For each platform declaration, there must be at least one impl declaration in
the module with the compatible signature; similarly, for each impl declaration,
there must be at least one platform declaration with the compatible signature.
Note that currently the presence of the 'impl' modifier is not checked yet.
Also, the sad fact is that if you have platform and impl declarations which are
not compatible, you get two errors: on the platform delcaration and on the impl
declaration. This needs to be addressed as well
Because of incomplete infrastructure, introduce a few dirty hacks to make
diagnostic tests analyze the common code via DefaultAnalyzerFacade, and also
add common module sources to analysis of platform-specific modules.
Also do not render 'platform' declarations from platform-specific modules to
the .txt file, since they are very likely to be exact copies of the same
declaration in the common module (see RecursiveDescriptorComparator)
Because PackageViewDescriptor may consist of several package fragments from
different modules (see LazyPackageViewDescriptorImpl#fragments), we now filter
out fragments from irrelevant modules before rendering them into the .txt
Basically they're built upon basic coroutine tests, but for each of them
different interceptResume implementation are injected
(currently there are 12 of them).
It might be more simple just to generated additional testData, but I see this
more problematic in a sense of further maintenance
Note that all tests add idempotent 'interceptRun' operators,
which just execute given lambda in the current thread
#KT-14891 Fixed
In AbstractPositionManagerTest, built-ins could no longer be resolved because
JvmIDEVirtualFileFinder is used in IDE tests (even when the compilation is
called through JvmResolveUtil/GenerationUtils as in compiler tests), which is
not yet able to locate .kotlin_builtins files
- In tests on built-ins with no sources, just call
JvmResolve.analyze(environment) and inspect the resulting module
- In AbstractLocalClassProtoTest, create container via
TDAForJVM.createContainer
- Inline single module container creation into AbstractDiagnosticsTest
To simplify migration from 1.0 to 1.1, do not allow data classes to
automatically implement abstract equals/hashCode/toString declared in
super-interfaces (KT-11306) if "-language-version 1.0" is specified
This is needed because otherwise with separate modules turned on in the
compiler, these tests do not display the contents of classes from the
dependency module in the rendered output
Currently behavior is unchanged because the "all project" scope is passed,
however in the future this will allow to implement separate modules in the
compiler properly
- Document what exactly should findPackageParts return
- Make EMPTY a named object instead of a val in the companion
- Do not use JvmPackagePartProvider in tests where Empty works fine
- Add a couple default values to arguments of setupResolverForProject
Add an assertion to prevent duplicate initialization of CodegenTestFiles
instance in CodegenTestCase, which was happening in ScriptGenTest with multiple
invocation of loadFile()
The only place where the logic has changed is in AbstractDiagnosticsTest, where
modules are already created and sealed before the analysis. Copy-paste the
container creation logic there (it's almost fine because it's also present in a
bunch of other tests), and simplify it: get rid of incremental compilation and
other stuff irrelevant for diagnostic tests.
This is needed to make analyzeFilesWithJavaIntegration configure the module
properly before sealing it
Ilya Gorbunov