Parser does not play well with Umlaut characters.

[josephbirkner]

Currently, Umlaut characters seem to throw off the parser. E.g.:

Or:

[johannes-wolf]

Maybe we can use https://github.com/uni-algo/uni-algo for codepoint properties & normalization.

[johannes-wolf]

Clarification: The parser has no problem handling umlaut characters, but does not allow any [^_a-zA-Z0-9] characters in identifiers.

Improving the error message or allowing Unicode identifiers would fix this minor issue.