Coq inhabited errors #1470
Merged
Coq inhabited errors #1470
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
eddywestbrook
approved these changes
Oct 8, 2021
There was a problem hiding this comment.
Well, I'm still a little on the fence about this approach overall, because the isort stuff is not checked by the SAW core type-checker, and so there are now SAW core terms that will type-check but really shouldn't. This "shouldn't" only comes out in the translation to Coq, and I'm worried that this could lead to problems down the road. OTOH, it's an interesting idea for how to get around the problem of making error translate to valid Coq without changing our current approaches too much. Also, if we do run into issues, this wasn't that big of a change, and we can always reverse it and try something else. So I feel like this is an experiment, and why not?
can be passed to the `error` axiom. This makes the axiom realizable. Add instances for the major types of interest. This messes up extraction for most of the Cryptol prelude, as the necessary `Inhabited` instances must be passed along through polymorpic definitions.
…d to be inhabited. This adds a boolean flag to to the SAWCore AST node for sorts. For most purposes (typechecking and such), this annotation is purely advisory and does not directly affect the SAWCore metatheory. However, for Coq export, it is used to identify types arguments to functions that are required to be inhabited. When appearing in a Pi type, lambda or definition, such type arguments will be accompanied by an implicit argument demanding a typeclass instance for the `Inhabited` class. In concrete SAWCore syntax, this annotaion is attached when using the new `isort` keyword, which is otherwise identical to `sort`. This patch only adds the new annotation flag, but does not yet plumb through the Coq translation changes.
annotation into implicit parameters mentioning the `Inhabited` typeclass.
For example, the vector `at` function is defined as so:
```
at : (n : Nat) -> (a : isort 0) -> Vec n a -> Nat -> a;
at n a v i = atWithDefault n a (error a "at: index out of bounds") v i;
```
It is translated into a Coq term with type:
```
forall (n : Nat) (a : Type) {Inh_a : Inhabited a}, Vec n a -> Nat -> a
```
necessary to propigate inhabited information. Regenerate the Coq prelude and ensure that inhabited class instances are avaliable where necessary.
This declaration is ignored for most purposes, but is intended to allow the user to directly influence translation passes. In particular, we use it here to emit typeclass instance declarations in the Coq translation for witnessing the inhabitedness of certain types.
and use explicit references for `multiFixM`.
and remove an unused argument to `DefRename`.
robdockins
force-pushed
the
coq-inhabited
branch
from
1891d4b to
e298ac0
Compare
added to sort terms.
robdockins
added
the
PR: ready to merge
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
In the Coq prelude, weaken the previously-unsound
erroraxiom by insisting that it only be applied at inhabited types. This requires some additional machinery to pipe around the necessary facts about types being inhabited.