bb2diaspora', $text);
2105
2106                 return $text;
2107         }
2108 }