Describe Type Alias Impl Trait (TAIT) Inference Algorithm #1297
Conversation
JohnTitor
added
the
S-waiting-on-review
|
Regarding |
|
Can you post a screenshot of what the mdbook-mermaid output looks like? cc @rust-lang/wg-rustc-dev-guide since this adds a new build tool |
| * Explain how it invokes `type_of` | ||
| * We look at the bounds, we are able to type check it as is | ||
|
|
||
| ```mermaid |
| The `type_of` query, when applied to an opaque type O, returns the hidden type. That hidden type is computed by combining the results from each constraining function within defining scope of O. | ||
|
|
||
| ```mermaid | ||
| flowchart TD |
| ### Within the ordinary type check of a single function | ||
|
|
||
| ```mermaid | ||
| flowchart TD |
| When queries encounter a `opaque_ty_obligation`, they do not try to process them, but instead just store the constraints into the infcx. | ||
|
|
||
| ```mermaid | ||
| graph TD |
|
Sorry for the delay, adding an additional seems fine, @nikomatsakis do you know the best person to review this PR? Maybe oli-obk? |
|
@oli-obk is this alright |
|
needs another sync with the hackmd, but yea, we should merge this once it passes CI again. We can iterate on the details later |
JohnTitor
added
S-waiting-on-author
|
@nikomatsakis Could you fix the CI failure? |
|
@JohnTitor CI says logs are deleted... are you able to see what the error is, and how does one restart CI? |
I cannot either as it's too old, but CI stopped at "Check line lengths" so there should be a long line. Actually, L3 on src/opaque-types-type-alias-impl-trait-inference.md has more than 100 characters. |
|
will merge this after CI succeeds, then sync with hackmd |
This PR contains a draft writeup explaining rust-lang/rust#92007. Mostly at this point it has diagrams that I wrote. The idea was to help myself understand that PR.
cc @oli-obk, who is going to help me finish it up
This hackmd is linked to the branch so that we can edit it live.