The main problem is that inside a state machine for a named suspend
function parameters of it's owner are available as a usual captured
closure parameters (i.e. through synthetic fields), while
TailRecursion codegen expects that parameters are straight local
variables.
So, the solution is just to define local var for each of real parameters
(all but the last continuation parameters) for tailrec functions.
#KT-15759 Fixed
Since now `suspend (Int) -> String` will be serialized as `(Int, Continuation<String>) -> Any?` + suspend flag.
Before this change such type serialized like this: Function2<Int, String> + suspend flag. And yes, type `Function2<Int, String>` isn't correct, because Function2 expect 3 type arguments.
We have special logic for this case and we deserialize such error-written types correctly.
(cherry picked from commit 3518cbe)
General effect will be the following:
- all member scopes for libraries and JDK will be constructed with -language-version/-api-version specified in project settings
- for modules with another (not like in project settings) -api-version or -language-version we will have not correct member scope -- for example we will see typealiases from such libraries.
#KT-15979 Fixed
(cherry picked from commit 0001865)
Array instatiation code should handle type alias constructors properly.
So far, we don't have constructors with type parameters
different from the type parameters of the resulting type,
so we can safely take type arguments of the underlying type.
This change fixes an SOE in isCastErased:
@JvmStatic
fun isCastErased(supertype: KotlinType, subtype: KotlinType, typeChecker: KotlinTypeChecker): Boolean {
if (supertype.isMarkedNullable || subtype.isMarkedNullable) {
return isCastErased(TypeUtils.makeNotNullable(supertype), TypeUtils.makeNotNullable(subtype), typeChecker)
}
TypeUtils.makeNotNullable(TypeTemplate) should not return the same object
if isMarkedNullable returned true on the instance
#KT-15516 Fixed
#KT-15915 fixed
Annotation processor can access a classes directory provided by AP environment.
Previously kapt 3 was using kotlin-classes directory as a a classes directory.
However compileKotlin task does not expect this.
Also having multiple tasks with the same output dir is a bad practise in Gradle.
This change introduces a separate directory for classes generated by kapt 3.
Its output is copied to a resulting classes dir (just as a kotlin-classes dir).
If @Input property is added in new plugin version
and this property has a new type (a class/enum which is not presented in previous plugin)
then downgrading leads to an exception,
because Gradle tries to deserialize the property value from its caches,
but the type of value does not exist.
Workaround: add new String property.
#KT-15500 fixed
GString is an object that represents string literal like `"${project.name}"`
in Groovy. It is not an instance of string.
Groovy automatically converts GString to String when it is passed where String is expected.
However freeCompilerArgs is a List<String>, so type parameter info is lost at runtime.
When iterating freeCompilerArgs in Kotlin as a list of string, an exception
is thrown because GString cannot be casted to String (toString should be called instead).
Denis Zharkov