go to bug id or search bugs for
It uses RFC 1738 (escape ' ' as +). It should use RFC 3986. i.e. escape ' ' as %20.
Since urldecode() decodes %20 to ' '. There is no BC issue.
Add a Patch
Add a Pull Request
Then wouldn't it be same as rawurlencode, I am assuming that there is a reason to have two of those with slight differences.