3 * @copyright Copyright (C) 2010-2024, the Friendica project
5 * @license GNU AGPL version 3 or any later version
7 * This program is free software: you can redistribute it and/or modify
8 * it under the terms of the GNU Affero General Public License as
9 * published by the Free Software Foundation, either version 3 of the
10 * License, or (at your option) any later version.
12 * This program is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 * GNU Affero General Public License for more details.
17 * You should have received a copy of the GNU Affero General Public License
18 * along with this program. If not, see <https://www.gnu.org/licenses/>.
22 namespace Friendica\Content\Text;
27 use Friendica\Content\ContactSelector;
28 use Friendica\Content\Item;
29 use Friendica\Content\OEmbed;
30 use Friendica\Content\PageInfo;
31 use Friendica\Content\Smilies;
32 use Friendica\Core\Hook;
33 use Friendica\Core\Logger;
34 use Friendica\Core\Protocol;
35 use Friendica\Core\Renderer;
37 use Friendica\Model\Contact;
38 use Friendica\Model\Event;
39 use Friendica\Model\Post;
40 use Friendica\Model\Tag;
41 use Friendica\Network\HTTPClient\Client\HttpClientAccept;
42 use Friendica\Network\HTTPClient\Client\HttpClientOptions;
43 use Friendica\Util\Map;
44 use Friendica\Util\ParseUrl;
45 use Friendica\Util\Proxy;
46 use Friendica\Util\Strings;
47 use Friendica\Util\XML;
51 // Update this value to the current date whenever changes are made to BBCode::convert
52 const VERSION = '2021-07-28';
56 const MASTODON_API = 2;
59 const TWITTER_API = 5;
64 const ACTIVITYPUB = 9;
67 const SHARED_ANCHOR = '<hr class="shared-anchor">';
68 const TOP_ANCHOR = '<br class="top-anchor">';
69 const BOTTOM_ANCHOR = '<br class="button-anchor">';
71 const PREVIEW_NONE = 0;
72 const PREVIEW_NO_IMAGE = 1;
73 const PREVIEW_LARGE = 2;
74 const PREVIEW_SMALL = 3;
77 * Fetches attachment data that were generated with the "attachment" element
79 * @param string $body Message body
81 * 'type' -> Message type ('link', 'video', 'photo')
82 * 'text' -> Text before the shared message
83 * 'after' -> Text after the shared message
84 * 'image' -> Preview image of the message
85 * 'url' -> Url to the attached message
86 * 'title' -> Title of the attachment
87 * 'description' -> Description of the attachment
88 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
90 public static function getAttachmentData(string $body): array
92 DI::profiler()->startRecording('rendering');
101 'provider_name' => '',
102 'provider_url' => '',
107 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
108 DI::profiler()->stopRecording();
112 $attributes = $match[2];
114 $data['text'] = trim($match[1]);
116 foreach (['type', 'url', 'title', 'image', 'preview', 'publisher_name', 'publisher_url', 'author_name', 'author_url'] as $field) {
117 preg_match('/' . preg_quote($field, '/') . '=("|\')(.*?)\1/ism', $attributes, $matches);
118 $value = $matches[2] ?? '';
122 case 'publisher_name':
123 $data['provider_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
126 case 'publisher_url':
127 $data['provider_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
131 $data['author_name'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
132 if ($data['provider_name'] == $data['author_name']) {
133 $data['author_name'] = '';
138 $data['author_url'] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
139 if ($data['provider_url'] == $data['author_url']) {
140 $data['author_url'] = '';
145 $value = self::toPlaintext(html_entity_decode($value, ENT_QUOTES, 'UTF-8'));
146 $value = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
147 $value = str_replace(['[', ']'], ['[', ']'], $value);
148 $data['title'] = $value;
151 $data[$field] = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
157 if (!in_array($data['type'], ['link', 'audio', 'photo', 'video'])) {
158 DI::profiler()->stopRecording();
162 $data['description'] = trim($match[3]);
164 $data['after'] = trim($match[4]);
166 $parts = parse_url($data['url']);
167 if (!empty($parts['scheme']) && !empty($parts['host'])) {
168 if (empty($data['provider_name'])) {
169 $data['provider_name'] = $parts['host'];
171 if (empty($data['provider_url']) || empty(parse_url($data['provider_url'], PHP_URL_SCHEME))) {
172 $data['provider_url'] = $parts['scheme'] . '://' . $parts['host'];
174 if (!empty($parts['port'])) {
175 $data['provider_url'] .= ':' . $parts['port'];
180 DI::profiler()->stopRecording();
185 * Remove [attachment] BBCode and replaces it with a regular [url]
187 * @param string $body
188 * @param boolean $no_link_desc No link description
189 * @return string with replaced body
191 public static function replaceAttachment(string $body, bool $no_link_desc = false): string
193 return preg_replace_callback(
194 "/\s*\[attachment (.*?)\](.*?)\[\/attachment\]\s*/ism",
195 function ($match) use ($body, $no_link_desc) {
196 $attach_data = self::getAttachmentData($match[0]);
197 if (empty($attach_data['url'])) {
199 } elseif (strpos(str_replace($match[0], '', $body), $attach_data['url']) !== false) {
201 } elseif (empty($attach_data['title']) || $no_link_desc) {
202 return " \n[url]" . $attach_data['url'] . "[/url]\n";
204 return " \n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
212 * Remove [attachment] BBCode
214 * @param string $body
215 * @return string with removed attachment
217 public static function removeAttachment(string $body): string
219 return trim(preg_replace("/\s*\[attachment .*?\].*?\[\/attachment\]\s*/ism", '', $body));
223 * Converts a BBCode text into plaintext
225 * @param string $text
226 * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
229 public static function toPlaintext(string $text, bool $keep_urls = true): string
231 DI::profiler()->startRecording('rendering');
232 // Remove pictures in advance to avoid unneeded proxy calls
233 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", ' ', $text);
234 $text = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $2 ', $text);
235 $text = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $text);
238 $text = self::replaceAttachment($text);
240 $naked_text = HTML::toPlaintext(self::convert($text, false, self::EXTERNAL, true), 0, !$keep_urls);
242 DI::profiler()->stopRecording();
247 * Converts text into a format that can be used for the channel search and the language detection.
249 * @param string $text
250 * @param integer $uri_id
253 public static function toSearchText(string $text, int $uri_id): string
255 // Removes attachments
256 $text = self::removeAttachment($text);
258 // Add images because of possible alt texts
259 if (!empty($uri_id)) {
260 $text = Post\Media::addAttachmentsToBody($uri_id, $text, [Post\Media::IMAGE]);
267 // Remove links without a link description
268 $text = preg_replace("~\[url\=.*\]https?:.*\[\/url\]~", ' ', $text);
271 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", ' ', $text);
273 // Replace picture with the alt description
274 $text = preg_replace("/\[img\=.*?\](.*?)\[\/img\]/ism", ' $1 ', $text);
276 // Remove the other pictures
277 $text = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $text);
279 // Removes mentions, remove links from hashtags
280 $text = preg_replace('/[@!]\[url\=.*?\].*?\[\/url\]/ism', ' ', $text);
281 $text = preg_replace('/[#]\[url\=.*?\](.*?)\[\/url\]/ism', ' #$1 ', $text);
282 $text = preg_replace('/[@!#]?\[url.*?\[\/url\]/ism', ' ', $text);
283 $text = preg_replace("/\[url=[^\[\]]*\](.*)\[\/url\]/Usi", ' $1 ', $text);
285 // Convert it to plain text
286 $text = self::toPlaintext($text, false);
288 // Remove possibly remaining links
289 $text = preg_replace(Strings::autoLinkRegEx(), '', $text);
291 // Remove all unneeded white space
294 $text = str_replace([' ', "\n", "\r", '"'], ' ', $text);
295 } while ($oldtext != $text);
300 private static function proxyUrl(string $image, int $simplehtml = self::INTERNAL, int $uriid = 0, string $size = ''): string
302 // Only send proxied pictures to API and for internal display
303 if (!in_array($simplehtml, [self::INTERNAL, self::MASTODON_API, self::TWITTER_API])) {
305 } elseif ($uriid > 0) {
306 return Post\Link::getByLink($uriid, $image, $size);
308 return Proxy::proxifyUrl($image, $size);
313 * Truncates imported message body string length to max_import_size
315 * The purpose of this function is to apply system message length limits to
316 * imported messages without including any embedded photos in the length
318 * @param string $body
320 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
322 public static function limitBodySize(string $body): string
324 DI::profiler()->startRecording('rendering');
325 $maxlen = DI::config()->get('config', 'max_import_size', 0);
327 // If the length of the body, including the embedded images, is smaller
328 // than the maximum, then don't waste time looking for the images
329 if ($maxlen && (strlen($body) > $maxlen)) {
331 Logger::info('the total body length exceeds the limit', ['maxlen' => $maxlen, 'body_len' => strlen($body)]);
337 $img_start = strpos($orig_body, '[img');
338 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
339 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
340 while (($img_st_close !== false) && ($img_end !== false)) {
342 $img_st_close++; // make it point to AFTER the closing bracket
343 $img_end += $img_start;
344 $img_end += strlen('[/img]');
346 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
347 // This is an embedded image
349 if (($textlen + $img_start) > $maxlen) {
350 if ($textlen < $maxlen) {
351 Logger::debug('the limit happens before an embedded image');
352 $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
356 $new_body = $new_body . substr($orig_body, 0, $img_start);
357 $textlen += $img_start;
360 $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
363 if (($textlen + $img_end) > $maxlen) {
364 if ($textlen < $maxlen) {
365 Logger::debug('the limit happens before the end of a non-embedded image');
366 $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
370 $new_body = $new_body . substr($orig_body, 0, $img_end);
371 $textlen += $img_end;
374 $orig_body = substr($orig_body, $img_end);
376 if ($orig_body === false) {
377 // in case the body ends on a closing image tag
381 $img_start = strpos($orig_body, '[img');
382 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
383 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
386 if (($textlen + strlen($orig_body)) > $maxlen) {
387 if ($textlen < $maxlen) {
388 Logger::debug('the limit happens after the end of the last image');
389 $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
392 Logger::debug('the text size with embedded images extracted did not violate the limit');
393 $new_body = $new_body . $orig_body;
396 DI::profiler()->stopRecording();
399 DI::profiler()->stopRecording();
405 * Processes [attachment] tags
407 * Note: Can produce a [bookmark] tag in the returned string
409 * @param string $text
410 * @param integer $simplehtml
411 * @param bool $tryoembed
415 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
417 public static function convertAttachment(string $text, int $simplehtml = self::INTERNAL, bool $tryoembed = true, array $data = [], int $uriid = 0, int $preview_mode = self::PREVIEW_LARGE): string
419 DI::profiler()->startRecording('rendering');
420 $data = $data ?: self::getAttachmentData($text);
421 if (empty($data) || empty($data['url'])) {
422 DI::profiler()->stopRecording();
426 if (isset($data['title'])) {
427 $data['title'] = strip_tags($data['title']);
428 $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
433 if (((strpos($data['text'], '[img=') !== false) || (strpos($data['text'], '[img]') !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) {
434 $data['preview'] = $data['image'];
440 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
441 $return = OEmbed::getHTML($data['url'], $data['title']);
443 throw new Exception('OEmbed is disabled for this attachment.');
445 } catch (Exception $e) {
446 $data['title'] = ($data['title'] ?? '') ?: $data['url'];
448 if ($simplehtml != self::CONNECTORS) {
449 $return = sprintf('<div class="type-%s">', $data['type']);
452 if ($preview_mode == self::PREVIEW_NO_IMAGE) {
453 unset($data['image']);
454 unset($data['preview']);
457 if (!empty($data['title']) && !empty($data['url'])) {
458 $preview_class = $preview_mode == self::PREVIEW_LARGE ? 'attachment-image' : 'attachment-preview';
459 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
460 $return .= sprintf('<a href="%s" target="_blank" rel="noopener noreferrer"><img src="%s" alt="" title="%s" class="' . $preview_class . '" /></a>', $data['url'], self::proxyUrl($data['image'], $simplehtml, $uriid), $data['title']);
462 if (!empty($data['image'])) {
463 $return .= sprintf('<a href="%s" target="_blank" rel="noopener noreferrer"><img src="%s" alt="" title="%s" class="' . $preview_class . '" /></a><br>', $data['url'], self::proxyUrl($data['image'], $simplehtml, $uriid), $data['title']);
464 } elseif (!empty($data['preview'])) {
465 $return .= sprintf('<a href="%s" target="_blank" rel="noopener noreferrer"><img src="%s" alt="" title="%s" class="attachment-preview" /></a><br>', $data['url'], self::proxyUrl($data['preview'], $simplehtml, $uriid), $data['title']);
467 $return .= sprintf('<h4><a href="%s" target="_blank" rel="noopener noreferrer">%s</a></h4>', $data['url'], $data['title']);
471 if (!empty($data['description']) && $data['description'] != $data['title']) {
473 $return .= sprintf('<blockquote>%s</blockquote>', trim(HTML::purify($data['description'])));
476 if (!empty($data['provider_url']) && !empty($data['provider_name'])) {
477 if (!empty($data['author_name'])) {
478 $return .= sprintf('<sup><a href="%s" target="_blank" rel="noopener noreferrer">%s (%s)</a></sup>', $data['provider_url'], $data['author_name'], $data['provider_name']);
480 $return .= sprintf('<sup><a href="%s" target="_blank" rel="noopener noreferrer">%s</a></sup>', $data['provider_url'], $data['provider_name']);
484 if ($simplehtml != self::CONNECTORS) {
489 DI::profiler()->stopRecording();
490 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
493 public static function removeShareInformation(string $text, bool $plaintext = false, bool $nolink = false): string
495 DI::profiler()->startRecording('rendering');
496 $data = self::getAttachmentData($text);
499 DI::profiler()->stopRecording();
502 DI::profiler()->stopRecording();
503 return $data['text'] . ($data['after'] ?? '');
506 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
507 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
508 if ($plaintext || (($title != '') && strstr($text, $title))) {
509 $data['title'] = $data['url'];
510 } elseif (($text != '') && strstr($title, $text)) {
511 $data['text'] = $data['title'];
512 $data['title'] = $data['url'];
515 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
516 DI::profiler()->stopRecording();
517 return $data['title'] . $data['after'];
520 // If the link already is included in the post, don't add it again
521 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
522 DI::profiler()->stopRecording();
523 return $data['text'] . $data['after'];
526 $text = $data['text'];
528 if (!empty($data['url']) && !empty($data['title'])) {
529 $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
530 } elseif (!empty($data['url'])) {
531 $text .= "\n[url]" . $data['url'] . '[/url]';
534 DI::profiler()->stopRecording();
535 return $text . "\n" . $data['after'];
539 * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
541 * @param array $match Array with the matching values
542 * @return string reformatted link including HTML codes
544 private static function convertUrlForActivityPubCallback(array $match): string
548 if (isset($match[2]) && ($match[1] != $match[2])) {
552 $parts = parse_url($url);
553 if (!isset($parts['scheme'])) {
557 return self::convertUrlForActivityPub($url);
561 * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
563 * @param string $url URL that is about to be reformatted
564 * @return string reformatted link including HTML codes
566 private static function convertUrlForActivityPub(string $url): string
568 return sprintf('<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>', $url, Strings::getStyledURL($url));
572 * [noparse][i]italic[/i][/noparse] turns into
573 * [noparse][ i ]italic[ /i ][/noparse],
574 * to hide them from parser.
576 * @param array $match
579 private static function escapeNoparseCallback(array $match): string
581 $whole_match = $match[0];
582 $captured = $match[1];
583 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
584 $new_str = str_replace($captured, $spacefied, $whole_match);
589 * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
590 * now turns back and the [noparse] tags are trimmed
591 * returning [i]italic[/i]
593 * @param array $match
596 private static function unescapeNoparseCallback(array $match): string
598 $captured = $match[1];
599 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
604 * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
607 * @param string $text Text to search
608 * @param string $name Tag name
609 * @param int $occurrences Number of first occurrences to skip
610 * @return boolean|array
612 public static function getTagPosition(string $text, string $name, int $occurrences = 0)
614 DI::profiler()->startRecording('rendering');
615 if ($occurrences < 0) {
620 for ($i = 0; $i <= $occurrences; $i++) {
621 if ($start_open !== false) {
622 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
626 if ($start_open === false) {
627 DI::profiler()->stopRecording();
631 $start_equal = strpos($text, '=', $start_open);
632 $start_close = strpos($text, ']', $start_open);
634 if ($start_close === false) {
635 DI::profiler()->stopRecording();
641 $end_open = strpos($text, '[/' . $name . ']', $start_close);
643 if ($end_open === false) {
644 DI::profiler()->stopRecording();
650 'open' => $start_open,
651 'close' => $start_close
655 'close' => $end_open + strlen('[/' . $name . ']')
659 if ($start_equal !== false) {
660 $res['start']['equal'] = $start_equal + 1;
663 DI::profiler()->stopRecording();
668 * Performs a preg_replace within the boundaries of all named BBCode tags in a text
670 * @param string $pattern Preg pattern string
671 * @param string $replace Preg replace string
672 * @param string $name BBCode tag name
673 * @param string $text Text to search
676 public static function pregReplaceInTag(string $pattern, string $replace, string $name, string $text): string
678 DI::profiler()->startRecording('rendering');
680 $pos = self::getTagPosition($text, $name, $occurrences);
681 while ($pos !== false && $occurrences++ < 1000) {
682 $start = substr($text, 0, $pos['start']['open']);
683 $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
684 $end = substr($text, $pos['end']['close']);
685 if ($end === false) {
689 $subject = preg_replace($pattern, $replace, $subject);
690 $text = $start . $subject . $end;
692 $pos = self::getTagPosition($text, $name, $occurrences);
695 DI::profiler()->stopRecording();
699 private static function extractImagesFromItemBody(string $body): array
706 $img_start = strpos($orig_body, '[img');
707 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
708 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
709 while (($img_st_close !== false) && ($img_end !== false)) {
710 $img_st_close++; // make it point to AFTER the closing bracket
711 $img_end += $img_start;
713 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
714 // This is an embedded image
715 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
716 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
720 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
723 $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
725 if ($orig_body === false) {
726 // in case the body ends on a closing image tag
730 $img_start = strpos($orig_body, '[img');
731 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
732 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
735 $new_body = $new_body . $orig_body;
737 return ['body' => $new_body, 'images' => $saved_image];
740 private static function interpolateSavedImagesIntoItemBody(int $uriid, string $body, array $images): string
745 foreach ($images as $image) {
746 // We're depending on the property of 'foreach' (specified on the PHP website) that
747 // it loops over the array starting from the first element and going sequentially
748 // to the last element
749 $newbody = str_replace(
750 '[$#saved_image' . $cnt . '#$]',
751 '<img src="' . self::proxyUrl($image, self::INTERNAL, $uriid) . '" alt="' . DI::l10n()->t('Image/photo') . '" />',
761 * @param string $text A BBCode string
762 * @return array Empty array if no share tag is present or the following array, missing attributes end up empty strings:
763 * - comment : Text before the opening share tag
764 * - shared : Text inside the share tags
765 * - author : (Optional) Display name of the shared author
766 * - profile : (Optional) Profile page URL of the shared author
767 * - avatar : (Optional) Profile picture URL of the shared author
768 * - link : (Optional) Canonical URL of the shared post
769 * - posted : (Optional) Date the shared post was initially posted ("Y-m-d H:i:s" in GMT)
770 * - message_id: (Optional) Shared post URI if any
771 * - guid : (Optional) Shared post GUID if any
773 public static function fetchShareAttributes(string $text): array
775 DI::profiler()->startRecording('rendering');
776 if (preg_match('~(.*?)\[share](.*)\[/share]~ism', $text, $matches)) {
777 DI::profiler()->stopRecording();
785 'message_id' => trim($matches[2]),
786 'comment' => trim($matches[1]),
790 // See Issue https://github.com/friendica/friendica/issues/10454
791 // Hashtags in usernames are expanded to links. This here is a quick fix.
792 $text = preg_replace('~([@!#])\[url=.*?](.*?)\[/url]~ism', '$1$2', $text);
794 if (!preg_match('~(.*?)\[share(.*?)](.*)\[/share]~ism', $text, $matches)) {
795 DI::profiler()->stopRecording();
799 $attributes = self::extractShareAttributes($matches[2]);
801 $attributes['comment'] = trim($matches[1]);
802 $attributes['shared'] = trim($matches[3]);
804 DI::profiler()->stopRecording();
809 * @see BBCode::fetchShareAttributes()
810 * @param string $shareString Internal opening share tag string matched by the regular expression
811 * @return array A fixed attribute array where missing attribute are represented by empty strings
813 private static function extractShareAttributes(string $shareString): array
816 foreach (['author', 'profile', 'avatar', 'link', 'posted', 'guid', 'message_id'] as $field) {
817 preg_match("/$field=(['\"])(.+?)\\1/ism", $shareString, $matches);
818 $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
825 * Remove the share block
827 * @param string $body
830 public static function removeSharedData(string $body): string
832 return trim(preg_replace("/\s*\[share.*?\].*?\[\/share\]\s*/ism", '', $body));
836 * This function converts a [share] block to text according to a provided callback function whose signature is:
838 * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
841 * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
843 * - $author_contact is a contact record array
844 * - $content is the inner content of the [share] block
845 * - $is_quote_share indicates whether there's any content before the [share] block
846 * - Return value is the string that should replace the [share] block in the provided text
848 * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
850 * @param string $text A BBCode string
851 * @param callable $callback
852 * @return string The BBCode string with all [share] blocks replaced
854 public static function convertShare(string $text, callable $callback, int $uriid = 0): string
856 DI::profiler()->startRecording('rendering');
857 $return = preg_replace_callback(
858 '~(.*?)\[share(.*?)](.*)\[/share]~ism',
859 function ($match) use ($callback, $uriid) {
860 $attributes = self::extractShareAttributes($match[2]);
862 $author_contact = Contact::getByURL($attributes['profile'], false, ['id', 'url', 'addr', 'name', 'micro']);
863 $author_contact['url'] = ($author_contact['url'] ?? $attributes['profile']);
864 $author_contact['addr'] = ($author_contact['addr'] ?? '');
866 $attributes['author'] = ($author_contact['name'] ?? '') ?: $attributes['author'];
867 $attributes['avatar'] = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
868 $attributes['profile'] = ($author_contact['url'] ?? '') ?: $attributes['profile'];
870 if (!empty($author_contact['id'])) {
871 $attributes['avatar'] = Contact::getAvatarUrlForId($author_contact['id'], Proxy::SIZE_THUMB);
872 } elseif ($attributes['avatar']) {
873 $attributes['avatar'] = self::proxyUrl($attributes['avatar'], self::INTERNAL, $uriid, Proxy::SIZE_THUMB);
876 $content = preg_replace(Strings::autoLinkRegEx(), '<a href="$1">$1</a>', $match[3]);
878 return $match[1] . $callback($attributes, $author_contact, $content, trim($match[1]) != '');
883 DI::profiler()->stopRecording();
884 return trim($return);
888 * Convert complex IMG and ZMG elements
890 * @param [type] $text
891 * @param integer $simplehtml
892 * @param integer $uriid
895 private static function convertImages(string $text, int $simplehtml, int $uriid = 0): string
897 DI::profiler()->startRecording('rendering');
898 $return = preg_replace_callback(
899 "/\[[zi]mg(.*?)\]([^\[\]]*)\[\/[zi]mg\]/ism",
900 function ($match) use ($simplehtml, $uriid) {
901 $attribute_string = $match[1];
903 foreach (['alt', 'width', 'height'] as $field) {
904 preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
905 $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
908 $img_str = '<img src="' . self::proxyUrl($match[2], $simplehtml, $uriid) . '"';
909 foreach ($attributes as $key => $value) {
910 if (!empty($value)) {
911 $img_str .= ' ' . $key . '="' . htmlspecialchars($value, ENT_COMPAT) . '"';
914 return $img_str . '>';
919 DI::profiler()->stopRecording();
924 * Default [share] tag conversion callback
926 * Note: Can produce a [bookmark] tag in the output
928 * @see BBCode::convertShare()
929 * @param array $attributes [share] block attribute values
930 * @param array $author_contact Contact row of the shared author
931 * @param string $content Inner content of the [share] block
932 * @param boolean $is_quote_share Whether there is content before the [share] block
933 * @param integer $simplehtml Mysterious integer value depending on the target network/formatting style
935 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
937 private static function convertShareCallback(array $attributes, array $author_contact, string $content, bool $is_quote_share, int $simplehtml): string
939 DI::profiler()->startRecording('rendering');
940 $mention = $attributes['author'] . ' (' . ($author_contact['addr'] ?? '') . ')';
942 switch ($simplehtml) {
943 case self::MASTODON_API:
944 case self::TWITTER_API:
945 $text = ($is_quote_share ? '<br>' : '') .
946 '<b><a href="' . $attributes['link'] . '">' . html_entity_decode('♲', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . "</a>:</b><br>\n" .
947 '<blockquote class="shared_content" dir="auto">' . $content . '</blockquote>';
950 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
951 $text = ($is_quote_share ? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
953 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
955 if (!empty($attributes['posted']) && !empty($attributes['link'])) {
956 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
959 $text = ($is_quote_share ? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
961 if (empty($attributes['posted']) && !empty($attributes['link'])) {
962 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
967 case self::CONNECTORS:
968 $headline = '<p><b>' . html_entity_decode('♲ ', ENT_QUOTES, 'UTF-8');
969 $headline .= DI::l10n()->t('<a href="%1$s" target="_blank" rel="noopener noreferrer">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
970 $headline .= ':</b></p>' . "\n";
972 $text = ($is_quote_share ? '<hr />' : '') . $headline . '<blockquote class="shared_content" dir="auto">' . trim($content) . '</blockquote>' . "\n";
976 $text = ($is_quote_share ? '<br>' : '') . '<p>' . html_entity_decode('♲ ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
978 case self::ACTIVITYPUB:
979 $author = '@<span class="vcard"><a href="' . $author_contact['url'] . '" class="url u-url mention" title="' . $author_contact['addr'] . '"><span class="fn nickname mention">' . $author_contact['addr'] . '</span></a>:</span>';
980 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('♲', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
983 $text = ($is_quote_share ? "\n" : '');
985 $contact = Contact::getByURL($attributes['profile'], false, ['network']);
986 $network = $contact['network'] ?? Protocol::PHANTOM;
988 $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
989 $text .= self::SHARED_ANCHOR . Renderer::replaceMacros($tpl, [
990 '$profile' => $attributes['profile'],
991 '$avatar' => $attributes['avatar'],
992 '$author' => $attributes['author'],
993 '$link' => $attributes['link'],
994 '$link_title' => DI::l10n()->t('Link to source'),
995 '$posted' => $attributes['posted'],
996 '$guid' => $attributes['guid'],
997 '$network_name' => ContactSelector::networkToName($network, $attributes['profile']),
998 '$network_icon' => ContactSelector::networkToIcon($network, $attributes['profile']),
999 '$content' => self::TOP_ANCHOR . self::setMentions(trim($content), 0, $network) . self::BOTTOM_ANCHOR,
1007 private static function removePictureLinksCallback(array $match): string
1009 $cache_key = 'remove:' . $match[1];
1010 $text = DI::cache()->get($cache_key);
1012 if (is_null($text)) {
1013 $curlResult = DI::httpClient()->head($match[1], [HttpClientOptions::TIMEOUT => DI::config()->get('system', 'xrd_timeout')]);
1014 if ($curlResult->isSuccess()) {
1015 $mimetype = $curlResult->getHeader('Content-Type')[0] ?? '';
1020 if (substr($mimetype, 0, 6) == 'image/') {
1021 $text = '[url=' . $match[1] . ']' . $match[1] . '[/url]';
1023 $text = '[url=' . $match[2] . ']' . $match[2] . '[/url]';
1025 // if its not a picture then look if its a page that contains a picture link
1026 $body = DI::httpClient()->fetch($match[1], HttpClientAccept::HTML, 0);
1028 DI::cache()->set($cache_key, $text);
1032 $doc = new DOMDocument();
1033 @$doc->loadHTML($body);
1034 $xpath = new DOMXPath($doc);
1035 $list = $xpath->query('//meta[@name]');
1036 foreach ($list as $node) {
1039 if ($node->attributes->length) {
1040 foreach ($node->attributes as $attribute) {
1041 $attr[$attribute->name] = $attribute->value;
1045 if (strtolower($attr['name']) == 'twitter:image') {
1046 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1050 DI::cache()->set($cache_key, $text);
1057 * Callback: Expands links from given $match array
1059 * @param array $match Array with link match
1060 * @return string BBCode
1062 private static function expandLinksCallback(array $match): string
1064 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1065 return ($match[1] . '[url]' . $match[2] . '[/url]');
1067 return ($match[1] . $match[3] . ' [url]' . $match[2] . '[/url]');
1072 * Callback: Cleans picture links
1074 * @param array $match Array with link match
1075 * @return string BBCode
1077 private static function cleanPictureLinksCallback(array $match): string
1079 // When the picture link is the own photo path then we can avoid fetching the link
1080 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()) . '/photos/');
1081 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1082 if (!empty($match[3])) {
1083 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1085 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1090 $cache_key = 'clean:' . $match[1];
1091 $text = DI::cache()->get($cache_key);
1092 if (!is_null($text)) {
1096 $curlResult = DI::httpClient()->head($match[1], [HttpClientOptions::TIMEOUT => DI::config()->get('system', 'xrd_timeout')]);
1097 if ($curlResult->isSuccess()) {
1098 $mimetype = $curlResult->getHeader('Content-Type')[0] ?? '';
1103 // if its a link to a picture then embed this picture
1104 if (substr($mimetype, 0, 6) == 'image/') {
1105 $text = '[img]' . $match[1] . '[/img]';
1107 if (!empty($match[3])) {
1108 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1110 $text = '[img]' . $match[2] . '[/img]';
1113 // if its not a picture then look if its a page that contains a picture link
1114 $body = DI::httpClient()->fetch($match[1], HttpClientAccept::HTML, 0);
1116 DI::cache()->set($cache_key, $text);
1120 $doc = new DOMDocument();
1121 @$doc->loadHTML($body);
1122 $xpath = new DOMXPath($doc);
1123 $list = $xpath->query('//meta[@name]');
1124 foreach ($list as $node) {
1126 if ($node->attributes->length) {
1127 foreach ($node->attributes as $attribute) {
1128 $attr[$attribute->name] = $attribute->value;
1132 if (strtolower($attr['name']) == "twitter:image") {
1133 if (!empty($match[3])) {
1134 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1136 $text = "[img]" . $attr['content'] . "[/img]";
1141 DI::cache()->set($cache_key, $text);
1147 * Cleans picture links
1149 * @param string $text HTML/BBCode string
1150 * @return string Cleaned HTML/BBCode
1152 public static function cleanPictureLinks(string $text): string
1154 DI::profiler()->startRecording('rendering');
1155 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", [self::class, 'cleanPictureLinksCallback'], $text);
1156 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", [self::class, 'cleanPictureLinksCallback'], $return);
1157 DI::profiler()->stopRecording();
1164 * @param string $text HTML/BBCode string
1165 * @return string Cleaned HTML/BBCode
1167 public static function removeLinks(string $bbcode): string
1169 DI::profiler()->startRecording('rendering');
1170 $bbcode = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", ' ', $bbcode);
1171 $bbcode = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $1 ', $bbcode);
1172 $bbcode = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $bbcode);
1174 $bbcode = preg_replace('/[@!#]\[url\=.*?\].*?\[\/url\]/ism', '', $bbcode);
1175 $bbcode = preg_replace("/\[url=[^\[\]]*\](.*)\[\/url\]/Usi", ' $1 ', $bbcode);
1176 $bbcode = preg_replace('/[@!#]?\[url.*?\[\/url\]/ism', '', $bbcode);
1177 DI::profiler()->stopRecording();
1182 * Replace names in mentions with nicknames
1184 * @param string $body HTML/BBCode
1185 * @return string Body with replaced mentions
1187 public static function setMentionsToNicknames(string $body): string
1189 DI::profiler()->startRecording('rendering');
1190 $regexp = "/([@!])\[url\=([^\[\]]*)\].*?\[\/url\]/ism";
1191 $body = preg_replace_callback($regexp, [self::class, 'mentionCallback'], $body);
1192 DI::profiler()->stopRecording();
1197 * Callback function to replace a Friendica style mention in a mention with the nickname
1199 * @param array $match Matching values for the callback
1200 * @return string Replaced mention or empty string
1201 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1203 private static function mentionCallback(array $match): string
1205 if (empty($match[2])) {
1209 $data = Contact::getByURL($match[2], false, ['url', 'nick']);
1210 if (empty($data['nick'])) {
1214 return $match[1] . '[url=' . $data['url'] . ']' . $data['nick'] . '[/url]';
1218 * Normalize links to Youtube and Vimeo to a unified format.
1220 * @param string $text
1223 private static function normalizeVideoLinks(string $text): string
1225 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1226 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1227 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/shorts\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1228 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1230 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1231 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1237 * Expand Youtube and Vimeo links to
1239 * @param string $text
1242 public static function expandVideoLinks(string $text): string
1244 $text = self::normalizeVideoLinks($text);
1245 $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism", '[url=https://www.youtube.com/watch?v=$1]https://www.youtube.com/watch?v=$1[/url]', $text);
1246 $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism", '[url=https://vimeo.com/$1]https://vimeo.com/$1[/url]', $text);
1252 * Converts a BBCode message for a given URI-ID to a HTML message
1254 * BBcode 2 HTML was written by WAY2WEB.net
1255 * extended to work with Mistpark/Friendica - Mike Macgirvin
1257 * Simple HTML values meaning:
1258 * - 0: Friendica display
1260 * - 2: Used for Windows Phone push, Friendica API
1261 * - 3: Used before converting to Markdown in bb2diaspora.php
1262 * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1265 * - 7: Used for dfrn, OStatus
1266 * - 8: Used for Twitter, WP backlink text setting
1270 * @param string $text
1271 * @param int $simple_html
1273 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1275 public static function convertForUriId(int $uriid = null, string $text = null, int $simple_html = self::INTERNAL): string
1277 $try_oembed = ($simple_html == self::INTERNAL);
1279 return self::convert($text ?? '', $try_oembed, $simple_html, false, $uriid ?? 0);
1283 * Converts a BBCode message to HTML message
1285 * BBcode 2 HTML was written by WAY2WEB.net
1286 * extended to work with Mistpark/Friendica - Mike Macgirvin
1288 * Simple HTML values meaning:
1289 * - 0: Friendica display
1291 * - 2: Used for Windows Phone push, Friendica API
1292 * - 3: Used before converting to Markdown in bb2diaspora.php
1293 * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1296 * - 7: Used for dfrn, OStatus
1297 * - 8: Used for Twitter, WP backlink text setting
1300 * @param string $text
1301 * @param bool $try_oembed
1302 * @param int $simple_html
1303 * @param bool $for_plaintext
1305 * @return string Converted code or empty string
1306 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1308 public static function convert(string $text = null, bool $try_oembed = true, int $simple_html = self::INTERNAL, bool $for_plaintext = false, int $uriid = 0): string
1310 // Accounting for null default column values
1311 if (is_null($text) || $text === '') {
1315 DI::profiler()->startRecording('rendering');
1317 Hook::callAll('bbcode', $text);
1321 $text = self::performWithEscapedTags($text, ['code'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a, $uriid) {
1322 $text = self::performWithEscapedTags($text, ['noparse', 'nobb', 'pre'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a, $uriid) {
1324 * preg_match_callback function to replace potential Oembed tags with Oembed content
1326 * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1328 * $match[2] = $title or absent
1330 $try_oembed_callback = function (array $match) {
1332 $title = $match[2] ?? '';
1335 $return = OEmbed::getHTML($url, $title);
1336 } catch (Exception $ex) {
1337 $return = $match[0];
1343 // Remove the abstract element. It is a non visible element.
1344 $text = self::stripAbstract($text);
1346 // Line ending normalisation
1347 $text = str_replace("\r\n", "\n", $text);
1349 // Move new lines outside of tags
1350 $text = preg_replace("#\[(\w*)](\n*)#ism", '$2[$1]', $text);
1351 $text = preg_replace("#(\n*)\[/(\w*)]#ism", '[/$2]$1', $text);
1353 // Extract the private images which use data urls since preg has issues with
1354 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1355 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1357 $extracted = self::extractImagesFromItemBody($text);
1358 $text = $extracted['body'];
1359 $saved_image = $extracted['images'];
1361 // If we find any event code, turn it into an event.
1362 // After we're finished processing the bbcode we'll
1363 // replace all of the event code with a reformatted version.
1365 $ev = Event::fromBBCode($text);
1367 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1368 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1370 $text = str_replace("<", "<", $text);
1371 $text = str_replace(">", ">", $text);
1373 // remove some newlines before the general conversion
1374 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1]$2[/share]\n", $text);
1375 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "\n[quote$1]$2[/quote]\n", $text);
1377 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1379 $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1382 // Remove linefeeds inside of the table elements. See issue #6799
1384 "\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1385 "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1386 "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1387 "\n[hr]", "[hr]\n", " [hr]", "[hr] ",
1388 "\n[attachment ", " [attachment ", "\n[/attachment]", "[/attachment]\n", " [/attachment]", "[/attachment] ",
1389 "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] ",
1390 " \n", "\t\n", "[/li]\n", "\n[li]", "\n[*]",
1393 "[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1394 "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1395 "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1396 "[hr]", "[hr]", "[hr]", "[hr]",
1397 "[attachment ", "[attachment ", "[/attachment]", "[/attachment]", "[/attachment]", "[/attachment]",
1398 "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]",
1399 "\n", "\n", "[/li]", "[li]", "[*]",
1403 $text = str_replace($search, $replace, $text);
1404 } while ($oldtext != $text);
1406 // Replace these here only once
1407 $search = ["\n[table]", "[/table]\n"];
1408 $replace = ["[table]", "[/table]"];
1409 $text = str_replace($search, $replace, $text);
1411 // Trim new lines regardless of the system.remove_multiplicated_lines config value
1412 $text = trim($text, "\n");
1414 // removing multiplicated newlines
1415 if (DI::config()->get('system', 'remove_multiplicated_lines')) {
1417 "\n\n\n", "[/quote]\n\n", "\n[/quote]", "\n[ul]", "[/ul]\n", "\n[ol]", "[/ol]\n", "\n\n[share ", "[/attachment]\n",
1418 "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"
1421 "\n\n", "[/quote]\n", "[/quote]", "[ul]", "[/ul]", "[ol]", "[/ol]", "\n[share ", "[/attachment]",
1422 "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"
1426 $text = str_replace($search, $replace, $text);
1427 } while ($oldtext != $text);
1430 /// @todo Have a closer look at the different html modes
1431 // Handle attached links or videos
1432 if ($simple_html == self::NPF) {
1433 $text = self::removeAttachment($text);
1434 } elseif (in_array($simple_html, [self::MASTODON_API, self::TWITTER_API, self::ACTIVITYPUB])) {
1435 $text = self::replaceAttachment($text);
1436 } elseif (!in_array($simple_html, [self::INTERNAL, self::EXTERNAL, self::CONNECTORS])) {
1437 $text = self::replaceAttachment($text, true);
1439 $text = self::convertAttachment($text, $simple_html, $try_oembed, [], $uriid);
1442 $nosmile = strpos($text, '[nosmile]') !== false;
1443 $text = str_replace('[nosmile]', '', $text);
1445 // Replace non graphical smilies for external posts
1447 $text = self::performWithEscapedTags($text, ['img'], function ($text) use ($simple_html, $for_plaintext) {
1448 return Smilies::replace($text, ($simple_html != self::INTERNAL) || $for_plaintext);
1452 // leave open the possibility of [map=something]
1453 // this is replaced in Item::prepareBody() which has knowledge of the item location
1454 if (strpos($text, '[/map]') !== false) {
1455 $text = preg_replace_callback(
1456 "/\[map\](.*?)\[\/map\]/ism",
1457 function ($match) use ($simple_html) {
1458 return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1464 if (strpos($text, '[map=') !== false) {
1465 $text = preg_replace_callback(
1466 "/\[map=(.*?)\]/ism",
1467 function ($match) use ($simple_html) {
1468 return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1474 if (strpos($text, '[map]') !== false) {
1475 $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1478 // Check for headers
1480 if ($simple_html == self::INTERNAL) {
1481 //Ensure to always start with <h4> if possible
1483 for ($level = 6; $level > 0; $level--) {
1484 if (preg_match("(\[h$level\].*?\[\/h$level\])ism", $text)) {
1488 if ($heading_count > 0) {
1489 $heading = min($heading_count + 3, 6);
1490 for ($level = 6; $level > 0; $level--) {
1491 if (preg_match("(\[h$level\].*?\[\/h$level\])ism", $text)) {
1492 $text = preg_replace("(\[h$level\](.*?)\[\/h$level\])ism", "</p><h$heading>$1</h$heading><p>", $text);
1498 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '</p><h1>$1</h1><p>', $text);
1499 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '</p><h2>$1</h2><p>', $text);
1500 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '</p><h3>$1</h3><p>', $text);
1501 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '</p><h4>$1</h4><p>', $text);
1502 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '</p><h5>$1</h5><p>', $text);
1503 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '</p><h6>$1</h6><p>', $text);
1506 // Check for paragraph
1507 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1509 // Check for bold text
1510 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1512 // Check for Italics text
1513 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1515 // Check for Underline text
1516 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1518 // Check for strike-through text
1519 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1521 // Check for over-line text
1522 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1524 // Check for colored text
1525 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1527 // Check for sized text
1528 // [size=50] --> font-size: 50px (with the unit).
1529 if ($simple_html != self::DIASPORA) {
1530 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1px;line-height:normal;">$2</span>', $text);
1531 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", '<span style="font-size:$1;line-height:normal;">$2</span>', $text);
1533 // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1534 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1538 // Check for centered text
1539 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", '<div style="text-align:center;">$1</div>', $text);
1541 // Check for list text
1542 $text = str_replace("[*]", "<li>", $text);
1544 // Check for block-level custom CSS
1545 $text = preg_replace('#(?<=^|\n)\[style=(.*?)](.*?)\[/style](?:\n|$)#ism', '<div style="$1">$2</div>', $text);
1547 // Check for inline custom CSS
1548 $text = preg_replace("(\[style=(.*?)\](.*?)\[\/style\])ism", '<span style="$1">$2</span>', $text);
1550 // Mastodon Emoji (internal tag, do not document for users)
1551 if ($simple_html == self::MASTODON_API) {
1552 $text = preg_replace("(\[emoji=(.*?)](.*?)\[/emoji])ism", '$2', $text);
1554 $text = preg_replace("(\[emoji=(.*?)](.*?)\[/emoji])ism", '<span class="mastodon emoji"><img src="$1" alt="$2" title="$2"/></span>', $text);
1557 // Check for CSS classes
1558 // @deprecated since 2021.12, left for backward-compatibility reasons
1559 $text = preg_replace("(\[class=(.*?)\](.*?)\[\/class\])ism", '<span class="$1">$2</span>', $text);
1560 // Add HTML new lines
1561 $text = str_replace("\n\n", '</p><p>', $text);
1562 $text = str_replace("\n", '<br>', $text);
1564 // handle nested lists
1567 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1568 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1569 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1570 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1571 $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '</p><ul class="listbullet" style="list-style-type: circle;">$1</ul><p>', $text);
1572 $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '</p><ul class="listnone" style="list-style-type: none;">$1</ul><p>', $text);
1573 $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '</p><ul class="listdecimal" style="list-style-type: decimal;">$1</ul><p>', $text);
1574 $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '</p><ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul><p>', $text);
1575 $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '</p><ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul><p>', $text);
1576 $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '</p><ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul><p>', $text);
1577 $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '</p><ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul><p>', $text);
1578 $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '</p><ul>$1</ul><p>', $text);
1579 $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '</p><ol>$1</ol><p>', $text);
1580 $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1583 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1584 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1585 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1586 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '</p><table>$1</table><p>', $text);
1588 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '</p><table border="1" >$1</table><p>', $text);
1589 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '</p><table border="0" >$1</table><p>', $text);
1591 $text = str_replace('[hr]', '</p><hr /><p>', $text);
1593 if (!$for_plaintext) {
1594 $text = self::performWithEscapedTags($text, ['url', 'img', 'audio', 'video', 'youtube', 'vimeo', 'share', 'attachment', 'iframe', 'bookmark'], function ($text) {
1595 return preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1599 // Check for font change text
1600 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1602 // Declare the format for [spoiler] layout
1603 $SpoilerLayout = '<details class="spoiler"><summary>' . DI::l10n()->t('Click to open/close') . '</summary>$1</details>';
1605 // Check for [spoiler] text
1606 // handle nested quotes
1608 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1609 $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1612 // Check for [spoiler=Title] text
1614 // handle nested quotes
1616 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1617 $text = preg_replace(
1618 "/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1619 '<details class="spoiler"><summary>$1</summary>$2</details>',
1624 // Declare the format for [quote] layout
1625 $QuoteLayout = '</p><blockquote>$1</blockquote><p>';
1627 // Check for [quote] text
1628 // handle nested quotes
1630 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1631 $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1634 // Check for [quote=Author] text
1636 $t_wrote = DI::l10n()->t('$1 wrote:');
1638 // handle nested quotes
1640 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1641 $text = preg_replace(
1642 "/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1643 "<p><strong class=" . '"author"' . ">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1649 // [img=widthxheight]image source[/img]
1650 $text = preg_replace_callback(
1651 "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1652 function ($matches) use ($simple_html, $uriid) {
1653 if (strpos($matches[3], "data:image/") === 0) {
1657 $matches[3] = self::proxyUrl($matches[3], $simple_html, $uriid);
1658 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1663 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1664 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1666 $text = preg_replace_callback(
1667 "/\[[iz]mg\=(.*?)\](.*?)\[\/[iz]mg\]/ism",
1668 function ($matches) use ($simple_html, $uriid) {
1669 $matches[1] = self::proxyUrl($matches[1], $simple_html, $uriid);
1670 $alt = htmlspecialchars($matches[2], ENT_COMPAT);
1671 // Fix for Markdown problems with Diaspora, see issue #12701
1672 if (($simple_html != self::DIASPORA) || strpos($matches[2], '"') === false) {
1673 return '<img src="' . $matches[1] . '" alt="' . $alt . '" title="' . $alt . '">';
1675 return '<img src="' . $matches[1] . '" alt="' . $alt . '">';
1682 // [img]pathtoimage[/img]
1683 $text = preg_replace_callback(
1684 "/\[[iz]mg\](.*?)\[\/[iz]mg\]/ism",
1685 function ($matches) use ($simple_html, $uriid) {
1686 if (strpos($matches[1], "data:image/") === 0) {
1690 $matches[1] = self::proxyUrl($matches[1], $simple_html, $uriid);
1691 return "[img]" . $matches[1] . "[/img]";
1696 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1697 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1699 $text = self::convertImages($text, $simple_html, $uriid);
1701 $text = preg_replace("/\[crypt\](.*?)\[\/crypt\]/ism", '<br><img src="' . DI::baseUrl() . '/images/lock_icon.gif" alt="' . DI::l10n()->t('Encrypted content') . '" title="' . DI::l10n()->t('Encrypted content') . '" /><br>', $text);
1702 $text = preg_replace("/\[crypt(.*?)\](.*?)\[\/crypt\]/ism", '<br><img src="' . DI::baseUrl() . '/images/lock_icon.gif" alt="' . DI::l10n()->t('Encrypted content') . '" title="' . '$1' . ' ' . DI::l10n()->t('Encrypted content') . '" /><br>', $text);
1703 //$text = preg_replace("/\[crypt=(.*?)\](.*?)\[\/crypt\]/ism", '<br><img src="' .DI::baseUrl() . '/images/lock_icon.gif" alt="' . DI::l10n()->t('Encrypted content') . '" title="' . '$1' . ' ' . DI::l10n()->t('Encrypted content') . '" /><br>', $text);
1705 // Simplify "video" element
1706 $text = preg_replace('(\[video[^\]]*?\ssrc\s?=\s?([^\s\]]+)[^\]]*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1708 if ($simple_html == self::NPF) {
1709 $text = preg_replace(
1710 "/\[video\](.*?)\[\/video\]/ism",
1711 '</p><video src="$1" controls width="100%" height="auto">$1</video><p>',
1714 $text = preg_replace(
1715 "/\[audio\](.*?)\[\/audio\]/ism",
1716 '</p><audio src="$1" controls>$1">$1</audio><p>',
1719 } elseif ($try_oembed) {
1720 // html5 video and audio
1721 $text = preg_replace(
1722 "/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4).*?)\[\/video\]/ism",
1723 '<video src="$1" controls width="100%" height="auto"><a href="$1">$1</a></video>',
1727 $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1728 $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1730 $text = preg_replace(
1731 "/\[video\](.*?)\[\/video\]/ism",
1732 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>',
1735 $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '<audio src="$1" controls><a href="$1">$1</a></audio>', $text);
1737 $text = preg_replace(
1738 "/\[video\](.*?)\[\/video\]/ism",
1739 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>',
1742 $text = preg_replace(
1743 "/\[audio\](.*?)\[\/audio\]/ism",
1744 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>',
1749 // Backward compatibility, [iframe] support has been removed in version 2020.12
1750 $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1752 $text = self::normalizeVideoLinks($text);
1754 // Youtube extensions
1756 $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism", '<iframe width="' . $a->getThemeInfoValue('videowidth') . '" height="' . $a->getThemeInfoValue('videoheight') . '" src="https://www.youtube.com/embed/$1" frameborder="0" ></iframe>', $text);
1758 $text = preg_replace(
1759 "/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1760 '<a href="https://www.youtube.com/watch?v=$1" target="_blank" rel="noopener noreferrer">https://www.youtube.com/watch?v=$1</a>',
1767 $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism", '<iframe width="' . $a->getThemeInfoValue('videowidth') . '" height="' . $a->getThemeInfoValue('videoheight') . '" src="https://player.vimeo.com/video/$1" frameborder="0" ></iframe>', $text);
1769 $text = preg_replace(
1770 "/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1771 '<a href="https://vimeo.com/$1" target="_blank" rel="noopener noreferrer">https://vimeo.com/$1</a>',
1777 $text = OEmbed::BBCode2HTML($text);
1779 // Avoid triple linefeeds through oembed
1780 $text = str_replace("<br style='clear:left'></span><br><br>", "<br style='clear:left'></span><br>", $text);
1782 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1783 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1784 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1785 // start which is always required). Allow desc with a missing summary for compatibility.
1787 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1788 $sub = Event::getHTML($ev, $simple_html, $uriid);
1790 $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1791 $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1792 $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1793 $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1794 $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1795 $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1798 if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA) && Smilies::isEmojiPost($text)) {
1799 $text = '<span style="font-size: xx-large; line-height: normal;">' . $text . '</span>';
1802 // Handle mentions and hashtag links
1803 if ($simple_html == self::DIASPORA) {
1804 // The ! is converted to @ since Diaspora only understands the @
1805 $text = preg_replace(
1806 "/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1807 '@<a href="$2">$3</a>',
1810 } elseif (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1811 $text = preg_replace(
1812 "/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1813 '<span class="h-card"><a href="$2" class="u-url mention">$1<span>$3</span></a></span>',
1816 $text = preg_replace(
1817 "/([#])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1818 '<a href="$2" class="mention hashtag" rel="tag">$1<span>$3</span></a>',
1821 } elseif (in_array($simple_html, [self::INTERNAL, self::EXTERNAL, self::TWITTER_API])) {
1822 $text = preg_replace(
1823 "/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1824 '<bdi>$1<a href="$2" class="userinfo mention" title="$3">$3</a></bdi>',
1827 } elseif ($simple_html == self::MASTODON_API) {
1828 $text = preg_replace(
1829 "/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1830 '<a class="u-url mention status-link" href="$2" rel="nofollow noopener noreferrer" target="_blank" title="$3">$1<span>$3</span></a>',
1833 $text = preg_replace(
1834 "/([#])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1835 '<a class="mention hashtag status-link" href="$2" rel="tag">$1<span>$3</span></a>',
1839 $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1842 if (!$for_plaintext) {
1843 if (in_array($simple_html, [self::OSTATUS, self::MASTODON_API, self::TWITTER_API, self::ACTIVITYPUB])) {
1844 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", [self::class, 'convertUrlForActivityPubCallback'], $text);
1845 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", [self::class, 'convertUrlForActivityPubCallback'], $text);
1848 $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1849 $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", [self::class, 'removePictureLinksCallback'], $text);
1852 // Bookmarks in red - will be converted to bookmarks in friendica
1853 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1854 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1855 $text = preg_replace(
1856 "/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1857 "[bookmark=$1]$2[/bookmark]",
1861 if (in_array($simple_html, [self::OSTATUS, self::TWITTER, self::BLUESKY])) {
1862 $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", [self::class, 'expandLinksCallback'], $text);
1863 //$text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $text);
1864 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]', $text);
1867 // Perform URL Search
1869 $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1872 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1874 // Handle Diaspora posts
1875 $text = preg_replace_callback(
1876 "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1878 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1883 $text = preg_replace_callback(
1884 "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1886 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1891 // Server independent link to posts and comments
1892 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1893 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1894 $text = preg_replace($expression, DI::baseUrl() . "/display/$1", $text);
1898 * - #[url=<anything>]<term>[/url]
1899 * - [url=<anything>]#<term>[/url]
1901 self::performWithEscapedTags($text, ['url', 'share'], function ($text) use ($simple_html) {
1902 $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function ($matches) use ($simple_html) {
1903 if ($simple_html == self::ACTIVITYPUB) {
1904 return '<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1905 . '" data-tag="' . XML::escape($matches[1]) . '" rel="tag ugc">#'
1906 . XML::escape($matches[1]) . '</a>';
1908 return '#<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1909 . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1910 . XML::escape($matches[1]) . '</a>';
1916 // We need no target="_blank" rel="noopener noreferrer" for local links
1917 // convert links start with DI::baseUrl() as local link without the target="_blank" rel="noopener noreferrer" attribute
1918 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
1919 $text = preg_replace("/\[url\](" . $escapedBaseUrl . ".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1920 $text = preg_replace("/\[url\=(" . $escapedBaseUrl . ".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1922 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1923 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1925 // Red compatibility, though the link can't be authenticated on Friendica
1926 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1929 // we may need to restrict this further if it picks up too many strays
1930 // link acct:user@host to a webfinger profile redirector
1932 $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="' . DI::baseUrl() . '/acctlink?addr=$1@$2" target="extlink">acct:$1@$2</a>', $text);
1934 // Perform MAIL Search
1935 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1936 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1938 /// @todo What is the meaning of these lines?
1939 $text = preg_replace('/\[\&\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1940 $text = preg_replace('/\&\#039\;/', '\'', $text);
1942 // Currently deactivated, it made problems with " inside of alt texts.
1943 //$text = preg_replace('/\"\;/', '"', $text);
1945 // fix any escaped ampersands that may have been converted into links
1946 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1948 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1949 $allowed_src_protocols = ['//', 'http://', 'https://', 'contact/redir/', 'cid:'];
1951 array_walk($allowed_src_protocols, function (&$value) {
1952 $value = preg_quote($value, '#');
1955 $text = preg_replace(
1956 '#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1957 '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">',
1961 // sanitize href attributes (only allowlisted protocols URLs)
1962 // default value for backward compatibility
1963 $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
1965 // Always allowed protocol even if config isn't set or not including it
1966 $allowed_link_protocols[] = '//';
1967 $allowed_link_protocols[] = 'http://';
1968 $allowed_link_protocols[] = 'https://';
1969 $allowed_link_protocols[] = 'contact/redir/';
1971 array_walk($allowed_link_protocols, function (&$value) {
1972 $value = preg_quote($value, '#');
1975 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1976 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text);
1979 $text = self::convertShare(
1981 function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1982 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1987 $text = self::interpolateSavedImagesIntoItemBody($uriid, $text, $saved_image);
1990 }); // Escaped noparse, nobb, pre
1992 // Remove escaping tags and replace new lines that remain
1993 $text = preg_replace_callback('/\[(noparse|nobb)](.*?)\[\/\1]/ism', function ($match) {
1994 return str_replace("\n", "<br>", $match[2]);
1997 // Additionally, [pre] tags preserve spaces
1998 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", function ($match) {
1999 return str_replace([' ', "\n"], [' ', "<br>"], htmlentities($match[1], ENT_NOQUOTES, 'UTF-8'));
2005 $text = preg_replace_callback(
2006 "#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
2007 function ($matches) {
2008 if (strpos($matches[2], "\n") !== false) {
2009 $return = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlentities(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
2011 $return = '<code>' . htmlentities($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
2019 // Default iframe allowed domains/path
2020 $allowedIframeDomains = [
2021 DI::baseUrl()->getHost()
2022 . (DI::baseUrl()->getPath() ? '/' . DI::baseUrl()->getPath() : '')
2023 . '/oembed/', # The path part has to change with the source in Content\Oembed::iframe
2024 'www.youtube.com/embed/',
2025 'player.vimeo.com/video/',
2028 $allowedIframeDomains = array_merge(
2029 $allowedIframeDomains,
2030 DI::config()->get('system', 'allowed_oembed') ?
2031 explode(',', DI::config()->get('system', 'allowed_oembed'))
2035 if (strpos($text, '<p>') !== false || strpos($text, '</p>') !== false) {
2036 $text = '<p>' . $text . '</p>';
2039 $text = HTML::purify($text, $allowedIframeDomains);
2040 DI::profiler()->stopRecording();
2046 * Strips the "abstract" tag from the provided text
2048 * @param string $text The text with BBCode
2049 * @return string The same text - but without "abstract" element
2051 public static function stripAbstract(string $text): string
2053 DI::profiler()->startRecording('rendering');
2055 $text = self::performWithEscapedTags($text, ['code', 'noparse', 'nobb', 'pre'], function ($text) {
2056 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", ' ', $text);
2057 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", ' ', $text);
2061 DI::profiler()->stopRecording();
2066 * Returns the value of the "abstract" element
2068 * @param string $text The text that maybe contains the element
2069 * @param string $addon The addon for which the abstract is meant for
2070 * @return string The abstract
2072 public static function getAbstract(string $text, string $addon = ''): string
2074 DI::profiler()->startRecording('rendering');
2075 $addon = strtolower($addon);
2077 $abstract = self::performWithEscapedTags($text, ['code', 'noparse', 'nobb', 'pre'], function ($text) use ($addon) {
2078 if ($addon && preg_match('#\[abstract=' . preg_quote($addon, '#') . '](.*?)\[/abstract]#ism', $text, $matches)) {
2082 if (preg_match("#\[abstract](.*?)\[/abstract]#ism", $text, $matches)) {
2089 DI::profiler()->stopRecording();
2094 * Callback function to replace a Friendica style mention in a mention for Diaspora
2096 * @param array $match Matching values for the callback
2097 * [1] = Mention type (! or @)
2100 * @return string Replaced mention
2101 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2102 * @throws \ImagickException
2104 private static function bbCodeMention2DiasporaCallback(array $match): string
2106 $contact = Contact::getByURL($match[3], false, ['addr']);
2107 if (empty($contact['addr'])) {
2111 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
2116 * Converts a BBCode text into Markdown
2118 * This function converts a BBCode item body to be sent to Markdown-enabled
2119 * systems like Diaspora and Libertree
2121 * @param string $text
2122 * @param bool $for_diaspora Diaspora requires more changes than Libertree
2124 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2126 public static function toMarkdown(string $text, bool $for_diaspora = true): string
2128 DI::profiler()->startRecording('rendering');
2129 $original_text = $text;
2131 // Since Diaspora is creating a summary for links, this function removes them before posting
2132 if ($for_diaspora) {
2133 $text = self::removeShareInformation($text);
2137 * Transform #tags, strip off the [url] and replace spaces with underscore
2139 $url_search_string = "^\[\]";
2140 $text = preg_replace_callback(
2141 "/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2142 function ($matches) {
2143 return '#' . str_replace(' ', '_', $matches[2]);
2148 // Converting images with size parameters to simple images. Markdown doesn't know it.
2149 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2151 // Convert it to HTML - don't try oembed
2152 if ($for_diaspora) {
2153 $text = self::convertForUriId(0, $text, self::DIASPORA);
2155 // Add all tags that maybe were removed
2156 if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2158 foreach ($tags[2] as $tag) {
2159 $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2160 if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2161 $tagline .= '#' . $tag . ' ';
2164 $text = $text . ' ' . $tagline;
2167 $text = self::convertForUriId(0, $text, self::CONNECTORS);
2170 // If a link is followed by a quote then there should be a newline before it
2171 // Maybe we should make this newline at every time before a quote.
2172 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2174 // The converter doesn't convert these elements
2175 $text = str_replace(['<div>', '</div>'], ['<p>', '</p>'], $text);
2177 // Now convert HTML to Markdown
2178 $text = HTML::toMarkdown($text);
2180 // Libertree has a problem with escaped hashtags.
2181 $text = str_replace(['\#'], ['#'], $text);
2183 // Remove any leading or trailing whitespace, as this will mess up
2184 // the Diaspora signature verification and cause the item to disappear
2185 $text = trim($text);
2187 if ($for_diaspora) {
2188 $url_search_string = "^\[\]";
2189 $text = preg_replace_callback(
2190 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2191 [self::class, 'bbCodeMention2DiasporaCallback'],
2196 Hook::callAll('bb2diaspora', $text);
2198 DI::profiler()->stopRecording();
2203 * Pull out all #hashtags and @person tags from $string.
2205 * We also get @person@domain.com - which would make
2206 * the regex quite complicated as tags can also
2207 * end a sentence. So we'll run through our results
2208 * and strip the period from any tags which end with one.
2209 * Returns array of tags found, or empty array.
2211 * @param string $string Post content
2212 * @return array List of tag and person names
2214 public static function getTags(string $string): array
2216 DI::profiler()->startRecording('rendering');
2219 self::performWithEscapedTags($string, ['noparse', 'pre', 'code', 'img', 'attachment'], function ($string) use (&$ret) {
2220 // Convert hashtag links to hashtags
2221 $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2 ', $string);
2223 // Force line feeds at bbtags
2224 $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2226 // ignore anything in a bbtag
2227 $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2229 // Match full names against @tags including the space between first and last
2230 // We will look these up afterward to see if they are full names or not recognisable.
2232 if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2233 foreach ($matches[1] as $match) {
2234 if (strstr($match, ']')) {
2235 // we might be inside a bbcode color tag - leave it alone
2239 if (substr($match, -1, 1) === '.') {
2240 $ret[] = substr($match, 0, -1);
2247 // Otherwise pull out single word tags. These can be @nickname, @first_last
2250 if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?\']*[^\^ \x0D\x0A,;:?!\'.])/', $string, $matches)) {
2251 foreach ($matches[1] as $match) {
2252 if (strstr($match, ']')) {
2253 // we might be inside a bbcode color tag - leave it alone
2257 // try not to catch url fragments
2258 if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {
2267 DI::profiler()->stopRecording();
2268 return array_unique($ret);
2272 * Expand tags to URLs, checks the tag is at the start of a line or preceded by a non-word character
2274 * @param string $body HTML/BBCode
2275 * @return string body with expanded tags
2277 public static function expandTags(string $body): string
2279 return preg_replace_callback(
2280 "/(?<=\W|^)([!#@])([^\^ \x0D\x0A,;:?'\"]*[^\^ \x0D\x0A,;:?!'\".])/",
2281 function (array $match) {
2282 switch ($match[1]) {
2285 $contact = Contact::getByURL($match[2]);
2286 if (!empty($contact)) {
2287 return $match[1] . '[url=' . $contact['url'] . ']' . $contact['name'] . '[/url]';
2289 return $match[1] . $match[2];
2295 return $match[1] . '[url=' . DI::baseUrl() . '/search?tag=' . $match[2] . ']' . $match[2] . '[/url]';
2303 * Perform a custom function on a text after having escaped blocks enclosed in the provided tag list.
2305 * @param string $text HTML/BBCode
2306 * @param array $tagList A list of tag names, e.g ['noparse', 'nobb', 'pre']
2307 * @param callable $callback
2309 * @see Strings::performWithEscapedBlocks
2311 public static function performWithEscapedTags(string $text, array $tagList, callable $callback): string
2313 $tagList = array_map('preg_quote', $tagList);
2315 return Strings::performWithEscapedBlocks($text, '#\[(?:' . implode('|', $tagList) . ').*?\[/(?:' . implode('|', $tagList) . ')]#ism', $callback);
2319 * Replaces mentions in the provided message body in BBCode links for the provided user and network if any
2321 * @param string $body HTML/BBCode
2322 * @param int $profile_uid Profile user id
2323 * @param string $network Network name
2324 * @return string HTML/BBCode with inserted images
2325 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2326 * @throws \ImagickException
2328 public static function setMentions(string $body, $profile_uid = 0, $network = '')
2330 DI::profiler()->startRecording('rendering');
2331 $body = self::performWithEscapedTags($body, ['noparse', 'pre', 'code', 'img'], function ($body) use ($profile_uid, $network) {
2332 $tags = self::getTags($body);
2336 foreach ($tags as $tag) {
2337 $tag_type = substr($tag, 0, 1);
2339 if ($tag_type == Tag::TAG_CHARACTER[Tag::HASHTAG]) {
2344 * If we already tagged 'Robert Johnson', don't try and tag 'Robert'.
2345 * Robert Johnson should be first in the $tags array
2347 foreach ($tagged as $nextTag) {
2348 if (stristr($nextTag, $tag . ' ')) {
2353 if (($success = Item::replaceTag($body, $profile_uid, $tag, $network)) && $success['replaced']) {
2361 DI::profiler()->stopRecording();
2366 * @param string $author Author display name
2367 * @param string $profile Author profile URL
2368 * @param string $avatar Author profile picture URL
2369 * @param string $link Post source URL
2370 * @param string $posted Post created date
2371 * @param string|null $guid Post guid (if any)
2372 * @param string|null $uri Post uri (if any)
2374 * @TODO Rewrite to handle over whole record array
2376 public static function getShareOpeningTag(string $author, string $profile, string $avatar, string $link, string $posted, string $guid = null, string $uri = null): string
2378 DI::profiler()->startRecording('rendering');
2379 $header = "[share author='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $author) .
2380 "' profile='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $profile) .
2381 "' avatar='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $avatar) .
2382 "' link='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $link) .
2383 "' posted='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $posted);
2386 $header .= "' guid='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $guid);
2390 $header .= "' message_id='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $uri);
2395 DI::profiler()->stopRecording();
2400 * Returns the BBCode relevant to embed the provided URL in a post body.
2401 * For media type, it will return [img], [video] and [audio] tags.
2402 * For regular web pages, it will either output a [bookmark] tag if title and description were provided,
2403 * an [attachment] tag or a simple [url] tag depending on $tryAttachment.
2405 * @param string $url
2406 * @param bool $tryAttachment
2407 * @param string|null $title
2408 * @param string|null $description
2409 * @param string|null $tags
2411 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2412 * @see ParseUrl::getSiteinfoCached
2414 public static function embedURL(string $url, bool $tryAttachment = true, string $title = null, string $description = null, string $tags = null): string
2416 DI::profiler()->startRecording('rendering');
2417 DI::logger()->info($url);
2419 // If there is already some content information submitted we don't
2420 // need to parse the url for content.
2421 if (!empty($title) && !empty($description)) {
2422 $title = str_replace(["\r", "\n"], ['', ''], $title);
2424 $description = '[quote]' . trim($description) . '[/quote]' . "\n";
2427 if (!empty($tags)) {
2428 $arr_tags = ParseUrl::convertTagsToArray($tags);
2429 if (count($arr_tags)) {
2430 $str_tags = "\n" . implode(' ', $arr_tags) . "\n";
2434 $result = sprintf('[bookmark=%s]%s[/bookmark]%s', $url, ($title) ? $title : $url, $description) . $str_tags;
2436 DI::logger()->info('(unparsed): returns: ' . $result);
2438 DI::profiler()->stopRecording();
2442 $siteinfo = ParseUrl::getSiteinfoCached($url);
2444 if (in_array($siteinfo['type'], ['image', 'video', 'audio'])) {
2445 switch ($siteinfo['type']) {
2447 $bbcode = "\n" . '[video]' . $url . '[/video]' . "\n";
2450 $bbcode = "\n" . '[audio]' . $url . '[/audio]' . "\n";
2453 $bbcode = "\n" . '[img]' . $url . '[/img]' . "\n";
2457 DI::profiler()->stopRecording();
2461 unset($siteinfo['keywords']);
2463 // Bypass attachment if parse url for a comment
2464 if (!$tryAttachment) {
2465 DI::profiler()->stopRecording();
2466 return "\n" . '[url=' . $url . ']' . ($siteinfo['title'] ?? $url) . '[/url]';
2469 // Format it as BBCode attachment
2470 $bbcode = "\n" . PageInfo::getFooterFromData($siteinfo);
2471 DI::profiler()->stopRecording();