- As mentioned in KT-60570, memory consumption in cached symbol names
providers is caused partially by lots of instances of empty mutable
sets. This commit replaces such sets with the `emptySet()` singleton.
- This improvement also makes sense for callable names because the IDE
"package names with top-level callables" set currently contains all
package names in the project.
- In my local tests of the `setUp` Find Usages performance test, the
retained size of `FirDelegatingCachedSymbolNamesProvider` is cut down
from 208MB to 170MB by the empty class name set optimization, and from
170MB to 115MB by the empty callable name set optimization.
UNNECESSARY_SAFE_CALL is a warning,
UNEXPECTED_SAFE_CALL is an error, thus
it's a breaking change.
Also see KT-60695.
^KT-59860 Fixed
Merge-request: KT-MR-11210
Merged-by: Nikolay Lunyak <Nikolay.Lunyak@jetbrains.com>
This change is necessary in order to make `FirElement` abstract class. To do it we need to remove all `accept` and `transform` methods from fir interfaces. So we have to make sure that all fir element implementations have their `accept` and `transform` implementations in their super classes. In particular this change makes 'accept' and 'transform' methods in `FirConstructedClassTypeParameterRef` and `FirConstructedClassTypeParameterRef` base classes.
Now the stub-based symbol provider is reused for common source sets,
which use metadata dependencies. Filtering should be done on scope
level, then declaration provider won't yield them in the first place.
KT-58769
Use existing stub-based JVM library symbol provider for .knm and
.kotlin_metadata files. The only real difference is the scope filtering
by file types
KT-58769
This implementation only checks annotations set on expect/actual
declarations and requires further refinement (e.g. checking of other
annotation targets, class scopes within typealiases).
^KT-58551
From `ConcurrentMap#compute`
>The entire method invocation is performed atomically.
>Some attempted update operations on this map by other threads may
>be blocked while computation is in progress, so the computation
>should be short and simple
And we can call resolution (`reanalyze()`) under this synchronized
block that can take unpredictable time.
This fix drops all heavy operations from synchronization
Previously, when no file annotations were present, the FIR element
didn't have a source.
By making it nullable, it will only be created when appropriate and the
source will never be null.
#KT-55835
We cannot always return null, because in such case some expressions
would become not fully explorable from the types perspective (see
the documentation on the `getCorrespondingTypeIfPossible`)
`FirNamedReference` might appear when resolving method references (like
`foo::bar`), but also when IJ Platform tries to resolve other parts of
the Kotlin PSI, notably a `KtNameReferenceExpression` in a
function call (`bar` in `foo.bar(baz)` expression).
N.B. FE10 implementation does not support returning `null` as a type -
currently it always returns `Unit` type in case it cannot figure out
the actual type. This issue should probably be tackled together with
KT-60166, so that both implementations are more or less aligned
^KT-60341 Fixed
^KT-59077 Fixed
^KTIJ-25745 Fixed
This allows us to properly complete array literals arguments of
annotation calls fixing several false-negative type mismatch errors
as well as enabling the inference of generic type arguments.
#KT-59581 Fixed
#KT-58883 Fixed
Even when the aliased import comes from the same package, it cannot be
removed, because it will break the code
Also, add more tests for the imports coming from the same package
^KTIJ-26276 Fixed
It is required to avoid leakage of annotations instance from
the original type
It should be enough to just create a new instance of an annotation
without a deep copy, because transformer shouldn't touch it
^KT-60387 Fixed
Marco Pennekamp