This is a hack to work around the fact that type mappings should not be
inherited by inlining contexts for lambdas called from anonymous
objects. As the lambda can call the inline function again, this could
produce a reference to the original object, which is remapped to a new
type in the parent context. Unfortunately, there are many redundant
`MethodRemapper`s between the lambda and the class file, so simply
editing `TypeRemapper` does not work. Hence, this hack. For now.
(Issue found by compiling IntelliJ IDEA BTW.)
Do not report same set of diagnostics for variable call if actual
error was happened with a function candidate. Here the candidate is
invoke function on DeepRecursiveFunction
^KT-40991 Fixed
^KT-41491 Fixed
^KT-40926 In Progress
Type of a block is a kind of irrelevant for lambdas: their type is much
more complicated and defined via FirDataFlowAnalyzer#returnExpressionsOfAnonymousFunction
at at FirCallCompleter.LambdaAnalyzerImpl#analyzeAndGetLambdaReturnArguments
It might be necessary for cases like
run @l1{
run {
if (...) return@l1 1
}
}
"if" is a synthetic call, but without candidate since there's just one branch
But return@l1 1 is incomplete because it's resolved in dependent context
and "1" is hang incomplete integer literal
When resolving inline class methods in binary dependencies we look for
methods matching both the new and the old mangling scheme. On the IR
backend the method signature has to be computed for the inline class
replacement, since the logic for signature mangling is not contained in
the method signature mapping, unlike in the old backend.
Take local/anonymous classes and their inner classes into account.
Simplify approximation: use first available super classifier instead
of first super class. This approximation should only happen for
private declarations that were not previously approximated by frontend.
So basically the only requirement for the approximated types is to be
denotable. Note that this only works if the types are not used later.
JVM uses a different string table implementatin as it needs exact
types of private members for reflection.
^KT-20996 Fixed
Before this commit, during the resolve of companion objects we used
the same context than for any nested class. However, during companion
object resolve we should not have companion object receiver itself
accessible in any case (in particular, it should not be accessible
in constructor). So in this commit we introduced separate context
for this purpose.
Mikhail Glukhikh