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.

[johannes-wolf]

Can we close this? This is not a bug, as simfil can handle non-ASCIIp characters in strings but does not allow them in identifiers.

[josephbirkner]

If we close this, I would create a follow-up issue to allow unicode in the identifiers, as this might reasonably happen, or explicitely disallow adding unicode-using field names in simfil Objects. But I would prefer the former, because for some information it will be more useful to show it in a "dictionary" way. E.g. a list of translations of a POI name into different languages is much more convenient if it is shown as a dictionary of language: translation rather than array-of-objects-with-two-fields.