TypeResolver: Resolve Module.Type as qualified type name
#69855
Conversation
AnthonyLatsis
requested review from
hborla,
slavapestov and
xedin
as code owners
|
@swift-ci please smoke test Linux |
| auto ty = resolveType(repr->getBase(), options); | ||
| baseOptions |= TypeResolutionFlags::DisallowOpaqueTypes; | ||
|
|
||
| // The parser assumes 'X.Type' is a metatype representation, but it might be |
There was a problem hiding this comment.
Can we get rid of metatype representations and model them as a MemberTypeRepr?
There was a problem hiding this comment.
Oh, but we will need two bits to disambiguate between X.Type/X.Protocol and X.`Type`/X.`Protocol`. Does that sound alright?
There was a problem hiding this comment.
Or just one isEscaped bit for SimpleIdentTypeRepr really.
There was a problem hiding this comment.
We absolutely need that bit to disambiguate in expression context anyway. Currently, both of these type-check as Int.Type.Type instances:
extension Int {
struct `Type` {}
}
_ = Int.`Type`.self
_ = Int.Type.selfThere was a problem hiding this comment.
Let’s start with #70467.
| auto *memberTR = | ||
| MemberTypeRepr::create(ctx, {simpleIdentBase, memberComponent}); | ||
|
|
||
| auto ty = resolveType(memberTR, options); |
There was a problem hiding this comment.
Nit: Local ty variable seems kind of ambiguous here as it has the same name as the one we resolve base type, wonder if would be better name then differently?
Resolves #67378.