Sanitize links before storing them

This commit is contained in:
Michael 2024-02-12 05:21:13 +00:00
parent fe00a3893d
commit 061f43788c
4 changed files with 31 additions and 28 deletions

View file

@ -41,6 +41,7 @@ use Friendica\Model\Tag;
use Friendica\Network\HTTPClient\Client\HttpClientAccept;
use Friendica\Network\HTTPClient\Client\HttpClientOptions;
use Friendica\Util\Map;
use Friendica\Util\Network;
use Friendica\Util\ParseUrl;
use Friendica\Util\Proxy;
use Friendica\Util\Strings;
@ -434,7 +435,7 @@ class BBCode
return $text;
}
$data['url'] = self::sanitizeLink($data['url']);
$data['url'] = Network::sanitizeUrl($data['url']);
if (isset($data['title'])) {
$data['title'] = strip_tags($data['title']);
@ -487,7 +488,7 @@ class BBCode
}
if (!empty($data['provider_url']) && !empty($data['provider_name'])) {
$data['provider_url'] = self::sanitizeLink($data['provider_url']);
$data['provider_url'] = Network::sanitizeUrl($data['provider_url']);
if (!empty($data['author_name'])) {
$return .= sprintf('<sup><a href="%s" target="_blank" rel="noopener noreferrer">%s (%s)</a></sup>', $data['provider_url'], $data['author_name'], $data['provider_name']);
} else {
@ -1067,29 +1068,6 @@ class BBCode
return $text;
}
/**
* Remove invalid parts from an URL
*
* @param string $url
* @return string sanitized URL
*/
private static function sanitizeLink(string $url): string
{
$sanitzed = $url = trim($url);
foreach (['"', ' '] as $character) {
$pos = strpos($sanitzed, $character);
if ($pos !== false) {
$sanitzed = trim(substr($sanitzed, 0, $pos));
}
}
if ($sanitzed != $url) {
Logger::debug('Link got sanitized', ['url' => $url, 'sanitzed' => $sanitzed]);
}
return $sanitzed;
}
/**
* Callback: Sanitize links from given $match array
*
@ -1099,9 +1077,9 @@ class BBCode
private static function sanitizeLinksCallback(array $match): string
{
if (count($match) == 3) {
return '[' . $match[1] . ']' . self::sanitizeLink($match[2]) . '[/' . $match[1] . ']';
return '[' . $match[1] . ']' . Network::sanitizeUrl($match[2]) . '[/' . $match[1] . ']';
} else {
return '[' . $match[1] . '=' . self::sanitizeLink($match[2]) . ']' . $match[3] . '[/' . $match[1] . ']';
return '[' . $match[1] . '=' . Network::sanitizeUrl($match[2]) . ']' . $match[3] . '[/' . $match[1] . ']';
}
}

View file

@ -31,6 +31,7 @@ use Friendica\Util\HTTPSignature;
use Friendica\Util\Images;
use Friendica\Util\Proxy;
use Friendica\Object\Image;
use Friendica\Util\Network;
/**
* Class Link
@ -77,7 +78,7 @@ class Link
} else {
$fields = self::fetchMimeType($url);
$fields['uri-id'] = $uriId;
$fields['url'] = $url;
$fields['url'] = Network::sanitizeUrl($url);
DBA::insert('post-link', $fields, Database::INSERT_IGNORE);
$id = DBA::lastInsertId();

View file

@ -96,6 +96,7 @@ class Media
return false;
}
$media['url'] = Network::sanitizeUrl($media['url']);
$media = self::unsetEmptyFields($media);
$media = DI::dbaDefinition()->truncateFieldsForTable('post-media', $media);

View file

@ -659,6 +659,29 @@ class Network
return !empty($scheme) && in_array($scheme, ['http', 'https']) && parse_url($url, PHP_URL_HOST);
}
/**
* Remove invalid parts from an URL
*
* @param string $url
* @return string sanitized URL
*/
public static function sanitizeUrl(string $url): string
{
$sanitized = $url = trim($url);
foreach (['"', ' '] as $character) {
$pos = strpos($sanitized, $character);
if ($pos !== false) {
$sanitized = trim(substr($sanitized, 0, $pos));
}
}
if ($sanitized != $url) {
Logger::debug('Link got sanitized', ['url' => $url, 'sanitzed' => $sanitized]);
}
return $sanitized;
}
/**
* Creates an Uri object out of a given Uri string
*