An proof refinement logic for computational type theory. Inspired by Nuprl. [For up-to-date development, see JonPRL's successor, RedPRL: https://github.com/redprl/sml-redprl]
I'd like to add canonical-forms tests like Nuprl has... But rather than adding one for each canonical operator, it would be cool to do it all at once, as in match-operator{θ}(M; L; R), where the parameter θ is a canonical operator. This expression only has a value in case θ is known to be a canonical operator.
I'd like to add canonical-forms tests like Nuprl has... But rather than adding one for each canonical operator, it would be cool to do it all at once, as in
match-operator{θ}(M; L; R)
, where the parameterθ
is a canonical operator. This expression only has a value in caseθ
is known to be a canonical operator.The operational semantics would be: