AST: Allow non-canonical opaque archetypes #84299
Merged
Conversation
This file contains hidden or 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
We cannot in general assume the substitution map's replacement types are written in terms of the generic signautre of the owner declaration. This was only necessary because this substitution map did not preserve type sugar, but now that it does, we no longer need this hack.
slavapestov
force-pushed
the
non-canonical-archetypes
branch
from
95b582f to
d7726dc
Compare
slavapestov
requested review from
AnthonyLatsis,
hborla,
rjmccall,
tshortli,
xedin and
xymus
as code owners
|
@swift-ci Please smoke test |
|
@swift-ci Please test source compatibility |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.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.
An opaque archetype contains a substitution map. Previously, we required that all replacement types in this substitution map were canonical. This requires the ASTPrinter to re-sugaring generic parameter types in order to get something that round trips, and this re-sugaring was hacky and not fully correct.
Instead, let's just say that opaque archetypes can be non-canonical if their replacement types are non-canonical. Computing the canonical type of an opaque archetype then canonicalizes this substitution map.
This PR only changes opaque archetypes, but existential and pack element archetypes also store substitution maps, and they should receive the same treatment eventually too.