. * */ namespace Friendica\Content\Text; use DOMDocument; use DOMXPath; use Exception; use Friendica\Content\ContactSelector; use Friendica\Content\Item; use Friendica\Content\OEmbed; use Friendica\Content\Smilies; use Friendica\Core\Hook; use Friendica\Core\Logger; use Friendica\Core\Protocol; use Friendica\Core\Renderer; use Friendica\Core\System; use Friendica\DI; use Friendica\Model\Contact; use Friendica\Model\Event; use Friendica\Model\Photo; use Friendica\Model\Tag; use Friendica\Network\Probe; use Friendica\Object\Image; use Friendica\Protocol\Activity; use Friendica\Util\Images; use Friendica\Util\Map; use Friendica\Util\Network; use Friendica\Util\ParseUrl; use Friendica\Util\Proxy as ProxyUtils; use Friendica\Util\Strings; use Friendica\Util\XML; class BBCode { const INTERNAL = 0; const API = 2; const DIASPORA = 3; const CONNECTORS = 4; const OSTATUS = 7; const TWITTER = 8; const BACKLINK = 8; const ACTIVITYPUB = 9; /** * Fetches attachment data that were generated the old way * * @param string $body Message body * @return array * 'type' -> Message type ('link', 'video', 'photo') * 'text' -> Text before the shared message * 'after' -> Text after the shared message * 'image' -> Preview image of the message * 'url' -> Url to the attached message * 'title' -> Title of the attachment * 'description' -> Description of the attachment * @throws \Friendica\Network\HTTPException\InternalServerErrorException */ private static function getOldAttachmentData($body) { $post = []; // Simplify image codes $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body); if (preg_match_all("(\[class=(.*?)\](.*?)\[\/class\])ism", $body, $attached, PREG_SET_ORDER)) { foreach ($attached as $data) { if (!in_array($data[1], ['type-link', 'type-video', 'type-photo'])) { continue; } $post['type'] = substr($data[1], 5); $pos = strpos($body, $data[0]); if ($pos > 0) { $post['text'] = trim(substr($body, 0, $pos)); $post['after'] = trim(substr($body, $pos + strlen($data[0]))); } else { $post['text'] = trim(str_replace($data[0], '', $body)); $post['after'] = ''; } $attacheddata = $data[2]; if (preg_match("/\[img\](.*?)\[\/img\]/ism", $attacheddata, $matches)) { $picturedata = Images::getInfoFromURLCached($matches[1]); if ($picturedata) { if (($picturedata[0] >= 500) && ($picturedata[0] >= $picturedata[1])) { $post['image'] = $matches[1]; } else { $post['preview'] = $matches[1]; } } } if (preg_match("/\[bookmark\=(.*?)\](.*?)\[\/bookmark\]/ism", $attacheddata, $matches)) { $post['url'] = $matches[1]; $post['title'] = $matches[2]; } if (!empty($post['url']) && (in_array($post['type'], ['link', 'video'])) && preg_match("/\[url\=(.*?)\](.*?)\[\/url\]/ism", $attacheddata, $matches)) { $post['url'] = $matches[1]; } // Search for description if (preg_match("/\[quote\](.*?)\[\/quote\]/ism", $attacheddata, $matches)) { $post['description'] = $matches[1]; } } } return $post; } /** * Fetches attachment data that were generated with the "attachment" element * * @param string $body Message body * @return array * 'type' -> Message type ('link', 'video', 'photo') * 'text' -> Text before the shared message * 'after' -> Text after the shared message * 'image' -> Preview image of the message * 'url' -> Url to the attached message * 'title' -> Title of the attachment * 'description' -> Description of the attachment * @throws \Friendica\Network\HTTPException\InternalServerErrorException */ public static function getAttachmentData($body) { $data = [ 'type' => '', 'text' => '', 'after' => '', 'image' => null, 'url' => '', 'title' => '', 'description' => '', ]; if (!preg_match("/(.*)\[attachment(.*?)\](.*?)\[\/attachment\](.*)/ism", $body, $match)) { return self::getOldAttachmentData($body); } $attributes = $match[2]; $data['text'] = trim($match[1]); $type = ''; preg_match("/type='(.*?)'/ism", $attributes, $matches); if (!empty($matches[1])) { $type = strtolower($matches[1]); } preg_match('/type="(.*?)"/ism', $attributes, $matches); if (!empty($matches[1])) { $type = strtolower($matches[1]); } if ($type == '') { return []; } if (!in_array($type, ['link', 'audio', 'photo', 'video'])) { return []; } if ($type != '') { $data['type'] = $type; } $url = ''; preg_match("/url='(.*?)'/ism", $attributes, $matches); if (!empty($matches[1])) { $url = $matches[1]; } preg_match('/url="(.*?)"/ism', $attributes, $matches); if (!empty($matches[1])) { $url = $matches[1]; } if ($url != '') { $data['url'] = html_entity_decode($url, ENT_QUOTES, 'UTF-8'); } $title = ''; preg_match("/title='(.*?)'/ism", $attributes, $matches); if (!empty($matches[1])) { $title = $matches[1]; } preg_match('/title="(.*?)"/ism', $attributes, $matches); if (!empty($matches[1])) { $title = $matches[1]; } if ($title != '') { $title = self::convert(html_entity_decode($title, ENT_QUOTES, 'UTF-8'), false, true); $title = html_entity_decode($title, ENT_QUOTES, 'UTF-8'); $title = str_replace(['[', ']'], ['[', ']'], $title); $data['title'] = $title; } $image = ''; preg_match("/image='(.*?)'/ism", $attributes, $matches); if (!empty($matches[1])) { $image = $matches[1]; } preg_match('/image="(.*?)"/ism', $attributes, $matches); if (!empty($matches[1])) { $image = $matches[1]; } if ($image != '') { $data['image'] = html_entity_decode($image, ENT_QUOTES, 'UTF-8'); } $preview = ''; preg_match("/preview='(.*?)'/ism", $attributes, $matches); if (!empty($matches[1])) { $preview = $matches[1]; } preg_match('/preview="(.*?)"/ism', $attributes, $matches); if (!empty($matches[1])) { $preview = $matches[1]; } if ($preview != '') { $data['preview'] = html_entity_decode($preview, ENT_QUOTES, 'UTF-8'); } $data['description'] = trim($match[3]); $data['after'] = trim($match[4]); return $data; } public static function getAttachedData($body, $item = []) { /* - text: - type: link, video, photo - title: - url: - image: - description: - (thumbnail) */ $has_title = !empty($item['title']); $plink = $item['plink'] ?? ''; $post = self::getAttachmentData($body); // Get all linked images with alternative image description if (preg_match_all("/\[img=([^\[\]]*)\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) { foreach ($pictures as $picture) { if (Photo::isLocal($picture[1])) { $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => $picture[2]]; } } if (!empty($post['images']) && !empty($post['images'][0]['description'])) { $post['image_description'] = $post['images'][0]['description']; } } if (preg_match_all("/\[img\]([^\[\]]*)\[\/img\]/Usi", $body, $pictures, PREG_SET_ORDER)) { foreach ($pictures as $picture) { if (Photo::isLocal($picture[1])) { $post['images'][] = ['url' => str_replace('-1.', '-0.', $picture[1]), 'description' => '']; } } } // if nothing is found, it maybe having an image. if (!isset($post['type'])) { // Simplify image codes $body = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $body); $body = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", '[img]$1[/img]', $body); $post['text'] = $body; if (preg_match_all("#\[url=([^\]]+?)\]\s*\[img\]([^\[]+?)\[/img\]\s*\[/url\]#ism", $body, $pictures, PREG_SET_ORDER)) { if ((count($pictures) == 1) && !$has_title) { if (!empty($item['object-type']) && ($item['object-type'] == Activity\ObjectType::IMAGE)) { // Replace the preview picture with the real picture $url = str_replace('-1.', '-0.', $pictures[0][2]); $data = ['url' => $url, 'type' => 'photo']; } else { // Checking, if the link goes to a picture $data = ParseUrl::getSiteinfoCached($pictures[0][1], true); } // Workaround: // Sometimes photo posts to the own album are not detected at the start. // So we seem to cannot use the cache for these cases. That's strange. if (($data['type'] != 'photo') && strstr($pictures[0][1], "/photos/")) { $data = ParseUrl::getSiteinfo($pictures[0][1], true); } if ($data['type'] == 'photo') { $post['type'] = 'photo'; if (isset($data['images'][0])) { $post['image'] = $data['images'][0]['src']; $post['url'] = $data['url']; } else { $post['image'] = $data['url']; } $post['preview'] = $pictures[0][2]; $post['text'] = trim(str_replace($pictures[0][0], '', $body)); } else { $imgdata = Images::getInfoFromURLCached($pictures[0][1]); if ($imgdata && substr($imgdata['mime'], 0, 6) == 'image/') { $post['type'] = 'photo'; $post['image'] = $pictures[0][1]; $post['preview'] = $pictures[0][2]; $post['text'] = trim(str_replace($pictures[0][0], '', $body)); } } } elseif (count($pictures) > 0) { $post['type'] = 'link'; $post['url'] = $plink; $post['image'] = $pictures[0][2]; $post['text'] = $body; foreach ($pictures as $picture) { $post['text'] = trim(str_replace($picture[0], '', $post['text'])); } } } elseif (preg_match_all("(\[img\](.*?)\[\/img\])ism", $body, $pictures, PREG_SET_ORDER)) { if ((count($pictures) == 1) && !$has_title) { $post['type'] = 'photo'; $post['image'] = $pictures[0][1]; $post['text'] = str_replace($pictures[0][0], '', $body); } elseif (count($pictures) > 0) { $post['type'] = 'link'; $post['url'] = $plink; $post['image'] = $pictures[0][1]; $post['text'] = $body; foreach ($pictures as $picture) { $post['text'] = trim(str_replace($picture[0], '', $post['text'])); } } } // Test for the external links preg_match_all("(\[url\](.*?)\[\/url\])ism", $post['text'], $links1, PREG_SET_ORDER); preg_match_all("(\[url\=(.*?)\].*?\[\/url\])ism", $post['text'], $links2, PREG_SET_ORDER); $links = array_merge($links1, $links2); // If there is only a single one, then use it. // This should cover link posts via API. if ((count($links) == 1) && !isset($post['preview']) && !$has_title) { $post['type'] = 'link'; $post['url'] = $links[0][1]; } // Simplify "video" element $post['text'] = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $post['text']); // Now count the number of external media links preg_match_all("(\[vimeo\](.*?)\[\/vimeo\])ism", $post['text'], $links1, PREG_SET_ORDER); preg_match_all("(\[youtube\\](.*?)\[\/youtube\\])ism", $post['text'], $links2, PREG_SET_ORDER); preg_match_all("(\[video\\](.*?)\[\/video\\])ism", $post['text'], $links3, PREG_SET_ORDER); preg_match_all("(\[audio\\](.*?)\[\/audio\\])ism", $post['text'], $links4, PREG_SET_ORDER); // Add them to the other external links $links = array_merge($links, $links1, $links2, $links3, $links4); // Are there more than one? if (count($links) > 1) { // The post will be the type "text", which means a blog post unset($post['type']); $post['url'] = $plink; } if (!isset($post['type'])) { $post['type'] = "text"; $post['text'] = trim($body); } } elseif (isset($post['url']) && ($post['type'] == 'video')) { $data = ParseUrl::getSiteinfoCached($post['url'], true); if (isset($data['images'][0])) { $post['image'] = $data['images'][0]['src']; } } return $post; } /** * Remove [attachment] BBCode and replaces it with a regular [url] * * @param string $body * @param boolean $no_link_desc No link description * * @return string with replaced body */ public static function removeAttachment($body, $no_link_desc = false) { return preg_replace_callback("/\s*\[attachment (.*)\](.*?)\[\/attachment\]\s*/ism", function ($match) use ($no_link_desc) { $attach_data = self::getAttachmentData($match[0]); if (empty($attach_data['url'])) { return $match[0]; } elseif (empty($attach_data['title']) || $no_link_desc) { return "\n[url]" . $attach_data['url'] . "[/url]\n"; } else { return "\n[url=" . $attach_data['url'] . ']' . $attach_data['title'] . "[/url]\n"; } }, $body); } /** * Converts a BBCode text into plaintext * * @param $text * @param bool $keep_urls Whether to keep URLs in the resulting plaintext * * @return string */ public static function toPlaintext($text, $keep_urls = true) { $naked_text = HTML::toPlaintext(self::convert($text, false, 0, true), 0, !$keep_urls); return $naked_text; } private static function proxyUrl($image, $simplehtml = self::INTERNAL) { // Only send proxied pictures to API and for internal display if (in_array($simplehtml, [self::INTERNAL, self::API])) { return ProxyUtils::proxifyUrl($image); } else { return $image; } } /** * This function changing the visual size (not the real size) of images. * The function does not work for pictures with an alternate text description. * This could only be changed by using some new "img" BBCode format. * * @param string $srctext The body with images * @return string The body with possibly scaled images */ public static function scaleExternalImages(string $srctext) { $s = $srctext; // Simplify image links $s = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $s); $matches = null; $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER); if ($c) { foreach ($matches as $mtch) { Logger::log('scale_external_image: ' . $mtch[1]); $hostname = str_replace('www.', '', substr(DI::baseUrl(), strpos(DI::baseUrl(), '://') + 3)); if (stristr($mtch[1], $hostname)) { continue; } $curlResult = Network::curl($mtch[1], true); if (!$curlResult->isSuccess()) { continue; } $i = $curlResult->getBody(); $type = $curlResult->getContentType(); $type = Images::getMimeTypeByData($i, $mtch[1], $type); if ($i) { $Image = new Image($i, $type); if ($Image->isValid()) { $orig_width = $Image->getWidth(); $orig_height = $Image->getHeight(); if ($orig_width > 640 || $orig_height > 640) { $Image->scaleDown(640); $new_width = $Image->getWidth(); $new_height = $Image->getHeight(); Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG); $s = str_replace( $mtch[0], '[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]' . "\n", $s ); Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG); } } } } } return $s; } /** * Truncates imported message body string length to max_import_size * * The purpose of this function is to apply system message length limits to * imported messages without including any embedded photos in the length * * @param string $body * @return string * @throws \Friendica\Network\HTTPException\InternalServerErrorException */ public static function limitBodySize($body) { $maxlen = DI::config()->get('config', 'max_import_size', 0); // If the length of the body, including the embedded images, is smaller // than the maximum, then don't waste time looking for the images if ($maxlen && (strlen($body) > $maxlen)) { Logger::log('the total body length exceeds the limit', Logger::DEBUG); $orig_body = $body; $new_body = ''; $textlen = 0; $img_start = strpos($orig_body, '[img'); $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false); $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false); while (($img_st_close !== false) && ($img_end !== false)) { $img_st_close++; // make it point to AFTER the closing bracket $img_end += $img_start; $img_end += strlen('[/img]'); if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) { // This is an embedded image if (($textlen + $img_start) > $maxlen) { if ($textlen < $maxlen) { Logger::log('the limit happens before an embedded image', Logger::DEBUG); $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen); $textlen = $maxlen; } } else { $new_body = $new_body . substr($orig_body, 0, $img_start); $textlen += $img_start; } $new_body = $new_body . substr($orig_body, $img_start, $img_end - $img_start); } else { if (($textlen + $img_end) > $maxlen) { if ($textlen < $maxlen) { Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG); $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen); $textlen = $maxlen; } } else { $new_body = $new_body . substr($orig_body, 0, $img_end); $textlen += $img_end; } } $orig_body = substr($orig_body, $img_end); if ($orig_body === false) { // in case the body ends on a closing image tag $orig_body = ''; } $img_start = strpos($orig_body, '[img'); $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false); $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false); } if (($textlen + strlen($orig_body)) > $maxlen) { if ($textlen < $maxlen) { Logger::log('the limit happens after the end of the last image', Logger::DEBUG); $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen); } } else { Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG); $new_body = $new_body . $orig_body; } return $new_body; } else { return $body; } } /** * Processes [attachment] tags * * Note: Can produce a [bookmark] tag in the returned string * * @param string $text * @param integer $simplehtml * @param bool $tryoembed * @return string * @throws \Friendica\Network\HTTPException\InternalServerErrorException */ private static function convertAttachment($text, $simplehtml = self::INTERNAL, $tryoembed = true) { $data = self::getAttachmentData($text); if (empty($data) || empty($data['url'])) { return $text; } if (isset($data['title'])) { $data['title'] = strip_tags($data['title']); $data['title'] = str_replace(['http://', 'https://'], '', $data['title']); } else { $data['title'] = null; } if (((strpos($data['text'], "[img=") !== false) || (strpos($data['text'], "[img]") !== false) || DI::config()->get('system', 'always_show_preview')) && !empty($data['image'])) { $data['preview'] = $data['image']; $data['image'] = ''; } $return = ''; try { if ($tryoembed && OEmbed::isAllowedURL($data['url'])) { $return = OEmbed::getHTML($data['url'], $data['title']); } else { throw new Exception('OEmbed is disabled for this attachment.'); } } catch (Exception $e) { $data['title'] = ($data['title'] ?? '') ?: $data['url']; if ($simplehtml != self::CONNECTORS) { $return = sprintf('
', $data['type']); } if (!empty($data['title']) && !empty($data['url'])) { if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) { $return .= sprintf('', $data['url'], self::proxyUrl($data['image'], $simplehtml), $data['title']); } else { if (!empty($data['image'])) { $return .= sprintf('
', $data['url'], self::proxyUrl($data['image'], $simplehtml), $data['title']); } elseif (!empty($data['preview'])) { $return .= sprintf('
', $data['url'], self::proxyUrl($data['preview'], $simplehtml), $data['title']); } $return .= sprintf('

%s

', $data['url'], $data['title']); } } if (!empty($data['description']) && $data['description'] != $data['title']) { // Sanitize the HTML by converting it to BBCode $bbcode = HTML::toBBCode($data['description']); $return .= sprintf('
%s
', trim(self::convert($bbcode))); } if (!empty($data['url'])) { $return .= sprintf('%s', $data['url'], parse_url($data['url'], PHP_URL_HOST)); } if ($simplehtml != self::CONNECTORS) { $return .= '
'; } } return trim(($data['text'] ?? '') . ' ' . $return . ' ' . ($data['after'] ?? '')); } public static function removeShareInformation($Text, $plaintext = false, $nolink = false) { $data = self::getAttachmentData($Text); if (!$data) { return $Text; } elseif ($nolink) { return $data['text'] . ($data['after'] ?? ''); } $title = htmlentities($data['title'] ?? '', ENT_QUOTES, 'UTF-8', false); $text = htmlentities($data['text'], ENT_QUOTES, 'UTF-8', false); if ($plaintext || (($title != '') && strstr($text, $title))) { $data['title'] = $data['url']; } elseif (($text != '') && strstr($title, $text)) { $data['text'] = $data['title']; $data['title'] = $data['url']; } if (empty($data['text']) && !empty($data['title']) && empty($data['url'])) { return $data['title'] . $data['after']; } // If the link already is included in the post, don't add it again if (!empty($data['url']) && strpos($data['text'], $data['url'])) { return $data['text'] . $data['after']; } $text = $data['text']; if (!empty($data['url']) && !empty($data['title'])) { $text .= "\n[url=" . $data['url'] . ']' . $data['title'] . '[/url]'; } elseif (!empty($data['url'])) { $text .= "\n[url]" . $data['url'] . '[/url]'; } return $text . "\n" . $data['after']; } /** * Converts [url] BBCodes in a format that looks fine on Mastodon. (callback function) * * @param array $match Array with the matching values * @return string reformatted link including HTML codes */ private static function convertUrlForActivityPubCallback($match) { $url = $match[1]; if (isset($match[2]) && ($match[1] != $match[2])) { return $match[0]; } $parts = parse_url($url); if (!isset($parts['scheme'])) { return $match[0]; } return self::convertUrlForActivityPub($url); } /** * Converts [url] BBCodes in a format that looks fine on ActivityPub systems. * * @param string $url URL that is about to be reformatted * @return string reformatted link including HTML codes */ private static function convertUrlForActivityPub($url) { $html = '%s'; return sprintf($html, $url, self::getStyledURL($url)); } /** * Converts an URL in a nicer format (without the scheme and possibly shortened) * * @param string $url URL that is about to be reformatted * @return string reformatted link */ private static function getStyledURL($url) { $parts = parse_url($url); $scheme = $parts['scheme'] . '://'; $styled_url = str_replace($scheme, '', $url); if (strlen($styled_url) > 30) { $styled_url = substr($styled_url, 0, 30) . "…"; } return $styled_url; } /* * [noparse][i]italic[/i][/noparse] turns into * [noparse][ i ]italic[ /i ][/noparse], * to hide them from parser. */ private static function escapeNoparseCallback($match) { $whole_match = $match[0]; $captured = $match[1]; $spacefied = preg_replace("/\[(.*?)\]/", "[ $1 ]", $captured); $new_str = str_replace($captured, $spacefied, $whole_match); return $new_str; } /* * The previously spacefied [noparse][ i ]italic[ /i ][/noparse], * now turns back and the [noparse] tags are trimed * returning [i]italic[/i] */ private static function unescapeNoparseCallback($match) { $captured = $match[1]; $unspacefied = preg_replace("/\[ (.*?)\ ]/", "[$1]", $captured); return $unspacefied; } /** * Returns the bracket character positions of a set of opening and closing BBCode tags, optionally skipping first * occurrences * * @param string $text Text to search * @param string $name Tag name * @param int $occurrences Number of first occurrences to skip * @return boolean|array */ public static function getTagPosition($text, $name, $occurrences = 0) { if ($occurrences < 0) { $occurrences = 0; } $start_open = -1; for ($i = 0; $i <= $occurrences; $i++) { if ($start_open !== false) { $start_open = strpos($text, '[' . $name, $start_open + 1); // allow [name= type tags } } if ($start_open === false) { return false; } $start_equal = strpos($text, '=', $start_open); $start_close = strpos($text, ']', $start_open); if ($start_close === false) { return false; } $start_close++; $end_open = strpos($text, '[/' . $name . ']', $start_close); if ($end_open === false) { return false; } $res = [ 'start' => [ 'open' => $start_open, 'close' => $start_close ], 'end' => [ 'open' => $end_open, 'close' => $end_open + strlen('[/' . $name . ']') ], ]; if ($start_equal !== false) { $res['start']['equal'] = $start_equal + 1; } return $res; } /** * Performs a preg_replace within the boundaries of all named BBCode tags in a text * * @param string $pattern Preg pattern string * @param string $replace Preg replace string * @param string $name BBCode tag name * @param string $text Text to search * @return string */ public static function pregReplaceInTag($pattern, $replace, $name, $text) { $occurrences = 0; $pos = self::getTagPosition($text, $name, $occurrences); while ($pos !== false && $occurrences++ < 1000) { $start = substr($text, 0, $pos['start']['open']); $subject = substr($text, $pos['start']['open'], $pos['end']['close'] - $pos['start']['open']); $end = substr($text, $pos['end']['close']); if ($end === false) { $end = ''; } $subject = preg_replace($pattern, $replace, $subject); $text = $start . $subject . $end; $pos = self::getTagPosition($text, $name, $occurrences); } return $text; } private static function extractImagesFromItemBody($body) { $saved_image = []; $orig_body = $body; $new_body = ''; $cnt = 0; $img_start = strpos($orig_body, '[img'); $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false); $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false); while (($img_st_close !== false) && ($img_end !== false)) { $img_st_close++; // make it point to AFTER the closing bracket $img_end += $img_start; if (!strcmp(substr($orig_body, $img_start + $img_st_close, 5), 'data:')) { // This is an embedded image $saved_image[$cnt] = substr($orig_body, $img_start + $img_st_close, $img_end - ($img_start + $img_st_close)); $new_body = $new_body . substr($orig_body, 0, $img_start) . '[$#saved_image' . $cnt . '#$]'; $cnt++; } else { $new_body = $new_body . substr($orig_body, 0, $img_end + strlen('[/img]')); } $orig_body = substr($orig_body, $img_end + strlen('[/img]')); if ($orig_body === false) { // in case the body ends on a closing image tag $orig_body = ''; } $img_start = strpos($orig_body, '[img'); $img_st_close = ($img_start !== false ? strpos(substr($orig_body, $img_start), ']') : false); $img_end = ($img_start !== false ? strpos(substr($orig_body, $img_start), '[/img]') : false); } $new_body = $new_body . $orig_body; return ['body' => $new_body, 'images' => $saved_image]; } private static function interpolateSavedImagesIntoItemBody($body, array $images) { $newbody = $body; $cnt = 0; foreach ($images as $image) { // We're depending on the property of 'foreach' (specified on the PHP website) that // it loops over the array starting from the first element and going sequentially // to the last element $newbody = str_replace('[$#saved_image' . $cnt . '#$]', '' . DI::l10n()->t('Image/photo') . '', $newbody); $cnt++; } return $newbody; } /** * This function converts a [share] block to text according to a provided callback function whose signature is: * * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string * * Where: * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact * data lookup * - $author_contact is a contact record array * - $content is the inner content of the [share] block * - $is_quote_share indicates whether there's any content before the [share] block * - Return value is the string that should replace the [share] block in the provided text * * This function is intended to be used by addon connector to format a share block like the target network is expecting it. * * @param string $text A BBCode string * @param callable $callback * @return string The BBCode string with all [share] blocks replaced */ public static function convertShare($text, callable $callback) { $return = preg_replace_callback( "/(.*?)\[share(.*?)\](.*)\[\/share\]/ism", function ($match) use ($callback) { $attribute_string = $match[2]; $attributes = []; foreach (['author', 'profile', 'avatar', 'link', 'posted'] as $field) { preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches); $attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8'); } // We only call this so that a previously unknown contact can be added. // This is important for the function "Model\Contact::getDetailsByURL()". // This function then can fetch an entry from the contact table. $default['url'] = $attributes['profile']; if (!empty($attributes['author'])) { $default['name'] = $attributes['author']; } if (!empty($attributes['avatar'])) { $default['photo'] = $attributes['avatar']; } Contact::getIdForURL($attributes['profile'], 0, true, $default); $author_contact = Contact::getDetailsByURL($attributes['profile']); $author_contact['url'] = ($author_contact['url'] ?? $attributes['profile']); $author_contact['addr'] = ($author_contact['addr'] ?? '') ?: Protocol::getAddrFromProfileUrl($attributes['profile']); $attributes['author'] = ($author_contact['name'] ?? '') ?: $attributes['author']; $attributes['avatar'] = ($author_contact['micro'] ?? '') ?: $attributes['avatar']; $attributes['profile'] = ($author_contact['url'] ?? '') ?: $attributes['profile']; if ($attributes['avatar']) { $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB); } return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != ''); }, $text ); return $return; } /** * Default [share] tag conversion callback * * Note: Can produce a [bookmark] tag in the output * * @see BBCode::convertShare() * @param array $attributes [share] block attribute values * @param array $author_contact Contact row of the shared author * @param string $content Inner content of the [share] block * @param boolean $is_quote_share Whether there is content before the [share] block * @param integer $simplehtml Mysterious integer value depending on the target network/formatting style * @return string * @throws \Friendica\Network\HTTPException\InternalServerErrorException */ private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml) { $mention = Protocol::formatMention($attributes['profile'], $attributes['author']); switch ($simplehtml) { case self::API: $text = ($is_quote_share? '
' : '') . '

' . html_entity_decode('♲ ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ':

' . "\n" . $content; break; case self::DIASPORA: if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) { $text = ($is_quote_share? '
' : '') . '

' . $attributes['link'] . '

' . "\n"; } else { $headline = '

' . $attributes['author'] . ':

' . "\n"; if (!empty($attributes['posted']) && !empty($attributes['link'])) { $headline = '

' . $attributes['author'] . ' - ' . $attributes['posted'] . ' GMT

' . "\n"; } $text = ($is_quote_share? '
' : '') . $headline . '
' . trim($content) . '
' . "\n"; if (empty($attributes['posted']) && !empty($attributes['link'])) { $text .= '

[Source]

' . "\n"; } } break; case self::CONNECTORS: $headline = '

' . html_entity_decode('♲ ', ENT_QUOTES, 'UTF-8'); $headline .= DI::l10n()->t('%2$s %3$s', $attributes['link'], $mention, $attributes['posted']); $headline .= ':

' . "\n"; $text = ($is_quote_share? '
' : '') . $headline . '
' . trim($content) . '
' . "\n"; break; case self::OSTATUS: $text = ($is_quote_share? '
' : '') . '

' . html_entity_decode('♲ ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '

' . "\n"; break; case self::ACTIVITYPUB: $author = '@' . $author_contact['addr'] . ':'; $text = '
' . html_entity_decode('♲', ENT_QUOTES, 'UTF-8') . ' ' . $author . '
' . $content . '
' . "\n"; break; default: $text = ($is_quote_share? "\n" : ''); $authorId = Contact::getIdForURL($attributes['profile']); $contact = Contact::getById($authorId, ['network']); $tpl = Renderer::getMarkupTemplate('shared_content.tpl'); $text .= Renderer::replaceMacros($tpl, [ '$profile' => $attributes['profile'], '$avatar' => $attributes['avatar'], '$author' => $attributes['author'], '$link' => $attributes['link'], '$link_title' => DI::l10n()->t('link to source'), '$posted' => $attributes['posted'], '$network_name' => ContactSelector::networkToName($contact['network'], $attributes['profile']), '$network_icon' => ContactSelector::networkToIcon($contact['network'], $attributes['profile']), '$content' => self::setMentions(trim($content), 0, $contact['network']), ]); break; } return $text; } private static function removePictureLinksCallback($match) { $cache_key = 'remove:' . $match[1]; $text = DI::cache()->get($cache_key); if (is_null($text)) { $a = DI::app(); $stamp1 = microtime(true); $ch = @curl_init($match[1]); @curl_setopt($ch, CURLOPT_NOBODY, true); @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true); @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent()); @curl_exec($ch); $curl_info = @curl_getinfo($ch); DI::profiler()->saveTimestamp($stamp1, "network", System::callstack()); if (substr($curl_info['content_type'], 0, 6) == 'image/') { $text = "[url=" . $match[1] . ']' . $match[1] . "[/url]"; } else { $text = "[url=" . $match[2] . ']' . $match[2] . "[/url]"; // if its not a picture then look if its a page that contains a picture link $body = Network::fetchUrl($match[1]); $doc = new DOMDocument(); @$doc->loadHTML($body); $xpath = new DOMXPath($doc); $list = $xpath->query("//meta[@name]"); foreach ($list as $node) { $attr = []; if ($node->attributes->length) { foreach ($node->attributes as $attribute) { $attr[$attribute->name] = $attribute->value; } } if (strtolower($attr['name']) == 'twitter:image') { $text = '[url=' . $attr['content'] . ']' . $attr['content'] . '[/url]'; } } } DI::cache()->set($cache_key, $text); } return $text; } private static function expandLinksCallback($match) { if (($match[3] == '') || ($match[2] == $match[3]) || stristr($match[2], $match[3])) { return ($match[1] . "[url]" . $match[2] . "[/url]"); } else { return ($match[1] . $match[3] . " [url]" . $match[2] . "[/url]"); } } private static function cleanPictureLinksCallback($match) { $a = DI::app(); // When the picture link is the own photo path then we can avoid fetching the link $own_photo_url = preg_quote(Strings::normaliseLink(DI::baseUrl()->get()) . '/photos/'); if (preg_match('|' . $own_photo_url . '.*?/image/|', Strings::normaliseLink($match[1]))) { if (!empty($match[3])) { $text = '[img=' . str_replace('-1.', '-0.', $match[2]) . ']' . $match[3] . '[/img]'; } else { $text = '[img]' . str_replace('-1.', '-0.', $match[2]) . '[/img]'; } return $text; } $cache_key = 'clean:' . $match[1]; $text = DI::cache()->get($cache_key); if (!is_null($text)) { return $text; } // Only fetch the header, not the content $stamp1 = microtime(true); $ch = @curl_init($match[1]); @curl_setopt($ch, CURLOPT_NOBODY, true); @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true); @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent()); @curl_exec($ch); $curl_info = @curl_getinfo($ch); DI::profiler()->saveTimestamp($stamp1, "network", System::callstack()); // if its a link to a picture then embed this picture if (substr($curl_info['content_type'], 0, 6) == 'image/') { $text = '[img]' . $match[1] . '[/img]'; } else { if (!empty($match[3])) { $text = '[img=' . $match[2] . ']' . $match[3] . '[/img]'; } else { $text = '[img]' . $match[2] . '[/img]'; } // if its not a picture then look if its a page that contains a picture link $body = Network::fetchUrl($match[1]); $doc = new DOMDocument(); @$doc->loadHTML($body); $xpath = new DOMXPath($doc); $list = $xpath->query("//meta[@name]"); foreach ($list as $node) { $attr = []; if ($node->attributes->length) { foreach ($node->attributes as $attribute) { $attr[$attribute->name] = $attribute->value; } } if (strtolower($attr['name']) == "twitter:image") { if (!empty($match[3])) { $text = "[img=" . $attr['content'] . "]" . $match[3] . "[/img]"; } else { $text = "[img]" . $attr['content'] . "[/img]"; } } } } DI::cache()->set($cache_key, $text); return $text; } public static function cleanPictureLinks($text) { $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img=(.*)\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $text); $return = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::cleanPictureLinksCallback', $return); return $return; } /** * Converts a BBCode message to HTML message * * BBcode 2 HTML was written by WAY2WEB.net * extended to work with Mistpark/Friendica - Mike Macgirvin * * Simple HTML values meaning: * - 0: Friendica display * - 1: Unused * - 2: Used for Windows Phone push, Friendica API * - 3: Used before converting to Markdown in bb2diaspora.php * - 4: Used for WordPress, Libertree (before Markdown), pump.io and tumblr * - 5: Unused * - 6: Unused * - 7: Used for dfrn, OStatus * - 8: Used for Twitter, WP backlink text setting * - 9: ActivityPub * * @param string $text * @param bool $try_oembed * @param int $simple_html * @param bool $for_plaintext * @return string * @throws \Friendica\Network\HTTPException\InternalServerErrorException */ public static function convert(string $text = null, $try_oembed = true, $simple_html = self::INTERNAL, $for_plaintext = false) { // Accounting for null default column values if (is_null($text) || $text === '') { return ''; } $a = DI::app(); $text = self::performWithEscapedTags($text, ['code'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) { $text = self::performWithEscapedTags($text, ['noparse', 'nobb', 'pre'], function ($text) use ($try_oembed, $simple_html, $for_plaintext, $a) { /* * preg_match_callback function to replace potential Oembed tags with Oembed content * * $match[0] = [tag]$url[/tag] or [tag=$url]$title[/tag] * $match[1] = $url * $match[2] = $title or absent */ $try_oembed_callback = function ($match) { $url = $match[1]; $title = $match[2] ?? null; try { $return = OEmbed::getHTML($url, $title); } catch (Exception $ex) { $return = $match[0]; } return $return; }; // Remove the abstract element. It is a non visible element. $text = self::stripAbstract($text); // Move all spaces out of the tags $text = preg_replace("/\[(\w*)\](\s*)/ism", '$2[$1]', $text); $text = preg_replace("/(\s*)\[\/(\w*)\]/ism", '[/$2]$1', $text); // Extract the private images which use data urls since preg has issues with // large data sizes. Stash them away while we do bbcode conversion, and then put them back // in after we've done all the regex matching. We cannot use any preg functions to do this. $extracted = self::extractImagesFromItemBody($text); $text = $extracted['body']; $saved_image = $extracted['images']; // If we find any event code, turn it into an event. // After we're finished processing the bbcode we'll // replace all of the event code with a reformatted version. $ev = Event::fromBBCode($text); // Replace any html brackets with HTML Entities to prevent executing HTML or script // Don't use strip_tags here because it breaks [url] search by replacing & with amp $text = str_replace("<", "<", $text); $text = str_replace(">", ">", $text); // remove some newlines before the general conversion $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text); $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text); // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems if (!$try_oembed) { $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text); } // Convert new line chars to html
tags // nlbr seems to be hopelessly messed up // $Text = nl2br($Text); // We'll emulate it. $text = trim($text); $text = str_replace("\r\n", "\n", $text); // Remove linefeeds inside of the table elements. See issue #6799 $search = ["\n[th]", "[th]\n", " [th]", "\n[/th]", "[/th]\n", "[/th] ", "\n[td]", "[td]\n", " [td]", "\n[/td]", "[/td]\n", "[/td] ", "\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ", "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "]; $replace = ["[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]", "[td]", "[td]", "[td]", "[/td]", "[/td]", "[/td]", "[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]", "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"]; do { $oldtext = $text; $text = str_replace($search, $replace, $text); } while ($oldtext != $text); // Replace these here only once $search = ["\n[table]", "[/table]\n"]; $replace = ["[table]", "[/table]"]; $text = str_replace($search, $replace, $text); // removing multiplicated newlines if (DI::config()->get('system', 'remove_multiplicated_lines')) { $search = ["\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n", "\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"]; $replace = ["\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[ul]", "[/ul]", "\n[share ", "[/attachment]", "[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"]; do { $oldtext = $text; $text = str_replace($search, $replace, $text); } while ($oldtext != $text); } /// @todo Have a closer look at the different html modes // Handle attached links or videos if ($simple_html == self::ACTIVITYPUB) { $text = self::removeAttachment($text); } elseif (!in_array($simple_html, [self::INTERNAL, self::CONNECTORS])) { $text = self::removeAttachment($text, true); } else { $text = self::convertAttachment($text, $simple_html, $try_oembed); } // leave open the posibility of [map=something] // this is replaced in Item::prepareBody() which has knowledge of the item location if (strpos($text, '[/map]') !== false) { $text = preg_replace_callback( "/\[map\](.*?)\[\/map\]/ism", function ($match) use ($simple_html) { return str_replace($match[0], '

' . Map::byLocation($match[1], $simple_html) . '

', $match[0]); }, $text ); } if (strpos($text, '[map=') !== false) { $text = preg_replace_callback( "/\[map=(.*?)\]/ism", function ($match) use ($simple_html) { return str_replace($match[0], '

' . Map::byCoordinates(str_replace('/', ' ', $match[1]), $simple_html) . '

', $match[0]); }, $text ); } if (strpos($text, '[map]') !== false) { $text = preg_replace("/\[map\]/", '

', $text); } // Check for headers $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '

$1

', $text); $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '

$1

', $text); $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '

$1

', $text); $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '

$1

', $text); $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '
$1
', $text); $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '
$1
', $text); // Check for paragraph $text = preg_replace("(\[p\](.*?)\[\/p\])ism", '

$1

', $text); // Check for bold text $text = preg_replace("(\[b\](.*?)\[\/b\])ism", '$1', $text); // Check for Italics text $text = preg_replace("(\[i\](.*?)\[\/i\])ism", '$1', $text); // Check for Underline text $text = preg_replace("(\[u\](.*?)\[\/u\])ism", '$1', $text); // Check for strike-through text $text = preg_replace("(\[s\](.*?)\[\/s\])ism", '$1', $text); // Check for over-line text $text = preg_replace("(\[o\](.*?)\[\/o\])ism", '$1', $text); // Check for colored text $text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])ism", "$2", $text); // Check for sized text // [size=50] --> font-size: 50px (with the unit). if ($simple_html != self::DIASPORA) { $text = preg_replace("(\[size=(\d*?)\](.*?)\[\/size\])ism", "$2", $text); $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text); } else { // Issue 2199: Diaspora doesn't interpret the construct above, nor the or element $text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])ism", "$2", $text); } // Check for centered text $text = preg_replace("(\[center\](.*?)\[\/center\])ism", "
$1
", $text); // Check for list text $text = str_replace("[*]", "
  • ", $text); // Check for style sheet commands $text = preg_replace_callback( "(\[style=(.*?)\](.*?)\[\/style\])ism", function ($match) { return "" . $match[2] . ""; }, $text ); // Check for CSS classes $text = preg_replace_callback( "(\[class=(.*?)\](.*?)\[\/class\])ism", function ($match) { return "" . $match[2] . ""; }, $text ); // handle nested lists $endlessloop = 0; while ((((strpos($text, "[/list]") !== false) && (strpos($text, "[list") !== false)) || ((strpos($text, "[/ol]") !== false) && (strpos($text, "[ol]") !== false)) || ((strpos($text, "[/ul]") !== false) && (strpos($text, "[ul]") !== false)) || ((strpos($text, "[/li]") !== false) && (strpos($text, "[li]") !== false))) && (++$endlessloop < 20)) { $text = preg_replace("/\[list\](.*?)\[\/list\]/ism", '
      $1
    ', $text); $text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '
      $1
    ', $text); $text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '
      $1
    ', $text); $text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '
      $2
    ', $text); $text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '
      $2
    ', $text); $text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '
      $2
    ', $text); $text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '
      $2
    ', $text); $text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '
      $1
    ', $text); $text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '
      $1
    ', $text); $text = preg_replace("/\[li\](.*?)\[\/li\]/ism", '
  • $1
  • ', $text); } $text = preg_replace("/\[th\](.*?)\[\/th\]/sm", '$1', $text); $text = preg_replace("/\[td\](.*?)\[\/td\]/sm", '$1', $text); $text = preg_replace("/\[tr\](.*?)\[\/tr\]/sm", '$1', $text); $text = preg_replace("/\[table\](.*?)\[\/table\]/sm", '$1
    ', $text); $text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '$1
    ', $text); $text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '$1
    ', $text); $text = str_replace('[hr]', '
    ', $text); if (!$for_plaintext) { $escaped = []; // Escaping BBCodes susceptible to contain rogue URL we don'' want the autolinker to catch $text = preg_replace_callback('#\[(url|img|audio|video|youtube|vimeo|share|attachment|iframe|bookmark).+?\[/\1\]#ism', function ($matches) use (&$escaped) { $return = '{escaped-' . count($escaped) . '}'; $escaped[] = $matches[0]; return $return; }, $text ); // Autolinker for isolated URLs $text = preg_replace(Strings::autoLinkRegEx(), '[url]$1[/url]', $text); // Restoring escaped blocks $text = preg_replace_callback('/{escaped-([0-9]+)}/iU', function ($matches) use ($escaped) { return $escaped[intval($matches[1])] ?? $matches[0]; }, $text ); } // This is actually executed in Item::prepareBody() $nosmile = strpos($text, '[nosmile]') !== false; $text = str_replace('[nosmile]', '', $text); // Check for font change text $text = preg_replace("/\[font=(.*?)\](.*?)\[\/font\]/sm", "$2", $text); // Declare the format for [spoiler] layout $SpoilerLayout = '
    ' . DI::l10n()->t('Click to open/close') . '$1
    '; // Check for [spoiler] text // handle nested quotes $endlessloop = 0; while ((strpos($text, "[/spoiler]") !== false) && (strpos($text, "[spoiler]") !== false) && (++$endlessloop < 20)) { $text = preg_replace("/\[spoiler\](.*?)\[\/spoiler\]/ism", $SpoilerLayout, $text); } // Check for [spoiler=Title] text // handle nested quotes $endlessloop = 0; while ((strpos($text, "[/spoiler]")!== false) && (strpos($text, "[spoiler=") !== false) && (++$endlessloop < 20)) { $text = preg_replace("/\[spoiler=[\"\']*(.*?)[\"\']*\](.*?)\[\/spoiler\]/ism", '
    $1$2
    ', $text); } // Declare the format for [quote] layout $QuoteLayout = '
    $1
    '; // Check for [quote] text // handle nested quotes $endlessloop = 0; while ((strpos($text, "[/quote]") !== false) && (strpos($text, "[quote]") !== false) && (++$endlessloop < 20)) { $text = preg_replace("/\[quote\](.*?)\[\/quote\]/ism", "$QuoteLayout", $text); } // Check for [quote=Author] text $t_wrote = DI::l10n()->t('$1 wrote:'); // handle nested quotes $endlessloop = 0; while ((strpos($text, "[/quote]")!== false) && (strpos($text, "[quote=") !== false) && (++$endlessloop < 20)) { $text = preg_replace("/\[quote=[\"\']*(.*?)[\"\']*\](.*?)\[\/quote\]/ism", "

    " . $t_wrote . "

    $2
    ", $text); } // [img=widthxheight]image source[/img] $text = preg_replace_callback( "/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", function ($matches) use ($simple_html) { if (strpos($matches[3], "data:image/") === 0) { return $matches[0]; } $matches[3] = self::proxyUrl($matches[3], $simple_html); return "[img=" . $matches[1] . "x" . $matches[2] . "]" . $matches[3] . "[/img]"; }, $text ); $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '', $text); $text = preg_replace("/\[zmg\=([0-9]*)x([0-9]*)\](.*?)\[\/zmg\]/ism", '', $text); $text = preg_replace_callback("/\[img\=(.*?)\](.*?)\[\/img\]/ism", function ($matches) use ($simple_html) { $matches[1] = self::proxyUrl($matches[1], $simple_html); $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT); return '' . $matches[2] . ''; }, $text); // Images // [img]pathtoimage[/img] $text = preg_replace_callback( "/\[img\](.*?)\[\/img\]/ism", function ($matches) use ($simple_html) { if (strpos($matches[1], "data:image/") === 0) { return $matches[0]; } $matches[1] = self::proxyUrl($matches[1], $simple_html); return "[img]" . $matches[1] . "[/img]"; }, $text ); $text = preg_replace("/\[img\](.*?)\[\/img\]/ism", '' . DI::l10n()->t('Image/photo') . '', $text); $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '' . DI::l10n()->t('Image/photo') . '', $text); $text = preg_replace("/\[crypt\](.*?)\[\/crypt\]/ism", '
    ' . DI::l10n()->t('Encrypted content') . '
    ', $text); $text = preg_replace("/\[crypt(.*?)\](.*?)\[\/crypt\]/ism", '
    ' . DI::l10n()->t('Encrypted content') . '
    ', $text); //$Text = preg_replace("/\[crypt=(.*?)\](.*?)\[\/crypt\]/ism", '
    ' . DI::l10n()->t('Encrypted content') . '
    ', $Text); // Simplify "video" element $text = preg_replace('(\[video.*?\ssrc\s?=\s?([^\s\]]+).*?\].*?\[/video\])ism', '[video]$1[/video]', $text); // Try to Oembed if ($try_oembed) { $text = preg_replace("/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4).*?)\[\/video\]/ism", '', $text); $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '', $text); $text = preg_replace_callback("/\[video\](.*?)\[\/video\]/ism", $try_oembed_callback, $text); $text = preg_replace_callback("/\[audio\](.*?)\[\/audio\]/ism", $try_oembed_callback, $text); } else { $text = preg_replace("/\[video\](.*?)\[\/video\]/ism", '$1', $text); $text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '$1', $text); } // html5 video and audio if ($try_oembed) { $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '', $text); } else { $text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '$1', $text); } // Youtube extensions if ($try_oembed) { $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text); $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text); $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text); } $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text); $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text); $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text); if ($try_oembed) { $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism", '', $text); } else { $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism", 'https://www.youtube.com/watch?v=$1', $text); } if ($try_oembed) { $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text); $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text); } $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text); $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text); if ($try_oembed) { $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism", '', $text); } else { $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism", 'https://vimeo.com/$1', $text); } // oembed tag $text = OEmbed::BBCode2HTML($text); // Avoid triple linefeeds through oembed $text = str_replace("


    ", "

    ", $text); // If we found an event earlier, strip out all the event code and replace with a reformatted version. // Replace the event-start section with the entire formatted event. The other bbcode is stripped. // Summary (e.g. title) is required, earlier revisions only required description (in addition to // start which is always required). Allow desc with a missing summary for compatibility. if ((!empty($ev['desc']) || !empty($ev['summary'])) && !empty($ev['start'])) { $sub = Event::getHTML($ev, $simple_html); $text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/ism", '', $text); $text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/ism", '', $text); $text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/ism", $sub, $text); $text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/ism", '', $text); $text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/ism", '', $text); $text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/ism", '', $text); $text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text); } // Replace non graphical smilies for external posts if (!$nosmile && !$for_plaintext) { $text = Smilies::replace($text); } if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA)) { $conv = html_entity_decode(str_replace([' ', "\n", "\r"], '', $text)); // Emojis are always 4 byte Unicode characters if (!empty($conv) && (strlen($conv) / mb_strlen($conv) == 4)) { $text = '' . $text . ''; } } if (!$for_plaintext) { if (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) { $text = preg_replace_callback("/\[url\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text); $text = preg_replace_callback("/\[url\=(.*?)\](.*?)\[\/url\]/ism", 'self::convertUrlForActivityPubCallback', $text); } } else { $text = preg_replace("(\[url\](.*?)\[\/url\])ism", " $1 ", $text); $text = preg_replace_callback("&\[url=([^\[\]]*)\]\[img\](.*)\[\/img\]\[\/url\]&Usi", 'self::removePictureLinksCallback', $text); } $text = str_replace(["\r","\n"], ['
    ', '
    '], $text); // Remove all hashtag addresses if ($simple_html && !in_array($simple_html, [self::DIASPORA, self::OSTATUS, self::ACTIVITYPUB])) { $text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text); } elseif ($simple_html == self::DIASPORA) { // The ! is converted to @ since Diaspora only understands the @ $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '@$3', $text); } elseif (in_array($simple_html, [self::OSTATUS, self::ACTIVITYPUB])) { $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text); } elseif (!$simple_html) { $text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text); } // Bookmarks in red - will be converted to bookmarks in friendica $text = preg_replace("/#\^\[url\](.*?)\[\/url\]/ism", '[bookmark=$1]$1[/bookmark]', $text); $text = preg_replace("/#\^\[url\=(.*?)\](.*?)\[\/url\]/ism", '[bookmark=$1]$2[/bookmark]', $text); $text = preg_replace("/#\[url\=.*?\]\^\[\/url\]\[url\=(.*?)\](.*?)\[\/url\]/i", "[bookmark=$1]$2[/bookmark]", $text); if (in_array($simple_html, [self::API, self::OSTATUS, self::TWITTER])) { $text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", "self::expandLinksCallback", $text); //$Text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $Text); $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]',$text); } // Perform URL Search if ($try_oembed) { $text = preg_replace_callback("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", $try_oembed_callback, $text); } $text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", '[url=$1]$2[/url]', $text); // Handle Diaspora posts $text = preg_replace_callback( "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi", function ($match) { return "[url=" . DI::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]"; }, $text ); $text = preg_replace_callback( "&\[url=/people\?q\=(.*)\](.*)\[\/url\]&Usi", function ($match) { return "[url=" . DI::baseUrl() . "/search?search=%40" . $match[1] . "]" . $match[2] . "[/url]"; }, $text ); // Server independent link to posts and comments // See issue: https://github.com/diaspora/diaspora_federation/issues/75 $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism"; $text = preg_replace($expression, DI::baseUrl()."/display/$1", $text); /* Tag conversion * Supports: * - #[url=][/url] * - [url=]#[/url] */ $text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) use ($simple_html) { if ($simple_html == BBCode::ACTIVITYPUB) { return '#' . XML::escape($matches[1]) . ''; } else { return '#'; } }, $text); // We need no target="_blank" rel="noopener noreferrer" for local links // convert links start with DI::baseUrl() as local link without the target="_blank" rel="noopener noreferrer" attribute $escapedBaseUrl = preg_quote(DI::baseUrl(), '/'); $text = preg_replace("/\[url\](".$escapedBaseUrl.".*?)\[\/url\]/ism", '$1', $text); $text = preg_replace("/\[url\=(".$escapedBaseUrl.".*?)\](.*?)\[\/url\]/ism", '$2', $text); $text = preg_replace("/\[url\](.*?)\[\/url\]/ism", '$1', $text); $text = preg_replace("/\[url\=(.*?)\](.*?)\[\/url\]/ism", '$2', $text); // Red compatibility, though the link can't be authenticated on Friendica $text = preg_replace("/\[zrl\=(.*?)\](.*?)\[\/zrl\]/ism", '$2', $text); // we may need to restrict this further if it picks up too many strays // link acct:user@host to a webfinger profile redirector $text = preg_replace('/acct:([^@]+)@((?!\-)(?:[a-zA-Z\d\-]{0,62}[a-zA-Z\d]\.){1,126}(?!\d+)[a-zA-Z\d]{1,63})/', 'acct:$1@$2', $text); // Perform MAIL Search $text = preg_replace("/\[mail\](.*?)\[\/mail\]/", '$1', $text); $text = preg_replace("/\[mail\=(.*?)\](.*?)\[\/mail\]/", '$2', $text); /// @todo What is the meaning of these lines? $text = preg_replace('/\[\&\;([#a-z0-9]+)\;\]/', '&$1;', $text); $text = preg_replace('/\&\#039\;/', '\'', $text); // Currently deactivated, it made problems with " inside of alt texts. //$text = preg_replace('/\"\;/', '"', $text); // fix any escaped ampersands that may have been converted into links $text = preg_replace('/\<([^>]*?)(src|href)=(.*?)\&\;(.*?)\>/ism', '<$1$2=$3&$4>', $text); // sanitizes src attributes (http and redir URLs for displaying in a web page, cid used for inline images in emails) $allowed_src_protocols = ['//', 'http://', 'https://', 'redir/', 'cid:']; array_walk($allowed_src_protocols, function(&$value) { $value = preg_quote($value, '#');}); $text = preg_replace('#<([^>]*?)(src)="(?!' . implode('|', $allowed_src_protocols) . ')(.*?)"(.*?)>#ism', '<$1$2=""$4 data-original-src="$3" class="invalid-src" title="' . DI::l10n()->t('Invalid source protocol') . '">', $text); // sanitize href attributes (only allowlisted protocols URLs) // default value for backward compatibility $allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []); // Always allowed protocol even if config isn't set or not including it $allowed_link_protocols[] = '//'; $allowed_link_protocols[] = 'http://'; $allowed_link_protocols[] = 'https://'; $allowed_link_protocols[] = 'redir/'; array_walk($allowed_link_protocols, function(&$value) { $value = preg_quote($value, '#');}); $regex = '#<([^>]*?)(href)="(?!' . implode('|', $allowed_link_protocols) . ')(.*?)"(.*?)>#ism'; $text = preg_replace($regex, '<$1$2="javascript:void(0)"$4 data-original-href="$3" class="invalid-href" title="' . DI::l10n()->t('Invalid link protocol') . '">', $text); // Shared content $text = self::convertShare( $text, function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) { return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html); } ); $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image); return $text; }); // Escaped noparse, nobb, pre // Remove escaping tags $text = preg_replace("/\[noparse\](.*?)\[\/noparse\]/ism", '\1', $text); $text = preg_replace("/\[nobb\](.*?)\[\/nobb\]/ism", '\1', $text); $text = preg_replace("/\[pre\](.*?)\[\/pre\]/ism", '\1', $text); return $text; }); // Escaped code $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#ism", function ($matches) { if (strpos($matches[2], "\n") !== false) { $return = '
    ' . htmlspecialchars(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '
    '; } else { $return = '' . htmlspecialchars($matches[2], ENT_NOQUOTES, 'UTF-8') . ''; } return $return; }, $text ); // Clean up the HTML by loading and saving the HTML with the DOM. // Bad structured html can break a whole page. // For performance reasons do it only with activated item cache or at export. if (!$try_oembed || (get_itemcachepath() != '')) { $doc = new DOMDocument(); $doc->preserveWhiteSpace = false; $text = mb_convert_encoding($text, 'HTML-ENTITIES', "UTF-8"); $doctype = ''; $encoding = ''; @$doc->loadHTML($encoding . $doctype . '' . $text . ''); $doc->encoding = 'UTF-8'; $text = $doc->saveHTML(); $text = str_replace(['', '', $doctype, $encoding], ['', '', '', ''], $text); $text = str_replace('
    ', '', $text); //$Text = mb_convert_encoding($Text, "UTF-8", 'HTML-ENTITIES'); } // Clean up some useless linebreaks in lists //$Text = str_replace('

    ', '', $Text); //$Text = str_replace('
    ', '', $Text); //$Text = str_replace('
  • ', '
  • ', $Text); //$Text = str_replace('
    '], ['
    '], $text); $stamp1 = microtime(true); // Now convert HTML to Markdown $text = HTML::toMarkdown($text); DI::profiler()->saveTimestamp($stamp1, "parser", System::callstack()); // Libertree has a problem with escaped hashtags. $text = str_replace(['\#'], ['#'], $text); // Remove any leading or trailing whitespace, as this will mess up // the Diaspora signature verification and cause the item to disappear $text = trim($text); if ($for_diaspora) { $url_search_string = "^\[\]"; $text = preg_replace_callback( "/([@!])\[(.*?)\]\(([$url_search_string]*?)\)/ism", ['self', 'bbCodeMention2DiasporaCallback'], $text ); } Hook::callAll('bb2diaspora', $text); return $text; } /** * Pull out all #hashtags and @person tags from $string. * * We also get @person@domain.com - which would make * the regex quite complicated as tags can also * end a sentence. So we'll run through our results * and strip the period from any tags which end with one. * Returns array of tags found, or empty array. * * @param string $string Post content * * @return array List of tag and person names */ public static function getTags($string) { $ret = []; BBCode::performWithEscapedTags($string, ['noparse', 'pre', 'code'], function ($string) use (&$ret) { // Convert hashtag links to hashtags $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2 ', $string); // Force line feeds at bbtags $string = str_replace(['[', ']'], ["\n[", "]\n"], $string); // ignore anything in a bbtag $string = preg_replace('/\[(.*?)\]/sm', '', $string); // Match full names against @tags including the space between first and last // We will look these up afterward to see if they are full names or not recognisable. if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) { foreach ($matches[1] as $match) { if (strstr($match, ']')) { // we might be inside a bbcode color tag - leave it alone continue; } if (substr($match, -1, 1) === '.') { $ret[] = substr($match, 0, -1); } else { $ret[] = $match; } } } // Otherwise pull out single word tags. These can be @nickname, @first_last // and #hash tags. if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?\']*[^\^ \x0D\x0A,;:?!\'.])/', $string, $matches)) { foreach ($matches[1] as $match) { if (strstr($match, ']')) { // we might be inside a bbcode color tag - leave it alone continue; } // ignore strictly numeric tags like #1 if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) { continue; } // try not to catch url fragments if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) { continue; } $ret[] = $match; } } }); return array_unique($ret); } /** * Perform a custom function on a text after having escaped blocks enclosed in the provided tag list. * * @param string $text * @param array $tagList A list of tag names, e.g ['noparse', 'nobb', 'pre'] * @param callable $callback * @return string * @throws Exception *@see Strings::performWithEscapedBlocks * */ public static function performWithEscapedTags(string $text, array $tagList, callable $callback) { $tagList = array_map('preg_quote', $tagList); return Strings::performWithEscapedBlocks($text, '#\[(?:' . implode('|', $tagList) . ').*?\[/(?:' . implode('|', $tagList) . ')]#ism', $callback); } /** * Replaces mentions in the provided message body for the provided user and network if any * * @param $body * @param $profile_uid * @param $network * @return string * @throws \Friendica\Network\HTTPException\InternalServerErrorException * @throws \ImagickException */ public static function setMentions($body, $profile_uid = 0, $network = '') { BBCode::performWithEscapedTags($body, ['noparse', 'pre', 'code', 'img'], function ($body) use ($profile_uid, $network) { $tags = BBCode::getTags($body); $tagged = []; $inform = ''; foreach ($tags as $tag) { $tag_type = substr($tag, 0, 1); if ($tag_type == Tag::TAG_CHARACTER[Tag::HASHTAG]) { continue; } /* * If we already tagged 'Robert Johnson', don't try and tag 'Robert'. * Robert Johnson should be first in the $tags array */ foreach ($tagged as $nextTag) { if (stristr($nextTag, $tag . ' ')) { continue 2; } } $success = Item::replaceTag($body, $inform, $profile_uid, $tag, $network); if ($success['replaced']) { $tagged[] = $tag; } } return $body; }); return $body; } /** * @param $author * @param $profile * @param $avatar * @param $guid * @param $posted * @param $link * @return string * @TODO Rewrite to handle over whole record array */ public static function getShareOpeningTag($author, $profile, $avatar, $guid, $posted, $link) { $header = "[share author='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $author). "' profile='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $profile). "' avatar='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $avatar); if ($guid) { $header .= "' guid='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $guid); } if ($posted) { $header .= "' posted='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $posted); } $header .= "' link='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $link)."']"; return $header; } }