feat: Feature a "type annotations needed" diagnostic#22144
Draft
ChayimFriedman2 wants to merge 8 commits intorust-lang:masterfrom
Draft
feat: Feature a "type annotations needed" diagnostic#22144ChayimFriedman2 wants to merge 8 commits intorust-lang:masterfrom
ChayimFriedman2 wants to merge 8 commits intorust-lang:masterfrom
Conversation
This comment has been minimized.
This comment has been minimized.
And make `InferenceTable` used by inference only.
Some usages were converted into `demand_eqtype()` which is the name in rustc and also reports diagnostics, and others should use a different method.
The obligation `Type: ParentTrait` for a path used to be registered three(!) times. Leave it registered implicitly as one of the generic predicates of the item.
- Do not eagerly normalize the `IntoFuture::Output` projection - Store `IntoFuture` and `IntoFuture::Output` in `LangItems`, to save computing them every time. They are not really lang items, but are similar in nature.
This is what rustc does. It's more performant, but we couldn't do that so far because old code relied on also restoring the obligations, i.e. it was registering maybe-incorrect obligations in the `InferenceTable` without first checking them in a separate `ObligationCtxt`. That code is now gone.
There are two kinds of this diagnostic: 1 Some constructs, like calling or indexing, require the type to be known *immediately*. 2. Otherwise, if we cannot resolve the variable at the end of inference, it's an error. This PR implements only the first kind, which is also simpler to implement. We just need to push an error in `structurally_resolve_type()` (as opposed to `try_structurally_resolve_type()` that never errors) if the type cannot be resolved, and of course also make sure we only call this function where rustc does. The second case needs to track the origin of infer vars, and it's for another time.
ChayimFriedman2
force-pushed
the
type-must-be-known
branch
from
9fae888 to
169f880
Compare
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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.
There are two kinds of this diagnostic:
1 Some constructs, like calling or indexing, require the type to be known immediately.
2. Otherwise, if we cannot resolve the variable at the end of inference, it's an error.
This PR implements only the first kind, which is also simpler to implement. We just need to push an error in
structurally_resolve_type()(as opposed totry_structurally_resolve_type()that never errors) if the type cannot be resolved, and of course also make sure we only call this function where rustc does.The second case needs to track the origin of infer vars, and it's for another time.
Blocked on #22141.