If we do, the local variable table will not make sense. As as
example:
```
inline fun foo(getString: () -> String = { "OK" }) {
println(getString())
}
inline fun bar() {
}
fun main() {
bar()
foo()
}
```
leads to the following bytecode:
```
public static final void main();
descriptor: ()V
flags: ACC_PUBLIC, ACC_STATIC, ACC_FINAL
Code:
stack=2, locals=4, args_size=0
0: iconst_0
1: istore_0
2: nop
3: nop
4: iconst_0
5: istore_1
6: nop
7: ldc #53 // String OK
9: astore_2
10: iconst_0
11: istore_3
12: getstatic #30 // Field java/lang/System.out:Ljava/io/PrintStream;
15: aload_2
16: invokevirtual #36 // Method java/io/PrintStream.println:(Ljava/lang/Object;)V
19: nop
20: return
LineNumberTable:
line 9: 0
line 13: 2
line 10: 3
line 14: 4
line 15: 6
line 16: 7
line 17: 19
line 11: 20
LocalVariableTable:
Start Length Slot Name Signature
2 1 0 $i$f$bar I
6 14 1 $i$f$foo I
4 16 0 getString$iv Lkotlin/jvm/functions/Function0;
```
The `getString$iv` local should not be there. It has been inlined away.
Leaving it in the local variable table leads to inconsistent locals
info. Local 0 contains an int but we declare a local of type
Function0.
[JS IR] Move tests into compiler/testData
[JS IR] Add cyclic dependencies with lazy property initialization
[JS IR] Add test on not initialization in case of call non properties (classed, objects, enum classes, const vals)
[JS IR] Add initialization through top level
[JS IR] Ignore enum getInstance function in property lazy initialization
[JS IR] Use let function with useful result instead of pure apply and also
[JS IR] Remove duplicated tests in js.translator
... in case `-Xno-optimized-callable-references` is enabled. Before this
change, the generated abstract equals/hashCode methods were considered
as accidental overrides because they did not have equals/hashCode from
the supertype in the overriddenSymbols list.
#KT-43666 Fixed
[JS IR] Add tests with empty vararg and default arg before it
[JS IR] Ignore backend in tests instead of target backend
[JS IR] Add function with named spread operator
[JS IR] Remove ignoring of current js backend in jsExternalVarargFun
[JS IR] Add with arguments after vararg
^KT-42357 fixed
if it is called using parens and not by calling 'invoke' method.
Use underlying type when calling continuation constructor if suspend
function is method inside inline class.
#KT-43505 Fixed
#KT-39437 Fixed
inside 'create' if 'create' overrides 'create' from
BaseContinuationImpl. In other words, unbox the parameter if 'create'
accepts only one parameter.
#KT-43249 Fixed
#KT-43533 Fixed
inside 'invoke' if 'create' does not override 'create' from
BaseContinuationImpl. In other words, when suspend lambda accepts more
than one parameter (including receiver).
Do that only if we do not generate bridge 'invoke' method, since
inline classes are unboxed in the bridge.
Use mangled name for 'create' function in this case inside 'invoke'.
#KT-43249 In progress
#KT-39847 Fixed
#KT-38937 Fixed
Like function arguments, they are context-dependent, but unlike function
arguments, callable references should be resolved eagerly as if they are
explicit receivers.
- rename fileToPurenessInitializers onto fileToInitializerPureness
- remove redundant check on top-level property
[JS IR] Rename initialis* to initializ* for consistency
[JS IR] Move propertyLazyInitialization property to context from configuration
[JS IR] Add test on lazy initialization properties order
[JS IR] Add multi module for lazy initialization of properties
[JS IR] Move tests onto js.translator
[JS IR] Rename fileToInitializerPureness according to context name
^KT-43222 fixed
1. Use 'x' for each parameter, which is not an inline class, every
possible clash is handled by signature rather than name. This change
makes more API changes binary-compatible. So, the changes are in line
with the vision of inline classes are value classes, like primitives.
2. Take return type into account when mangling a function if the return
type is inline class. Otherwise, boxing bridge will not be generated,
which leads to CCE at runtime.
Dmitry Petrov