If a function uses a type imported as implementationOnly (or similar), it cannot be serialized.
I added a new API in ModuleDecl (canBeUsedForCrossModuleOptimization), which performs this check.
rdar://72864719
When importing Clang types.
This is not an option with Explicit Module Builds. If the module being built does not (directly or transitively) depend on `Foundation`, then attempting to load it will produce an error because Implicit module loading is no longer allowed..
This change addresses a small number of cases where ClangImporter relies on being able to load `Foundation` on-demand:
- When importing a single Decl from a clang module, we check whether it has certain conformances by checking all extensions of the NominalTypeDecl of the Decl in question, to see if any of the extensions contain the conformance we are looking for, but we only check extensions whose parent module is either the original module of the NominalTypeDecl or the overlay module of the NominalTypeDecl or Foundation. It seems that we do not need to actually import `Foundation` here, just checking the module Identifier should be sufficient.
- In `maybeImportNSErrorOutParameter`, change the behavior to have an exit condition based on whether `Foundation` can be imported, before attempting to load it.
- When checking whether or not we are allowed to bridge an Objective-C type, it also looks sufficient the query whether or not `Foundation` *can* be imported, without loading it.
```
@_specialize(exported: true, spi: SPIGroupName, where T == Int)
public func myFunc() { }
```
The specialized entry point is only visible for modules that import
using `_spi(SPIGroupName) import ModuleDefiningMyFunc `.
rdar://64993425
Rename ImportedModuleDesc to AttributedImport and make it a template that’s parameterized on the representation of the module. This will allow us to reduce duplicated representations of “abstract” ImportDecls.
To help consolidate our various types describing imports, this commit moves the following types and methods to Import.h:
* ImplicitImports
* ImplicitStdlibKind
* ImplicitImportInfo
* ModuleDecl::ImportedModule
* ModuleDecl::OrderImportedModules (as ImportedModule::Order)
* ModuleDecl::removeDuplicateImports() (as ImportedModule::removeDuplicates())
* SourceFile::ImportFlags
* SourceFile::ImportOptions
* SourceFile::ImportedModuleDesc
This commit is large and intentionally kept mechanical—nothing interesting to see here.
The lack of clarity manifested as unexpected behavior when using
getImportedModules to create the module import graph. The new behavior
makes SPI-ness and Shadowing-ness behave similarly in terms of
filtering. We also check if a filter is well-formed to avoid
accidental empty import lists.
Remove the fast path skipping evaluating SPI groups for non-public
decls. This knowledge is still required to allow the use of SPI types in
the signatures of `@usableFromInline` declarations and in internal
properties of structs in non library evolution compilation.
rdar://68530659
rdar://68527580
Extensions to implementation-only types are accepted at type-checking
only if they don't define any public members. However if they declared a
conformance to a public type they were also printed in the
swiftinterface, making it unparsable because of an unknown type.
Still accept such extensions but don't print them.
rdar://problem/67516588
LLVM, as of 77e0e9e17daf0865620abcd41f692ab0642367c4, now builds with
-Wsuggest-override. Let's clean up the swift sources rather than disable
the warning locally.
* Re-create `ASTScope` for each completion
* Add generic params and where clause scope even without missing body
* Use `getOriginalBodySourceRange()` for `AbstractFunctionBodyScope`
* Source range translations for replaced ranges when finding scopes
* Bypass source range checks when the completion happens in the replaced
range
* Be lenient with ASTScope / DeclContext mismatch in code completion
…and modify resolveFileIDConflicts() to diagnose any such violations instead of asserting.
Swift does not allow any two files in the same module to have the same filename, even if they are in different directories. However, this is enforced in the driver, so tests that invoke the frontend directly can violate it. Turns out that a couple of those snuck into the test suite at various points.
This commit updates those tests. It also causes the frontend to diagnose the duplicate filename error just as the driver would have, which should help us understand what happened more easily if this crops up again in the future.
NFC, since invoking the frontend directly is unsupported.
Lift the `DisablePoundIfEvaluation` parsing option
into `LangOptions` to subsume the need for the
`EvaluateConditionals` parameter, and sink the
computation of `CanDelayBodies` down into
`createSourceFileForMainModule`.
