Instead of failing constraint generation by returning `nullptr` for an `ErrorExpr` or returning a null type when a type fails to be resolved, return a fresh type variable. This allows the constraint solver to continue further and produce more meaningful diagnostics.
Most importantly, it allows us to produce a solution where previously constraint generation for a syntactic element had failed, which is required to type check multi-statement closures in result builders inside the constraint system.
When computing type relations of code completion items with respect to the context they are used in, we add a `ConstraintKind::Conversion` constraint with an empty constraint locator. This currently crashes in `hasPreconcurrencyCallee`, but actually having a null anchor doesn’t produce any issues.
When determining the linkage of protocol witness table lazy access
functions and their cache variables, look through opaque types to find
the underlying decls.
rdar://96194366
To make this test work, fix an issue in `ConstraintSystem::salvage` where a
threshold breach during solving went unnoticed due to exiting on ambiguity
before reaching the `isTooComplex` check. Address this by moving the
`isTooComplex` check to before we start processing solutions, and stick another
one in `findBestSolution` for short-circuiting while we're here.
This change adds basic helper protocols and structs that are going to be used for making C++ sequences and collection safe and Swifty by adding conformances to `Swift.Sequence`, `Swift.Collection`, etc.
This is not meant to be a final design.
If the type of an `ExplicitCastExpr` is not valid, it is a null type, which causes a crash when checking if the cast type has parameterized existentials.
rdar://95629905
* Implement String.WordView
* Add isWordAligned bit
* Hide WordView for now (also separate Index type)
add bidirectional conformance
Fix tests
* Address comments from Karoy and Michael
* Remove word view, use index methods
* Address Karoy's comments
aaa
If the sequence passed to `Set.intersection` has exactly as many duplicate items as items missing from `self`, then the new implementation in 5.6 will mistakenly return `self` unchanged. (It counts duplicate hits as distinct items in the result.)
rdar://94803458
