As discussed in this StackOverflow question, when encoding/decoding an url it may happen that some combination of characters ends up being converted.
I got tricked while encoding an URL with a “currency” GET parameter:
http://www.example.com¤cy=EU -> http://www.example.com¤cy=EU
In HTML, the
¤ combination is interpreted as the
¤ (currency) character.
Thus the first part “¤” was converted into “¤” and the remaining part of the parameter “cy” was left as is.
A silly issue indeed 🙂