Special characters in QUrl will be transformed to punycode

Question

I'm having a textfield where the user can add an URL with I'm having a textfield where the user can add an URL with QUrl::fromUserInput() and it then will be put into a list. If I use special characters in the URL like "http://blöd.de/" it will be shown as "http://blöd.de/" but if I only type in "ö" it will get converted to the punycode "http://xn--nda/".

I tried every QUrl::FormattingOptions and every QUrl::ParsingMode

qDebug() << QUrl::fromUserInput("blöd.de"); // results in: http://blöd.de
qDebug() << QUrl::fromUserInput("ö");       // results in: http://xn--nda

Does somebody have an idea how I can convert this punycode back to the special character? And why is it only not converted when I have a top level domain?

Answer 1

The reason some urls are shown with Unicode characters and others with punnycode is to prevent homograph attacks.

One way to decide how to behave for a specific url is by the mean of a TLD whitelist. In Qt you can see and edit the whitelist using QUrl::idnWhitelist() and QUrl::setIdnWhitelist(const QStringList &list).

In your example .de is in the whitelist, but .ö is not. That is why you can see a difference in behaviour.