Fiets\Util\Names::extractParts fails when name contains multibyte/utf8 chars

[jorisleker]

When processing names that start with an uppercase char and have a multibyte char in the second position (ie Jöris), the regex in Fiets\Util\Names::ExtractParts matches on the [A-Z]{1}\.?[\W] pattern, because [\W] matches any char not in the set [a-z0-9]. This results in us treating the J as an initial, and replacing the first half of the multibyte ö char with a.