The “global” lookup of operators means that we can find operators in
inheriting protocols, causing type-check failures. Don’t consider those
overrides when removing overrides from name lookup results.
In Swift 4, we only gave custom overload types to properties defined in extensions of generic types, using the null type for any other var decl. This meant that a property defined in such an extension would never shadow a property not defined in such an extension. As a result, this permitted cross-module overloads of properties of different types on generic types in certain cases.
This patch adds an exception to the shadowing rules for properties defined in generic type extensions under Swift 4 mode. Permitting cross-module property overloads in the general case would also be source breaking (causing ambiguity errors), so this can be handled in a follow-up Swift 5 mode PR if desired.
Resolves SR-7341.
- getAsDeclOrDeclExtensionContext -> getAsDecl
This is basically the same as a dyn_cast, so it should use a 'getAs'
name like TypeBase does.
- getAsNominalTypeOrNominalTypeExtensionContext -> getSelfNominalTypeDecl
- getAsClassOrClassExtensionContext -> getSelfClassDecl
- getAsEnumOrEnumExtensionContext -> getSelfEnumDecl
- getAsStructOrStructExtensionContext -> getSelfStructDecl
- getAsProtocolOrProtocolExtensionContext -> getSelfProtocolDecl
- getAsTypeOrTypeExtensionContext -> getSelfTypeDecl (private)
These do /not/ return some form of 'this'; instead, they get the
extended types when 'this' is an extension. They started off life with
'is' names, which makes sense, but changed to this at some point. The
names I went with match up with getSelfInterfaceType and
getSelfTypeInContext, even though strictly speaking they're closer to
what getDeclaredInterfaceType does. But it didn't seem right to claim
that an extension "declares" the ClassDecl here.
- getAsProtocolExtensionContext -> getExtendedProtocolDecl
Like the above, this didn't return the ExtensionDecl; it returned its
extended type.
This entire commit is a mechanical change: find-and-replace, followed
by manual reformatted but no code changes.
This patch removes the need for Request objects to provide a default
cycle-breaking value, instead opting to return llvm::Expected so clients
must handle a cycle failure explicitly.
Currently, all clients do the 'default' behavior, but this opens the
possibility for future requests to handle failures explicitly.
The name-lookup behavior that avoids looking for members of a nominal type
or extension therefore when resolving the inheritance clause is now
handled by UnqualifiedLookup, so remove it from the type checker itself.
Fixes rdar://problem/39130543.
Without this, we were firing off way more InheritedDeclsReferenced-
Requests than were actually needed. This drops it way down in the
profile, for what I /think/ is minor wins in type-checking the
stdlib...but it's hard to tell, since the time varies from run to run.
Rather than require clients of lookupQualified() to resolve
their type declarations to nominal type declarations (and
separately cope with modules), have lookupQualified() accept
an array of TypeDecls and handle the resolution to nominal
type declarations (where it can look directly) and module
declarations, combining the results.
Rather than validating the signature of any declaration found by
name lookup, first check whether there is a collision on the full name
of the declaration. This should result in fewer declaration validations.
Name lookup within a protocol extension also looks into protocols and
superclasses on the right-hand side of Self constraints in the where
clause, e.g., "Self: Foo". Turn that function into a request to avoid
infinite recursion on invalid code.
Within unqualified name lookup, replace uses of Type-based lookup
(formed from calls to getSelfTypeInContext()) with declaration-based
name lookup, so that name lookup doesn't depend directly on the type
checker.
The main oddity here is that we need to consider "Self: Foo"
requirements within a protocol extension during name lookup, so that
we will also look into "Foo". This is handled via the basic
name-lookup facilities that allow us to resolve TypeReprs to
declarations without going through the type checker.
Use the declaration-based name lookup facilities to re-implement
ProtocolDecl::getInheritedProtocols(), rather than dynamically selecting
between the requirement signature and the inherited types. This reduces
dependencies for this computation down to basic name lookup (no semantic
analysis) and gives us a stable result.
Add API to get all of the nominal types directly referenced from the
inheritance clause of a given declaration. Use that to find the protocols
to enter into the conformance lookup table based on a given declaration,
without going through the type checker [*].
[*] Except for unqualified lookup still needing to use the type checker.
Use ExtensionDecl::getExtendedNominal() to wire up extensions to their
nominal types early in type checking (the bindExtensions()) operation,
rather than going through type validation to do so.
Introduce a request for ExtensionDecl::getExtendedNominal() that
uses TypeRepr-based resolution to find the extended nominal
type declaration without going through type resolution.
Introduce ExtensionDecl::getExtendedNominal() to provide the nominal
type declaration that the extension declaration extends. Move most
of the existing callers of the callers to getExtendedType() over to
getExtendedNominal(), because they don’t need the full type information.
ExtensionDecl::getExtendedNominal() is itself not very interesting yet,
because it depends on getExtendedType().
Introduce three new requests for name lookup operations that avoid performing
type checking/semantic analysis. They work using syntactic information
(e.g., TypeReprs) and AST-level name lookup operations that will (eventually)
avoid and calls back into type checking. The new requests are:
* Retrieve the superclass declaration of a protocol or class declaration. Use
this request for ClassDecl::getSuperclassDecl() and
ProtocolDecl::getSuperclassDecl().
* Retrieve the types “directly referenced” by a particular location in
an inheritance clause. This query is based on looking at the TypeReprs
and performing fairly-minimal lookup, so it does not involve any Type
computations.
* Retrieve the types “directly referenced” by the underlying type of
a typealias. This query allows us to desugar a typealias without forming
a type.
Along with these is a core operation to transform a set of TypeDecl*s
into a set of NominalTypeDecl*s, looking through typealiases, and
without involving Type at all. The superclass-decl request does this
to find a ClassDecl; other requests will eventually do this to (e.g.)
find all of the protocols mentioned in an inheritance clause.
Switch a number of callers of the Type-based lookupQualified() over to
the newer (and preferred) declaration-based lookupQualified(). These are
the easy ones; NFC.
Make the core lookupQualified() API accept an array of TypeDecls in which
it should look, rather than looking into a Type. This is in preparation
for breaking more type-checker dependencies in the name lookup code.
Add null type check for `Self` type in `UnqualifiedLookup`. It seems this
doesn't happen in normal compilation. But when code-completion happens in
inactive conditional compilation block, since surrounding context hasn't
been typechecked, `Self` type can be null.
rdar://problem/41234606
This function (actually checkAccess) was relying on some implicit
assumptions that aren't actually valid in all cases. When they're not,
just fall back to a slower but more correct implementation; when they
are, assert that the two implementations get the same answer. This
allows us to get rid of adjustAccessLevelForProtocolExtension (see
previous commit), though unfortunately not all of the associated hack.
The diff is bigger than I'd like because it includes moving functions
from NameLookup.cpp into Decl.cpp, but most of those didn't change.
- checkAccess only changed in the one if branch for protocols
- ValueDecl::isAccessibleFrom just added the assertion
- AbstractStorageDecl::isSetterAccessibleFrom did not change
No expected functionality change.
This gets adjustAccessLevelForProtocolExtension, a hack of sorts to
begin with, out of ValueDecl's general API, and down to a helper for
isAccessibleFrom and isSetterAccessibleFrom. (The only reason these
two don't go through access scopes is as an optimization.)
...to push people towards getFormalAccessScope. The one use case that
isn't covered by that is checking whether a declaration behaves as
'open' in the current file; I've added ValueDecl::hasOpenAccess to
handle that specific case.
No intended functionality change.
Make the TypeChecker’s name-lookup routines static, so they don’t depend
on the TypeCheck instance… except for one pesky place where we jump back
into protocol conformance checking. This is part of teasing apart the type
checker.
More groundwork for protocols with superclass constraints.
In several places we need to distinguish between existential
types that have a superclass term (MyClass & Proto) and
existential types containing a protocol with a superclass
constraint.
This is similar to how I can write 'AnyObject & Proto', or
write 'Proto1 & Proto2' where Proto1 has an ': AnyObject'
in its inheritance clause.
Note that some of the usages will be revisited later as
I do more refactoring and testing. This is just a first pass.
