Allow construction of `RegExp` from invalid UTF-8 strings

Question

Closed this issue 2 years ago · 2 comments

Since Automa is byte-oriented, we probably need a way to do e.g. re"[\x80-\x81]"

Answer 1 · 2021-08-20T10:22:53.000Z

We also need to be able to do re"[α-γ]"

Answer 2 · 2022-04-30T15:14:07.000Z

Also, it looks like Automa handles some Unicode characters poorly inside parse_class:

julia> re"[ø]".args
1-element Vector{Any}:
 0xf8:0xf8