The C# highlighter (and possibly others) chokes on non-ASCII characters in class names, as can be seen in this post:
A comment in the code the default parser marks it as a known shortcoming in Prettify:
fallthroughStylePatterns.push( // TODO(mikesamuel): recognize non-latin letters and numerals in idents [PR_LITERAL, /^@[a-z_$][a-z_$@0-9]*/i, null], [PR_TYPE, /^(?:[@_]?[A-Z]+[a-z][A-Za-z_$@0-9]*|\w+_t\b)/, null], [PR_PLAIN, /^[a-z_$][a-z_$@0-9]*/i, null], ...
By "non-latin," I think it really means non-ASCII. Fixing it would involve changing not just the statement quoted above, but several places where patterns like
a-zA-Z appear. Prettify does some case-folding as part of its processing of regular expressions, and I imagine that contributes to making this a non-trivial problem to solve.
While I agree that's a bug (and I agree with the comments...that's bad code as well), it's one that would need to be fixed on the prettify side. Given this is the first time I've seen this though, I wouldn't wager they'll fix it unless they can do so in a way that doesn't suffer performance-wise for this edge case.