This proved to be a fragile technique, which probably doesn't even improve
performance in most cases but has lots of unexpected problems: unconditional
initialization of reflection classes, increasing the size of the bytecode, bugs
with <clinit> in annotations on JVM 6, inability to support conversion of a
class from Kotlin to Java without recompiling clients which use it
reflectively, etc.
0. Such properties are called special because their accessor JVM name differs from usual one
1. When making call to such property, always choose special name
2. When generating Kotlin class inheriting such property generate `final bridge int size() { return this.getSize(); }`
3. If there is no `size` declaration in current class generate `bridge int getSize() { // super-call }`
This format of the string table allows to reduce the size of the Kotlin
metadata in JVM class files by reusing constants already present in the
constant pool. Currently the string table produced by JvmStringTable is not
fully optimized in serialization (in particular, the 'substring' operation
which will be used to extract type names out of generic signature, is not used
at all), but the format and its complete support in the deserialization
(JvmNameResolver) allows future improvement without changing the binary version
Delete the old ones in package kotlin.reflect.jvm because otherwise the code
using those functions will become red in a lot less meaningful way (overload
resolution ambiguity) than if they're deleted (unresolved import)
Based on the work originally done by @dnpetrov
#KT-8380 Fixed
Annotation arguments with default values are rendered as ... if renderDefaultAnnotationArguments is true.
Tests: java retention does not taken into account by Descriptor comparator.
Java retentinon changed to kotlin retention in some tests + one new test with java retention added.
More accurate tests for intentions in byte code (visibility controlled).
Get rid of all classes except kotlin.reflect.KFunction, which will be used to
represent all kinds of simple functions.
Lots of changes to test data are related to the fact that KFunction is not an
extension function (as opposed to KMemberFunction and KExtensionFunction who
were) and so a member or an extension function reference now requires all
arguments be passed to it in the parentheses, including receivers. This is
probably temporary until we support calling any function both as a free
function and as an extension. In JS, functions and extension functions are not
interchangeable, so tests on this behavior are removed until this is supported
Parse TryCatchBlockNode's in generated bytecode, infer stack save/restore points.
Save stack to local variables before 'try'.
Restore stack after the beginning of try-block, catch-block, and default handler.
Integrate before/after inline markers rewriting (otherwise it'll break our stacks).
#KT-3309 Fixed
A set of tests provided. Some external tests fixed accordingly.
Companion object creation changed accordingly.
Derived classes now can use base class with the private constructor.
Refactoring of AccessorForFunctionDescriptor.
For each non-abstract non-declared (i.e. inherited from supertypes) method in
an interface we generate its static form to the TImpl, which calls the TImpl
method from the corresponding supertype.
The accidental override tests changed because we're now trying to generate the
delegate for the super method, not knowing that it will clash with the declared
method
#KT-2888 Fixed
#KT-5393 Fixed
