Describe your feature or enhancement suggestion
Currently, when the URL decoder encounters a character that is already decoded, Cryptii throws an error. Instead, Cryptii should ignore decoded characters. In the following example, %2F should be converted to /, but the : causes an error.

Describe a concrete use case
https://duckduckgo.com/l/?uddg=https%3A%2F%2Fduckduckgo.com%2F%3Fq=this%2520is%2520a%2520redirect&ia%3Dweb
redirects to:
https://duckduckgo.com/?q=this%20is%20a%20redirect

The first URL above, similar to what I might get in an email, is evidently a redirect, but the portion of the first URL after uddg= is hard to understand because encoded characters are mixed in with the decoded characters. I'd like be able to decode the encoded characters in "mixed" URLs to make sure that the URLs lead to where I expect and to remove any unnecessary portions, such as tracking identifiers.

A URL decoder wouldn't know whether %20 is part of a URL or just part of a random piece of text, so a toggle to leave %20 as a %20, rather than as a space (as Cryptii currently does), would also be helpful.