Tim Disney
Tim Disney
> But I think I will make a new repo for the readtable/reader/charstream stuff under my account if that's alright with you. Sounds good! > regex literal detection relies on...
Thinking about `Term`/`Syntax` unification more I've become less happy about having methods like `is*` and `from*` on those objects. I think it causes more breakage for not much value. The...
The downside of that approach is you need the additional wrapping pass. But if there's not a clean way to separate the tokenization from the wrapping we can go with...
> What regex stuff are you thinking would not need changing if `Token | List` were put into the prefix? Oh, all the regex prefix stuff would need to change....
Sounds good!
It's actually here: https://github.com/mozilla/sweet.js/blob/master/src/token-expander.js#L23-L34 But its a bit subtle so I need to think through it.
Agreed, in general the ambiguous subsets error needs to be more descriptive (say what and where the binding was) and this case in particular the error should mention re-declaration.
Right, as you've noticed the "real" `lineNumber` etc. have to be adjusted. Not sure if this is obvious from that code but the reason we have to do the adjustment...
An unfortunate and glaring omission, we don't have explicit grammar support for putting contracts on classes yet. I really should have added them before, it's now at the top of...
Interesting... > 1. The -> wont work for classes, ==> need to be used. This is definitely a bug. The `==>` contract is supposed to restrict the function it is...