Another pitfall : A RDN containing "OU = Exemple \\C3\\A4\\" A filter containing "uid=r\u00e9dacteur1"
Both are legit. The RDN \\C3\\A4 encodes for 'ä' and a filter accepts UTF-8 chars. Sadly, at the end, we use the same normalization and PrepareString process, which leads to wrong results :/ This has to be addressed...
