.
*
*/
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\PageInfo;
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\Object\Image;
use Friendica\Protocol\Activity;
use Friendica\Util\Images;
use Friendica\Util\Map;
use Friendica\Util\ParseUrl;
use Friendica\Util\Proxy as ProxyUtils;
use Friendica\Util\Strings;
use Friendica\Util\XML;
class BBCode
{
// Update this value to the current date whenever changes are made to BBCode::convert
const VERSION = '2020-12-18-video-embeds';
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' => '',
'provider_name' => '',
'provider_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]);
$parts = parse_url($data['url']);
if (!empty($parts['scheme']) && !empty($parts['host'])) {
$data['provider_name'] = $parts['host'];
$data['provider_url'] = $parts['scheme'] . '://' . $parts['host'];
if (!empty($parts['port'])) {
$data['provider_url'] .= ':' . $parts['port'];
}
}
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;
}
$curlResult = DI::httpRequest()->get($mtch[1]);
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::info('External images scaled', ['orig_width' => $orig_width, 'new_width' => $new_width, 'orig_height' => $orig_height, 'new_height' => $new_height, 'match' => $mtch[0]]);
$s = str_replace(
$mtch[0],
'[img=' . $new_width . 'x' . $new_height. ']' . $mtch[1] . '[/img]'
. "\n",
$s
);
Logger::info('New string', ['image' => $s]);
}
}
}
}
}
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::info('the total body length exceeds the limit', ['maxlen' => $maxlen, 'body_len' => strlen($body)]);
$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::info('the limit happens before an embedded image');
$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::info('the limit happens before the end of a non-embedded image');
$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::info('the limit happens after the end of the last image');
$new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen);
}
} else {
Logger::info('the text size with embedded images extracted did not violate the limit');
$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('
', $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) {
$return .= '
%s ';
return sprintf($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' . $cnt . '#$]',
'$1 ', $match[3]);
return $match[1] . $callback($attributes, $author_contact, $content, trim($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? '' . 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? '' . 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 = '@' . $author_contact['addr'] . ' ' . Map::byLocation($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("(\[style=(.*?)\](.*?)\[\/style\])ism", '$2 ', $text);
// Check for CSS classes
$text = preg_replace("(\[class=(.*?)\](.*?)\[\/class\])ism", '$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", '', $text);
$text = preg_replace("/\[list=\](.*?)\[\/list\]/ism", '', $text);
$text = preg_replace("/\[list=1\](.*?)\[\/list\]/ism", '', $text);
$text = preg_replace("/\[list=((?-i)i)\](.*?)\[\/list\]/ism", '', $text);
$text = preg_replace("/\[list=((?-i)I)\](.*?)\[\/list\]/ism", '', $text);
$text = preg_replace("/\[list=((?-i)a)\](.*?)\[\/list\]/ism", '', $text);
$text = preg_replace("/\[list=((?-i)A)\](.*?)\[\/list\]/ism", '', $text);
$text = preg_replace("/\[ul\](.*?)\[\/ul\]/ism", '', $text);
$text = preg_replace("/\[ol\](.*?)\[\/ol\]/ism", '', $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", '', $text);
$text = preg_replace("/\[table border=1\](.*?)\[\/table\]/sm", '', $text);
$text = preg_replace("/\[table border=0\](.*?)\[\/table\]/sm", '', $text);
$text = str_replace('[hr]', '$2 ", $text);
// Declare the format for [spoiler] layout
$SpoilerLayout = '' . DI::l10n()->t('Click to open/close') . ' $1$1 $2$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", '$1 $1 ', $text);
$text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism", '$1 $1 ', $text);
$text = preg_replace("/\[audio\](.*?)\[\/audio\]/ism",
'$1 ', $text);
}
// Backward compatibility, [iframe] support has been removed in version 2020.12
$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", 'VIDEO ', $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("' . $text . ' ';
}
}
if (!$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', $text);
}
// Remove all hashtag addresses
if ($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 $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 '#'
. XML::escape($matches[1]) . ' ';
}
}, $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 and replace new lines that remain
$text = preg_replace_callback('/\[(noparse|nobb)](.*?)\[\/\1]/ism', function ($match) {
return str_replace("\n", "' . htmlentities(trim($matches[2], "\n\r"), ENT_NOQUOTES, 'UTF-8') . '' . htmlentities($matches[2], ENT_NOQUOTES, 'UTF-8') . '';
}
return $return;
},
$text
);
$config = \HTMLPurifier_HTML5Config::createDefault();
$config->set('HTML.Doctype', 'HTML5');
$config->set('HTML.SafeIframe', true);
$config->set('URI.SafeIframeRegexp', '%^(?:
https://www.youtube.com/embed/
|
https://player.vimeo.com/video/
|
' . DI::baseUrl() . '/oembed/ # Has to change with the source in Content\Oembed::iframe
)%xi');
$config->set('Attr.AllowedRel', [
'noreferrer' => true,
'noopener' => true,
]);
$config->set('Attr.AllowedFrameTargets', [
'_blank' => true,
]);
$HTMLPurifier = new \HTMLPurifier($config);
$text = $HTMLPurifier->purify($text);
return $text;
}
/**
* Strips the "abstract" tag from the provided text
*
* @param string $text The text with BBCode
* @return string The same text - but without "abstract" element
*/
public static function stripAbstract($text)
{
$text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", '', $text);
$text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", '', $text);
return $text;
}
/**
* Returns the value of the "abstract" element
*
* @param string $text The text that maybe contains the element
* @param string $addon The addon for which the abstract is meant for
* @return string The abstract
*/
public static function getAbstract($text, $addon = '')
{
$abstract = '';
$abstracts = [];
$addon = strtolower($addon);
if (preg_match_all("/\[abstract=(.*?)\](.*?)\[\/abstract\]/ism", $text, $results, PREG_SET_ORDER)) {
foreach ($results AS $result) {
$abstracts[strtolower($result[1])] = $result[2];
}
}
if (isset($abstracts[$addon])) {
$abstract = $abstracts[$addon];
}
if ($abstract == '' && preg_match("/\[abstract\](.*?)\[\/abstract\]/ism", $text, $result)) {
$abstract = $result[1];
}
return $abstract;
}
/**
* Callback function to replace a Friendica style mention in a mention for Diaspora
*
* @param array $match Matching values for the callback
* [1] = Mention type (! or @)
* [2] = Name
* [3] = Address
* @return string Replaced mention
* @throws \Friendica\Network\HTTPException\InternalServerErrorException
* @throws \ImagickException
*/
private static function bbCodeMention2DiasporaCallback($match)
{
$contact = Contact::getByURL($match[3], false, ['addr']);
if (empty($contact['addr'])) {
return $match[0];
}
$mention = $match[1] . '{' . $match[2] . '; ' . $contact['addr'] . '}';
return $mention;
}
/**
* Converts a BBCode text into Markdown
*
* This function converts a BBCode item body to be sent to Markdown-enabled
* systems like Diaspora and Libertree
*
* @param string $text
* @param bool $for_diaspora Diaspora requires more changes than Libertree
* @return string
* @throws \Friendica\Network\HTTPException\InternalServerErrorException
*/
public static function toMarkdown($text, $for_diaspora = true)
{
$original_text = $text;
// Since Diaspora is creating a summary for links, this function removes them before posting
if ($for_diaspora) {
$text = self::removeShareInformation($text);
}
/**
* Transform #tags, strip off the [url] and replace spaces with underscore
*/
$url_search_string = "^\[\]";
$text = preg_replace_callback("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/i",
function ($matches) {
return '#' . str_replace(' ', '_', $matches[2]);
},
$text
);
// Converting images with size parameters to simple images. Markdown doesn't know it.
$text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text);
// Convert it to HTML - don't try oembed
if ($for_diaspora) {
$text = self::convert($text, false, self::DIASPORA);
// Add all tags that maybe were removed
if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
$tagline = '';
foreach ($tags[2] as $tag) {
$tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
if (!strpos(html_entity_decode($text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
$tagline .= '#' . $tag . ' ';
}
}
$text = $text . " " . $tagline;
}
} else {
$text = self::convert($text, false, self::CONNECTORS);
}
// If a link is followed by a quote then there should be a newline before it
// Maybe we should make this newline at every time before a quote.
$text = str_replace([''], [''], $text);
$stamp1 = microtime(true);
// Now convert HTML to Markdown
$text = HTML::toMarkdown($text);
DI::profiler()->saveTimestamp($stamp1, "parser");
// 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', 'img'], 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;
}
// 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 string $author Author display name
* @param string $profile Author profile URL
* @param string $avatar Author profile picture URL
* @param string $link Post source URL
* @param string $posted Post created date
* @param string|null $guid Post guid (if any)
* @return string
* @TODO Rewrite to handle over whole record array
*/
public static function getShareOpeningTag(string $author, string $profile, string $avatar, string $link, string $posted, string $guid = null)
{
$header = "[share author='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $author) .
"' profile='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $profile) .
"' avatar='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $avatar) .
"' link='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $link) .
"' posted='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $posted);
if ($guid) {
$header .= "' guid='" . str_replace(["'", "[", "]"], ["'", "[", "]"], $guid);
}
$header .= "']";
return $header;
}
/**
* Returns the BBCode relevant to embed the provided URL in a post body.
* For media type, it will return [img], [video] and [audio] tags.
* For regular web pages, it will either output a [bookmark] tag if title and description were provided,
* an [attachment] tag or a simple [url] tag depending on $tryAttachment.
*
* @param string $url
* @param bool $tryAttachment
* @param string|null $title
* @param string|null $description
* @param string|null $tags
* @return string
* @throws \Friendica\Network\HTTPException\InternalServerErrorException
*@see ParseUrl::getSiteinfoCached
*
*/
public static function embedURL(string $url, bool $tryAttachment = true, string $title = null, string $description = null, string $tags = null): string
{
DI::logger()->info($url);
// If there is already some content information submitted we don't
// need to parse the url for content.
if (!empty($title) && !empty($description)) {
$title = str_replace(["\r", "\n"], ['', ''], $title);
$description = '[quote]' . trim($description) . '[/quote]' . "\n";
$str_tags = '';
if (!empty($tags)) {
$arr_tags = ParseUrl::convertTagsToArray($tags);
if (count($arr_tags)) {
$str_tags = "\n" . implode(' ', $arr_tags) . "\n";
}
}
$result = sprintf('[bookmark=%s]%s[/bookmark]%s', $url, ($title) ? $title : $url, $description) . $str_tags;
DI::logger()->info('(unparsed): returns: ' . $result);
return $result;
}
$siteinfo = ParseUrl::getSiteinfoCached($url);
if (in_array($siteinfo['type'], ['image', 'video', 'audio'])) {
switch ($siteinfo['type']) {
case 'video':
$bbcode = "\n" . '[video]' . $url . '[/video]' . "\n";
break;
case 'audio':
$bbcode = "\n" . '[audio]' . $url . '[/audio]' . "\n";
break;
default:
$bbcode = "\n" . '[img]' . $url . '[/img]' . "\n";
break;
}
return $bbcode;
}
unset($siteinfo['keywords']);
// Bypass attachment if parse url for a comment
if (!$tryAttachment) {
return "\n" . '[url=' . $url . ']' . $siteinfo['title'] . '[/url]';
}
// Format it as BBCode attachment
return "\n" . PageInfo::getFooterFromData($siteinfo);
}
}
 . '/images/lock_icon.gif "' . DI::l10n()->t('Encrypted content') . '"){kind=icon}