3 * @copyright Copyright (C) 2020, Friendica
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\OEmbed;
28 use Friendica\Content\Smilies;
29 use Friendica\Core\Hook;
30 use Friendica\Core\Logger;
31 use Friendica\Core\Protocol;
32 use Friendica\Core\Renderer;
33 use Friendica\Core\System;
35 use Friendica\Model\Contact;
36 use Friendica\Model\Event;
37 use Friendica\Model\Photo;
38 use Friendica\Network\Probe;
39 use Friendica\Object\Image;
40 use Friendica\Protocol\Activity;
41 use Friendica\Util\Images;
42 use Friendica\Util\Map;
43 use Friendica\Util\Network;
44 use Friendica\Util\ParseUrl;
45 use Friendica\Util\Proxy as ProxyUtils;
46 use Friendica\Util\Strings;
47 use Friendica\Util\XML;
58 const ACTIVITYPUB = 9;
61 * Fetches attachment data that were generated the old way
63 * @param string $body Message body
65 * 'type' -> Message type ('link', 'video', 'photo')
66 * 'text' -> Text before the shared message
67 * 'after' -> Text after the shared message
68 * 'image' -> Preview image of the message
69 * 'url' -> Url to the attached message
70 * 'title' -> Title of the attachment
71 * 'description' -> Description of the attachment
72 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
74 private static function getOldAttachmentData($body)
78 // Simplify image codes
79 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
81 if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) {
82 foreach ($attached as $data) {
83 if (!in_array($data[1], ['type-link', 'type-video', 'type-photo'])) {
87 $post['type'] = substr($data[1], 5);
89 $pos = strpos($body, $data[0]);
91 $post['text'] = trim(substr($body, 0, $pos));
92 $post['after'] = trim(substr($body, $pos + strlen($data[0])));
94 $post['text'] = trim(str_replace($data[0], '', $body));
98 $attacheddata = $data[2];
100 if (preg_match("/\[img\](.*?)\[\/img\]/ism", $attacheddata, $matches)) {
102 $picturedata = Images::getInfoFromURLCached($matches[1]);
105 if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) {
106 $post['image'] = $matches[1];
108 $post['preview'] = $matches[1];
113 if (preg_match("/\[bookmark\=(.*?)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) {
114 $post['url'] = $matches[1];
115 $post['title'] = $matches[2];
117 if (!empty($post['url']) && (in_array($post['type'], ['link', 'video']))
118 && preg_match("/\[url\=(.*?)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) {
119 $post['url'] = $matches[1];
122 // Search for description
123 if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) {
124 $post['description'] = $matches[1];
132 * Fetches attachment data that were generated with the "attachment" element
134 * @param string $body Message body
136 * 'type' -> Message type ('link', 'video', 'photo')
137 * 'text' -> Text before the shared message
138 * 'after' -> Text after the shared message
139 * 'image' -> Preview image of the message
140 * 'url' -> Url to the attached message
141 * 'title' -> Title of the attachment
142 * 'description' -> Description of the attachment
143 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
145 public static function getAttachmentData($body)
157 if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) {
158 return self::getOldAttachmentData($body);
161 $attributes = $match[2];
163 $data['text'] = trim($match[1]);
166 preg_match("/type='(.*?)'/ism", $attributes, $matches);
167 if (!empty($matches[1])) {
168 $type = strtolower($matches[1]);
171 preg_match('/type="(.*?)"/ism', $attributes, $matches);
172 if (!empty($matches[1])) {
173 $type = strtolower($matches[1]);
180 if (!in_array($type, ['link', 'audio', 'photo', 'video'])) {
185 $data['type'] = $type;
189 preg_match("/url='(.*?)'/ism", $attributes, $matches);
190 if (!empty($matches[1])) {
194 preg_match('/url="(.*?)"/ism', $attributes, $matches);
195 if (!empty($matches[1])) {
200 $data['url'] = html_entity_decode($url, ENT_QUOTES, 'UTF-8');
204 preg_match("/title='(.*?)'/ism", $attributes, $matches);
205 if (!empty($matches[1])) {
206 $title = $matches[1];
209 preg_match('/title="(.*?)"/ism', $attributes, $matches);
210 if (!empty($matches[1])) {
211 $title = $matches[1];
215 $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true);
216 $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8');
217 $title = str_replace(['[', ']'], ['[', ']'], $title);
218 $data['title'] = $title;
222 preg_match("/image='(.*?)'/ism", $attributes, $matches);
223 if (!empty($matches[1])) {
224 $image = $matches[1];
227 preg_match('/image="(.*?)"/ism', $attributes, $matches);
228 if (!empty($matches[1])) {
229 $image = $matches[1];
233 $data['image'] = html_entity_decode($image, ENT_QUOTES, 'UTF-8');
237 preg_match("/preview='(.*?)'/ism", $attributes, $matches);
238 if (!empty($matches[1])) {
239 $preview = $matches[1];
242 preg_match('/preview="(.*?)"/ism', $attributes, $matches);
243 if (!empty($matches[1])) {
244 $preview = $matches[1];
247 if ($preview != '') {
248 $data['preview'] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8');
251 $data['description'] = trim($match[3]);
253 $data['after'] = trim($match[4]);
258 public static function getAttachedData($body, $item = [])
262 - type: link, video, photo
270 $has_title = !empty($item['title']);
271 $plink = $item['plink'] ?? '';
272 $post = self::getAttachmentData($body);
274 // Get all linked images with alternative image description
275 if (preg_match_all("/\[img=([^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
276 foreach ($pictures as $picture) {
277 if (Photo::isLocal($picture[1])) {
278 $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]];
281 if (!empty($post['images']) && !empty($post['images'][0]['description'])) {
282 $post['image_description'] = $post['images'][0]['description'];
286 if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) {
287 foreach ($pictures as $picture) {
288 if (Photo::isLocal($picture[1])) {
289 $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => ''];
294 // if nothing is found, it maybe having an image.
295 if (!isset($post['type'])) {
296 // Simplify image codes
297 $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body);
298 $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body);
299 $post['text'] = $body;
301 if (preg_match_all("#\[url=([^\]]+?)\]\s*\[img\]([^\[]+?)\[/img\]\s*\[/url\]#ism", $body, $pictures, PREG_SET_ORDER)) {
302 if ((count($pictures) == 1) && !$has_title) {
303 if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) {
304 // Replace the preview picture with the real picture
305 $url = str_replace('-1.', '-0.', $pictures[0][2]);
306 $data = ['url' => $url, 'type' => 'photo'];
308 // Checking, if the link goes to a picture
309 $data = ParseUrl::getSiteinfoCached($pictures[0][1], true);
313 // Sometimes photo posts to the own album are not detected at the start.
314 // So we seem to cannot use the cache for these cases. That's strange.
315 if (($data['type'] != 'photo') && strstr($pictures[0][1], "/photos/")) {
316 $data = ParseUrl::getSiteinfo($pictures[0][1], true);
319 if ($data['type'] == 'photo') {
320 $post['type'] = 'photo';
321 if (isset($data['images'][0])) {
322 $post['image'] = $data['images'][0]['src'];
323 $post['url'] = $data['url'];
325 $post['image'] = $data['url'];
328 $post['preview'] = $pictures[0][2];
329 $post['text'] = trim(str_replace($pictures[0][0], '', $body));
331 $imgdata = Images::getInfoFromURLCached($pictures[0][1]);
332 if ($imgdata && substr($imgdata['mime'], 0, 6) == 'image/') {
333 $post['type'] = 'photo';
334 $post['image'] = $pictures[0][1];
335 $post['preview'] = $pictures[0][2];
336 $post['text'] = trim(str_replace($pictures[0][0], '', $body));
339 } elseif (count($pictures) > 0) {
340 $post['type'] = 'link';
341 $post['url'] = $plink;
342 $post['image'] = $pictures[0][2];
343 $post['text'] = $body;
345 foreach ($pictures as $picture) {
346 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
349 } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) {
350 if ((count($pictures) == 1) && !$has_title) {
351 $post['type'] = 'photo';
352 $post['image'] = $pictures[0][1];
353 $post['text'] = str_replace($pictures[0][0], '', $body);
354 } elseif (count($pictures) > 0) {
355 $post['type'] = 'link';
356 $post['url'] = $plink;
357 $post['image'] = $pictures[0][1];
358 $post['text'] = $body;
360 foreach ($pictures as $picture) {
361 $post['text'] = trim(str_replace($picture[0], '', $post['text']));
366 // Test for the external links
367 preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER);
368 preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER);
370 $links = array_merge($links1, $links2);
372 // If there is only a single one, then use it.
373 // This should cover link posts via API.
374 if ((count($links) == 1) && !isset($post['preview']) && !$has_title) {
375 $post['type'] = 'link';
376 $post['url'] = $links[0][1];
379 // Simplify "video" element
380 $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']);
382 // Now count the number of external media links
383 preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER);
384 preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER);
385 preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER);
386 preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER);
388 // Add them to the other external links
389 $links = array_merge($links, $links1, $links2, $links3, $links4);
391 // Are there more than one?
392 if (count($links) > 1) {
393 // The post will be the type "text", which means a blog post
394 unset($post['type']);
395 $post['url'] = $plink;
398 if (!isset($post['type'])) {
399 $post['type'] = "text";
400 $post['text'] = trim($body);
402 } elseif (isset($post['url']) && ($post['type'] == 'video')) {
403 $data = ParseUrl::getSiteinfoCached($post['url'], true);
405 if (isset($data['images'][0])) {
406 $post['image'] = $data['images'][0]['src'];
414 * Remove [attachment] BBCode and replaces it with a regular [url]
416 * @param string $body
417 * @param boolean $no_link_desc No link description
419 * @return string with replaced body
421 public static function removeAttachment($body, $no_link_desc = false)
423 return preg_replace_callback("/\s*\[attachment (.*)\](.*?)\[\/attachment\]\s*/ism",
424 function ($match) use ($no_link_desc) {
425 $attach_data = self::getAttachmentData($match[0]);
426 if (empty($attach_data['url'])) {
428 } elseif (empty($attach_data['title']) || $no_link_desc) {
429 return "\n[url]" . $attach_data['url'] . "[/url]\n";
431 return "\n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n";
437 * Converts a BBCode text into plaintext
440 * @param bool $keep_urls Whether to keep URLs in the resulting plaintext
444 public static function toPlaintext($text, $keep_urls = true)
446 $naked_text = HTML::toPlaintext(self::convert($text, false, 0, true), 0, !$keep_urls);
451 private static function proxyUrl($image, $simplehtml = self::INTERNAL)
453 // Only send proxied pictures to API and for internal display
454 if (in_array($simplehtml, [self::INTERNAL, self::API])) {
455 return ProxyUtils::proxifyUrl($image);
462 * This function changing the visual size (not the real size) of images.
463 * The function does not work for pictures with an alternate text description.
464 * This could only be changed by using some new "img" BBCode format.
466 * @param string $srctext The body with images
467 * @return string The body with possibly scaled images
469 public static function scaleExternalImages(string $srctext)
473 // Simplify image links
474 $s = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $s);
477 $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER);
479 foreach ($matches as $mtch) {
480 Logger::log('scale_external_image: ' . $mtch[1]);
482 $hostname = str_replace('www.', '', substr(DI::baseUrl(), strpos(DI::baseUrl(), '://') + 3));
483 if (stristr($mtch[1], $hostname)) {
487 $curlResult = Network::curl($mtch[1], true);
488 if (!$curlResult->isSuccess()) {
492 $i = $curlResult->getBody();
493 $type = $curlResult->getContentType();
494 $type = Images::getMimeTypeByData($i, $mtch[1], $type);
497 $Image = new Image($i, $type);
498 if ($Image->isValid()) {
499 $orig_width = $Image->getWidth();
500 $orig_height = $Image->getHeight();
502 if ($orig_width > 640 || $orig_height > 640) {
503 $Image->scaleDown(640);
504 $new_width = $Image->getWidth();
505 $new_height = $Image->getHeight();
506 Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG);
509 '[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]'
513 Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG);
524 * Truncates imported message body string length to max_import_size
526 * The purpose of this function is to apply system message length limits to
527 * imported messages without including any embedded photos in the length
529 * @param string $body
531 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
533 public static function limitBodySize($body)
535 $maxlen = DI::config()->get('config', 'max_import_size', 0);
537 // If the length of the body, including the embedded images, is smaller
538 // than the maximum, then don't waste time looking for the images
539 if ($maxlen && (strlen($body) > $maxlen)) {
541 Logger::log('the total body length exceeds the limit', Logger::DEBUG);
547 $img_start = strpos($orig_body, '[img');
548 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
549 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
550 while (($img_st_close !== false) && ($img_end !== false)) {
552 $img_st_close++; // make it point to AFTER the closing bracket
553 $img_end += $img_start;
554 $img_end += strlen('[/img]');
556 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
557 // This is an embedded image
559 if (($textlen + $img_start) > $maxlen) {
560 if ($textlen < $maxlen) {
561 Logger::log('the limit happens before an embedded image', Logger::DEBUG);
562 $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
566 $new_body = $new_body . substr($orig_body, 0, $img_start);
567 $textlen += $img_start;
570 $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start);
573 if (($textlen + $img_end) > $maxlen) {
574 if ($textlen < $maxlen) {
575 Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG);
576 $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
580 $new_body = $new_body . substr($orig_body, 0, $img_end);
581 $textlen += $img_end;
584 $orig_body = substr($orig_body, $img_end);
586 if ($orig_body === false) {
587 // in case the body ends on a closing image tag
591 $img_start = strpos($orig_body, '[img');
592 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
593 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
596 if (($textlen + strlen($orig_body)) > $maxlen) {
597 if ($textlen < $maxlen) {
598 Logger::log('the limit happens after the end of the last image', Logger::DEBUG);
599 $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
602 Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG);
603 $new_body = $new_body . $orig_body;
613 * Processes [attachment] tags
615 * Note: Can produce a [bookmark] tag in the returned string
617 * @param string $text
618 * @param integer $simplehtml
619 * @param bool $tryoembed
621 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
623 private static function convertAttachment($text, $simplehtml = self::INTERNAL, $tryoembed = true)
625 $data = self::getAttachmentData($text);
626 if (empty($data) || empty($data['url'])) {
630 if (isset($data['title'])) {
631 $data['title'] = strip_tags($data['title']);
632 $data['title'] = str_replace(['http://', 'https://'], '', $data['title']);
634 $data['title'] = null;
637 if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) {
638 $data['preview'] = $data['image'];
644 if ($tryoembed && OEmbed::isAllowedURL($data['url'])) {
645 $return = OEmbed::getHTML($data['url'], $data['title']);
647 throw new Exception('OEmbed is disabled for this attachment.');
649 } catch (Exception $e) {
650 $data['title'] = ($data['title'] ?? '') ?: $data['url'];
652 if ($simplehtml != self::CONNECTORS) {
653 $return = sprintf('<div class="type-%s">', $data['type']);
656 if (!empty($data['title']) && !empty($data['url'])) {
657 if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
658 $return .= sprintf('<a href="%s" target="_blank" rel="noopener noreferrer"><img src="%s" alt="" title="%s" class="attachment-image" /></a>', $data['url'], self::proxyUrl($data['image'], $simplehtml), $data['title']);
660 if (!empty($data['image'])) {
661 $return .= sprintf('<a href="%s" target="_blank" rel="noopener noreferrer"><img src="%s" alt="" title="%s" class="attachment-image" /></a><br />', $data['url'], self::proxyUrl($data['image'], $simplehtml), $data['title']);
662 } elseif (!empty($data['preview'])) {
663 $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), $data['title']);
665 $return .= sprintf('<h4><a href="%s">%s</a></h4>', $data['url'], $data['title']);
669 if (!empty($data['description']) && $data['description'] != $data['title']) {
670 // Sanitize the HTML by converting it to BBCode
671 $bbcode = HTML::toBBCode($data['description']);
672 $return .= sprintf('<blockquote>%s</blockquote>', trim(self::convert($bbcode)));
675 if (!empty($data['url'])) {
676 $return .= sprintf('<sup><a href="%s">%s</a></sup>', $data['url'], parse_url($data['url'], PHP_URL_HOST));
679 if ($simplehtml != self::CONNECTORS) {
684 return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? ''));
687 public static function removeShareInformation($Text, $plaintext = false, $nolink = false)
689 $data = self::getAttachmentData($Text);
694 return $data['text'] . ($data['after'] ?? '');
697 $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false);
698 $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false);
699 if ($plaintext || (($title != '') && strstr($text, $title))) {
700 $data['title'] = $data['url'];
701 } elseif (($text != '') && strstr($title, $text)) {
702 $data['text'] = $data['title'];
703 $data['title'] = $data['url'];
706 if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) {
707 return $data['title'] . $data['after'];
710 // If the link already is included in the post, don't add it again
711 if (!empty($data['url']) && strpos($data['text'], $data['url'])) {
712 return $data['text'] . $data['after'];
715 $text = $data['text'];
717 if (!empty($data['url']) && !empty($data['title'])) {
718 $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]';
719 } elseif (!empty($data['url'])) {
720 $text .= "\n[url]" . $data['url'] . '[/url]';
723 return $text . "\n" . $data['after'];
727 * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function)
729 * @param array $match Array with the matching values
730 * @return string reformatted link including HTML codes
732 private static function convertUrlForActivityPubCallback($match)
736 if (isset($match[2]) && ($match[1] != $match[2])) {
740 $parts = parse_url($url);
741 if (!isset($parts['scheme'])) {
745 return self::convertUrlForActivityPub($url);
749 * Converts [url] BBCodes in a format that looks fine on ActivityPub systems.
751 * @param string $url URL that is about to be reformatted
752 * @return string reformatted link including HTML codes
754 private static function convertUrlForActivityPub($url)
756 $html = '<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>';
757 return sprintf($html, $url, self::getStyledURL($url));
761 * Converts an URL in a nicer format (without the scheme and possibly shortened)
763 * @param string $url URL that is about to be reformatted
764 * @return string reformatted link
766 private static function getStyledURL($url)
768 $parts = parse_url($url);
769 $scheme = $parts['scheme'] . '://';
770 $styled_url = str_replace($scheme, '', $url);
772 if (strlen($styled_url) > 30) {
773 $styled_url = substr($styled_url, 0, 30) . "…";
780 * [noparse][i]italic[/i][/noparse] turns into
781 * [noparse][ i ]italic[ /i ][/noparse],
782 * to hide them from parser.
784 private static function escapeNoparseCallback($match)
786 $whole_match = $match[0];
787 $captured = $match[1];
788 $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured);
789 $new_str = str_replace($captured, $spacefied, $whole_match);
794 * The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
795 * now turns back and the [noparse] tags are trimed
796 * returning [i]italic[/i]
798 private static function unescapeNoparseCallback($match)
800 $captured = $match[1];
801 $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured);
806 * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first
809 * @param string $text Text to search
810 * @param string $name Tag name
811 * @param int $occurrences Number of first occurrences to skip
812 * @return boolean|array
814 public static function getTagPosition($text, $name, $occurrences = 0)
816 if ($occurrences < 0) {
821 for ($i = 0; $i <= $occurrences; $i++) {
822 if ($start_open !== false) {
823 $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags
827 if ($start_open === false) {
831 $start_equal = strpos($text, '=', $start_open);
832 $start_close = strpos($text, ']', $start_open);
834 if ($start_close === false) {
840 $end_open = strpos($text, '[/' . $name . ']', $start_close);
842 if ($end_open === false) {
848 'open' => $start_open,
849 'close' => $start_close
853 'close' => $end_open + strlen('[/' . $name . ']')
857 if ($start_equal !== false) {
858 $res['start']['equal'] = $start_equal + 1;
865 * Performs a preg_replace within the boundaries of all named BBCode tags in a text
867 * @param string $pattern Preg pattern string
868 * @param string $replace Preg replace string
869 * @param string $name BBCode tag name
870 * @param string $text Text to search
873 public static function pregReplaceInTag($pattern, $replace, $name, $text)
876 $pos = self::getTagPosition($text, $name, $occurrences);
877 while ($pos !== false && $occurrences++ < 1000) {
878 $start = substr($text, 0, $pos['start']['open']);
879 $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']);
880 $end = substr($text, $pos['end']['close']);
881 if ($end === false) {
885 $subject = preg_replace($pattern, $replace, $subject);
886 $text = $start . $subject . $end;
888 $pos = self::getTagPosition($text, $name, $occurrences);
894 private static function extractImagesFromItemBody($body)
901 $img_start = strpos($orig_body, '[img');
902 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
903 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
904 while (($img_st_close !== false) && ($img_end !== false)) {
905 $img_st_close++; // make it point to AFTER the closing bracket
906 $img_end += $img_start;
908 if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) {
909 // This is an embedded image
910 $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close));
911 $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]';
915 $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]'));
918 $orig_body = substr($orig_body, $img_end + strlen('[/img]'));
920 if ($orig_body === false) {
921 // in case the body ends on a closing image tag
925 $img_start = strpos($orig_body, '[img');
926 $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false);
927 $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false);
930 $new_body = $new_body . $orig_body;
932 return ['body' => $new_body, 'images' => $saved_image];
935 private static function interpolateSavedImagesIntoItemBody($body, array $images)
940 foreach ($images as $image) {
941 // We're depending on the property of 'foreach' (specified on the PHP website) that
942 // it loops over the array starting from the first element and going sequentially
943 // to the last element
944 $newbody = str_replace('[$#saved_image' . $cnt . '#$]',
945 '<img src="' . self::proxyUrl($image) . '" alt="' . DI::l10n()->t('Image/photo') . '" />', $newbody);
953 * This function converts a [share] block to text according to a provided callback function whose signature is:
955 * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string
958 * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact
960 * - $author_contact is a contact record array
961 * - $content is the inner content of the [share] block
962 * - $is_quote_share indicates whether there's any content before the [share] block
963 * - Return value is the string that should replace the [share] block in the provided text
965 * This function is intended to be used by addon connector to format a share block like the target network is expecting it.
967 * @param string $text A BBCode string
968 * @param callable $callback
969 * @return string The BBCode string with all [share] blocks replaced
971 public static function convertShare($text, callable $callback)
973 $return = preg_replace_callback(
974 "/(.*?)\[share(.*?)\](.*)\[\/share\]/ism",
975 function ($match) use ($callback) {
976 $attribute_string = $match[2];
978 foreach (['author', 'profile', 'avatar', 'link', 'posted'] as $field) {
979 preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
980 $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
983 // We only call this so that a previously unknown contact can be added.
984 // This is important for the function "Model\Contact::getDetailsByURL()".
985 // This function then can fetch an entry from the contact table.
986 $default['url'] = $attributes['profile'];
988 if (!empty($attributes['author'])) {
989 $default['name'] = $attributes['author'];
992 if (!empty($attributes['avatar'])) {
993 $default['photo'] = $attributes['avatar'];
996 Contact::getIdForURL($attributes['profile'], 0, true, $default);
998 $author_contact = Contact::getDetailsByURL($attributes['profile']);
999 $author_contact['url'] = ($author_contact['url'] ?? $attributes['profile']);
1000 $author_contact['addr'] = ($author_contact['addr'] ?? '') ?: Protocol::getAddrFromProfileUrl($attributes['profile']);
1002 $attributes['author'] = ($author_contact['name'] ?? '') ?: $attributes['author'];
1003 $attributes['avatar'] = ($author_contact['micro'] ?? '') ?: $attributes['avatar'];
1004 $attributes['profile'] = ($author_contact['url'] ?? '') ?: $attributes['profile'];
1006 if ($attributes['avatar']) {
1007 $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB);
1010 return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != '');
1019 * Default [share] tag conversion callback
1021 * Note: Can produce a [bookmark] tag in the output
1023 * @see BBCode::convertShare()
1024 * @param array $attributes [share] block attribute values
1025 * @param array $author_contact Contact row of the shared author
1026 * @param string $content Inner content of the [share] block
1027 * @param boolean $is_quote_share Whether there is content before the [share] block
1028 * @param integer $simplehtml Mysterious integer value depending on the target network/formatting style
1030 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1032 private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml)
1034 $mention = Protocol::formatMention($attributes['profile'], $attributes['author']);
1036 switch ($simplehtml) {
1038 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('♲ ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ': </p>' . "\n" . $content;
1040 case self::DIASPORA:
1041 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) {
1042 $text = ($is_quote_share? '<hr />' : '') . '<p><a href="' . $attributes['link'] . '">' . $attributes['link'] . '</a></p>' . "\n";
1044 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a>:</b></p>' . "\n";
1046 if (!empty($attributes['posted']) && !empty($attributes['link'])) {
1047 $headline = '<p><b>♲ <a href="' . $attributes['profile'] . '">' . $attributes['author'] . '</a></b> - <a href="' . $attributes['link'] . '">' . $attributes['posted'] . ' GMT</a></p>' . "\n";
1050 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote>' . trim($content) . '</blockquote>' . "\n";
1052 if (empty($attributes['posted']) && !empty($attributes['link'])) {
1053 $text .= '<p><a href="' . $attributes['link'] . '">[Source]</a></p>' . "\n";
1058 case self::CONNECTORS:
1059 $headline = '<p><b>' . html_entity_decode('♲ ', ENT_QUOTES, 'UTF-8');
1060 $headline .= DI::l10n()->t('<a href="%1$s" target="_blank" rel="noopener noreferrer">%2$s</a> %3$s', $attributes['link'], $mention, $attributes['posted']);
1061 $headline .= ':</b></p>' . "\n";
1063 $text = ($is_quote_share? '<hr />' : '') . $headline . '<blockquote class="shared_content">' . trim($content) . '</blockquote>' . "\n";
1067 $text = ($is_quote_share? '<br />' : '') . '<p>' . html_entity_decode('♲ ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '</p>' . "\n";
1069 case self::ACTIVITYPUB:
1070 $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>';
1071 $text = '<div><a href="' . $attributes['link'] . '">' . html_entity_decode('♲', ENT_QUOTES, 'UTF-8') . '</a> ' . $author . '<blockquote>' . $content . '</blockquote></div>' . "\n";
1074 // Transforms quoted tweets in rich attachments to avoid nested tweets
1075 if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($attributes['link'])) {
1077 $text = ($is_quote_share? '<br />' : '') . OEmbed::getHTML($attributes['link']);
1078 } catch (Exception $e) {
1079 $text = ($is_quote_share? '<br />' : '') . sprintf('[bookmark=%s]%s[/bookmark]', $attributes['link'], $content);
1082 $text = ($is_quote_share? "\n" : '');
1084 $tpl = Renderer::getMarkupTemplate('shared_content.tpl');
1085 $text .= Renderer::replaceMacros($tpl, [
1086 '$profile' => $attributes['profile'],
1087 '$avatar' => $attributes['avatar'],
1088 '$author' => $attributes['author'],
1089 '$link' => $attributes['link'],
1090 '$posted' => $attributes['posted'],
1091 '$content' => trim($content)
1100 private static function removePictureLinksCallback($match)
1102 $cache_key = 'remove:' . $match[1];
1103 $text = DI::cache()->get($cache_key);
1105 if (is_null($text)) {
1108 $stamp1 = microtime(true);
1110 $ch = @curl_init($match[1]);
1111 @curl_setopt($ch, CURLOPT_NOBODY, true);
1112 @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1113 @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1115 $curl_info = @curl_getinfo($ch);
1117 DI::profiler()->saveTimestamp($stamp1, "network", System::callstack());
1119 if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1120 $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]";
1122 $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]";
1124 // if its not a picture then look if its a page that contains a picture link
1125 $body = Network::fetchUrl($match[1]);
1127 $doc = new DOMDocument();
1128 @$doc->loadHTML($body);
1129 $xpath = new DOMXPath($doc);
1130 $list = $xpath->query("//meta[@name]");
1131 foreach ($list as $node) {
1134 if ($node->attributes->length) {
1135 foreach ($node->attributes as $attribute) {
1136 $attr[$attribute->name] = $attribute->value;
1140 if (strtolower($attr['name']) == 'twitter:image') {
1141 $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]';
1145 DI::cache()->set($cache_key, $text);
1151 private static function expandLinksCallback($match)
1153 if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) {
1154 return ($match[1] . "[url]" . $match[2] . "[/url]");
1156 return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]");
1160 private static function cleanPictureLinksCallback($match)
1164 // When the picture link is the own photo path then we can avoid fetching the link
1165 $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()->get()) . '/photos/');
1166 if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) {
1167 if (!empty($match[3])) {
1168 $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]';
1170 $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]';
1175 $cache_key = 'clean:' . $match[1];
1176 $text = DI::cache()->get($cache_key);
1177 if (!is_null($text)) {
1181 // Only fetch the header, not the content
1182 $stamp1 = microtime(true);
1184 $ch = @curl_init($match[1]);
1185 @curl_setopt($ch, CURLOPT_NOBODY, true);
1186 @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
1187 @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent());
1189 $curl_info = @curl_getinfo($ch);
1191 DI::profiler()->saveTimestamp($stamp1, "network", System::callstack());
1193 // if its a link to a picture then embed this picture
1194 if (substr($curl_info['content_type'], 0, 6) == 'image/') {
1195 $text = '[img]' . $match[1] . '[/img]';
1197 if (!empty($match[3])) {
1198 $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]';
1200 $text = '[img]' . $match[2] . '[/img]';
1203 // if its not a picture then look if its a page that contains a picture link
1204 $body = Network::fetchUrl($match[1]);
1206 $doc = new DOMDocument();
1207 @$doc->loadHTML($body);
1208 $xpath = new DOMXPath($doc);
1209 $list = $xpath->query("//meta[@name]");
1210 foreach ($list as $node) {
1212 if ($node->attributes->length) {
1213 foreach ($node->attributes as $attribute) {
1214 $attr[$attribute->name] = $attribute->value;
1218 if (strtolower($attr['name']) == "twitter:image") {
1219 if (!empty($match[3])) {
1220 $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]";
1222 $text = "[img]" . $attr['content'] . "[/img]";
1227 DI::cache()->set($cache_key, $text);
1232 public static function cleanPictureLinks($text)
1234 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text);
1235 $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return);
1240 * Converts a BBCode message to HTML message
1242 * BBcode 2 HTML was written by WAY2WEB.net
1243 * extended to work with Mistpark/Friendica - Mike Macgirvin
1245 * Simple HTML values meaning:
1246 * - 0: Friendica display
1248 * - 2: Used for Windows Phone push, Friendica API
1249 * - 3: Used before converting to Markdown in bb2diaspora.php
1250 * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr
1253 * - 7: Used for dfrn, OStatus
1254 * - 8: Used for Twitter, WP backlink text setting
1257 * @param string $text
1258 * @param bool $try_oembed
1259 * @param int $simple_html
1260 * @param bool $for_plaintext
1262 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1264 public static function convert($text, $try_oembed = true, $simple_html = self::INTERNAL, $for_plaintext = false)
1269 * preg_match_callback function to replace potential Oembed tags with Oembed content
1271 * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag]
1273 * $match[2] = $title or absent
1275 $try_oembed_callback = function ($match)
1278 $title = $match[2] ?? null;
1281 $return = OEmbed::getHTML($url, $title);
1282 } catch (Exception $ex) {
1283 $return = $match[0];
1289 // Extracting code blocks before the whitespace processing and the autolinker
1292 $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism",
1293 function ($matches) use (&$codeblocks) {
1294 $return = '#codeblock-' . count($codeblocks) . '#';
1295 if (strpos($matches[2], "\n") !== false) {
1296 $codeblocks[] = '<pre><code class="language-' . trim($matches[1]) . '">' . htmlspecialchars(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '</code></pre>';
1298 $codeblocks[] = '<code>' . htmlspecialchars($matches[2], ENT_NOQUOTES, 'UTF-8') . '</code>';
1306 // Hide all [noparse] contained bbtags by spacefying them
1307 // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image?
1309 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::escapeNoparseCallback', $text);
1310 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::escapeNoparseCallback', $text);
1311 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::escapeNoparseCallback', $text);
1313 // Remove the abstract element. It is a non visible element.
1314 $text = self::stripAbstract($text);
1316 // Move all spaces out of the tags
1317 $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text);
1318 $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text);
1320 // Extract the private images which use data urls since preg has issues with
1321 // large data sizes. Stash them away while we do bbcode conversion, and then put them back
1322 // in after we've done all the regex matching. We cannot use any preg functions to do this.
1324 $extracted = self::extractImagesFromItemBody($text);
1325 $text = $extracted['body'];
1326 $saved_image = $extracted['images'];
1328 // If we find any event code, turn it into an event.
1329 // After we're finished processing the bbcode we'll
1330 // replace all of the event code with a reformatted version.
1332 $ev = Event::fromBBCode($text);
1334 // Replace any html brackets with HTML Entities to prevent executing HTML or script
1335 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
1337 $text = str_replace("<", "<", $text);
1338 $text = str_replace(">", ">", $text);
1340 // remove some newlines before the general conversion
1341 $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text);
1342 $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text);
1344 // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems
1346 $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text);
1349 // Convert new line chars to html <br /> tags
1351 // nlbr seems to be hopelessly messed up
1352 // $Text = nl2br($Text);
1354 // We'll emulate it.
1356 $text = trim($text);
1357 $text = str_replace("\r\n", "\n", $text);
1359 // Remove linefeeds inside of the table elements. See issue #6799
1360 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ",
1361 "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ",
1362 "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
1363 "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "];
1364 $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
1365 "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]",
1366 "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
1367 "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"];
1370 $text = str_replace($search, $replace, $text);
1371 } while ($oldtext != $text);
1373 // Replace these here only once
1374 $search = ["\n[table]", "[/table]\n"];
1375 $replace = ["[table]", "[/table]"];
1376 $text = str_replace($search, $replace, $text);
1378 // removing multiplicated newlines
1379 if (DI::config()->get('system', 'remove_multiplicated_lines')) {
1380 $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
1381 "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"];
1382 $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
1383 "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"];
1386 $text = str_replace($search, $replace, $text);
1387 } while ($oldtext != $text);
1390 /// @todo Have a closer look at the different html modes
1391 // Handle attached links or videos
1392 if ($simple_html == self::ACTIVITYPUB) {
1393 $text = self::removeAttachment($text);
1394 } elseif (!in_array($simple_html, [self::INTERNAL, self::CONNECTORS])) {
1395 $text = self::removeAttachment($text, true);
1397 $text = self::convertAttachment($text, $simple_html, $try_oembed);
1400 // leave open the posibility of [map=something]
1401 // this is replaced in Item::prepareBody() which has knowledge of the item location
1402 if (strpos($text, '[/map]') !== false) {
1403 $text = preg_replace_callback(
1404 "/\[map\](.*?)\[\/map\]/ism",
1405 function ($match) use ($simple_html) {
1406 return str_replace($match[0], '<p class="map">' . Map::byLocation($match[1], $simple_html) . '</p>', $match[0]);
1412 if (strpos($text, '[map=') !== false) {
1413 $text = preg_replace_callback(
1414 "/\[map=(.*?)\]/ism",
1415 function ($match) use ($simple_html) {
1416 return str_replace($match[0], '<p class="map">' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '</p>', $match[0]);
1422 if (strpos($text, '[map]') !== false) {
1423 $text = preg_replace("/\[map\]/", '<p class="map"></p>', $text);
1426 // Check for headers
1427 $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '<h1>$1</h1>', $text);
1428 $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '<h2>$1</h2>', $text);
1429 $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '<h3>$1</h3>', $text);
1430 $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '<h4>$1</h4>', $text);
1431 $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '<h5>$1</h5>', $text);
1432 $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '<h6>$1</h6>', $text);
1434 // Check for paragraph
1435 $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
1437 // Check for bold text
1438 $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '<strong>$1</strong>', $text);
1440 // Check for Italics text
1441 $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '<em>$1</em>', $text);
1443 // Check for Underline text
1444 $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '<u>$1</u>', $text);
1446 // Check for strike-through text
1447 $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '<s>$1</s>', $text);
1449 // Check for over-line text
1450 $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '<span class="overline">$1</span>', $text);
1452 // Check for colored text
1453 $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "<span style=\"color: $1;\">$2</span>", $text);
1455 // Check for sized text
1456 // [size=50] --> font-size: 50px (with the unit).
1457 if ($simple_html != self::DIASPORA) {
1458 $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1px; line-height: initial;\">$2</span>", $text);
1459 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "<span style=\"font-size: $1; line-height: initial;\">$2</span>", $text);
1461 // Issue 2199: Diaspora doesn't interpret the construct above, nor the <small> or <big> element
1462 $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text);
1466 // Check for centered text
1467 $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "<div style=\"text-align:center;\">$1</div>", $text);
1469 // Check for list text
1470 $text = str_replace("[*]", "<li>", $text);
1472 // Check for style sheet commands
1473 $text = preg_replace_callback(
1474 "(\[style=(.*?)\](.*?)\[\/style\])ism",
1476 return "<span style=\"" . HTML::sanitizeCSS($match[1]) . ";\">" . $match[2] . "</span>";
1481 // Check for CSS classes
1482 $text = preg_replace_callback(
1483 "(\[class=(.*?)\](.*?)\[\/class\])ism",
1485 return "<span class=\"" . HTML::sanitizeCSS($match[1]) . "\">" . $match[2] . "</span>";
1490 // handle nested lists
1493 while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) ||
1494 ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) ||
1495 ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) ||
1496 ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) {
1497 $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1498 $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '<ul class="listnone" style="list-style-type: none;">$1</ul>', $text);
1499 $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1500 $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '<ul class="listlowerroman" style="list-style-type: lower-roman;">$2</ul>', $text);
1501 $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '<ul class="listupperroman" style="list-style-type: upper-roman;">$2</ul>', $text);
1502 $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '<ul class="listloweralpha" style="list-style-type: lower-alpha;">$2</ul>', $text);
1503 $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '<ul class="listupperalpha" style="list-style-type: upper-alpha;">$2</ul>', $text);
1504 $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '<ul class="listbullet" style="list-style-type: circle;">$1</ul>', $text);
1505 $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '<ul class="listdecimal" style="list-style-type: decimal;">$1</ul>', $text);
1506 $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '<li>$1</li>', $text);
1509 $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '<th>$1</th>', $text);
1510 $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '<td>$1</td>', $text);
1511 $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '<tr>$1</tr>', $text);
1512 $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '<table>$1</table>', $text);
1514 $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '<table border="1" >$1</table>', $text);
1515 $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '<table border="0" >$1</table>', $text);
1517 $text = str_replace('[hr]', '<hr />', $text);
1519 if (!$for_plaintext) {
1522 // Escaping BBCodes susceptible to contain rogue URL we don'' want the autolinker to catch
1523 $text = preg_replace_callback('#\[(url|img|audio|video|youtube|vimeo|share|attachment|iframe|bookmark).+?\[/\1\]#ism',
1524 function ($matches) use (&$escaped) {
1525 $return = '{escaped-' . count($escaped) . '}';
1526 $escaped[] = $matches[0];
1533 // Autolinker for isolated URLs
1534 $text = preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text);
1536 // Restoring escaped blocks
1537 $text = preg_replace_callback('/{escaped-([0-9]+)}/iU',
1538 function ($matches) use ($escaped) {
1539 return $escaped[intval($matches[1])] ?? $matches[0];
1545 // This is actually executed in Item::prepareBody()
1547 $nosmile = strpos($text, '[nosmile]') !== false;
1548 $text = str_replace('[nosmile]', '', $text);
1550 // Check for font change text
1551 $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "<span style=\"font-family: $1;\">$2</span>", $text);
1553 // Declare the format for [spoiler] layout
1554 $SpoilerLayout = '<details class="spoiler"><summary>' . DI::l10n()->t('Click to open/close') . '</summary>$1</details>';
1556 // Check for [spoiler] text
1557 // handle nested quotes
1559 while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) {
1560 $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text);
1563 // Check for [spoiler=Title] text
1565 // handle nested quotes
1567 while ((strpos($text, "[/spoiler]")!== false) && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) {
1568 $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism",
1569 '<details class="spoiler"><summary>$1</summary>$2</details>',
1573 // Declare the format for [quote] layout
1574 $QuoteLayout = '<blockquote>$1</blockquote>';
1576 // Check for [quote] text
1577 // handle nested quotes
1579 while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) {
1580 $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text);
1583 // Check for [quote=Author] text
1585 $t_wrote = DI::l10n()->t('$1 wrote:');
1587 // handle nested quotes
1589 while ((strpos($text, "[/quote]")!== false) && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) {
1590 $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism",
1591 "<p><strong class=".'"author"'.">" . $t_wrote . "</strong></p><blockquote>$2</blockquote>",
1596 // [img=widthxheight]image source[/img]
1597 $text = preg_replace_callback(
1598 "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism",
1599 function ($matches) use ($simple_html) {
1600 if (strpos($matches[3], "data:image/") === 0) {
1604 $matches[3] = self::proxyUrl($matches[3], $simple_html);
1605 return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]";
1610 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '<img src="$3" style="width: $1px;" >', $text);
1611 $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '<img class="zrl" src="$3" style="width: $1px;" >', $text);
1613 $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism",
1614 function ($matches) use ($simple_html) {
1615 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1616 $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
1617 return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
1622 // [img]pathtoimage[/img]
1623 $text = preg_replace_callback(
1624 "/\[img\](.*?)\[\/img\]/ism",
1625 function ($matches) use ($simple_html) {
1626 if (strpos($matches[1], "data:image/") === 0) {
1630 $matches[1] = self::proxyUrl($matches[1], $simple_html);
1631 return "[img]" . $matches[1] . "[/img]";
1636 $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1637 $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '<img src="$1" alt="' . DI::l10n()->t('Image/photo') . '" />', $text);
1639 $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);
1640 $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 />', $text);
1641 //$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);
1643 // Simplify "video" element
1644 $text = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
1648 $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);
1649 $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '<audio src="$1" controls="controls"><a href="$1">$1</a></audio>', $text);
1651 $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text);
1652 $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text);
1654 $text = preg_replace("/\[video\](.*?)\[\/video\]/ism",
1655 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1656 $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
1657 '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1660 // html5 video and audio
1664 $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<iframe src="$1" width="' . $a->videowidth . '" height="' . $a->videoheight . '"><a href="$1">$1</a></iframe>', $text);
1666 $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
1669 // Youtube extensions
1671 $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1672 $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1673 $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
1676 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1677 $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1678 $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
1681 $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);
1683 $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism",
1684 '<a href="https://www.youtube.com/watch?v=$1" target="_blank" rel="noopener noreferrer">https://www.youtube.com/watch?v=$1</a>', $text);
1688 $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1689 $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
1692 $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1693 $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
1696 $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);
1698 $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism",
1699 '<a href="https://vimeo.com/$1" target="_blank" rel="noopener noreferrer">https://vimeo.com/$1</a>', $text);
1703 $text = OEmbed::BBCode2HTML($text);
1705 // Avoid triple linefeeds through oembed
1706 $text = str_replace("<br style='clear:left'></span><br /><br />", "<br style='clear:left'></span><br />", $text);
1708 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
1709 // Replace the event-start section with the entire formatted event. The other bbcode is stripped.
1710 // Summary (e.g. title) is required, earlier revisions only required description (in addition to
1711 // start which is always required). Allow desc with a missing summary for compatibility.
1713 if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) {
1714 $sub = Event::getHTML($ev, $simple_html);
1716 $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text);
1717 $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text);
1718 $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text);
1719 $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text);
1720 $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text);
1721 $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text);
1722 $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
1725 // Replace non graphical smilies for external posts
1726 if (!$nosmile && !$for_plaintext) {
1727 $text = Smilies::replace($text);
1730 if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA)) {
1731 $conv = html_entity_decode(str_replace([' ', "\n", "\r"], '', $text));
1732 // Emojis are always 4 byte Unicode characters
1733 if (!empty($conv) && (strlen($conv) / mb_strlen($conv) == 4)) {
1734 $text = '<span style="font-size: xx-large; line-height: initial;">' . $text . '</span>';
1738 if (!$for_plaintext) {
1739 if (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1740 $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1741 $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text);
1744 $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text);
1745 $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text);
1748 $text = str_replace(["\r","\n"], ['<br />', '<br />'], $text);
1750 // Remove all hashtag addresses
1751 if ($simple_html && !in_array($simple_html, [self::DIASPORA, self::OSTATUS, self::ACTIVITYPUB])) {
1752 $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
1753 } elseif ($simple_html == self::DIASPORA) {
1754 // The ! is converted to @ since Diaspora only understands the @
1755 $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1756 '@<a href="$2">$3</a>',
1758 } elseif (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) {
1759 $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1760 '$1<span class="vcard"><a href="$2" class="url u-url mention" title="$3"><span class="fn nickname mention">$3</span></a></span>',
1762 } elseif (!$simple_html) {
1763 $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
1764 '$1<a href="$2" class="userinfo mention" title="$3">$3</a>',
1768 // Bookmarks in red - will be converted to bookmarks in friendica
1769 $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text);
1770 $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text);
1771 $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i",
1772 "[bookmark=$1]$2[/bookmark]", $text);
1774 if (in_array($simple_html, [self::API, self::OSTATUS, self::TWITTER])) {
1775 $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text);
1776 //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text);
1777 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text);
1780 // Perform URL Search
1782 $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text);
1785 $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text);
1787 // Handle Diaspora posts
1788 $text = preg_replace_callback(
1789 "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi",
1791 return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]";
1795 $text = preg_replace_callback(
1796 "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi",
1798 return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]";
1802 // Server independent link to posts and comments
1803 // See issue: https://github.com/diaspora/diaspora_federation/issues/75
1804 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism";
1805 $text = preg_replace($expression, DI::baseUrl()."/display/$1", $text);
1809 * - #[url=<anything>]<term>[/url]
1810 * - [url=<anything>]#<term>[/url]
1812 $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) use ($simple_html) {
1813 if ($simple_html == BBCode::ACTIVITYPUB) {
1814 return '<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1815 . '" data-tag="' . XML::escape($matches[1]) . '" rel="tag ugc">#'
1816 . XML::escape($matches[1]) . '</a>';
1818 return '#<a href="' . DI::baseUrl() . '/search?tag=' . rawurlencode($matches[1])
1819 . '" class="tag" rel="tag" title="' . XML::escape($matches[1]) . '">'
1820 . XML::escape($matches[1]) . '</a>';
1824 // We need no target="_blank" rel="noopener noreferrer" for local links
1825 // convert links start with DI::baseUrl() as local link without the target="_blank" rel="noopener noreferrer" attribute
1826 $escapedBaseUrl = preg_quote(DI::baseUrl(), '/');
1827 $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '<a href="$1">$1</a>', $text);
1828 $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '<a href="$1">$2</a>', $text);
1830 $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$1</a>', $text);
1831 $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1833 // Red compatibility, though the link can't be authenticated on Friendica
1834 $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '<a href="$1" target="_blank" rel="noopener noreferrer">$2</a>', $text);
1837 // we may need to restrict this further if it picks up too many strays
1838 // link acct:user@host to a webfinger profile redirector
1840 $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);
1842 // Perform MAIL Search
1843 $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '<a href="mailto:$1">$1</a>', $text);
1844 $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '<a href="mailto:$1">$2</a>', $text);
1846 // Unhide all [noparse] contained bbtags unspacefying them
1847 // and triming the [noparse] tag.
1849 $text = preg_replace_callback("/\[noparse\](.*?)\[\/noparse\]/ism", 'self::unescapeNoparseCallback', $text);
1850 $text = preg_replace_callback("/\[nobb\](.*?)\[\/nobb\]/ism", 'self::unescapeNoparseCallback', $text);
1851 $text = preg_replace_callback("/\[pre\](.*?)\[\/pre\]/ism", 'self::unescapeNoparseCallback', $text);
1853 /// @todo What is the meaning of these lines?
1854 $text = preg_replace('/\[\&\;([#a-z0-9]+)\;\]/', '&$1;', $text);
1855 $text = preg_replace('/\&\#039\;/', '\'', $text);
1857 // Currently deactivated, it made problems with " inside of alt texts.
1858 //$text = preg_replace('/\"\;/', '"', $text);
1860 // fix any escaped ampersands that may have been converted into links
1861 $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&\;(.*?)\>/ism', '<$1$2=$3&$4>', $text);
1863 // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails)
1864 $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:'];
1866 array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');});
1868 $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism',
1869 '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">', $text);
1871 // sanitize href attributes (only whitelisted protocols URLs)
1872 // default value for backward compatibility
1873 $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
1875 // Always allowed protocol even if config isn't set or not including it
1876 $allowed_link_protocols[] = '//';
1877 $allowed_link_protocols[] = 'http://';
1878 $allowed_link_protocols[] = 'https://';
1879 $allowed_link_protocols[] = 'redir/';
1881 array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');});
1883 $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism';
1884 $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text);
1887 $text = self::convertShare(
1889 function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) {
1890 return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html);
1895 $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image);
1898 // Restore code blocks
1899 $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU',
1900 function ($matches) use ($codeblocks) {
1901 $return = $matches[0];
1902 if (isset($codeblocks[intval($matches[1])])) {
1903 $return = $codeblocks[$matches[1]];
1910 // Clean up the HTML by loading and saving the HTML with the DOM.
1911 // Bad structured html can break a whole page.
1912 // For performance reasons do it only with activated item cache or at export.
1913 if (!$try_oembed || (get_itemcachepath() != '')) {
1914 $doc = new DOMDocument();
1915 $doc->preserveWhiteSpace = false;
1917 $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8");
1919 $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">';
1920 $encoding = '<?xml encoding="UTF-8">';
1921 @$doc->loadHTML($encoding . $doctype . '<html><body>' . $text . '</body></html>');
1922 $doc->encoding = 'UTF-8';
1923 $text = $doc->saveHTML();
1924 $text = str_replace(['<html><body>', '</body></html>', $doctype, $encoding], ['', '', '', ''], $text);
1926 $text = str_replace('<br></li>', '</li>', $text);
1928 //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES');
1931 // Clean up some useless linebreaks in lists
1932 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1933 //$Text = str_replace('</ul><br />', '</ul>', $Text);
1934 //$Text = str_replace('</li><br />', '</li>', $Text);
1935 //$Text = str_replace('<br /><li>', '<li>', $Text);
1936 //$Text = str_replace('<br /><ul', '<ul ', $Text);
1938 Hook::callAll('bbcode', $text);
1944 * Strips the "abstract" tag from the provided text
1946 * @param string $text The text with BBCode
1947 * @return string The same text - but without "abstract" element
1949 public static function stripAbstract($text)
1951 $text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
1952 $text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
1958 * Returns the value of the "abstract" element
1960 * @param string $text The text that maybe contains the element
1961 * @param string $addon The addon for which the abstract is meant for
1962 * @return string The abstract
1964 public static function getAbstract($text, $addon = '')
1968 $addon = strtolower($addon);
1970 if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
1971 foreach ($results AS $result) {
1972 $abstracts[strtolower($result[1])] = $result[2];
1976 if (isset($abstracts[$addon])) {
1977 $abstract = $abstracts[$addon];
1980 if ($abstract == '' && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
1981 $abstract = $result[1];
1988 * Callback function to replace a Friendica style mention in a mention for Diaspora
1990 * @param array $match Matching values for the callback
1991 * [1] = Mention type (! or @)
1994 * @return string Replaced mention
1995 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
1996 * @throws \ImagickException
1998 private static function bbCodeMention2DiasporaCallback($match)
2000 $contact = Contact::getDetailsByURL($match[3]);
2002 if (empty($contact['addr'])) {
2003 $contact = Probe::uri($match[3]);
2006 if (empty($contact['addr'])) {
2010 $mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
2015 * Converts a BBCode text into Markdown
2017 * This function converts a BBCode item body to be sent to Markdown-enabled
2018 * systems like Diaspora and Libertree
2020 * @param string $text
2021 * @param bool $for_diaspora Diaspora requires more changes than Libertree
2023 * @throws \Friendica\Network\HTTPException\InternalServerErrorException
2025 public static function toMarkdown($text, $for_diaspora = true)
2027 $original_text = $text;
2029 // Since Diaspora is creating a summary for links, this function removes them before posting
2030 if ($for_diaspora) {
2031 $text = self::removeShareInformation($text);
2035 * Transform #tags, strip off the [url] and replace spaces with underscore
2037 $url_search_string = "^\[\]";
2038 $text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
2039 function ($matches) {
2040 return '#' . str_replace(' ', '_', $matches[2]);
2045 // Converting images with size parameters to simple images. Markdown doesn't know it.
2046 $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
2048 // Convert it to HTML - don't try oembed
2049 if ($for_diaspora) {
2050 $text = self::convert($text, false, self::DIASPORA);
2052 // Add all tags that maybe were removed
2053 if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
2055 foreach ($tags[2] as $tag) {
2056 $tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2057 if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
2058 $tagline .= '#' . $tag . ' ';
2061 $text = $text . " " . $tagline;
2064 $text = self::convert($text, false, self::CONNECTORS);
2067 // If a link is followed by a quote then there should be a newline before it
2068 // Maybe we should make this newline at every time before a quote.
2069 $text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
2071 $stamp1 = microtime(true);
2073 // Now convert HTML to Markdown
2074 $text = HTML::toMarkdown($text);
2076 DI::profiler()->saveTimestamp($stamp1, "parser", System::callstack());
2078 // Libertree has a problem with escaped hashtags.
2079 $text = str_replace(['\#'], ['#'], $text);
2081 // Remove any leading or trailing whitespace, as this will mess up
2082 // the Diaspora signature verification and cause the item to disappear
2083 $text = trim($text);
2085 if ($for_diaspora) {
2086 $url_search_string = "^\[\]";
2087 $text = preg_replace_callback(
2088 "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism",
2089 ['self', 'bbCodeMention2DiasporaCallback'],
2094 Hook::callAll('bb2diaspora', $text);
2100 * Pull out all #hashtags and @person tags from $string.
2102 * We also get @person@domain.com - which would make
2103 * the regex quite complicated as tags can also
2104 * end a sentence. So we'll run through our results
2105 * and strip the period from any tags which end with one.
2106 * Returns array of tags found, or empty array.
2108 * @param string $string Post content
2110 * @return array List of tag and person names
2112 public static function getTags($string)
2116 // Convert hashtag links to hashtags
2117 $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2 ', $string);
2119 // ignore anything in a code block
2120 $string = preg_replace('/\[code.*?\].*?\[\/code\]/sm', '', $string);
2122 // Force line feeds at bbtags
2123 $string = str_replace(['[', ']'], ["\n[", "]\n"], $string);
2125 // ignore anything in a bbtag
2126 $string = preg_replace('/\[(.*?)\]/sm', '', $string);
2128 // Match full names against @tags including the space between first and last
2129 // We will look these up afterward to see if they are full names or not recognisable.
2131 if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) {
2132 foreach ($matches[1] as $match) {
2133 if (strstr($match, ']')) {
2134 // we might be inside a bbcode color tag - leave it alone
2138 if (substr($match, -1, 1) === '.') {
2139 $ret[] = substr($match, 0, -1);
2146 // Otherwise pull out single word tags. These can be @nickname, @first_last
2149 if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?]+)([ \x0D\x0A,;:?]|$)/', $string, $matches)) {
2150 foreach ($matches[1] as $match) {
2151 if (strstr($match, ']')) {
2152 // we might be inside a bbcode color tag - leave it alone
2156 if (substr($match, -1, 1) === '.') {
2157 $match = substr($match,0,-1);
2160 // ignore strictly numeric tags like #1
2161 if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) {
2165 // try not to catch url fragments
2166 if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) {