[ASTPrinter] When printing a property wrapper attribute name for the fully annotated declaration, make sure that it is wrapped within the `syntaxtype.attribute.name` tag
This reverts commit beb8ecd8cc. Add a
workaround for the dependency issue.
It is unclear why `sourcekitd` is getting added improperly as a
dependency on `lib/sourcekitd.framework/sourcekitd`. This workaround
adjusts the dependency such that we end up with a dependency on
`lib/sourcekitd.framework/Versions/A/sourcekitd` as an order-only
dependency. This should fix the compile issue. I am unable to
reproduce this issue with the `add_library` usage for adding a Swift
library. This allows us to cleave the host and target libraries, and so
the workaround is sufficient to make progress and the problem will be
resolved with the migration towards CMake for handling the dependencies.
When a “separately imported overlay” is added to a SourceFile, two things happen:
1. The direct import of the underlying module is removed from getImports*() by default. It is only visible if the caller passes ImportFilterKind:: ShadowedBySeparateOverlay. This means that non-module-scoped lookups will search _OverlayModule before searching its re-export UnderlyingModule, allowing it to shadow underlying declarations.
2. When you ask for lookupInModule() to look in the underlying module in that source file, it looks in the overlays instead. This means that UnderlyingModule.foo() can find declarations in _OverlayModule.
This moves the swift-lang library out of the tools/SourceKit directory
and into the stdlib directory. As it stands, the library required that
the standard library and the SDK overlay was being built. This is
implicitly done when building the stdlib subdirectory. Furthermore, it
had an odd tie between the flags for Swift and SourceKit which now have
the logic separated. This clearly delineates the host and target
components in the repository.
We used to compute the mangled name in other cases, but document structure is
a syntactic request and can't guarantee that the class/protocol we're getting
the mangled name of is valid in any way so it often breaks assumptions in the
mangler and causes it to crash. It's not clear if the runtime_name is actually
being used anymore, so this change restricts reporting it to just the cases
where we don't need to mangle.
rdar://problem/40956377
Now
* NotApplicable: The result is not relevant for type relation (e.g.
keywords, and overloads)
* Unknown: the relation was not calculated (e.g. cached results), or the
context type is unknown.
* Invalid: The result type is invalid for this context (i.e. 'Void' for
non-'Void' context)
* Unrelated: The result type has no relation to the context type
* Convertible: The result type is convertible to the context type
* Identical: The result type is identical to the context type
This library is already supplied by the LLVM CMake config, so adding it here
sometimes causes the CommandLine option parser to fail, as it registers the
same option twice.
This should repair the Windows build after #29451. The quoting
behaviour was incorrect and was constructing an invalid compiler
invocation. Solve the issue by using `target_include_directories`
instead. However, since this needs the target, hoist the flag
computation to the local sites. This replicates more logic because of
the custom build trying to replicate the CMake build logic in CMake.
Use specific operations for setting the compile flags, link flags,
linked libraries, and library search paths. This allows us to use CMake
more effectively, simplifies the logic, and will ensure that flags are
not duplicated.
When doing header interface generation, we interpret clang command-line
arguments in `initInvocationByClangArguments` and attempt to setup a
matching Swift compiler invocation. One important argument is the module
cache, but clang will only interpret `-fmodules-cache-path` if modules
are enabled (typically with `-fmodules`). While the header itself might
not need modules, Swift will import its own shims module and during
testing this needs to honour lit's provided cache. So we add -fmodules
in sourcekitd-test anytime we add -fmodules-cache-path.
rdar://58836540
This moves the handling of the RPATH and the exported symbol to the
binary themselves. The exported symbol is needed due to the use of the
exported symbol list. This makes the small difference that `_main` is
always exported on Darwin which is not strictly needed in ASAN as ASAN
provides the entry point.
The RPATH is only setup on installation which is sufficient for testing
purposes as CMake ensures that the libraries are fully linked and will
be loaded properly when run from the build tree.
Add a global -module-cache-path option to `sourcekitd-test` and
`complete-test` and have lit provide the default module cache in its
substitutions. Previously many tests have explicitly provided the
`%mcp_opt` option, but this is easy to forget when writing new tests.
The module cache is inserted into the compiler arguments at the
beginning so that it's still possible for a test to override it with a
per-test cache if desired.
rdar://58752842
Add a platform kind and availability attributes for macCatalyst. macCatalyst
uses iOS version numbers and inherits availability from iOS attributes unless
a macCatalyst attribute is explicitly provided.
