]> git.mxchange.org Git - friendica.git/blob - src/Content/Text/BBCode.php
Merge pull request #4758 from annando/fix-regexp
[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 "mod/proxy.php";
32
33 class BBCode extends BaseObject
34 {
35         /**
36          * @brief Fetches attachment data that were generated the old way
37          *
38          * @param string $body Message body
39          * @return array
40          * 'type' -> Message type ("link", "video", "photo")
41          * 'text' -> Text before the shared message
42          * 'after' -> Text after the shared message
43          * 'image' -> Preview image of the message
44          * 'url' -> Url to the attached message
45          * 'title' -> Title of the attachment
46          * 'description' -> Description of the attachment
47          */
48         private static function getOldAttachmentData($body)
49         {
50                 $post = [];
51
52                 // Simplify image codes
53                 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
54
55                 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
56                         foreach ($attached as $data) {
57                                 if (!in_array($data[1], ["type-link", "type-video", "type-photo"])) {
58                                         continue;
59                                 }
60
61                                 $post["type"] = substr($data[1], 5);
62
63                                 $pos = strpos($body, $data[0]);
64                                 if ($pos > 0) {
65                                         $post["text"] = trim(substr($body, 0, $pos));
66                                         $post["after"] = trim(substr($body, $pos + strlen($data[0])));
67                                 } else {
68                                         $post["text"] = trim(str_replace($data[0], "", $body));
69                                 }
70
71                                 $attacheddata = $data[2];
72
73                                 $URLSearchString = "^\[\]";
74
75                                 if (preg_match("/\[img\]([$URLSearchString]*)\[\/img\]/ism", $attacheddata, $matches)) {
76
77                                         $picturedata = Image::getInfoFromURL($matches[1]);
78
79                                         if ($picturedata) {
80                                                 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
81                                                         $post["image"] = $matches[1];
82                                                 } else {
83                                                         $post["preview"] = $matches[1];
84                                                 }
85                                         }
86                                 }
87
88                                 if (preg_match("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
89                                         $post["url"] = $matches[1];
90                                         $post["title"] = $matches[2];
91                                 }
92                                 if (($post["url"] == "") && (in_array($post["type"], ["link", "video"]))
93                                         && preg_match("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
94                                         $post["url"] = $matches[1];
95                                 }
96
97                                 // Search for description
98                                 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
99                                         $post["description"] = $matches[1];
100                                 }
101                         }
102                 }
103                 return $post;
104         }
105
106         /**
107          * @brief Fetches attachment data that were generated with the "attachment" element
108          *
109          * @param string $body Message body
110          * @return array
111          * 'type' -> Message type ("link", "video", "photo")
112          * 'text' -> Text before the shared message
113          * 'after' -> Text after the shared message
114          * 'image' -> Preview image of the message
115          * 'url' -> Url to the attached message
116          * 'title' -> Title of the attachment
117          * 'description' -> Description of the attachment
118          */
119         public static function getAttachmentData($body)
120         {
121                 $data = [];
122
123                 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
124                         return self::getOldAttachmentData($body);
125                 }
126
127                 $attributes = $match[2];
128
129                 $data["text"] = trim($match[1]);
130
131                 $type = "";
132                 preg_match("/type='(.*?)'/ism", $attributes, $matches);
133                 if (x($matches, 1)) {
134                         $type = strtolower($matches[1]);
135                 }
136
137                 preg_match('/type="(.*?)"/ism', $attributes, $matches);
138                 if (x($matches, 1)) {
139                         $type = strtolower($matches[1]);
140                 }
141
142                 if ($type == "") {
143                         return [];
144                 }
145
146                 if (!in_array($type, ["link", "audio", "photo", "video"])) {
147                         return [];
148                 }
149
150                 if ($type != "") {
151                         $data["type"] = $type;
152                 }
153
154                 $url = "";
155                 preg_match("/url='(.*?)'/ism", $attributes, $matches);
156                 if (x($matches, 1)) {
157                         $url = $matches[1];
158                 }
159
160                 preg_match('/url="(.*?)"/ism', $attributes, $matches);
161                 if (x($matches, 1)) {
162                         $url = $matches[1];
163                 }
164
165                 if ($url != "") {
166                         $data["url"] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
167                 }
168
169                 $title = "";
170                 preg_match("/title='(.*?)'/ism", $attributes, $matches);
171                 if (x($matches, 1)) {
172                         $title = $matches[1];
173                 }
174
175                 preg_match('/title="(.*?)"/ism', $attributes, $matches);
176                 if (x($matches, 1)) {
177                         $title = $matches[1];
178                 }
179
180                 if ($title != "") {
181                         $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
182                         $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
183                         $title = str_replace(["[", "]"], ["&#91;", "&#93;"], $title);
184                         $data["title"] = $title;
185                 }
186
187                 $image = "";
188                 preg_match("/image='(.*?)'/ism", $attributes, $matches);
189                 if (x($matches, 1)) {
190                         $image = $matches[1];
191                 }
192
193                 preg_match('/image="(.*?)"/ism', $attributes, $matches);
194                 if (x($matches, 1)) {
195                         $image = $matches[1];
196                 }
197
198                 if ($image != "") {
199                         $data["image"] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
200                 }
201
202                 $preview = "";
203                 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
204                 if (x($matches, 1)) {
205                         $preview = $matches[1];
206                 }
207
208                 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
209                 if (x($matches, 1)) {
210                         $preview = $matches[1];
211                 }
212
213                 if ($preview != "") {
214                         $data["preview"] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
215                 }
216
217                 $data["description"] = trim($match[3]);
218
219                 $data["after"] = trim($match[4]);
220
221                 return $data;
222         }
223
224         public static function getAttachedData($body, $item = [])
225         {
226                 /*
227                 - text:
228                 - type: link, video, photo
229                 - title:
230                 - url:
231                 - image:
232                 - description:
233                 - (thumbnail)
234                 */
235
236                 $has_title = !empty($item['title']);
237                 $plink = (!empty($item['plink']) ? $item['plink'] : '');
238                 $post = self::getAttachmentData($body);
239
240                 // if nothing is found, it maybe having an image.
241                 if (!isset($post["type"])) {
242                         // Simplify image codes
243                         $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
244
245                         $URLSearchString = "^\[\]";
246
247                         $body = preg_replace("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
248
249                         if (preg_match_all("(\[url=([$URLSearchString]*)\]\s*\[img\]([$URLSearchString]*)\[\/img\]\s*\[\/url\])ism", $body, $pictures, PREG_SET_ORDER)) {
250                                 if ((count($pictures) == 1) && !$has_title) {
251                                         // Checking, if the link goes to a picture
252                                         $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
253
254                                         // Workaround:
255                                         // Sometimes photo posts to the own album are not detected at the start.
256                                         // So we seem to cannot use the cache for these cases. That's strange.
257                                         if (($data["type"] != "photo") && strstr($pictures[0][1], "/photos/")) {
258                                                 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
259                                         }
260
261                                         if ($data["type"] == "photo") {
262                                                 $post["type"] = "photo";
263                                                 if (isset($data["images"][0])) {
264                                                         $post["image"] = $data["images"][0]["src"];
265                                                         $post["url"] = $data["url"];
266                                                 } else {
267                                                         $post["image"] = $data["url"];
268                                                 }
269
270                                                 $post["preview"] = $pictures[0][2];
271                                                 $post["text"] = str_replace($pictures[0][0], "", $body);
272                                         } else {
273                                                 $imgdata = Image::getInfoFromURL($pictures[0][1]);
274                                                 if ($imgdata && substr($imgdata["mime"], 0, 6) == "image/") {
275                                                         $post["type"] = "photo";
276                                                         $post["image"] = $pictures[0][1];
277                                                         $post["preview"] = $pictures[0][2];
278                                                         $post["text"] = str_replace($pictures[0][0], "", $body);
279                                                 }
280                                         }
281                                 } elseif (count($pictures) > 0) {
282                                         $post["type"] = "link";
283                                         $post["url"] = $plink;
284                                         $post["image"] = $pictures[0][2];
285                                         $post["text"] = $body;
286                                 }
287                         } elseif (preg_match_all("(\[img\]([$URLSearchString]*)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
288                                 if ((count($pictures) == 1) && !$has_title) {
289                                         $post["type"] = "photo";
290                                         $post["image"] = $pictures[0][1];
291                                         $post["text"] = str_replace($pictures[0][0], "", $body);
292                                 } elseif (count($pictures) > 0) {
293                                         $post["type"] = "link";
294                                         $post["url"] = $plink;
295                                         $post["image"] = $pictures[0][1];
296                                         $post["text"] = $body;
297                                 }
298                         }
299
300                         // Test for the external links
301                         preg_match_all("(\[url\]([$URLSearchString]*)\[\/url\])ism", $body, $links1, PREG_SET_ORDER);
302                         preg_match_all("(\[url\=([$URLSearchString]*)\].*?\[\/url\])ism", $body, $links2, PREG_SET_ORDER);
303
304                         $links = array_merge($links1, $links2);
305
306                         // If there is only a single one, then use it.
307                         // This should cover link posts via API.
308                         if ((count($links) == 1) && !isset($post["preview"]) && !$has_title) {
309                                 $post["type"] = "link";
310                                 $post["text"] = trim($body);
311                                 $post["url"] = $links[0][1];
312                         }
313
314                         // Now count the number of external media links
315                         preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $body, $links1, PREG_SET_ORDER);
316                         preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $body, $links2, PREG_SET_ORDER);
317                         preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $body, $links3, PREG_SET_ORDER);
318                         preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $body, $links4, PREG_SET_ORDER);
319
320                         // Add them to the other external links
321                         $links = array_merge($links, $links1, $links2, $links3, $links4);
322
323                         // Are there more than one?
324                         if (count($links) > 1) {
325                                 // The post will be the type "text", which means a blog post
326                                 unset($post["type"]);
327                                 $post["url"] = $plink;
328                         }
329
330                         if (!isset($post["type"])) {
331                                 $post["type"] = "text";
332                                 $post["text"] = trim($body);
333                         }
334                 } elseif (isset($post["url"]) && ($post["type"] == "video")) {
335                         $data = ParseUrl::getSiteinfoCached($post["url"], true);
336
337                         if (isset($data["images"][0])) {
338                                 $post["image"] = $data["images"][0]["src"];
339                         }
340                 }
341
342                 return $post;
343         }
344
345         /**
346          * @brief Convert a message into plaintext for connectors to other networks
347          *
348          * @param array $b The message array that is about to be posted
349          * @param int $limit The maximum number of characters when posting to that network
350          * @param bool $includedlinks Has an attached link to be included into the message?
351          * @param int $htmlmode This triggers the behaviour of the bbcode conversion
352          * @param string $target_network Name of the network where the post should go to.
353          *
354          * @return string The converted message
355          */
356         public static function toPlaintext($b, $limit = 0, $includedlinks = false, $htmlmode = 2, $target_network = "")
357         {
358                 // Remove the hash tags
359                 $URLSearchString = "^\[\]";
360                 $body = preg_replace("/([#@])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $b["body"]);
361
362                 // Add an URL element if the text contains a raw link
363                 $body = preg_replace("/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url]$2[/url]', $body);
364
365                 // Remove the abstract
366                 $body = self::stripAbstract($body);
367
368                 // At first look at data that is attached via "type-..." stuff
369                 // This will hopefully replaced with a dedicated bbcode later
370                 //$post = self::getAttachedData($b["body"]);
371                 $post = self::getAttachedData($body, $b);
372
373                 if (($b["title"] != "") && ($post["text"] != "")) {
374                         $post["text"] = trim($b["title"]."\n\n".$post["text"]);
375                 } elseif ($b["title"] != "") {
376                         $post["text"] = trim($b["title"]);
377                 }
378
379                 $abstract = "";
380
381                 // Fetch the abstract from the given target network
382                 if ($target_network != "") {
383                         $default_abstract = self::getAbstract($b["body"]);
384                         $abstract = self::getAbstract($b["body"], $target_network);
385
386                         // If we post to a network with no limit we only fetch
387                         // an abstract exactly for this network
388                         if (($limit == 0) && ($abstract == $default_abstract)) {
389                                 $abstract = "";
390                         }
391                 } else {// Try to guess the correct target network
392                         switch ($htmlmode) {
393                                 case 8:
394                                         $abstract = self::getAbstract($b["body"], NETWORK_TWITTER);
395                                         break;
396                                 case 7:
397                                         $abstract = self::getAbstract($b["body"], NETWORK_STATUSNET);
398                                         break;
399                                 case 6:
400                                         $abstract = self::getAbstract($b["body"], NETWORK_APPNET);
401                                         break;
402                                 default: // We don't know the exact target.
403                                         // We fetch an abstract since there is a posting limit.
404                                         if ($limit > 0) {
405                                                 $abstract = self::getAbstract($b["body"]);
406                                         }
407                         }
408                 }
409
410                 if ($abstract != "") {
411                         $post["text"] = $abstract;
412
413                         if ($post["type"] == "text") {
414                                 $post["type"] = "link";
415                                 $post["url"] = $b["plink"];
416                         }
417                 }
418
419                 $html = self::convert($post["text"].$post["after"], false, $htmlmode);
420                 $msg = HTML::toPlaintext($html, 0, true);
421                 $msg = trim(html_entity_decode($msg, ENT_QUOTES, 'UTF-8'));
422
423                 $link = "";
424                 if ($includedlinks) {
425                         if ($post["type"] == "link") {
426                                 $link = $post["url"];
427                         } elseif ($post["type"] == "text") {
428                                 $link = $post["url"];
429                         } elseif ($post["type"] == "video") {
430                                 $link = $post["url"];
431                         } elseif ($post["type"] == "photo") {
432                                 $link = $post["image"];
433                         }
434
435                         if (($msg == "") && isset($post["title"])) {
436                                 $msg = trim($post["title"]);
437                         }
438
439                         if (($msg == "") && isset($post["description"])) {
440                                 $msg = trim($post["description"]);
441                         }
442
443                         // If the link is already contained in the post, then it neeedn't to be added again
444                         // But: if the link is beyond the limit, then it has to be added.
445                         if (($link != "") && strstr($msg, $link)) {
446                                 $pos = strpos($msg, $link);
447
448                                 // Will the text be shortened in the link?
449                                 // Or is the link the last item in the post?
450                                 if (($limit > 0) && ($pos < $limit) && (($pos + 23 > $limit) || ($pos + strlen($link) == strlen($msg)))) {
451                                         $msg = trim(str_replace($link, "", $msg));
452                                 } elseif (($limit == 0) || ($pos < $limit)) {
453                                         // The limit has to be increased since it will be shortened - but not now
454                                         // Only do it with Twitter (htmlmode = 8)
455                                         if (($limit > 0) && (strlen($link) > 23) && ($htmlmode == 8)) {
456                                                 $limit = $limit - 23 + strlen($link);
457                                         }
458
459                                         $link = "";
460
461                                         if ($post["type"] == "text") {
462                                                 unset($post["url"]);
463                                         }
464                                 }
465                         }
466                 }
467
468                 if ($limit > 0) {
469                         // Reduce multiple spaces
470                         // When posted to a network with limited space, we try to gain space where possible
471                         while (strpos($msg, "  ") !== false) {
472                                 $msg = str_replace("  ", " ", $msg);
473                         }
474
475                         // Twitter is using its own limiter, so we always assume that shortened links will have this length
476                         if (iconv_strlen($link, "UTF-8") > 0) {
477                                 $limit = $limit - 23;
478                         }
479
480                         if (iconv_strlen($msg, "UTF-8") > $limit) {
481                                 if (($post["type"] == "text") && isset($post["url"])) {
482                                         $post["url"] = $b["plink"];
483                                 } elseif (!isset($post["url"])) {
484                                         $limit = $limit - 23;
485                                         $post["url"] = $b["plink"];
486                                 // Which purpose has this line? It is now uncommented, but left as a reminder
487                                 //} elseif (strpos($b["body"], "[share") !== false) {
488                                 //      $post["url"] = $b["plink"];
489                                 } elseif (PConfig::get($b["uid"], "system", "no_intelligent_shortening")) {
490                                         $post["url"] = $b["plink"];
491                                 }
492                                 $msg = Plaintext::shorten($msg, $limit);
493                         }
494                 }
495
496                 $post["text"] = trim($msg);
497
498                 return($post);
499         }
500
501         public static function scaleExternalImages($srctext, $include_link = true, $scale_replace = false)
502         {
503                 // Suppress "view full size"
504                 if (intval(Config::get('system', 'no_view_full_size'))) {
505                         $include_link = false;
506                 }
507
508                 // Picture addresses can contain special characters
509                 $s = htmlspecialchars_decode($srctext);
510
511                 $matches = null;
512                 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
513                 if ($c) {
514                         foreach ($matches as $mtch) {
515                                 logger('scale_external_image: ' . $mtch[1]);
516
517                                 $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3));
518                                 if (stristr($mtch[1], $hostname)) {
519                                         continue;
520                                 }
521
522                                 // $scale_replace, if passed, is an array of two elements. The
523                                 // first is the name of the full-size image. The second is the
524                                 // name of a remote, scaled-down version of the full size image.
525                                 // This allows Friendica to display the smaller remote image if
526                                 // one exists, while still linking to the full-size image
527                                 if ($scale_replace) {
528                                         $scaled = str_replace($scale_replace[0], $scale_replace[1], $mtch[1]);
529                                 } else {
530                                         $scaled = $mtch[1];
531                                 }
532                                 $i = Network::fetchUrl($scaled);
533                                 if (!$i) {
534                                         return $srctext;
535                                 }
536
537                                 // guess mimetype from headers or filename
538                                 $type = Image::guessType($mtch[1], true);
539
540                                 if ($i) {
541                                         $Image = new Image($i, $type);
542                                         if ($Image->isValid()) {
543                                                 $orig_width = $Image->getWidth();
544                                                 $orig_height = $Image->getHeight();
545
546                                                 if ($orig_width > 640 || $orig_height > 640) {
547                                                         $Image->scaleDown(640);
548                                                         $new_width = $Image->getWidth();
549                                                         $new_height = $Image->getHeight();
550                                                         logger('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], LOGGER_DEBUG);
551                                                         $s = str_replace(
552                                                                 $mtch[0],
553                                                                 '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]'
554                                                                 . "\n" . (($include_link)
555                                                                         ? '[url=' . $mtch[1] . ']' . L10n::t('view full size') . '[/url]' . "\n"
556                                                                         : ''),
557                                                                 $s
558                                                         );
559                                                         logger('scale_external_images: new string: ' . $s, LOGGER_DEBUG);
560                                                 }
561                                         }
562                                 }
563                         }
564                 }
565
566                 // replace the special char encoding
567                 $s = htmlspecialchars($s, ENT_NOQUOTES, 'UTF-8');
568                 return $s;
569         }
570
571         /**
572          * The purpose of this function is to apply system message length limits to
573          * imported messages without including any embedded photos in the length
574          *
575          * @brief Truncates imported message body string length to max_import_size
576          * @param string $body
577          * @return string
578          */
579         public static function limitBodySize($body)
580         {
581                 $maxlen = get_max_import_size();
582
583                 // If the length of the body, including the embedded images, is smaller
584                 // than the maximum, then don't waste time looking for the images
585                 if ($maxlen && (strlen($body) > $maxlen)) {
586
587                         logger('the total body length exceeds the limit', LOGGER_DEBUG);
588
589                         $orig_body = $body;
590                         $new_body = '';
591                         $textlen = 0;
592
593                         $img_start = strpos($orig_body, '[img');
594                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
595                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
596                         while (($img_st_close !== false) && ($img_end !== false)) {
597
598                                 $img_st_close++; // make it point to AFTER the closing bracket
599                                 $img_end += $img_start;
600                                 $img_end += strlen('[/img]');
601
602                                 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
603                                         // This is an embedded image
604
605                                         if (($textlen + $img_start) > $maxlen) {
606                                                 if ($textlen < $maxlen) {
607                                                         logger('the limit happens before an embedded image', LOGGER_DEBUG);
608                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
609                                                         $textlen = $maxlen;
610                                                 }
611                                         } else {
612                                                 $new_body = $new_body . substr($orig_body, 0, $img_start);
613                                                 $textlen += $img_start;
614                                         }
615
616                                         $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
617                                 } else {
618
619                                         if (($textlen + $img_end) > $maxlen) {
620                                                 if ($textlen < $maxlen) {
621                                                         logger('the limit happens before the end of a non-embedded image', LOGGER_DEBUG);
622                                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
623                                                         $textlen = $maxlen;
624                                                 }
625                                         } else {
626                                                 $new_body = $new_body . substr($orig_body, 0, $img_end);
627                                                 $textlen += $img_end;
628                                         }
629                                 }
630                                 $orig_body = substr($orig_body, $img_end);
631
632                                 if ($orig_body === false) {
633                                         // in case the body ends on a closing image tag
634                                         $orig_body = '';
635                                 }
636
637                                 $img_start = strpos($orig_body, '[img');
638                                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
639                                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
640                         }
641
642                         if (($textlen + strlen($orig_body)) > $maxlen) {
643                                 if ($textlen < $maxlen) {
644                                         logger('the limit happens after the end of the last image', LOGGER_DEBUG);
645                                         $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
646                                 }
647                         } else {
648                                 logger('the text size with embedded images extracted did not violate the limit', LOGGER_DEBUG);
649                                 $new_body = $new_body . $orig_body;
650                         }
651
652                         return $new_body;
653                 } else {
654                         return $body;
655                 }
656         }
657
658         /**
659          * Processes [attachment] tags
660          *
661          * Note: Can produce a [bookmark] tag in the returned string
662          *
663          * @brief Processes [attachment] tags
664          * @param string $return
665          * @param bool|int $simplehtml
666          * @param bool $tryoembed
667          * @return string
668          */
669         private static function convertAttachment($return, $simplehtml = false, $tryoembed = true)
670         {
671                 $data = self::getAttachmentData($return);
672                 if (!$data) {
673                         return $return;
674                 }
675
676                 if (isset($data["title"])) {
677                         $data["title"] = strip_tags($data["title"]);
678                         $data["title"] = str_replace(["http://", "https://"], "", $data["title"]);
679                 }
680
681                 if (((strpos($data["text"], "[img=") !== false) || (strpos($data["text"], "[img]") !== false) || Config::get('system', 'always_show_preview')) && ($data["image"] != "")) {
682                         $data["preview"] = $data["image"];
683                         $data["image"] = "";
684                 }
685
686                 $return = '';
687                 if ($simplehtml == 7) {
688                         $return = self::convertUrlForOStatus($data["url"]);
689                 } elseif (($simplehtml != 4) && ($simplehtml != 0)) {
690                         $return = sprintf('<a href="%s" target="_blank">%s</a><br>', $data["url"], $data["title"]);
691                 } else {
692                         try {
693                                 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
694                                         $return = OEmbed::getHTML($data['url'], $data['title']);
695                                 } else {
696                                         throw new Exception('OEmbed is disabled for this attachment.');
697                                 }
698                         } catch (Exception $e) {
699                                 if ($simplehtml != 4) {
700                                         $return = sprintf('<div class="type-%s">', $data["type"]);
701                                 }
702
703                                 if ($data["image"] != "") {
704                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-image" /></a><br />', $data["url"], proxy_url($data["image"]), $data["title"]);
705                                 } elseif ($data["preview"] != "") {
706                                         $return .= sprintf('<a href="%s" target="_blank"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br />', $data["url"], proxy_url($data["preview"]), $data["title"]);
707                                 }
708
709                                 if (($data["type"] == "photo") && ($data["url"] != "") && ($data["image"] != "")) {
710                                         $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"]);
711                                 } else {
712                                         $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
713                                 }
714
715                                 if ($data["description"] != "" && $data["description"] != $data["title"]) {
716                                         // Sanitize the HTML by converting it to BBCode
717                                         $bbcode = HTML::toBBCode($data["description"]);
718                                         $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
719                                 }
720                                 if ($data["type"] == "link") {
721                                         $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
722                                 }
723
724                                 if ($simplehtml != 4) {
725                                         $return .= '</div>';
726                                 }
727                         }
728                 }
729
730                 return trim($data["text"] . ' ' . $return . ' ' . $data["after"]);
731         }
732
733         public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
734         {
735                 $data = self::getAttachmentData($Text);
736
737                 if (!$data) {
738                         return $Text;
739                 } elseif ($nolink) {
740                         return $data["text"] . $data["after"];
741                 }
742
743                 $title = htmlentities($data["title"], ENT_QUOTES, 'UTF-8', false);
744                 $text = htmlentities($data["text"], ENT_QUOTES, 'UTF-8', false);
745                 if ($plaintext || (($title != "") && strstr($text, $title))) {
746                         $data["title"] = $data["url"];
747                 } elseif (($text != "") && strstr($title, $text)) {
748                         $data["text"] = $data["title"];
749                         $data["title"] = $data["url"];
750                 }
751
752                 if (($data["text"] == "") && ($data["title"] != "") && ($data["url"] == "")) {
753                         return $data["title"] . $data["after"];
754                 }
755
756                 // If the link already is included in the post, don't add it again
757                 if (($data["url"] != "") && strpos($data["text"], $data["url"])) {
758                         return $data["text"] . $data["after"];
759                 }
760
761                 $text = $data["text"];
762
763                 if (($data["url"] != "") && ($data["title"] != "")) {
764                         $text .= "\n[url=" . $data["url"] . "]" . $data["title"] . "[/url]";
765                 } elseif (($data["url"] != "")) {
766                         $text .= "\n[url]" . $data["url"] . "[/url]";
767                 }
768
769                 return $text . "\n" . $data["after"];
770         }
771
772         /**
773          * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
774          *
775          * @brief Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
776          * @param array $match Array with the matching values
777          * @return string reformatted link including HTML codes
778          */
779         private static function convertUrlForOStatusCallback($match)
780         {
781                 $url = $match[1];
782
783                 if (isset($match[2]) && ($match[1] != $match[2])) {
784                         return $match[0];
785                 }
786
787                 $parts = parse_url($url);
788                 if (!isset($parts['scheme'])) {
789                         return $match[0];
790                 }
791
792                 return self::convertUrlForOStatus($url);
793         }
794
795         /**
796          * @brief Converts [url] BBCodes in a format that looks fine on OStatus systems.
797          * @param string $url URL that is about to be reformatted
798          * @return string reformatted link including HTML codes
799          */
800         private static function convertUrlForOStatus($url)
801         {
802                 $parts = parse_url($url);
803                 $scheme = $parts['scheme'] . '://';
804                 $styled_url = str_replace($scheme, '', $url);
805
806                 if (strlen($styled_url) > 30) {
807                         $styled_url = substr($styled_url, 0, 30) . "…";
808                 }
809
810                 $html = '<a href="%s" target="_blank">%s</a>';
811
812                 return sprintf($html, $url, $styled_url);
813         }
814
815         /*
816          * [noparse][i]italic[/i][/noparse] turns into
817          * [noparse][ i ]italic[ /i ][/noparse],
818          * to hide them from parser.
819          */
820         private static function escapeNoparseCallback($match)
821         {
822                 $whole_match = $match[0];
823                 $captured = $match[1];
824                 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
825                 $new_str = str_replace($captured, $spacefied, $whole_match);
826                 return $new_str;
827         }
828
829         /*
830          * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
831          * now turns back and the [noparse] tags are trimed
832          * returning [i]italic[/i]
833          */
834         private static function unescapeNoparseCallback($match)
835         {
836                 $captured = $match[1];
837                 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
838                 return $unspacefied;
839         }
840
841         /**
842          * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
843          * occurrences
844          *
845          * @param string $text        Text to search
846          * @param string $name        Tag name
847          * @param int    $occurrences Number of first occurrences to skip
848          * @return boolean|array
849          */
850         public static function getTagPosition($text, $name, $occurrences = 0)
851         {
852                 if ($occurrences < 0) {
853                         $occurrences = 0;
854                 }
855
856                 $start_open = -1;
857                 for ($i = 0; $i <= $occurrences; $i++) {
858                         if ($start_open !== false) {
859                                 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
860                         }
861                 }
862
863                 if ($start_open === false) {
864                         return false;
865                 }
866
867                 $start_equal = strpos($text, '=', $start_open);
868                 $start_close = strpos($text, ']', $start_open);
869
870                 if ($start_close === false) {
871                         return false;
872                 }
873
874                 $start_close++;
875
876                 $end_open = strpos($text, '[/' . $name . ']', $start_close);
877
878                 if ($end_open === false) {
879                         return false;
880                 }
881
882                 $res = [
883                         'start' => [
884                                 'open' => $start_open,
885                                 'close' => $start_close
886                         ],
887                         'end' => [
888                                 'open' => $end_open,
889                                 'close' => $end_open + strlen('[/' . $name . ']')
890                         ],
891                 ];
892
893                 if ($start_equal !== false) {
894                         $res['start']['equal'] = $start_equal + 1;
895                 }
896
897                 return $res;
898         }
899
900         /**
901          * Performs a preg_replace within the boundaries of all named BBCode tags in a text
902          *
903          * @param type $pattern Preg pattern string
904          * @param type $replace Preg replace string
905          * @param type $name    BBCode tag name
906          * @param type $text    Text to search
907          * @return string
908          */
909         public static function pregReplaceInTag($pattern, $replace, $name, $text)
910         {
911                 $occurrences = 0;
912                 $pos = self::getTagPosition($text, $name, $occurrences);
913                 while ($pos !== false && $occurrences++ < 1000) {
914                         $start = substr($text, 0, $pos['start']['open']);
915                         $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
916                         $end = substr($text, $pos['end']['close']);
917                         if ($end === false) {
918                                 $end = '';
919                         }
920
921                         $subject = preg_replace($pattern, $replace, $subject);
922                         $text = $start . $subject . $end;
923
924                         $pos = self::getTagPosition($text, $name, $occurrences);
925                 }
926
927                 return $text;
928         }
929
930         private static function extractImagesFromItemBody($body)
931         {
932                 $saved_image = [];
933                 $orig_body = $body;
934                 $new_body = '';
935
936                 $cnt = 0;
937                 $img_start = strpos($orig_body, '[img');
938                 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
939                 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
940                 while (($img_st_close !== false) && ($img_end !== false)) {
941                         $img_st_close++; // make it point to AFTER the closing bracket
942                         $img_end += $img_start;
943
944                         if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
945                                 // This is an embedded image
946                                 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
947                                 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
948
949                                 $cnt++;
950                         } else {
951                                 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
952                         }
953
954                         $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
955
956                         if ($orig_body === false) {
957                                 // in case the body ends on a closing image tag
958                                 $orig_body = '';
959                         }
960
961                         $img_start = strpos($orig_body, '[img');
962                         $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
963                         $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
964                 }
965
966                 $new_body = $new_body . $orig_body;
967
968                 return ['body' => $new_body, 'images' => $saved_image];
969         }
970
971         private static function interpolateSavedImagesIntoItemBody($body, array $images)
972         {
973                 $newbody = $body;
974
975                 $cnt = 0;
976                 foreach ($images as $image) {
977                         // We're depending on the property of 'foreach' (specified on the PHP website) that
978                         // it loops over the array starting from the first element and going sequentially
979                         // to the last element
980                         $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
981                                 '<img src="' . proxy_url($image) . '" alt="' . L10n::t('Image/photo') . '" />', $newbody);
982                         $cnt++;
983                 }
984
985                 return $newbody;
986         }
987
988         /**
989          * Processes [share] tags
990          *
991          * Note: Can produce a [bookmark] tag in the output
992          *
993          * @brief Processes [share] tags
994          * @param array    $share      preg_match_callback result array
995          * @param bool|int $simplehtml
996          * @return string
997          */
998         private static function convertShare($share, $simplehtml)
999         {
1000                 $attributes = $share[2];
1001
1002                 $author = "";
1003                 preg_match("/author='(.*?)'/ism", $attributes, $matches);
1004                 if (x($matches, 1)) {
1005                         $author = html_entity_decode($matches[1], ENT_QUOTES, 'UTF-8');
1006                 }
1007
1008                 preg_match('/author="(.*?)"/ism', $attributes, $matches);
1009                 if (x($matches, 1)) {
1010                         $author = $matches[1];
1011                 }
1012
1013                 $profile = "";
1014                 preg_match("/profile='(.*?)'/ism", $attributes, $matches);
1015                 if (x($matches, 1)) {
1016                         $profile = $matches[1];
1017                 }
1018
1019                 preg_match('/profile="(.*?)"/ism', $attributes, $matches);
1020                 if (x($matches, 1)) {
1021                         $profile = $matches[1];
1022                 }
1023
1024                 $avatar = "";
1025                 preg_match("/avatar='(.*?)'/ism", $attributes, $matches);
1026                 if (x($matches, 1)) {
1027                         $avatar = $matches[1];
1028                 }
1029
1030                 preg_match('/avatar="(.*?)"/ism', $attributes, $matches);
1031                 if (x($matches, 1)) {
1032                         $avatar = $matches[1];
1033                 }
1034
1035                 $link = "";
1036                 preg_match("/link='(.*?)'/ism", $attributes, $matches);
1037                 if (x($matches, 1)) {
1038                         $link = $matches[1];
1039                 }
1040
1041                 preg_match('/link="(.*?)"/ism', $attributes, $matches);
1042                 if (x($matches, 1)) {
1043                         $link = $matches[1];
1044                 }
1045
1046                 $posted = "";
1047
1048                 preg_match("/posted='(.*?)'/ism", $attributes, $matches);
1049                 if (x($matches, 1)) {
1050                         $posted = $matches[1];
1051                 }
1052
1053                 preg_match('/posted="(.*?)"/ism', $attributes, $matches);
1054                 if (x($matches, 1)) {
1055                         $posted = $matches[1];
1056                 }
1057
1058                 // We only call this so that a previously unknown contact can be added.
1059                 // This is important for the function "Model\Contact::getDetailsByURL()".
1060                 // This function then can fetch an entry from the contact table.
1061                 Contact::getIdForURL($profile, 0, true);
1062
1063                 $data = Contact::getDetailsByURL($profile);
1064
1065                 if (x($data, "name") && x($data, "addr")) {
1066                         $userid_compact = $data["name"] . " (" . $data["addr"] . ")";
1067                 } else {
1068                         $userid_compact = Protocol::getAddrFromProfileUrl($profile, $author);
1069                 }
1070
1071                 if (x($data, "addr")) {
1072                         $userid = $data["addr"];
1073                 } else {
1074                         $userid = Protocol::formatMention($profile, $author);
1075                 }
1076
1077                 if (x($data, "name")) {
1078                         $author = $data["name"];
1079                 }
1080
1081                 if (x($data, "micro")) {
1082                         $avatar = $data["micro"];
1083                 }
1084
1085                 $preshare = trim($share[1]);
1086                 if ($preshare != "") {
1087                         $preshare .= "<br />";
1088                 }
1089
1090                 switch ($simplehtml) {
1091                         case 1:
1092                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' <a href="' . $profile . '">' . $userid . "</a>: <br />»" . $share[3] . "«";
1093                                 break;
1094                         case 2:
1095                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1096                                 break;
1097                         case 3: // Diaspora
1098                                 $headline = '<b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . $userid . ':</b><br />';
1099
1100                                 $text = trim($share[1]);
1101
1102                                 if ($text != "") {
1103                                         $text .= "<hr />";
1104                                 }
1105
1106                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0) {
1107                                         $text .= '<br /><a href="' . $link . '">' . $link . '</a>';
1108                                 } else {
1109                                         $text .= $headline . '<blockquote>' . trim($share[3]) . "</blockquote><br />";
1110
1111                                         if ($link != "") {
1112                                                 $text .= '<br /><a href="' . $link . '">[l]</a>';
1113                                         }
1114                                 }
1115
1116                                 break;
1117                         case 4:
1118                                 $headline = '<br /><b>' . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8');
1119                                 $headline .= L10n::t('<a href="%1$s" target="_blank">%2$s</a> %3$s', $link, $userid, $posted);
1120                                 $headline .= ":</b><br />";
1121
1122                                 $text = trim($share[1]);
1123
1124                                 if ($text != "") {
1125                                         $text .= "<hr />";
1126                                 }
1127
1128                                 $text .= $headline . '<blockquote class="shared_content">' . trim($share[3]) . "</blockquote><br />";
1129
1130                                 break;
1131                         case 5:
1132                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1133                                 break;
1134                         case 6: // app.net
1135                                 $text = $preshare . "&gt;&gt; @" . $userid_compact . ": <br />" . $share[3];
1136                                 break;
1137                         case 7: // statusnet/GNU Social
1138                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . " @" . $userid_compact . ": " . $share[3];
1139                                 break;
1140                         case 8: // twitter
1141                                 $text = $preshare . "RT @" . $userid_compact . ": " . $share[3];
1142                                 break;
1143                         case 9: // Google+/Facebook
1144                                 $text = $preshare . html_entity_decode("&#x2672; ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ": <br />" . $share[3];
1145
1146                                 if ($link != "") {
1147                                         $text .= "<br /><br />" . $link;
1148                                 }
1149                                 break;
1150                         default:
1151                                 // Transforms quoted tweets in rich attachments to avoid nested tweets
1152                                 if (stripos(normalise_link($link), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($link)) {
1153                                         try {
1154                                                 $oembed = OEmbed::getHTML($link, $preshare);
1155                                         } catch (Exception $e) {
1156                                                 $oembed = sprintf('[bookmark=%s]%s[/bookmark]', $link, $preshare);
1157                                         }
1158
1159                                         $text = $preshare . $oembed;
1160                                 } else {
1161                                         $text = trim($share[1]) . "\n";
1162
1163                                         $avatar = proxy_url($avatar, false, PROXY_SIZE_THUMB);
1164
1165                                         $tpl = get_markup_template('shared_content.tpl');
1166                                         $text .= replace_macros($tpl, [
1167                                                 '$profile' => $profile,
1168                                                 '$avatar' => $avatar,
1169                                                 '$author' => $author,
1170                                                 '$link' => $link,
1171                                                 '$posted' => $posted,
1172                                                 '$content' => trim($share[3])
1173                                         ]);
1174                                 }
1175                                 break;
1176                 }
1177
1178                 return $text;
1179         }
1180
1181         private static function removePictureLinksCallback($match)
1182         {
1183                 $text = Cache::get($match[1]);
1184
1185                 if (is_null($text)) {
1186                         $a = self::getApp();
1187
1188                         $stamp1 = microtime(true);
1189
1190                         $ch = @curl_init($match[1]);
1191                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1192                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1193                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1194                         @curl_exec($ch);
1195                         $curl_info = @curl_getinfo($ch);
1196
1197                         $a->save_timestamp($stamp1, "network");
1198
1199                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1200                                 $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]";
1201                         } else {
1202                                 $text = "[url=" . $match[2] . "]" . $match[2] . "[/url]";
1203
1204                                 // if its not a picture then look if its a page that contains a picture link
1205                                 $body = Network::fetchUrl($match[1]);
1206
1207                                 $doc = new DOMDocument();
1208                                 @$doc->loadHTML($body);
1209                                 $xpath = new DOMXPath($doc);
1210                                 $list = $xpath->query("//meta[@name]");
1211                                 foreach ($list as $node) {
1212                                         $attr = [];
1213
1214                                         if ($node->attributes->length) {
1215                                                 foreach ($node->attributes as $attribute) {
1216                                                         $attr[$attribute->name] = $attribute->value;
1217                                                 }
1218                                         }
1219
1220                                         if (strtolower($attr["name"]) == "twitter:image") {
1221                                                 $text = "[url=" . $attr["content"] . "]" . $attr["content"] . "[/url]";
1222                                         }
1223                                 }
1224                         }
1225                         Cache::set($match[1], $text);
1226                 }
1227
1228                 return $text;
1229         }
1230
1231         private static function expandLinksCallback($match)
1232         {
1233                 if (($match[3] == "") || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1234                         return ($match[1] . "[url]" . $match[2] . "[/url]");
1235                 } else {
1236                         return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1237                 }
1238         }
1239
1240         private static function cleanPictureLinksCallback($match)
1241         {
1242                 $text = Cache::get($match[1]);
1243
1244                 if (is_null($text)) {
1245                         $a = self::getApp();
1246
1247                         $stamp1 = microtime(true);
1248
1249                         $ch = @curl_init($match[1]);
1250                         @curl_setopt($ch, CURLOPT_NOBODY, true);
1251                         @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1252                         @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent());
1253                         @curl_exec($ch);
1254                         $curl_info = @curl_getinfo($ch);
1255
1256                         $a->save_timestamp($stamp1, "network");
1257
1258                         // if its a link to a picture then embed this picture
1259                         if (substr($curl_info["content_type"], 0, 6) == "image/") {
1260                                 $text = "[img]" . $match[1] . "[/img]";
1261                         } else {
1262                                 $text = "[img]" . $match[2] . "[/img]";
1263
1264                                 // if its not a picture then look if its a page that contains a picture link
1265                                 $body = Network::fetchUrl($match[1]);
1266
1267                                 $doc = new DOMDocument();
1268                                 @$doc->loadHTML($body);
1269                                 $xpath = new DOMXPath($doc);
1270                                 $list = $xpath->query("//meta[@name]");
1271                                 foreach ($list as $node) {
1272                                         $attr = [];
1273                                         if ($node->attributes->length) {
1274                                                 foreach ($node->attributes as $attribute) {
1275                                                         $attr[$attribute->name] = $attribute->value;
1276                                                 }
1277                                         }
1278
1279                                         if (strtolower($attr["name"]) == "twitter:image") {
1280                                                 $text = "[img]" . $attr["content"] . "[/img]";
1281                                         }
1282                                 }
1283                         }
1284                         Cache::set($match[1], $text);
1285                 }
1286
1287                 return $text;
1288         }
1289
1290         public static function cleanPictureLinks($text)
1291         {
1292                 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1293                 return $return;
1294         }
1295
1296         private static function textHighlightCallback($match)
1297         {
1298                 // Fallback in case the language doesn't exist
1299                 $return = '[code]' . $match[2] . '[/code]';
1300
1301                 if (in_array(strtolower($match[1]),
1302                                 ['php', 'css', 'mysql', 'sql', 'abap', 'diff', 'html', 'perl', 'ruby',
1303                                 'vbscript', 'avrc', 'dtd', 'java', 'xml', 'cpp', 'python', 'javascript', 'js', 'sh', 'bash'])
1304                 ) {
1305                         $return = text_highlight($match[2], strtolower($match[1]));
1306                 }
1307
1308                 return $return;
1309         }
1310
1311         /**
1312          * @brief Converts a BBCode message to HTML message
1313          *
1314          * BBcode 2 HTML was written by WAY2WEB.net
1315          * extended to work with Mistpark/Friendica - Mike Macgirvin
1316          *
1317          * Simple HTML values meaning:
1318          * - 0: Friendica display
1319          * - 1: Unused
1320          * - 2: Used for Facebook, Google+, Windows Phone push, Friendica API
1321          * - 3: Used before converting to Markdown in bb2diaspora.php
1322          * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1323          * - 5: Unused
1324          * - 6: Used for Appnet
1325          * - 7: Used for dfrn, OStatus
1326          * - 8: Used for WP backlink text setting
1327          *
1328          * @param string $text
1329          * @param bool   $try_oembed
1330          * @param int    $simple_html
1331          * @param bool   $for_plaintext
1332          * @return string
1333          */
1334         public static function convert($text, $try_oembed = true, $simple_html = false, $for_plaintext = false)
1335         {
1336                 $a = self::getApp();
1337
1338                 /*
1339                  * preg_match_callback function to replace potential Oembed tags with Oembed content
1340                  *
1341                  * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1342                  * $match[1] = $url
1343                  * $match[2] = $title or absent
1344                  */
1345                 $try_oembed_callback = function ($match)
1346                 {
1347                         $url = $match[1];
1348                         $title = defaults($match, 2, null);
1349
1350                         try {
1351                                 $return = OEmbed::getHTML($url, $title);
1352                         } catch (Exception $ex) {
1353                                 $return = $match[0];
1354                         }
1355
1356                         return $return;
1357                 };
1358
1359                 // Hide all [noparse] contained bbtags by spacefying them
1360                 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1361
1362                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1363                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1364                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1365
1366                 // Remove the abstract element. It is a non visible element.
1367                 $text = self::stripAbstract($text);
1368
1369                 // Move all spaces out of the tags
1370                 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1371                 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1372
1373                 // Extract the private images which use data urls since preg has issues with
1374                 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1375                 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1376
1377                 $extracted = self::extractImagesFromItemBody($text);
1378                 $text = $extracted['body'];
1379                 $saved_image = $extracted['images'];
1380
1381                 // If we find any event code, turn it into an event.
1382                 // After we're finished processing the bbcode we'll
1383                 // replace all of the event code with a reformatted version.
1384
1385                 $ev = Event::fromBBCode($text);
1386
1387                 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1388                 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1389
1390                 $text = str_replace("<", "&lt;", $text);
1391                 $text = str_replace(">", "&gt;", $text);
1392
1393                 // remove some newlines before the general conversion
1394                 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1395                 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1396
1397                 $text = preg_replace("/\n\[code\]/ism", "[code]", $text);
1398                 $text = preg_replace("/\[\/code\]\n/ism", "[/code]", $text);
1399
1400                 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1401                 if (!$try_oembed) {
1402                         $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1403                 }
1404
1405                 // Check for [code] text here, before the linefeeds are messed with.
1406                 // The highlighter will unescape and re-escape the content.
1407                 if (strpos($text, '[code=') !== false) {
1408                         $text = preg_replace_callback("/\[code=(.*?)\](.*?)\[\/code\]/ism", 'self::textHighlightCallback', $text);
1409                 }
1410                 // Convert new line chars to html <br /> tags
1411
1412                 // nlbr seems to be hopelessly messed up
1413                 //      $Text = nl2br($Text);
1414
1415                 // We'll emulate it.
1416
1417                 $text = trim($text);
1418                 $text = str_replace("\r\n", "\n", $text);
1419
1420                 // removing multiplicated newlines
1421                 if (Config::get("system", "remove_multiplicated_lines")) {
1422                         $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1423                                         "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1424                         $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1425                                         "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1426                         do {
1427                                 $oldtext = $text;
1428                                 $text = str_replace($search, $replace, $text);
1429                         } while ($oldtext != $text);
1430                 }
1431
1432                 // Set up the parameters for a URL search string
1433                 $URLSearchString = "^\[\]";
1434                 // Set up the parameters for a MAIL search string
1435                 $MAILSearchString = $URLSearchString;
1436
1437                 // if the HTML is used to generate plain text, then don't do this search, but replace all URL of that kind to text
1438                 if (!$for_plaintext) {
1439                         // Autolink feature (thanks to http://code.seebz.net/p/autolink-php/)
1440                         // Currently disabled, since the function is too greedy
1441                         // $autolink_regex = "`([^\]\=\"']|^)(https?\://[^\s<]+[^\s<\.\)])`ism";
1442                         $autolink_regex = "/([^\]\='".'"'."]|^)(https?\:\/\/[a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism";
1443                         $text = preg_replace($autolink_regex, '$1[url]$2[/url]', $text);
1444                         if ($simple_html == 7) {
1445                                 $text = preg_replace_callback("/\[url\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1446                                 $text = preg_replace_callback("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", 'self::convertUrlForOStatusCallback', $text);
1447                         }
1448                 } else {
1449                         $text = preg_replace("(\[url\]([$URLSearchString]*)\[\/url\])ism", " $1 ", $text);
1450                         $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1451                 }
1452
1453
1454                 // Handle attached links or videos
1455                 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1456
1457                 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1458
1459                 // Remove all hashtag addresses
1460                 if ((!$try_oembed || $simple_html) && !in_array($simple_html, [3, 7])) {
1461                         $text = preg_replace("/([#@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$1$3', $text);
1462                 } elseif ($simple_html == 3) {
1463                         // The ! is converted to @ since Diaspora only understands the @
1464                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1465                                 '@<a href="$2">$3</a>',
1466                                 $text);
1467                 } elseif ($simple_html == 7) {
1468                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1469                                 '$1<span class="vcard"><a href="$2" class="url" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1470                                 $text);
1471                 } elseif (!$simple_html) {
1472                         $text = preg_replace("/([@!])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1473                                 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1474                                 $text);
1475                 }
1476
1477                 // Bookmarks in red - will be converted to bookmarks in friendica
1478                 $text = preg_replace("/#\^\[url\]([$URLSearchString]*)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1479                 $text = preg_replace("/#\^\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1480                 $text = preg_replace("/#\[url\=[$URLSearchString]*\]\^\[\/url\]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i",
1481                                         "[bookmark=$1]$2[/bookmark]", $text);
1482
1483                 if (in_array($simple_html, [2, 6, 7, 8, 9])) {
1484                         $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1485                         //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1486                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1487                 }
1488
1489                 if ($simple_html == 5) {
1490                         $text = preg_replace("/[^#@!]\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[url]$1[/url]', $text);
1491                 }
1492
1493                 // Perform URL Search
1494                 if ($try_oembed) {
1495                         $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1496                 }
1497
1498                 if ($simple_html == 5) {
1499                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url]$1[/url]', $text);
1500                 } else {
1501                         $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1502                 }
1503
1504                 // Handle Diaspora posts
1505                 $text = preg_replace_callback(
1506                         "&\[url=/posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1507                         function ($match) {
1508                                 return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1509                         }, $text
1510                 );
1511
1512                 // Server independent link to posts and comments
1513                 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1514                 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1515                 $text = preg_replace($expression, System::baseUrl()."/display/$1", $text);
1516
1517                 $text = preg_replace("/([#])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism",
1518                                         '$1<a href="' . System::baseUrl() . '/search?tag=$3" class="tag" title="$3">$3</a>', $text);
1519
1520                 $text = preg_replace("/\[url\=([$URLSearchString]*)\]#(.*?)\[\/url\]/ism",
1521                                         '#<a href="' . System::baseUrl() . '/search?tag=$2" class="tag" title="$2">$2</a>', $text);
1522
1523                 $text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$1</a>', $text);
1524                 $text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1525                 //$Text = preg_replace("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", '<a href="$1" target="_blank">$2</a>', $Text);
1526
1527                 // Red compatibility, though the link can't be authenticated on Friendica
1528                 $text = preg_replace("/\[zrl\=([$URLSearchString]*)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank">$2</a>', $text);
1529
1530
1531                 // we may need to restrict this further if it picks up too many strays
1532                 // link acct:user@host to a webfinger profile redirector
1533
1534                 $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);
1535
1536                 // Perform MAIL Search
1537                 $text = preg_replace("/\[mail\]([$MAILSearchString]*)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1538                 $text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1539
1540                 // leave open the posibility of [map=something]
1541                 // this is replaced in prepare_body() which has knowledge of the item location
1542
1543                 if (strpos($text, '[/map]') !== false) {
1544                         $text = preg_replace_callback(
1545                                 "/\[map\](.*?)\[\/map\]/ism",
1546                                 function ($match) use ($simple_html) {
1547                                         return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1548                                 },
1549                                 $text
1550                         );
1551                 }
1552                 if (strpos($text, '[map=') !== false) {
1553                         $text = preg_replace_callback(
1554                                 "/\[map=(.*?)\]/ism",
1555                                 function ($match) use ($simple_html) {
1556                                         return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1557                                 },
1558                                 $text
1559                         );
1560                 }
1561                 if (strpos($text, '[map]') !== false) {
1562                         $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1563                 }
1564
1565                 // Check for headers
1566                 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1567                 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1568                 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1569                 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1570                 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1571                 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1572
1573                 // Check for paragraph
1574                 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1575
1576                 // Check for bold text
1577                 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1578
1579                 // Check for Italics text
1580                 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1581
1582                 // Check for Underline text
1583                 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1584
1585                 // Check for strike-through text
1586                 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<strike>$1</strike>', $text);
1587
1588                 // Check for over-line text
1589                 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1590
1591                 // Check for colored text
1592                 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1593
1594                 // Check for sized text
1595                 // [size=50] --> font-size: 50px (with the unit).
1596                 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1597                 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1598
1599                 // Check for centered text
1600                 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1601
1602                 // Check for list text
1603                 $text = str_replace("[*]", "<li>", $text);
1604
1605                 // Check for style sheet commands
1606                 $text = preg_replace_callback(
1607                         "(\[style=(.*?)\](.*?)\[\/style\])ism",
1608                         function ($match) {
1609                                 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1610                         },
1611                         $text
1612                 );
1613
1614                 // Check for CSS classes
1615                 $text = preg_replace_callback(
1616                         "(\[class=(.*?)\](.*?)\[\/class\])ism",
1617                         function ($match) {
1618                                 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1619                         },
1620                         $text
1621                 );
1622
1623                 // handle nested lists
1624                 $endlessloop = 0;
1625
1626                 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1627                            ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1628                            ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1629                            ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1630                         $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1631                         $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1632                         $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1633                         $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1634                         $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1635                         $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1636                         $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1637                         $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1638                         $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1639                         $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1640                 }
1641
1642                 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1643                 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1644                 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1645                 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1646
1647                 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1648                 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1649
1650                 $text = str_replace('[hr]', '<hr />', $text);
1651
1652                 // This is actually executed in prepare_body()
1653
1654                 $text = str_replace('[nosmile]', '', $text);
1655
1656                 // Check for font change text
1657                 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1658
1659                 // Declare the format for [code] layout
1660
1661                 $CodeLayout = '<code>$1</code>';
1662                 // Check for [code] text
1663                 $text = preg_replace("/\[code\](.*?)\[\/code\]/ism", "$CodeLayout", $text);
1664
1665                 // Declare the format for [spoiler] layout
1666                 $SpoilerLayout = '<blockquote class="spoiler">$1</blockquote>';
1667
1668                 // Check for [spoiler] text
1669                 // handle nested quotes
1670                 $endlessloop = 0;
1671                 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1672                         $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", "$SpoilerLayout", $text);
1673                 }
1674
1675                 // Check for [spoiler=Author] text
1676
1677                 $t_wrote = L10n::t('$1 wrote:');
1678
1679                 // handle nested quotes
1680                 $endlessloop = 0;
1681                 while ((strpos($text, "[/spoiler]")!== false)  && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1682                         $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1683                                                  "<br /><strong class=".'"spoiler"'.">" . $t_wrote . "</strong><blockquote class=".'"spoiler"'.">$2</blockquote>",
1684                                                  $text);
1685                 }
1686
1687                 // Declare the format for [quote] layout
1688                 $QuoteLayout = '<blockquote>$1</blockquote>';
1689
1690                 // Check for [quote] text
1691                 // handle nested quotes
1692                 $endlessloop = 0;
1693                 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1694                         $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1695                 }
1696
1697                 // Check for [quote=Author] text
1698
1699                 $t_wrote = L10n::t('$1 wrote:');
1700
1701                 // handle nested quotes
1702                 $endlessloop = 0;
1703                 while ((strpos($text, "[/quote]")!== false)  && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1704                         $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1705                                                  "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1706                                                  $text);
1707                 }
1708
1709
1710                 // [img=widthxheight]image source[/img]
1711                 $text = preg_replace_callback(
1712                         "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1713                         function ($matches) {
1714                                 if (strpos($matches[3], "data:image/") === 0) {
1715                                         return $matches[0];
1716                                 }
1717
1718                                 $matches[3] = proxy_url($matches[3]);
1719                                 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1720                         },
1721                         $text
1722                 );
1723
1724                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1725                 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1726
1727                 $text = preg_replace_callback("/\[img\=([$URLSearchString]*)\](.*?)\[\/img\]/ism",
1728                         function ($matches) {
1729                                 $matches[1] = proxy_url($matches[1]);
1730                                 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1731                                 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '">';
1732                         },
1733                         $text);
1734
1735                 // Images
1736                 // [img]pathtoimage[/img]
1737                 $text = preg_replace_callback(
1738                         "/\[img\](.*?)\[\/img\]/ism",
1739                         function ($matches) {
1740                                 if (strpos($matches[1], "data:image/") === 0) {
1741                                         return $matches[0];
1742                                 }
1743
1744                                 $matches[1] = proxy_url($matches[1]);
1745                                 return "[img]" . $matches[1] . "[/img]";
1746                         },
1747                         $text
1748                 );
1749
1750                 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1751                 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . L10n::t('Image/photo') . '" />', $text);
1752
1753                 // Shared content
1754                 $text = preg_replace_callback("/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism",
1755                         function ($match) use ($simple_html) {
1756                                 return self::convertShare($match, $simple_html);
1757                         }, $text);
1758
1759                 $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);
1760                 $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 />', $text);
1761                 //$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 />', $Text);
1762
1763                 // Try to Oembed
1764                 if ($try_oembed) {
1765                         $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);
1766                         $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);
1767
1768                         $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1769                         $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1770                 } else {
1771                         $text = preg_replace("/\[video\](.*?)\[\/video\]/",
1772                                                 '<a href="$1" target="_blank">$1</a>', $text);
1773                         $text = preg_replace("/\[audio\](.*?)\[\/audio\]/",
1774                                                 '<a href="$1" target="_blank">$1</a>', $text);
1775                 }
1776
1777                 // html5 video and audio
1778
1779
1780                 if ($try_oembed) {
1781                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1782                 } else {
1783                         $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1784                 }
1785
1786                 // Youtube extensions
1787                 if ($try_oembed) {
1788                         $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1789                         $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1790                         $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1791                 }
1792
1793                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1794                 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1795                 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1796
1797                 if ($try_oembed) {
1798                         $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);
1799                 } else {
1800                         $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1801                                                 '<a href="https://www.youtube.com/watch?v=$1" target="_blank">https://www.youtube.com/watch?v=$1</a>', $text);
1802                 }
1803
1804                 if ($try_oembed) {
1805                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1806                         $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1807                 }
1808
1809                 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1810                 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1811
1812                 if ($try_oembed) {
1813                         $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);
1814                 } else {
1815                         $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1816                                                 '<a href="https://vimeo.com/$1" target="_blank">https://vimeo.com/$1</a>', $text);
1817                 }
1818
1819                 // oembed tag
1820                 $text = OEmbed::BBCode2HTML($text);
1821
1822                 // Avoid triple linefeeds through oembed
1823                 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1824
1825                 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1826                 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1827                 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1828                 // start which is always required). Allow desc with a missing summary for compatibility.
1829
1830                 if ((x($ev, 'desc') || x($ev, 'summary')) && x($ev, 'start')) {
1831                         $sub = Event::getHTML($ev, $simple_html);
1832
1833                         $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1834                         $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1835                         $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1836                         $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1837                         $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1838                         $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1839                         $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1840                 }
1841
1842                 // Replace non graphical smilies for external posts
1843                 if ($simple_html) {
1844                         $text = Smilies::replace($text, false, true);
1845                 }
1846
1847                 // Replace inline code blocks
1848                 $text = preg_replace_callback("|(?!<br[^>]*>)<code>([^<]*)</code>(?!<br[^>]*>)|ism",
1849                         function ($match) use ($simple_html) {
1850                                 $return = '<key>' . $match[1] . '</key>';
1851                                 // Use <code> for Diaspora inline code blocks
1852                                 if ($simple_html === 3) {
1853                                         $return = '<code>' . $match[1] . '</code>';
1854                                 }
1855                                 return $return;
1856                         }
1857                 , $text);
1858
1859                 // Unhide all [noparse] contained bbtags unspacefying them
1860                 // and triming the [noparse] tag.
1861
1862                 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1863                 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1864                 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1865
1866                 /// @todo What is the meaning of these lines?
1867                 $text = preg_replace('/\[\&amp\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1868                 $text = preg_replace('/\&\#039\;/', '\'', $text);
1869
1870                 // Currently deactivated, it made problems with " inside of alt texts.
1871                 //$text = preg_replace('/\&quot\;/', '"', $text);
1872
1873                 // fix any escaped ampersands that may have been converted into links
1874                 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&amp\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1875
1876                 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1877                 $allowed_src_protocols = ['http', 'redir', 'cid'];
1878                 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1879                                          '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . L10n::t('Invalid source protocol') . '">', $text);
1880
1881                 // sanitize href attributes (only whitelisted protocols URLs)
1882                 // default value for backward compatibility
1883                 $allowed_link_protocols = Config::get('system', 'allowed_link_protocols', ['ftp', 'mailto', 'gopher', 'cid']);
1884
1885                 // Always allowed protocol even if config isn't set or not including it
1886                 $allowed_link_protocols[] = 'http';
1887                 $allowed_link_protocols[] = 'redir/';
1888
1889                 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1890                 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . L10n::t('Invalid link protocol') . '">', $text);
1891
1892                 if ($saved_image) {
1893                         $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1894                 }
1895
1896                 // Clean up the HTML by loading and saving the HTML with the DOM.
1897                 // Bad structured html can break a whole page.
1898                 // For performance reasons do it only with ativated item cache or at export.
1899                 if (!$try_oembed || (get_itemcachepath() != "")) {
1900                         $doc = new DOMDocument();
1901                         $doc->preserveWhiteSpace = false;
1902
1903                         $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1904
1905                         $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1906                         $encoding = '<?xml encoding="UTF-8">';
1907                         @$doc->loadHTML($encoding.$doctype."<html><body>".$text."</body></html>");
1908                         $doc->encoding = 'UTF-8';
1909                         $text = $doc->saveHTML();
1910                         $text = str_replace(["<html><body>", "</body></html>", $doctype, $encoding], ["", "", "", ""], $text);
1911
1912                         $text = str_replace('<br></li>', '</li>', $text);
1913
1914                         //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1915                 }
1916
1917                 // Clean up some useless linebreaks in lists
1918                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1919                 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1920                 //$Text = str_replace('</li><br />', '</li>', $Text);
1921                 //$Text = str_replace('<br /><li>', '<li>', $Text);
1922                 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1923
1924                 Addon::callHooks('bbcode', $text);
1925
1926                 return trim($text);
1927         }
1928
1929         /**
1930          * @brief Strips the "abstract" tag from the provided text
1931          *
1932          * @param string $text The text with BBCode
1933          * @return string The same text - but without "abstract" element
1934          */
1935         public static function stripAbstract($text)
1936         {
1937                 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1938                 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1939
1940                 return $text;
1941         }
1942
1943         /**
1944          * @brief Returns the value of the "abstract" element
1945          *
1946          * @param string $text The text that maybe contains the element
1947          * @param string $addon The addon for which the abstract is meant for
1948          * @return string The abstract
1949          */
1950         private static function getAbstract($text, $addon = "")
1951         {
1952                 $abstract = "";
1953                 $abstracts = [];
1954                 $addon = strtolower($addon);
1955
1956                 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1957                         foreach ($results AS $result) {
1958                                 $abstracts[strtolower($result[1])] = $result[2];
1959                         }
1960                 }
1961
1962                 if (isset($abstracts[$addon])) {
1963                         $abstract = $abstracts[$addon];
1964                 }
1965
1966                 if ($abstract == "" && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1967                         $abstract = $result[1];
1968                 }
1969
1970                 return $abstract;
1971         }
1972
1973         /**
1974          * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
1975          *
1976          * @param array $match Matching values for the callback
1977          * @return string Replaced mention
1978          */
1979         private static function bbCodeMention2DiasporaCallback($match)
1980         {
1981                 $contact = Contact::getDetailsByURL($match[3]);
1982
1983                 if (empty($contact['addr'])) {
1984                         $contact = Probe::uri($match[3]);
1985                 }
1986
1987                 if (empty($contact['addr'])) {
1988                         return $match[0];
1989                 }
1990
1991                 $mention = '@{' . $match[2] . '; ' . $contact['addr'] . '}';
1992                 return $mention;
1993         }
1994
1995         /**
1996          * @brief Converts a BBCode text into Markdown
1997          *
1998          * This function converts a BBCode item body to be sent to Markdown-enabled
1999          * systems like Diaspora and Libertree
2000          *
2001          * @param string $text
2002          * @param bool   $for_diaspora Diaspora requires more changes than Libertree
2003          * @return string
2004          */
2005         public static function toMarkdown($text, $for_diaspora = true)
2006         {
2007                 $a = self::getApp();
2008
2009                 $original_text = $text;
2010
2011                 // Since Diaspora is creating a summary for links, this function removes them before posting
2012                 if ($for_diaspora) {
2013                         $text = self::removeShareInformation($text);
2014                 }
2015
2016                 /**
2017                  * Transform #tags, strip off the [url] and replace spaces with underscore
2018                  */
2019                 $url_search_string = "^\[\]";
2020                 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2021                         function ($matches) {
2022                                 return '#' . str_replace(' ', '_', $matches[2]);
2023                         },
2024                         $text
2025                 );
2026
2027                 // Converting images with size parameters to simple images. Markdown doesn't know it.
2028                 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2029
2030                 // Extracting multi-line code blocks before the whitespace processing/code highlighter in self::convert()
2031                 $codeblocks = [];
2032
2033                 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#is",
2034                         function ($matches) use (&$codeblocks) {
2035                                 $return = $matches[0];
2036                                 if (strpos($matches[2], "\n") !== false) {
2037                                         $return = '#codeblock-' . count($codeblocks) . '#';
2038
2039                                         $prefix = '````' . $matches[1] . PHP_EOL;
2040                                         $codeblocks[] = $prefix . trim($matches[2]) . PHP_EOL . '````';
2041                                 }
2042                                 return $return;
2043                         },
2044                         $text
2045                 );
2046
2047                 // Convert it to HTML - don't try oembed
2048                 if ($for_diaspora) {
2049                         $text = self::convert($text, false, 3);
2050
2051                         // Add all tags that maybe were removed
2052                         if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2053                                 $tagline = "";
2054                                 foreach ($tags[2] as $tag) {
2055                                         $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2056                                         if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2057                                                 $tagline .= '#' . $tag . ' ';
2058                                         }
2059                                 }
2060                                 $text = $text . " " . $tagline;
2061                         }
2062                 } else {
2063                         $text = self::convert($text, false, 4);
2064                 }
2065
2066                 // mask some special HTML chars from conversation to markdown
2067                 $text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_lt_;', '&_gt_;', '&_amp_;'], $text);
2068
2069                 // If a link is followed by a quote then there should be a newline before it
2070                 // Maybe we should make this newline at every time before a quote.
2071                 $text = str_replace(["</a><blockquote>"], ["</a><br><blockquote>"], $text);
2072
2073                 $stamp1 = microtime(true);
2074
2075                 // Now convert HTML to Markdown
2076                 $converter = new HtmlConverter();
2077                 $text = $converter->convert($text);
2078
2079                 // unmask the special chars back to HTML
2080                 $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $text);
2081
2082                 $a->save_timestamp($stamp1, "parser");
2083
2084                 // Libertree has a problem with escaped hashtags.
2085                 $text = str_replace(['\#'], ['#'], $text);
2086
2087                 // Remove any leading or trailing whitespace, as this will mess up
2088                 // the Diaspora signature verification and cause the item to disappear
2089                 $text = trim($text);
2090
2091                 if ($for_diaspora) {
2092                         $url_search_string = "^\[\]";
2093                         $text = preg_replace_callback(
2094                                 "/([@]\[(.*?)\])\(([$url_search_string]*?)\)/ism",
2095                                 ['self', 'bbCodeMention2DiasporaCallback'],
2096                                 $text
2097                         );
2098                 }
2099
2100                 // Restore code blocks
2101                 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
2102                         function ($matches) use ($codeblocks) {
2103                                 $return = '';
2104                                 if (isset($codeblocks[intval($matches[1])])) {
2105                                         $return = $codeblocks[$matches[1]];
2106                                 }
2107                                 return $return;
2108                         },
2109                         $text
2110                 );
2111
2112                 Addon::callHooks('bb2diaspora', $text);
2113
2114                 return $text;
2115         }
2116 }