The error-level enhancement is kept as warning-level and a new
LanguageFeature is introduced to turn the warning into an error.
#KT-63208 Fixed
#KT-63209
(cherry picked from commit 371b1eb3d5)
This is required for EnhancedTypeForWarningAttribute because scopes
should not be reused between cone types with different values of
this attribute.
#KT-63208
(cherry picked from commit 9189154cae)
^KT-59074 Fixed
This solution is not ideal. Ideally, the allopen compiler plugin
shouldn't report `private` properites as `open` KT-64980, but it will
unpredictably break other things.
For some reason, codegen box test was not added in 082c337faa. The main
point of it now is to check that everything works correctly in the
IrFakeOverrideBuilder mode.
This fixes some type argument mismatch errors caused by a captured type
being approximated and then captured again.
Some places need to be adapted to work with captured types that
previously only worked with approximated types.
#KT-62959 Fixed
The inter-module interaction was partially supported, but the
DiagnosticCompilerTestFE10TestdataTestGenerated and
LFirPreresolvedReversedDiagnosticCompilerFE10TestDataTestGenerated were
failing. This is because the arguments of annotations were not fully
resolved in loadExperimentalitiesFromAnnotationTo function.
^KT-60262 Fixed
Now fake overrides declarations invisible for override
are totally ignored by FakeOverrideBuilder, instead of
creating fake overrides for them, and than filtering them
out later.
As side-effect it fixes KT-64972.
^KT-64974
This origin was set somewhat arbitrarily.
It actually only makes sense for `IrCall`s to property getters.
There were no places in lowerings that checked `IrGetField` expressions
for this origin value.
^KT-62500 Fixed
At the moment, SubclassOptInRequired is marked with the
ExperimentalSubclassOptIn annotation. However, it does not work
as expected due to a missing opt-in error. To use SubclassOptInRequired,
an explicit opt-in is necessary because SubclassOptInRequired is an
unstable feature now.
^KT-64739
In before, actualizer was copying attributeOwnerId to new
nodes. This leads to attributeOwnerId linked to non-actualized
node. It's unclear what it leads to, so we just stop doing that.
In particular, if attributeOwnerId references another node,
this information would be lost.
^KT-64895
...instead of those computed from a frontend representation
(`symbol.signature`).
This is more robust; besides, we are going to turn off building
signatures from FIR, so `symbol.signature` is going to return null soon.
The error-level enhancement is kept as warning-level and a new
LanguageFeature is introduced to turn the warning into an error.
#KT-63208 Fixed
#KT-63209
This is required for EnhancedTypeForWarningAttribute because scopes
should not be reused between cone types with different values of
this attribute.
#KT-63208
Previously, if the parent class was in Java code and the subclass in
Kotlin, the opt-in usage error would not be reported for the subclass.
The problem was that the extractClassFromArgument function couldn't get
the class type from Java code. It was looking for a ConeClassLikeType,
but found a ConeFlexibleType instead.
#KT-60258 Fixed
Kirill Rakhman