]> git.mxchange.org Git - friendica.git/blob - src/Content/Text/BBCode.php
Replace include/event function with method calls
[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\Model\Event;
24 use Friendica\Network\Probe;
25 use Friendica\Object\Image;
26 use Friendica\Util\Map;
27 use Friendica\Util\Network;
28 use Friendica\Util\ParseUrl;
29 use League\HTMLToMarkdown\HtmlConverter;
30
31 require_once "include/event.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                         $return = self::convertUrlForOStatus($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                                         // Sanitize the HTML by converting it to BBCode
713                                         $bbcode = HTML::toBBCode($data["description"]);
714                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
715                                 }
716                                 if ($data["type"] == "link") {
717                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
718                                 }
719
720                                 if ($simplehtml != 4) {
721                                         $return .= '</div>';
722                                 }
723                         }
724                 }
725
726                 return trim($data["text"] . ' ' . $return . ' ' . $data["after"]);
727         }
728
729         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
730         {
731                 $data = self::getAttachmentData($Text);
732
733                 if (!$data) {
734                         return $Text;
735                 } elseif ($nolink) {
736                         return $data["text"] . $data["after"];
737                 }
738
739                 $title = htmlentities($data["title"], ENT_QUOTES, 'UTF-8', false);
740                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
741                 if ($plaintext || (($title != "") && strstr($text, $title))) {
742                         $data["title"] = $data["url"];
743                 } elseif (($text != "") && strstr($title, $text)) {
744                         $data["text"] = $data["title"];
745                         $data["title"] = $data["url"];
746                 }
747
748                 if (($data["text"] == "") && ($data["title"] != "") && ($data["url"] == "")) {
749                         return $data["title"] . $data["after"];
750                 }
751
752                 // If the link already is included in the post, don't add it again
753                 if (($data["url"] != "") && strpos($data["text"], $data["url"])) {
754                         return $data["text"] . $data["after"];
755                 }
756
757                 $text = $data["text"];
758
759                 if (($data["url"] != "") && ($data["title"] != "")) {
760                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
761                 } elseif (($data["url"] != "")) {
762                         $text .= "\n[url]" . $data["url"] . "[/url]";
763                 }
764
765                 return $text . "\n" . $data["after"];
766         }
767
768         /**
769          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
770          *
771          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
772          * @param array $match Array with the matching values
773          * @return string reformatted link including HTML codes
774          */
775         private static function convertUrlForOStatusCallback($match)
776         {
777                 $url = $match[1];
778
779                 if (isset($match[2]) && ($match[1] != $match[2])) {
780                         return $match[0];
781                 }
782
783                 $parts = parse_url($url);
784                 if (!isset($parts['scheme'])) {
785                         return $match[0];
786                 }
787
788                 return self::convertUrlForOStatus($url);
789         }
790
791         /**
792          * @brief Converts [url] BBCodes in a format that looks fine on OStatus systems.
793          * @param string $url URL that is about to be reformatted
794          * @return string reformatted link including HTML codes
795          */
796         private static function convertUrlForOStatus($url)
797         {
798                 $parts = parse_url($url);
799                 $scheme = $parts['scheme'] . '://';
800                 $styled_url = str_replace($scheme, '', $url);
801
802                 if (strlen($styled_url) > 30) {
803                         $styled_url = substr($styled_url, 0, 30) . "…";
804                 }
805
806                 $html = '<a href="%s" target="_blank">%s</a>';
807
808                 return sprintf($html, $url, $styled_url);
809         }
810
811         /*
812          * [noparse][i]italic[/i][/noparse] turns into
813          * [noparse][ i ]italic[ /i ][/noparse],
814          * to hide them from parser.
815          */
816         private static function escapeNoparseCallback($match)
817         {
818                 $whole_match = $match[0];
819                 $captured = $match[1];
820                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
821                 $new_str = str_replace($captured, $spacefied, $whole_match);
822                 return $new_str;
823         }
824
825         /*
826          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
827          * now turns back and the [noparse] tags are trimed
828          * returning [i]italic[/i]
829          */
830         private static function unescapeNoparseCallback($match)
831         {
832                 $captured = $match[1];
833                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
834                 return $unspacefied;
835         }
836
837         /**
838          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
839          * occurrences
840          *
841          * @param string $text        Text to search
842          * @param string $name        Tag name
843          * @param int    $occurrences Number of first occurrences to skip
844          * @return boolean|array
845          */
846         public static function getTagPosition($text, $name, $occurrences = 0)
847         {
848                 if ($occurrences < 0) {
849                         $occurrences = 0;
850                 }
851
852                 $start_open = -1;
853                 for ($i = 0; $i <= $occurrences; $i++) {
854                         if ($start_open !== false) {
855                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
856                         }
857                 }
858
859                 if ($start_open === false) {
860                         return false;
861                 }
862
863                 $start_equal = strpos($text, '=', $start_open);
864                 $start_close = strpos($text, ']', $start_open);
865
866                 if ($start_close === false) {
867                         return false;
868                 }
869
870                 $start_close++;
871
872                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
873
874                 if ($end_open === false) {
875                         return false;
876                 }
877
878                 $res = [
879                         'start' => [
880                                 'open' => $start_open,
881                                 'close' => $start_close
882                         ],
883                         'end' => [
884                                 'open' => $end_open,
885                                 'close' => $end_open + strlen('[/' . $name . ']')
886                         ],
887                 ];
888
889                 if ($start_equal !== false) {
890                         $res['start']['equal'] = $start_equal + 1;
891                 }
892
893                 return $res;
894         }
895
896         /**
897          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
898          *
899          * @param type $pattern Preg pattern string
900          * @param type $replace Preg replace string
901          * @param type $name    BBCode tag name
902          * @param type $text    Text to search
903          * @return string
904          */
905         public static function pregReplaceInTag($pattern, $replace, $name, $text)
906         {
907                 $occurrences = 0;
908                 $pos = self::getTagPosition($text, $name, $occurrences);
909                 while ($pos !== false && $occurrences++ < 1000) {
910                         $start = substr($text, 0, $pos['start']['open']);
911                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
912                         $end = substr($text, $pos['end']['close']);
913                         if ($end === false) {
914                                 $end = '';
915                         }
916
917                         $subject = preg_replace($pattern, $replace, $subject);
918                         $text = $start . $subject . $end;
919
920                         $pos = self::getTagPosition($text, $name, $occurrences);
921                 }
922
923                 return $text;
924         }
925
926         private static function extractImagesFromItemBody($body)
927         {
928                 $saved_image = [];
929                 $orig_body = $body;
930                 $new_body = '';
931
932                 $cnt = 0;
933                 $img_start = strpos($orig_body, '[img');
934                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
935                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
936                 while (($img_st_close !== false) && ($img_end !== false)) {
937                         $img_st_close++; // make it point to AFTER the closing bracket
938                         $img_end += $img_start;
939
940                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
941                                 // This is an embedded image
942                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
943                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
944
945                                 $cnt++;
946                         } else {
947                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
948                         }
949
950                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
951
952                         if ($orig_body === false) {
953                                 // in case the body ends on a closing image tag
954                                 $orig_body = '';
955                         }
956
957                         $img_start = strpos($orig_body, '[img');
958                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
959                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
960                 }
961
962                 $new_body = $new_body . $orig_body;
963
964                 return ['body' => $new_body, 'images' => $saved_image];
965         }
966
967         private static function interpolateSavedImagesIntoItemBody($body, array $images)
968         {
969                 $newbody = $body;
970
971                 $cnt = 0;
972                 foreach ($images as $image) {
973                         // We're depending on the property of 'foreach' (specified on the PHP website) that
974                         // it loops over the array starting from the first element and going sequentially
975                         // to the last element
976                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
977                                 '<img src="' . proxy_url($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
978                         $cnt++;
979                 }
980
981                 return $newbody;
982         }
983
984         /**
985          * Processes [share] tags
986          *
987          * Note: Can produce a [bookmark] tag in the output
988          *
989          * @brief Processes [share] tags
990          * @param array    $share      preg_match_callback result array
991          * @param bool|int $simplehtml
992          * @return string
993          */
994         private static function convertShare($share, $simplehtml)
995         {
996                 $attributes = $share[2];
997
998                 $author = "";
999                 preg_match("/author='(.*?)'/ism", $attributes, $matches);
1000                 if (x($matches, 1)) {
1001                         $author = html_entity_decode($matches[1], ENT_QUOTES, 'UTF-8');
1002                 }
1003
1004                 preg_match('/author="(.*?)"/ism', $attributes, $matches);
1005                 if (x($matches, 1)) {
1006                         $author = $matches[1];
1007                 }
1008
1009                 $profile = "";
1010                 preg_match("/profile='(.*?)'/ism", $attributes, $matches);
1011                 if (x($matches, 1)) {
1012                         $profile = $matches[1];
1013                 }
1014
1015                 preg_match('/profile="(.*?)"/ism', $attributes, $matches);
1016                 if (x($matches, 1)) {
1017                         $profile = $matches[1];
1018                 }
1019
1020                 $avatar = "";
1021                 preg_match("/avatar='(.*?)'/ism", $attributes, $matches);
1022                 if (x($matches, 1)) {
1023                         $avatar = $matches[1];
1024                 }
1025
1026                 preg_match('/avatar="(.*?)"/ism', $attributes, $matches);
1027                 if (x($matches, 1)) {
1028                         $avatar = $matches[1];
1029                 }
1030
1031                 $link = "";
1032                 preg_match("/link='(.*?)'/ism", $attributes, $matches);
1033                 if (x($matches, 1)) {
1034                         $link = $matches[1];
1035                 }
1036
1037                 preg_match('/link="(.*?)"/ism', $attributes, $matches);
1038                 if (x($matches, 1)) {
1039                         $link = $matches[1];
1040                 }
1041
1042                 $posted = "";
1043
1044                 preg_match("/posted='(.*?)'/ism", $attributes, $matches);
1045                 if (x($matches, 1)) {
1046                         $posted = $matches[1];
1047                 }
1048
1049                 preg_match('/posted="(.*?)"/ism', $attributes, $matches);
1050                 if (x($matches, 1)) {
1051                         $posted = $matches[1];
1052                 }
1053
1054                 // We only call this so that a previously unknown contact can be added.
1055                 // This is important for the function "Model\Contact::getDetailsByURL()".
1056                 // This function then can fetch an entry from the contact table.
1057                 Contact::getIdForURL($profile, 0, true);
1058
1059                 $data = Contact::getDetailsByURL($profile);
1060
1061                 if (x($data, "name") && x($data, "addr")) {
1062                         $userid_compact = $data["name"] . " (" . $data["addr"] . ")";
1063                 } else {
1064                         $userid_compact = Protocol::getAddrFromProfileUrl($profile, $author);
1065                 }
1066
1067                 if (x($data, "addr")) {
1068                         $userid = $data["addr"];
1069                 } else {
1070                         $userid = Protocol::formatMention($profile, $author);
1071                 }
1072
1073                 if (x($data, "name")) {
1074                         $author = $data["name"];
1075                 }
1076
1077                 if (x($data, "micro")) {
1078                         $avatar = $data["micro"];
1079                 }
1080
1081                 $preshare = trim($share[1]);
1082                 if ($preshare != "") {
1083                         $preshare .= "<br />";
1084                 }
1085
1086                 switch ($simplehtml) {
1087                         case 1:
1088                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' <a href="' . $profile . '">' . $userid . "</a>: <br />»" . $share[3] . "«";
1089                                 break;
1090                         case 2:
1091                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1092                                 break;
1093                         case 3: // Diaspora
1094                                 $headline = '<b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . $userid . ':</b><br />';
1095
1096                                 $text = trim($share[1]);
1097
1098                                 if ($text != "") {
1099                                         $text .= "<hr />";
1100                                 }
1101
1102                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0) {
1103                                         $text .= '<br /><a href="' . $link . '">' . $link . '</a>';
1104                                 } else {
1105                                         $text .= $headline . '<blockquote>' . trim($share[3]) . "</blockquote><br />";
1106
1107                                         if ($link != "") {
1108                                                 $text .= '<br /><a href="' . $link . '">[l]</a>';
1109                                         }
1110                                 }
1111
1112                                 break;
1113                         case 4:
1114                                 $headline = '<br /><b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8');
1115                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $link, $userid, $posted);
1116                                 $headline .= ":</b><br />";
1117
1118                                 $text = trim($share[1]);
1119
1120                                 if ($text != "") {
1121                                         $text .= "<hr />";
1122                                 }
1123
1124                                 $text .= $headline . '<blockquote class="shared_content">' . trim($share[3]) . "</blockquote><br />";
1125
1126                                 break;
1127                         case 5:
1128                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1129                                 break;
1130                         case 6: // app.net
1131                                 $text = $preshare . "&gt;&gt; @" . $userid_compact . ": <br />" . $share[3];
1132                                 break;
1133                         case 7: // statusnet/GNU Social
1134                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . " @" . $userid_compact . ": " . $share[3];
1135                                 break;
1136                         case 8: // twitter
1137                                 $text = $preshare . "RT @" . $userid_compact . ": " . $share[3];
1138                                 break;
1139                         case 9: // Google+/Facebook
1140                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1141
1142                                 if ($link != "") {
1143                                         $text .= "<br /><br />" . $link;
1144                                 }
1145                                 break;
1146                         default:
1147                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1148                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($link)) {
1149                                         try {
1150                                                 $oembed = OEmbed::getHTML($link, $preshare);
1151                                         } catch (Exception $e) {
1152                                                 $oembed = sprintf('[bookmark=%s]%s[/bookmark]', $link, $preshare);
1153                                         }
1154
1155                                         $text = $preshare . $oembed;
1156                                 } else {
1157                                         $text = trim($share[1]) . "\n";
1158
1159                                         $avatar = proxy_url($avatar, false, PROXY_SIZE_THUMB);
1160
1161                                         $tpl = get_markup_template('shared_content.tpl');
1162                                         $text .= replace_macros($tpl, [
1163                                                 '$profile' => $profile,
1164                                                 '$avatar' => $avatar,
1165                                                 '$author' => $author,
1166                                                 '$link' => $link,
1167                                                 '$posted' => $posted,
1168                                                 '$content' => trim($share[3])
1169                                         ]);
1170                                 }
1171                                 break;
1172                 }
1173
1174                 return $text;
1175         }
1176
1177         private static function removePictureLinksCallback($match)
1178         {
1179                 $text = Cache::get($match[1]);
1180
1181                 if (is_null($text)) {
1182                         $a = self::getApp();
1183
1184                         $stamp1 = microtime(true);
1185
1186                         $ch = @curl_init($match[1]);
1187                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1188                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1189                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1190                         @curl_exec($ch);
1191                         $curl_info = @curl_getinfo($ch);
1192
1193                         $a->save_timestamp($stamp1, "network");
1194
1195                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1196                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1197                         } else {
1198                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1199
1200                                 // if its not a picture then look if its a page that contains a picture link
1201                                 $body = Network::fetchUrl($match[1]);
1202
1203                                 $doc = new DOMDocument();
1204                                 @$doc->loadHTML($body);
1205                                 $xpath = new DOMXPath($doc);
1206                                 $list = $xpath->query("//meta[@name]");
1207                                 foreach ($list as $node) {
1208                                         $attr = [];
1209
1210                                         if ($node->attributes->length) {
1211                                                 foreach ($node->attributes as $attribute) {
1212                                                         $attr[$attribute->name] = $attribute->value;
1213                                                 }
1214                                         }
1215
1216                                         if (strtolower($attr["name"]) == "twitter:image") {
1217                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1218                                         }
1219                                 }
1220                         }
1221                         Cache::set($match[1], $text);
1222                 }
1223
1224                 return $text;
1225         }
1226
1227         private static function expandLinksCallback($match)
1228         {
1229                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1230                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1231                 } else {
1232                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1233                 }
1234         }
1235
1236         private static function cleanPictureLinksCallback($match)
1237         {
1238                 $text = Cache::get($match[1]);
1239
1240                 if (is_null($text)) {
1241                         $a = self::getApp();
1242
1243                         $stamp1 = microtime(true);
1244
1245                         $ch = @curl_init($match[1]);
1246                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1247                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1248                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1249                         @curl_exec($ch);
1250                         $curl_info = @curl_getinfo($ch);
1251
1252                         $a->save_timestamp($stamp1, "network");
1253
1254                         // if its a link to a picture then embed this picture
1255                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1256                                 $text = "[img]" . $match[1] . "[/img]";
1257                         } else {
1258                                 $text = "[img]" . $match[2] . "[/img]";
1259
1260                                 // if its not a picture then look if its a page that contains a picture link
1261                                 $body = Network::fetchUrl($match[1]);
1262
1263                                 $doc = new DOMDocument();
1264                                 @$doc->loadHTML($body);
1265                                 $xpath = new DOMXPath($doc);
1266                                 $list = $xpath->query("//meta[@name]");
1267                                 foreach ($list as $node) {
1268                                         $attr = [];
1269                                         if ($node->attributes->length) {
1270                                                 foreach ($node->attributes as $attribute) {
1271                                                         $attr[$attribute->name] = $attribute->value;
1272                                                 }
1273                                         }
1274
1275                                         if (strtolower($attr["name"]) == "twitter:image") {
1276                                                 $text = "[img]" . $attr["content"] . "[/img]";
1277                                         }
1278                                 }
1279                         }
1280                         Cache::set($match[1], $text);
1281                 }
1282
1283                 return $text;
1284         }
1285
1286         public static function cleanPictureLinks($text)
1287         {
1288                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1289                 return $return;
1290         }
1291
1292         private static function textHighlightCallback($match)
1293         {
1294                 if (in_array(strtolower($match[1]),
1295                                 ['php', 'css', 'mysql', 'sql', 'abap', 'diff', 'html', 'perl', 'ruby',
1296                                 'vbscript', 'avrc', 'dtd', 'java', 'xml', 'cpp', 'python', 'javascript', 'js', 'sh'])
1297                 ) {
1298                         return text_highlight($match[2], strtolower($match[1]));
1299                 }
1300                 return $match[0];
1301         }
1302
1303         /**
1304          * @brief Converts a BBCode message to HTML message
1305          *
1306          * BBcode 2 HTML was written by WAY2WEB.net
1307          * extended to work with Mistpark/Friendica - Mike Macgirvin
1308          *
1309          * Simple HTML values meaning:
1310          * - 0: Friendica display
1311          * - 1: Unused
1312          * - 2: Used for Facebook, Google+, Windows Phone push, Friendica API
1313          * - 3: Used before converting to Markdown in bb2diaspora.php
1314          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1315          * - 5: Unused
1316          * - 6: Used for Appnet
1317          * - 7: Used for dfrn, OStatus
1318          * - 8: Used for WP backlink text setting
1319          *
1320          * @param string $text
1321          * @param bool   $try_oembed
1322          * @param int    $simple_html
1323          * @param bool   $for_plaintext
1324          * @return string
1325          */
1326         public static function convert($text, $try_oembed = true, $simple_html = false, $for_plaintext = false)
1327         {
1328                 $a = self::getApp();
1329
1330                 /*
1331                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1332                  *
1333                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1334                  * $match[1] = $url
1335                  * $match[2] = $title or absent
1336                  */
1337                 $try_oembed_callback = function ($match)
1338                 {
1339                         $url = $match[1];
1340                         $title = defaults($match, 2, null);
1341
1342                         try {
1343                                 $return = OEmbed::getHTML($url, $title);
1344                         } catch (Exception $ex) {
1345                                 $return = $match[0];
1346                         }
1347
1348                         return $return;
1349                 };
1350
1351                 // Hide all [noparse] contained bbtags by spacefying them
1352                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1353
1354                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1355                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1356                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1357
1358                 // Remove the abstract element. It is a non visible element.
1359                 $text = self::stripAbstract($text);
1360
1361                 // Move all spaces out of the tags
1362                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1363                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1364
1365                 // Extract the private images which use data urls since preg has issues with
1366                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1367                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1368
1369                 $extracted = self::extractImagesFromItemBody($text);
1370                 $text = $extracted['body'];
1371                 $saved_image = $extracted['images'];
1372
1373                 // If we find any event code, turn it into an event.
1374                 // After we're finished processing the bbcode we'll
1375                 // replace all of the event code with a reformatted version.
1376
1377                 $ev = Event::fromBBCode($text);
1378
1379                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1380                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1381
1382                 $text = str_replace("<", "&lt;", $text);
1383                 $text = str_replace(">", "&gt;", $text);
1384
1385                 // remove some newlines before the general conversion
1386                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1387                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1388
1389                 $text = preg_replace("/\n\[code\]/ism", "[code]", $text);
1390                 $text = preg_replace("/\[\/code\]\n/ism", "[/code]", $text);
1391
1392                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1393                 if (!$try_oembed) {
1394                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1395                 }
1396
1397                 // Check for [code] text here, before the linefeeds are messed with.
1398                 // The highlighter will unescape and re-escape the content.
1399                 if (strpos($text, '[code=') !== false) {
1400                         $text = preg_replace_callback("/\[code=(.*?)\](.*?)\[\/code\]/ism", 'self::textHighlightCallback', $text);
1401                 }
1402                 // Convert new line chars to html <br /> tags
1403
1404                 // nlbr seems to be hopelessly messed up
1405                 //      $Text = nl2br($Text);
1406
1407                 // We'll emulate it.
1408
1409                 $text = trim($text);
1410                 $text = str_replace("\r\n", "\n", $text);
1411
1412                 // removing multiplicated newlines
1413                 if (Config::get("system", "remove_multiplicated_lines")) {
1414                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1415                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1416                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1417                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1418                         do {
1419                                 $oldtext = $text;
1420                                 $text = str_replace($search, $replace, $text);
1421                         } while ($oldtext != $text);
1422                 }
1423
1424                 // Set up the parameters for a URL search string
1425                 $URLSearchString = "^\[\]";
1426                 // Set up the parameters for a MAIL search string
1427                 $MAILSearchString = $URLSearchString;
1428
1429                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1430                 if (!$for_plaintext) {
1431                         // Autolink feature (thanks to http://code.seebz.net/p/autolink-php/)
1432                         // Currently disabled, since the function is too greedy
1433                         // $autolink_regex = "`([^\]\=\"']|^)(https?\://[^\s<]+[^\s<\.\)])`ism";
1434                         $autolink_regex = "/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism";
1435                         $text = preg_replace($autolink_regex, '$1[url]$2[/url]', $text);
1436                         if ($simple_html == 7) {
1437                                 $text = preg_replace_callback("/\[url\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1438                                 $text = preg_replace_callback("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1439                         }
1440                 } else {
1441                         $text = preg_replace("(\[url\]([$URLSearchString]*)\[\/url\])ism", " $1 ", $text);
1442                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1443                 }
1444
1445
1446                 // Handle attached links or videos
1447                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1448
1449                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1450
1451                 // Remove all hashtag addresses
1452                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7])) {
1453                         $text = preg_replace("/([#@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $text);
1454                 } elseif ($simple_html == 3) {
1455                         // The ! is converted to @ since Diaspora only understands the @
1456                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1457                                 '@<a href="$2">$3</a>',
1458                                 $text);
1459                 } elseif ($simple_html == 7) {
1460                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1461                                 '$1<span class="vcard"><a href="$2" class="url" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1462                                 $text);
1463                 } elseif (!$simple_html) {
1464                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1465                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1466                                 $text);
1467                 }
1468
1469                 // Bookmarks in red - will be converted to bookmarks in friendica
1470                 $text = preg_replace("/#\^\[url\]([$URLSearchString]*)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1471                 $text = preg_replace("/#\^\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1472                 $text = preg_replace("/#\[url\=[$URLSearchString]*\]\^\[\/url\]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i",
1473                                         "[bookmark=$1]$2[/bookmark]", $text);
1474
1475                 if (in_array($simple_html, [2, 6, 7, 8, 9])) {
1476                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1477                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1478                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1479                 }
1480
1481                 if ($simple_html == 5) {
1482                         $text = preg_replace("/[^#@!]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1483                 }
1484
1485                 // Perform URL Search
1486                 if ($try_oembed) {
1487                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1488                 }
1489
1490                 if ($simple_html == 5) {
1491                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1492                 } else {
1493                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1494                 }
1495
1496                 // Handle Diaspora posts
1497                 $text = preg_replace_callback(
1498                         "&\[url=/posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1499                         function ($match) {
1500                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1501                         }, $text
1502                 );
1503
1504                 // Server independent link to posts and comments
1505                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1506                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1507                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1508
1509                 $text = preg_replace("/([#])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1510                                         '$1<a href="' . System::baseUrl() . '/search?tag=$3" class="tag" title="$3">$3</a>', $text);
1511
1512                 $text = preg_replace("/\[url\=([$URLSearchString]*)\]#(.*?)\[\/url\]/ism",
1513                                         '#<a href="' . System::baseUrl() . '/search?tag=$2" class="tag" title="$2">$2</a>', $text);
1514
1515                 $text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1516                 $text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1517                 //$Text = preg_replace("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $Text);
1518
1519                 // Red compatibility, though the link can't be authenticated on Friendica
1520                 $text = preg_replace("/\[zrl\=([$URLSearchString]*)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1521
1522
1523                 // we may need to restrict this further if it picks up too many strays
1524                 // link acct:user@host to a webfinger profile redirector
1525
1526                 $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);
1527
1528                 // Perform MAIL Search
1529                 $text = preg_replace("/\[mail\]([$MAILSearchString]*)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1530                 $text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1531
1532                 // leave open the posibility of [map=something]
1533                 // this is replaced in prepare_body() which has knowledge of the item location
1534
1535                 if (strpos($text, '[/map]') !== false) {
1536                         $text = preg_replace_callback(
1537                                 "/\[map\](.*?)\[\/map\]/ism",
1538                                 function ($match) use ($simple_html) {
1539                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1540                                 },
1541                                 $text
1542                         );
1543                 }
1544                 if (strpos($text, '[map=') !== false) {
1545                         $text = preg_replace_callback(
1546                                 "/\[map=(.*?)\]/ism",
1547                                 function ($match) use ($simple_html) {
1548                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1549                                 },
1550                                 $text
1551                         );
1552                 }
1553                 if (strpos($text, '[map]') !== false) {
1554                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1555                 }
1556
1557                 // Check for headers
1558                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1559                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1560                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1561                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1562                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1563                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1564
1565                 // Check for paragraph
1566                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1567
1568                 // Check for bold text
1569                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1570
1571                 // Check for Italics text
1572                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1573
1574                 // Check for Underline text
1575                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1576
1577                 // Check for strike-through text
1578                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<strike>$1</strike>', $text);
1579
1580                 // Check for over-line text
1581                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1582
1583                 // Check for colored text
1584                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1585
1586                 // Check for sized text
1587                 // [size=50] --> font-size: 50px (with the unit).
1588                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1589                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1590
1591                 // Check for centered text
1592                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1593
1594                 // Check for list text
1595                 $text = str_replace("[*]", "<li>", $text);
1596
1597                 // Check for style sheet commands
1598                 $text = preg_replace_callback(
1599                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1600                         function ($match) {
1601                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1602                         },
1603                         $text
1604                 );
1605
1606                 // Check for CSS classes
1607                 $text = preg_replace_callback(
1608                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1609                         function ($match) {
1610                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1611                         },
1612                         $text
1613                 );
1614
1615                 // handle nested lists
1616                 $endlessloop = 0;
1617
1618                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1619                            ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1620                            ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1621                            ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1622                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1623                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1624                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1625                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1626                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1627                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1628                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1629                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1630                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1631                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1632                 }
1633
1634                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1635                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1636                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1637                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1638
1639                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1640                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1641
1642                 $text = str_replace('[hr]', '<hr />', $text);
1643
1644                 // This is actually executed in prepare_body()
1645
1646                 $text = str_replace('[nosmile]', '', $text);
1647
1648                 // Check for font change text
1649                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1650
1651                 // Declare the format for [code] layout
1652
1653                 $CodeLayout = '<code>$1</code>';
1654                 // Check for [code] text
1655                 $text = preg_replace("/\[code\](.*?)\[\/code\]/ism", "$CodeLayout", $text);
1656
1657                 // Declare the format for [spoiler] layout
1658                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
1659
1660                 // Check for [spoiler] text
1661                 // handle nested quotes
1662                 $endlessloop = 0;
1663                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1664                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", "$SpoilerLayout", $text);
1665                 }
1666
1667                 // Check for [spoiler=Author] text
1668
1669                 $t_wrote = L10n::t('$1 wrote:');
1670
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",
1675                                                  "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1676                                                  $text);
1677                 }
1678
1679                 // Declare the format for [quote] layout
1680                 $QuoteLayout = '<blockquote>$1</blockquote>';
1681
1682                 // Check for [quote] text
1683                 // handle nested quotes
1684                 $endlessloop = 0;
1685                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1686                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1687                 }
1688
1689                 // Check for [quote=Author] text
1690
1691                 $t_wrote = L10n::t('$1 wrote:');
1692
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",
1697                                                  "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1698                                                  $text);
1699                 }
1700
1701
1702                 // [img=widthxheight]image source[/img]
1703                 $text = preg_replace_callback(
1704                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1705                         function ($matches) {
1706                                 if (strpos($matches[3], "data:image/") === 0) {
1707                                         return $matches[0];
1708                                 }
1709
1710                                 $matches[3] = proxy_url($matches[3]);
1711                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1712                         },
1713                         $text
1714                 );
1715
1716                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1717                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1718
1719                 // Images
1720                 // [img]pathtoimage[/img]
1721                 $text = preg_replace_callback(
1722                         "/\[img\](.*?)\[\/img\]/ism",
1723                         function ($matches) {
1724                                 if (strpos($matches[1], "data:image/") === 0) {
1725                                         return $matches[0];
1726                                 }
1727
1728                                 $matches[1] = proxy_url($matches[1]);
1729                                 return "[img]" . $matches[1] . "[/img]";
1730                         },
1731                         $text
1732                 );
1733
1734                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1735                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1736
1737                 // Shared content
1738                 $text = preg_replace_callback("/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
1739                         function ($match) use ($simple_html) {
1740                                 return self::convertShare($match, $simple_html);
1741                         }, $text);
1742
1743                 $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);
1744                 $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 />', $text);
1745                 //$Text = preg_replace("/\[crypt=(.*?)\](.*?)\[\/crypt\]/ism", '<br/><img src="' .System::baseUrl() . '/images/lock_icon.gif" alt="' . L10n::t('Encrypted content') . '" title="' . '$1' . ' ' . L10n::t('Encrypted content') . '" /><br />', $Text);
1746
1747                 // Try to Oembed
1748                 if ($try_oembed) {
1749                         $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);
1750                         $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);
1751
1752                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1753                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1754                 } else {
1755                         $text = preg_replace("/\[video\](.*?)\[\/video\]/",
1756                                                 '<a href="$1" target="_blank">$1</a>', $text);
1757                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/",
1758                                                 '<a href="$1" target="_blank">$1</a>', $text);
1759                 }
1760
1761                 // html5 video and audio
1762
1763
1764                 if ($try_oembed) {
1765                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1766                 } else {
1767                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1768                 }
1769
1770                 // Youtube extensions
1771                 if ($try_oembed) {
1772                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1773                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1774                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1775                 }
1776
1777                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1778                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1779                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1780
1781                 if ($try_oembed) {
1782                         $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);
1783                 } else {
1784                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1785                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1786                 }
1787
1788                 if ($try_oembed) {
1789                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1790                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1791                 }
1792
1793                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1794                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1795
1796                 if ($try_oembed) {
1797                         $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);
1798                 } else {
1799                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1800                                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1801                 }
1802
1803                 // oembed tag
1804                 $text = OEmbed::BBCode2HTML($text);
1805
1806                 // Avoid triple linefeeds through oembed
1807                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1808
1809                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1810                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1811                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1812                 // start which is always required). Allow desc with a missing summary for compatibility.
1813
1814                 if ((x($ev, 'desc') || x($ev, 'summary')) && x($ev, 'start')) {
1815                         $sub = Event::getHTML($ev, $simple_html);
1816
1817                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1818                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1819                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1820                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1821                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1822                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1823                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1824                 }
1825
1826                 // Replace non graphical smilies for external posts
1827                 if ($simple_html) {
1828                         $text = Smilies::replace($text, false, true);
1829                 }
1830
1831                 // Replace inline code blocks
1832                 $text = preg_replace_callback("|(?!<br[^>]*>)<code>([^<]*)</code>(?!<br[^>]*>)|ism",
1833                         function ($match) use ($simple_html) {
1834                                 $return = '<key>' . $match[1] . '</key>';
1835                                 // Use <code> for Diaspora inline code blocks
1836                                 if ($simple_html === 3) {
1837                                         $return = '<code>' . $match[1] . '</code>';
1838                                 }
1839                                 return $return;
1840                         }
1841                 , $text);
1842
1843                 // Unhide all [noparse] contained bbtags unspacefying them
1844                 // and triming the [noparse] tag.
1845
1846                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1847                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1848                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1849
1850
1851                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1852                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1853                 $text = preg_replace('/\&quot\;/', '"', $text);
1854
1855                 // fix any escaped ampersands that may have been converted into links
1856                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1857
1858                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1859                 $allowed_src_protocols = ['http', 'redir', 'cid'];
1860                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1861                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1862
1863                 // sanitize href attributes (only whitelisted protocols URLs)
1864                 // default value for backward compatibility
1865                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', ['ftp', 'mailto', 'gopher', 'cid']);
1866
1867                 // Always allowed protocol even if config isn't set or not including it
1868                 $allowed_link_protocols[] = 'http';
1869                 $allowed_link_protocols[] = 'redir/';
1870
1871                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1872                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1873
1874                 if ($saved_image) {
1875                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1876                 }
1877
1878                 // Clean up the HTML by loading and saving the HTML with the DOM.
1879                 // Bad structured html can break a whole page.
1880                 // For performance reasons do it only with ativated item cache or at export.
1881                 if (!$try_oembed || (get_itemcachepath() != "")) {
1882                         $doc = new DOMDocument();
1883                         $doc->preserveWhiteSpace = false;
1884
1885                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1886
1887                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1888                         $encoding = '<?xml encoding="UTF-8">';
1889                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1890                         $doc->encoding = 'UTF-8';
1891                         $text = $doc->saveHTML();
1892                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1893
1894                         $text = str_replace('<br></li>', '</li>', $text);
1895
1896                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1897                 }
1898
1899                 // Clean up some useless linebreaks in lists
1900                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1901                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1902                 //$Text = str_replace('</li><br />', '</li>', $Text);
1903                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1904                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1905
1906                 Addon::callHooks('bbcode', $text);
1907
1908                 return trim($text);
1909         }
1910
1911         /**
1912          * @brief Strips the "abstract" tag from the provided text
1913          *
1914          * @param string $text The text with BBCode
1915          * @return string The same text - but without "abstract" element
1916          */
1917         public static function stripAbstract($text)
1918         {
1919                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1920                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1921
1922                 return $text;
1923         }
1924
1925         /**
1926          * @brief Returns the value of the "abstract" element
1927          *
1928          * @param string $text The text that maybe contains the element
1929          * @param string $addon The addon for which the abstract is meant for
1930          * @return string The abstract
1931          */
1932         private static function getAbstract($text, $addon = "")
1933         {
1934                 $abstract = "";
1935                 $abstracts = [];
1936                 $addon = strtolower($addon);
1937
1938                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1939                         foreach ($results AS $result) {
1940                                 $abstracts[strtolower($result[1])] = $result[2];
1941                         }
1942                 }
1943
1944                 if (isset($abstracts[$addon])) {
1945                         $abstract = $abstracts[$addon];
1946                 }
1947
1948                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1949                         $abstract = $result[1];
1950                 }
1951
1952                 return $abstract;
1953         }
1954
1955         /**
1956          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1957          *
1958          * @param array $match Matching values for the callback
1959          * @return string Replaced mention
1960          */
1961         private static function bbCodeMention2DiasporaCallback($match)
1962         {
1963                 $contact = Contact::getDetailsByURL($match[3]);
1964
1965                 if (empty($contact['addr'])) {
1966                         $contact = Probe::uri($match[3]);
1967                 }
1968
1969                 if (empty($contact['addr'])) {
1970                         return $match[0];
1971                 }
1972
1973                 $mention = '@{' . $match[2] . '; ' . $contact['addr'] . '}';
1974                 return $mention;
1975         }
1976
1977         /**
1978          * @brief Converts a BBCode text into Markdown
1979          *
1980          * This function converts a BBCode item body to be sent to Markdown-enabled
1981          * systems like Diaspora and Libertree
1982          *
1983          * @param string $text
1984          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
1985          * @return string
1986          */
1987         public static function toMarkdown($text, $for_diaspora = true)
1988         {
1989                 $a = self::getApp();
1990
1991                 $original_text = $text;
1992
1993                 // Since Diaspora is creating a summary for links, this function removes them before posting
1994                 if ($for_diaspora) {
1995                         $text = self::removeShareInformation($text);
1996                 }
1997
1998                 /**
1999                  * Transform #tags, strip off the [url] and replace spaces with underscore
2000                  */
2001                 $url_search_string = "^\[\]";
2002                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2003                         function ($matches) {
2004                                 return '#' . str_replace(' ', '_', $matches[2]);
2005                         },
2006                         $text
2007                 );
2008
2009                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2010                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2011
2012                 // Extracting multi-line code blocks before the whitespace processing/code highlighter in self::convert()
2013                 $codeblocks = [];
2014
2015                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#is",
2016                         function ($matches) use (&$codeblocks) {
2017                                 $return = $matches[0];
2018                                 if (strpos($matches[2], "\n") !== false) {
2019                                         $return = '#codeblock-' . count($codeblocks) . '#';
2020
2021                                         $prefix = '````' . $matches[1] . PHP_EOL;
2022                                         $codeblocks[] = $prefix . trim($matches[2]) . PHP_EOL . '````';
2023                                 }
2024                                 return $return;
2025                         },
2026                         $text
2027                 );
2028
2029                 // Convert it to HTML - don't try oembed
2030                 if ($for_diaspora) {
2031                         $text = self::convert($text, false, 3);
2032
2033                         // Add all tags that maybe were removed
2034                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2035                                 $tagline = "";
2036                                 foreach ($tags[2] as $tag) {
2037                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2038                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2039                                                 $tagline .= '#' . $tag . ' ';
2040                                         }
2041                                 }
2042                                 $text = $text . " " . $tagline;
2043                         }
2044                 } else {
2045                         $text = self::convert($text, false, 4);
2046                 }
2047
2048                 // mask some special HTML chars from conversation to markdown
2049                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
2050
2051                 // If a link is followed by a quote then there should be a newline before it
2052                 // Maybe we should make this newline at every time before a quote.
2053                 $text = str_replace(["</a><blockquote>"], ["</a><br><blockquote>"], $text);
2054
2055                 $stamp1 = microtime(true);
2056
2057                 // Now convert HTML to Markdown
2058                 $converter = new HtmlConverter();
2059                 $text = $converter->convert($text);
2060
2061                 // unmask the special chars back to HTML
2062                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
2063
2064                 $a->save_timestamp($stamp1, "parser");
2065
2066                 // Libertree has a problem with escaped hashtags.
2067                 $text = str_replace(['\#'], ['#'], $text);
2068
2069                 // Remove any leading or trailing whitespace, as this will mess up
2070                 // the Diaspora signature verification and cause the item to disappear
2071                 $text = trim($text);
2072
2073                 if ($for_diaspora) {
2074                         $url_search_string = "^\[\]";
2075                         $text = preg_replace_callback(
2076                                 "/([@]\[(.*?)\])\(([$url_search_string]*?)\)/ism",
2077                                 ['self', 'bbCodeMention2DiasporaCallback'],
2078                                 $text
2079                         );
2080                 }
2081
2082                 // Restore code blocks
2083                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
2084                         function ($matches) use ($codeblocks) {
2085                                 $return = '';
2086                                 if (isset($codeblocks[intval($matches[1])])) {
2087                                         $return = $codeblocks[$matches[1]];
2088                                 }
2089                                 return $return;
2090                         },
2091                         $text
2092                 );
2093
2094                 Addon::callHooks('bb2diaspora', $text);
2095
2096                 return $text;
2097         }
2098 }