wraps up SE-0004 and SE-0029.
I consider the diagnostic changes in Constraints/lvalues.swift to be
indicative of a QoI regression, but I'll deal with that separately.
Type level lookups can fail because the lookup is on an existential
metatype, like `MyProtocol.staticMethod(_:)` is invalid; however the
error message is unclear: “static member 'staticMethod(_:)' cannot be
used on instance of type ‘MyProtocol.Protocol’”.
This fix checks the base of member lookups that failed with the reason
UR_TypeMemberOnInstance for being existential metatypes. It produces
the clearer message “static member ‘staticMethod(_:)’ cannot be used on
protocol metatype ‘MyProtocol.Protocol’”. This change makes it clear
that the use of a static member on the *existential* metatype is the
problem.
The AvailabilityWalker was creating a new AvailabilityWalker instance whenever it
recursed through an assignexpr or memberrefexpr, which produced a new ExprStack.
This caused the fixit mechanics for migrating ++/-- to think that the ++/-- was
at the top level, when it wasn't.
When the selector named by Selector("foo") does not map to a known
Objective-C method, allow one to suppress the warning by wrapping the
string literal in an extra set of parentheses, e.g.,
Selector(("foo"))
Suggest this via a Fix-It on a note so it's discoverable. Addresses
rdar://problem/24791200.
This lets us eliminate the _getObjectiveCType() value witness, which
was working around the lack of proper type witness metadata in witness
tables. Boilerplate -= 1.
Parse 'var [behavior] x: T', and when we see it, try to instantiate the property's
implementation in terms of the given behavior. To start out, behaviors are modeled
as protocols. If the protocol follows this pattern:
```
protocol behavior {
associatedtype Value
}
extension behavior {
var value: Value { ... }
}
```
then the property is instantiated by forming a conformance to `behavior` where
`Self` is bound to the enclosing type and `Value` is bound to the property's
declared type, and invoking the accessors of the `value` implementation:
```
struct Foo {
var [behavior] foo: Int
}
/* behaves like */
extension Foo: private behavior {
@implements(behavior.Value)
private typealias `[behavior].Value` = Int
var foo: Int {
get { return value }
set { value = newValue }
}
}
```
If the protocol requires a `storage` member, and provides an `initStorage` method
to provide an initial value to the storage:
```
protocol storageBehavior {
associatedtype Value
var storage: Something<Value> { ... }
}
extension storageBehavior {
var value: Value { ... }
static func initStorage() -> Something<Value> { ... }
}
```
then a stored property of the appropriate type is instantiated to witness the
requirement, using `initStorage` to initialize:
```
struct Foo {
var [storageBehavior] foo: Int
}
/* behaves like */
extension Foo: private storageBehavior {
@implements(storageBehavior.Value)
private typealias `[storageBehavior].Value` = Int
@implements(storageBehavior.storage)
private var `[storageBehavior].storage`: Something<Int> = initStorage()
var foo: Int {
get { return value }
set { value = newValue }
}
}
```
In either case, the `value` and `storage` properties should support any combination
of get-only/settable and mutating/nonmutating modifiers. The instantiated property
follows the settability and mutating-ness of the `value` implementation. The
protocol can also impose requirements on the `Self` and `Value` types.
Bells and whistles such as initializer expressions, accessors,
out-of-line initialization, etc. are not implemented. Additionally, behaviors
that instantiate storage are currently only supported on instance properties.
This also hasn't been tested past sema yet; SIL and IRGen will likely expose
additional issues.
My previous commit here didn’t work correctly for nested tuples, both
because it didn’t recurse into them to propagate access kind correctly
and because an outer TupleIndex overload (when indexing into the nested
tuple) could still be expecting an lvalue type.
This fix is much better. ConstraintSystem::resolveOverload now
correctly always expects rvalue types from rvalue tuples. And during
applyMemberRefExpr, if the overload expects an rvalue but the tuple
contains lvalues, coerceToType() correctly does any recursive munging
of the tuple expr required.
The issue here is that the constraint solver was deciding on
FixKind::RelabelCallTuple as the fix for the problem and emitting the
diagnostic, even though there were two different fixes possible.
CSDiags has the infrastructure to support doing doing the right thing
here, but is only being used for ApplyExprs, not SubscriptExprs.
The solution is to fix both problems: remove FixKind::RelabelCallTuple,
to let CSDiags handle the problem, and enhance CSDiags to treat
SubscriptExpr more commonly with ApplyExpr. This improves several cases
where the solver was picking one solution randomly and suggesting that
as a fix, instead of listing that there are multiple different solutions.
Now that we have expressions that start with #, the [# introducer for
object literals is no longer guaranteed to indicate an object
literal. For example:
[#line, #column]
is an array literal and
[#line : #column]
is a dictionary literal. Use additional lookahead in the parser to
disambiguate these cases from object literals. Fixes
rdar://problem/24533081.
