Friendica/friendica-github
1
0
Fork 0
mirror of https://github.com/friendica/friendica synced 2025-05-01 09:04:24 +02:00
Code Issues Projects Releases Packages Wiki Activity

Some more warnings and erors are fixed

Browse source
This commit is contained in:
Michael 2022-08-28 19:27:21 +00:00
parent 61cbcf85a1
commit 3a840aa22d
4 changed files with 43 additions and 19 deletions
Download patch file Download diff file Expand all files Collapse all files

28
src/Util/ParseUrl.php
View file

@ -287,22 +287,22 @@ class ParseUrl
// Expected form: Content-Type: text/html; charset=ISO-8859-4
if (preg_match('/charset=([a-z0-9-_.\/]+)/i', $curlResult->getContentType(), $matches)) {
$charset = trim(trim(trim(array_pop($matches)), ';,'));
} else {
// Then in body that gets precedence
// Expected forms:
// - <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
// - <meta charset="utf-8">
// - <meta charset=utf-8>
// - <meta charSet="utf-8">
// We escape <style> and <script> tags since they can contain irrelevant charset information
// (see https://github.com/friendica/friendica/issues/9251#issuecomment-698636806)
Strings::performWithEscapedBlocks($body, '#<(?:style|script).*?</(?:style|script)>#ism', function ($body) use (&$charset) {
if (preg_match('/charset=["\']?([a-z0-9-_.\/]+)/i', $body, $matches)) {
$charset = trim(trim(trim(array_pop($matches)), ';,'));
}
});
}
// Then in body that gets precedence
// Expected forms:
// - <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
// - <meta charset="utf-8">
// - <meta charset=utf-8>
// - <meta charSet="utf-8">
// We escape <style> and <script> tags since they can contain irrelevant charset information
// (see https://github.com/friendica/friendica/issues/9251#issuecomment-698636806)
Strings::performWithEscapedBlocks($body, '#<(?:style|script).*?</(?:style|script)>#ism', function ($body) use (&$charset) {
if (preg_match('/charset=["\']?([a-z0-9-_.\/]+)/i', $body, $matches)) {
$charset = trim(trim(trim(array_pop($matches)), ';,'));
}
});
$siteinfo['charset'] = $charset;
if ($charset && strtoupper($charset) != 'UTF-8') {