Add Futhark lexer

[Munksgaard]

The purpose of this PR is to add support for the Futhark programming language.

I've tried to follow the guide on https://pygments.org/docs/lexerdevelopment/, but please let me know if there's something else I need to do.

[Munksgaard]

Oh my, perhaps I overdid it by adding so many examples...

[Anteru]

Hi, yes, I think that's a bit much on the example side. Could you please try to condense it to 1-2 files at most? Please keep in mind that all code you submit here must be yours or BSD compatible, and I do get nervous when I see large amounts of code getting submitted :)

[Munksgaard]

Hi @Anteru, thanks for your comment. I have added just a single example, that I've written myself for the occasion. Hopefully, this is more to your liking :)

[birkenfeld]

Thanks for the update! Unfortunately quite a few of the tokens generated by the examplefile are Errors :)

(The current way examplefiles are handled is pretty new, which is why this wasn't caught by the test. We'll get a fix whipped up for that.)

[Munksgaard]

Thanks for the update! Unfortunately quite a few of the tokens generated by the examplefile are Errors :)

Ah, yes, I can see that. I didn't know that it needed to classify everything correctly. I'll try to find some time to address this.

[Munksgaard]

@birkenfeld I've tried to update my lexer a bit, but now I'm getting an error in the linter: overlap in character class. I don't know what that means, perhaps you can help me understand?

[birkenfeld]

@Munksgaard it means that you have a single character multiple times in a character class. One way to do that is simply something like [aabc], the other way is what seems to have happened here:

(r'[+-/%=!><|&*^][+-/%=!><|&*^.]*', Operator),

The +-/ is a range from + to / which includes +,-./. - is special in character classes unless it is escaped with \ or put at the beginning or end of the class. So [-+... instead of [+-... will fix it.

[Munksgaard]

@Munksgaard it means that you have a single character multiple times in a character class. One way to do that is simply something like [aabc], the other way is what seems to have happened here:

(r'[+-/%=!><|&*^][+-/%=!><|&*^.]*', Operator),

The +-/ is a range from + to / which includes +,-./. - is special in character classes unless it is escaped with \ or put at the beginning or end of the class. So [-+... instead of [+-... will fix it.

Thank you, I've tried that now :)

[Anteru]

Merged, thanks!