The main changes are in jvm_package_table.proto and ModuleMapping.kt.
With JvmPackageName, package parts can now have a JVM package name that
differs from their Kotlin name. So, in addition to the old package parts
which were stored as short names + short name of multifile facade (we
can't change this because of compatibility with old compilers), we now
store separately those package parts, which have a different JVM package
name. The format is optimized to avoid storing any package name more
than once as a string.
Another notable change is in KotlinCliJavaFileManagerImpl, where we now
load .kotlin_module files when determining whether or not a package
exists. Before this change, no PsiPackage (and thus, no JavaPackage and
eventually, no LazyJavaPackageFragment) was created unless there was at
least one file in the corresponding directory. Now we also create
packages if they are "mapped" to other JVM packages, i.e. if all package
parts in them have been annotated with JvmPackageName.
Most of the other changes are refactorings to allow internal names of
package parts/multifile classes where previously there were only short
names.
For sake of working without jsr305.jar in the classpath `resolveTypeQualifierAnnotation`
may return javax.annotation.CheckForNull (although the latter is nickname itself)
#KT-19985 Fixed
In the case the single parameter of override has `Integer` type instead
of `int` type (while in common case it would be just `int`)
See the comment inside forceSingleValueParameterBoxing for clarification
#KT-19892 Fixed
Rename addDeclarations -> registerModuleComponents
Use it to provide SamWithReceiverResolver extensions instead
Post construction on container composition can be achieved
but manually inserting injections where it seems appropriate
is bug prone
This fixes a bug where SamWithReceiverPlugin extension was not registered
for some containers in IDE which led to incorrect highlighting in IDE
Add IDE test for applying SamWithReceiver plugin
#KT-18062 Fixed
So (for the most often reproduction case) #KT-19433 Fixed
Before this commit, internal names for nested classes were written as test/Foo/Bar (comparing to test/Foo$Bar in the normal mode), as getting qualified names from such internal names was trivial. But, because of IC, we needed to write class files to the disk, so our decompiler could find such "broken" classes and read it in a wrong way.
Prior to the 1.1.4, nullability related annotations were stored
in types that became hard to maintain at some moment and
we got rid of it (see 57b7b91444)
But enhancement for properties overrides stopped working
because there were effectively no annotations in the
resulting descriptor
#KT-19409 Fixed
#KT-19409 Fixed
Except AnnotationDescriptorImpl, which is refactored in the subsequent
commit.
Note that we no longer check the presence of parameters with the
corresponding names in the annotation class in
LazyJavaAnnotationDescriptor, this is why test data changed
Turns out, only the parameter's name is needed at all usages of this
method. Such a map is both easier to use (no need to call
ValueParameterDescriptor.getName) and easier to construct (no need to
resolve annotation class, its constructor, its parameters). In this
commit, only usages have changed but the implementations are still using
the old logic, this is going to be refactored in subsequent commits
It's quite a hot point and using local functions with default
parameters seem to affect negatively to overall performance
(at least some of compiler benchmark have shown it)
The case is when we have a simple type constructor
(not array nor collection) and type is not raw:
A<T1, >..A<T1>?
Actually these types are almost equal besides of nullability,
but we create and resolve two different simple types,
they have different arguments' lists, etc.
The idea is to add NullableSimpleType subclass with delegate
to another simple type
It should help a lot both with common cases and with corner ones:
flexibles types in spark for Function22 having exponential size
because of its flexibility
#KT-14375 Fixed
#KT-14323 Fixed
No package annotations are going to be loaded, and
TypeQualifierDefault/TypeQualifierNickname are no longer recognized by
default. Use the CLI argument "-Xload-jsr305-annotations" to enable this
behavior back
#KT-10942
It was only used for type-related nullability/mutability
annotations and it was necessary to remove them
in the descriptor renderer (duplicating their fqnames there).
At the same time they're only needed for types enhancement
where they can be simply restored from type owners' descriptors
The testData changes are more or less correct: this kind of annotations
is bound both to types themselves and their use because of their targets
As the type is anyway replaced with not-nullable version
explicitly, the only thing that changes is what type is loaded
for String[][].class:
- before it would be Array<Array<String?>?>
- now it's Array<(out) Array<(out) String!>!>
It's both a minor change and new behaviour can be considered
as correct
Stanislav Erokhin