When printing the generated interface of a module, also print the decls from
any underscored cross-import overlays it is the direct, or indirect underlying
module of. Declarations are grouped by overlay, with a descriptive `MARK:`
comment introducing each overlay, and a regular comment above each decl listing
the required bystander modules that must be imported for the decl to be
available.
In addition in each overlay:
- import declarations of any underlying modules are filtered out, since they
are either other underscored cross-import overlays, or the target module they
are being presented as being part of.
- import declarations that are also in the target module are filtered out, since
the overlay is being presented as a conditional part of the target module.
Resolves rdar://problem/59445385
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
This adjusts the tests for the difference between line endings on
different platforms. Windows uses CRLF while most Unicies use LF. This
was exposed during the update to the new LLVM snapshot.
It looks like we recently started binding extensions to their nominals in order
to continue to compute access levels via ValueDecl::getFormalAccess() after an
assertion was added to enforce that bindExtensions had been called before
anything tried to call ExtensionDecl::getBoundNominal() - which
getFormalAccess() depends on. Sourcekitd's syntactic requests are made on every
keypress in the editor though, so we shouldn't do any name binding (which may
require module loading) to keep them as fast as possible.
This patch restores the old inferAccessLevel() functions we used prior to the
switch to ValueDecl::getFormalAccess() (plus a few fixes) that does as much as
it can syntactically, without any name binding, and simply doesn't report the
access level in cases where it couldn't be computed without name-binding.
This also fixes an assertion hit we were getting trying to bind extensions in
inactive ifconfig clauses, which ASTScope doesn't support.
Resolves rdar://problem/57202584
The VFS tests were using Unix absolute paths, which does not play well
when Windows see them as relative to the current drive letter.
By using the temporal directory, both Windows and Unix can use the same
paths and avoid the problem.
Additionally, a couple of inputs have to be transformed into the native
path format, because sourcekitd-test compares the inputs as strings, and
they need to match exactly. So the source file and the name of the VFS
entries are transformed into native using the helper from LLVM support.
Group info works by matching source filenames with groups, but in
module merging the decls in the module no longer have associated
SourceFiles. Long-term, maybe we should switch this to working on
filenames directly (using the new support provided by swiftsourceinfo
files), but for now just don't crash.
rdar://problem/56592085
When looking for the SyntaxNode corresponding to a type attribute (like
@escaping), ModelASTWalker would look for one whose range *started* at the type
attribute's source location. It never found one, though, because the
SyntaxNode's range included the @, while the type attribute's source location
pointed to the name *after* the @.
ModelASTWalker was previously constructing SyntaxNodes for EnumElementDecls
manually when visiting their associated EnumCaseDecl so that they would appear
as children rather than siblings. It wasn't actually walking these nodes
though, so missed handling some things, e.g. closures passed as default
argument values. These were also still being visited later, and because the
first visit consumed all the associated TokenNodes, this was triggering an
assertion due to the associated TokenNodes not matching expectations.
Previously 'isSystemModule()' returns true only if the module is:
- Standard library
- Clang module and that is `IsSystem`
- Swift overlay for clang `IsSystem` module
Now:
- Clang module and that is `IsSystem`; or
- Swift overlay for clang `IsSystem` module
- Swift module found in either of these directories:
- Runtime library directoris (including stdlib)
- Frameworks in `-Fsystem` directories
- Frameworks in `$SDKROOT/System/Library/Frameworks/` (Darwin)
- Frameworks in `$SDKROOT/Library/Frameworks/` (Darwin)
rdar://problem/50516314
...and remove the option. This is ~technically~ CLI-breaking because
Swift 5 shipped this as a hidden driver option, but it wouldn't have
/done/ anything in Swift 5, so I think it's okay to remove.
Note that if a parseable interface (.swiftinterface) and a binary
interface (.swiftmodule) are both present, the binary one will still
be preferred. This just /allows/ parseable interfaces to be used.
rdar://problem/36885834
In addition to capturing more detailed preprocessor info, the
DetailedPreprocessorRecord option sets the clang module format to 'raw'
rather than the default 'object'. Sourcekitd doesn't link the code
generation libs, which it looks like the default 'object' format requires,
so it sets this option to true. The subinvocation generated when loading a
module from a .swiftinterface file still used the default prior to this
change though, so it would end up crashing sourcekitd.
This change sets the DetailedProccessorRecord option if the DetailedRecord
option is set on the preprocessor options of parent context's clang module
loader. This fixes interface generation crashing for modules that only have
a .swiftinterface file.
rdar://problem/43906499
* Add swift_evolve feature and disable several tests
This change adds a swift_evolve feature to our lit configuration and uses it to mark several tests as unsupported by swift_evolve.
One of these—test/api-digester/stability-stdlib-abi.swift—is actually pretty bad; we would really like to have it. But the digester has known issues exposed by swift-evolve.
* Remove order dependency in another test
* Tweaks from Jordan’s review
* Distinguish between tests which are intentionally unsupported and temporarily disabled.
* Add an explanation for one unsupproted test.
* Code nitpick.
@effects is too low a level, and not meant for general usage outside
the standard library. Therefore it deserves to be underscored like
other such attributes.
If the Clang declrations are *types*, canonical declaration in Swift is
imported for newest version of Swift. In interface generation, if the
declaration is versioned and it's imported as a member in either or both
version of Swift, we have to take compatibility typealias into account.
* Fixed 'ClangModuleUnit::getTopLevelDecls' to take isCompatibilityAlias() into account
* Fixed bugs in ClangImporter where member-to-member versioned types aren't properly imported.
* Fixed 'SwiftDeclConverter::importFullName' to check equality of getEffectiveContext()
* Fixed 'importer::addEntryToLookupTable' to check equality of getEffectiveContext()
(moved 'ClangImporter::Implementation::forEachDistinctName' to 'NameImporter')
These tests were relying on sourcekitd parsing as frontend instead of
using the driver. Update them now to avoid churn when we fix command
line argument parsing in sourcekit.
The changes from clang-importer-sdk to clang-importer-sdk-nosource -I %t
are because clang-importer-sdk implies using -enable-source-import.
Rather than hack them up to use -Xfrontend, it is cleaner to just stop
using source import at all for these tests. Incidentally, this improved
fidelity in a few places. When using the generated swift modules we
also need to pass a target triple to sourcekit, which exposed some tests
that had mac-specific data. This is a systemic issue for sourcekit
tests, but for now just make those few specific tests that we had
problems with run only on mac.
