.
*
*/
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\DI;
use Friendica\Model\Contact;
use Friendica\Model\Event;
use Friendica\Model\Photo;
use Friendica\Model\Post;
use Friendica\Model\Tag;
use Friendica\Network\HTTPClient\Client\HttpClientAccept;
use Friendica\Network\HTTPClient\Client\HttpClientOptions;
use Friendica\Object\Image;
use Friendica\Protocol\Activity;
use Friendica\Util\Images;
use Friendica\Util\Map;
use Friendica\Util\ParseUrl;
use Friendica\Util\Proxy;
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 = '2021-07-28';
const INTERNAL = 0;
const EXTERNAL = 1;
const API = 2;
const DIASPORA = 3;
const CONNECTORS = 4;
const OSTATUS = 7;
const TWITTER = 8;
const BACKLINK = 8;
const ACTIVITYPUB = 9;
const TOP_ANCHOR = '', $data['type']);
}
if (!empty($data['title']) && !empty($data['url'])) {
if (!empty($data['image']) && empty($data['text']) && ($data['type'] == 'photo')) {
$return .= sprintf('
', $data['url'], self::proxyUrl($data['image'], $simplehtml, $uriid), $data['title']);
} else {
if (!empty($data['image'])) {
$return .= sprintf('
', $data['url'], self::proxyUrl($data['image'], $simplehtml, $uriid), $data['title']);
} elseif (!empty($data['preview'])) {
$return .= sprintf('
', $data['url'], self::proxyUrl($data['preview'], $simplehtml, $uriid), $data['title']);
}
$return .= sprintf('
', $data['url'], $data['title']);
}
}
if (!empty($data['description']) && $data['description'] != $data['title']) {
// Sanitize the HTML
$return .= sprintf('
%s ', trim(HTML::purify($data['description'])));
}
if (!empty($data['provider_url']) && !empty($data['provider_name'])) {
if (!empty($data['author_name'])) {
$return .= sprintf('
%s (%s) ', $data['provider_url'], $data['author_name'], $data['provider_name']);
} else {
$return .= sprintf('
%s ', $data['provider_url'], $data['provider_name']);
}
}
if ($simplehtml != self::CONNECTORS) {
$return .= '
%s ', $url, self::getStyledURL($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(string $url): string
{
$parts = parse_url($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.
*
* @param array $match
* @return string
*/
private static function escapeNoparseCallback(array $match): string
{
$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]
*
* @param array $match
* @return string
*/
private static function unescapeNoparseCallback(array $match): string
{
$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(string $text, string $name, int $occurrences = 0)
{
DI::profiler()->startRecording('rendering');
if ($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) {
DI::profiler()->stopRecording();
return false;
}
$start_equal = strpos($text, '=', $start_open);
$start_close = strpos($text, ']', $start_open);
if ($start_close === false) {
DI::profiler()->stopRecording();
return false;
}
$start_close++;
$end_open = strpos($text, '[/' . $name . ']', $start_close);
if ($end_open === false) {
DI::profiler()->stopRecording();
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;
}
DI::profiler()->stopRecording();
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(string $pattern, string $replace, string $name, string $text): string
{
DI::profiler()->startRecording('rendering');
$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);
}
DI::profiler()->stopRecording();
return $text;
}
private static function extractImagesFromItemBody(string $body): array
{
$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(int $uriid, string $body, array $images): string
{
$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
);
DI::profiler()->stopRecording();
return $return;
}
/**
* Convert complex IMG and ZMG elements
*
* @param [type] $text
* @param integer $simplehtml
* @param integer $uriid
* @return string
*/
private static function convertImages(string $text, int $simplehtml, int $uriid = 0): string
{
DI::profiler()->startRecording('rendering');
$return = preg_replace_callback(
"/\[[zi]mg(.*?)\]([^\[\]]*)\[\/[zi]mg\]/ism",
function ($match) use ($simplehtml, $uriid) {
$attribute_string = $match[1];
$attributes = [];
foreach (['alt', 'width', 'height'] as $field) {
preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches);
$attributes[$field] = html_entity_decode($matches[2] ?? '', ENT_QUOTES, 'UTF-8');
}
$img_str = '' . html_entity_decode('♲', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . " :' . $content . ' ';
break;
case self::DIASPORA:
if (stripos(Strings::normaliseLink($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 block-level custom CSS
$text = preg_replace('#(?<=^|\n)\[style=(.*?)](.*?)\[/style](?:\n|$)#ism', '$2
', $text);
// Check for inline custom CSS
$text = preg_replace("(\[style=(.*?)\](.*?)\[\/style\])ism", '$2 ', $text);
// Mastodon Emoji (internal tag, do not document for users)
$text = preg_replace("(\[emoji=(.*?)](.*?)\[/emoji])ism", '$2 ', $text);
// Add HTML new lines
$text = str_replace("\n", ' $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, $uriid) {
if (strpos($matches[3], "data:image/") === 0) {
return $matches[0];
}
$matches[3] = self::proxyUrl($matches[3], $simple_html, $uriid);
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("' . $text . ' ';
}
}
// Handle mentions and hashtag links
if ($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 $1$3 ',
$text);
} elseif (in_array($simple_html, [self::INTERNAL, self::EXTERNAL, self::API])) {
$text = preg_replace("/([@!])\[url\=(.*?)\](.*?)\[\/url\]/ism",
'$1$3 ',
$text);
} else {
$text = preg_replace("/([#@!])\[url\=(.*?)\](.*?)\[\/url\]/ism", '$1$3', $text);
}
if (!$for_plaintext) {
if (in_array($simple_html, [self::OSTATUS, self::API, 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);
}
// 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::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", $text);
/* Tag conversion
* Supports:
* - #[url=][/url]
* - [url=]#[/url]
*/
self::performWithEscapedTags($text, ['url', 'share'], function ($text) use ($simple_html) {
$text = preg_replace_callback("/(?:#\[url\=[^\[\]]*\]|\[url\=[^\[\]]*\]#)(.*?)\[\/url\]/ism", function($matches) use ($simple_html) {
if ($simple_html == self::ACTIVITYPUB) {
return '#'
. XML::escape($matches[1]) . ' ';
} else {
return '#'
. XML::escape($matches[1]) . ' ';
}
}, $text);
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);
}, $uriid
);
$text = self::interpolateSavedImagesIntoItemBody($uriid, $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
);
// Default iframe allowed domains/path
$allowedIframeDomains = [
DI::baseUrl()->getHostname()
. (DI::baseUrl()->getUrlPath() ? '/' . DI::baseUrl()->getUrlPath() : '')
. '/oembed/', # The path part has to change with the source in Content\Oembed::iframe
'www.youtube.com/embed/',
'player.vimeo.com/video/',
];
$allowedIframeDomains = array_merge(
$allowedIframeDomains,
DI::config()->get('system', 'allowed_oembed') ?
explode(',', DI::config()->get('system', 'allowed_oembed'))
: []
);
$text = HTML::purify($text, $allowedIframeDomains);
DI::profiler()->stopRecording();
return trim($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(string $text): string
{
DI::profiler()->startRecording('rendering');
$text = BBCode::performWithEscapedTags($text, ['code', 'noparse', 'nobb', 'pre'], function ($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;
});
DI::profiler()->stopRecording();
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(string $text, string $addon = ''): string
{
DI::profiler()->startRecording('rendering');
$addon = strtolower($addon);
$abstract = BBCode::performWithEscapedTags($text, ['code', 'noparse', 'nobb', 'pre'], function ($text) use ($addon) {
if ($addon && preg_match('#\[abstract=' . preg_quote($addon, '#') . '](.*?)\[/abstract]#ism', $text, $matches)) {
return $matches[1];
}
if (preg_match("#\[abstract](.*?)\[/abstract]#ism", $text, $matches)) {
return $matches[1];
}
return '';
});
DI::profiler()->stopRecording();
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(array $match): string
{
$contact = Contact::getByURL($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(string $text, bool $for_diaspora = true): string
{
DI::profiler()->startRecording('rendering');
$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);
// Now convert HTML to Markdown
$text = HTML::toMarkdown($text);
// 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);
DI::profiler()->stopRecording();
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 $string): array
{
DI::profiler()->startRecording('rendering');
$ret = [];
self::performWithEscapedTags($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[] = $match;
}
}
});
DI::profiler()->stopRecording();
return array_unique($ret);
}
/**
* Expand tags to URLs, checks the tag is at the start of a line or preceded by a non-word character
*
* @param string $body HTML/BBCode
* @return string body with expanded tags
*/
public static function expandTags(string $body): string
{
return preg_replace_callback("/(?<=\W|^)([!#@])([^\^ \x0D\x0A,;:?'\"]*[^\^ \x0D\x0A,;:?!'\".])/",
function (array $match) {
switch ($match[1]) {
case '!':
case '@':
$contact = Contact::getByURL($match[2]);
if (!empty($contact)) {
return $match[1] . '[url=' . $contact['url'] . ']' . $contact['name'] . '[/url]';
} else {
return $match[1] . $match[2];
}
break;
case '#':
default:
return $match[1] . '[url=' . DI::baseUrl() . '/search?tag=' . $match[2] . ']' . $match[2] . '[/url]';
}
}, $body);
}
/**
* Perform a custom function on a text after having escaped blocks enclosed in the provided tag list.
*
* @param string $text HTML/BBCode
* @param array $tagList A list of tag names, e.g ['noparse', 'nobb', 'pre']
* @param callable $callback
* @return string
* @see Strings::performWithEscapedBlocks
*/
public static function performWithEscapedTags(string $text, array $tagList, callable $callback): string
{
$tagList = array_map('preg_quote', $tagList);
return Strings::performWithEscapedBlocks($text, '#\[(?:' . implode('|', $tagList) . ').*?\[/(?:' . implode('|', $tagList) . ')]#ism', $callback);
}
/**
* Replaces mentions in the provided message body in BBCode links for the provided user and network if any
*
* @param string $body HTML/BBCode
* @param int $profile_uid Profile user id
* @param string $network Network name
* @return string HTML/BBCode with inserted images
* @throws \Friendica\Network\HTTPException\InternalServerErrorException
* @throws \ImagickException
*/
public static function setMentions(string $body, $profile_uid = 0, $network = '')
{
DI::profiler()->startRecording('rendering');
$body = self::performWithEscapedTags($body, ['noparse', 'pre', 'code', 'img'], function ($body) use ($profile_uid, $network) {
$tags = self::getTags($body);
$tagged = [];
foreach ($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, $tag . ' ')) {
continue 2;
}
}
if (($success = Item::replaceTag($body, $profile_uid, $tag, $network)) && $success['replaced']) {
$tagged[] = $tag;
}
}
return $body;
});
DI::profiler()->stopRecording();
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): string
{
DI::profiler()->startRecording('rendering');
$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 .= "']";
DI::profiler()->stopRecording();
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 $tags = null): string
{
DI::profiler()->startRecording('rendering');
DI::logger()->info($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);
DI::profiler()->stopRecording();
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;
}
DI::profiler()->stopRecording();
return $bbcode;
}
unset($siteinfo['keywords']);
// Bypass attachment if parse url for a comment
if (!$tryAttachment) {
DI::profiler()->stopRecording();
return "\n" . '[url=' . $url . ']' . $siteinfo['title'] . '[/url]';
}
// Format it as BBCode attachment
$bbcode = "\n" . PageInfo::getFooterFromData($siteinfo);
DI::profiler()->stopRecording();
return $bbcode;
}
}
 . '/images/lock_icon.gif "' . DI::l10n()->t('Encrypted content') . '"){kind=icon}