Closed andreasabel closed 2 years ago
Closed in favor of #200.
Actually, I see that #200 breaks #141, so I leave my solution in the race.
There is a problem with regexes like [01-89]
which mean one of 0
, 1-8
, 9
. If we already have numbers (NUM), then this is hard to parse. Maybe NUM
should only be lexed inside the multiplication brances {nnn,mmm}
, using lexer states.
Closing in favor of #202.
WIP #197: add rules to interpret number literals as character sequences
I think this fixes the problem, but this isn't the final test case yet, so, WIP.