Nathan Rebours
Nathan Rebours
That would be a kick-ass error message but how doable is it? (my guess being not doable since the error is reported by the compiler afterward)
The "generated by" is the cherry on top but even without it would be great and a massive improvement over the current situation. Especially in a world where the compiler...
For now there's no special support for those cases and the expansion function will only receive the partial application arguments indeed. We could eventually support the `(f x) y` case...
Yes you're right we should document these. For now you can consider that one should always prefer the most recent API, that is prefer the `Vn` submodule with the largest...
I'd like to give this a try in order to make my life easier with #61. I'm trying to understand what the scope of this actually is. My understanding so...
Yeah I think we should widen the use of `Expansion_context` to all context free rules indeed. If you'd like to give this a try I'm happy to point you in...
CC @diml @jsomers @jonludlam
It could definitely be extracted into odoc! @jonludlam would you accept such a PR? Atm MDX uses odoc parser so it would even further simplify the whole thing if it...
Sorry to jump in but what is the status of this PR? Is there anything I can do to help unblock this as merging this PR and releasing a new...
I'm not completely happy with the way internal parsers which don't need the `passphrase` argument deal with it. I gave it a default value so it can pass the test...