When resolving a class body for a class without a primary constructor
(e.g., an interface), no checks were performed for redeclarations
in the corresponding class header.
Creating & initializing a lexical scope of an appropriate kind will do it.
Note that since class has no primary constructor, only type parameters
could be redeclared (and that's KT-4960).
To get rid of pointless/confusing implementations in ModuleDescriptor,
PackageViewDescriptor, TypeParameterDescriptor and others.
Note that there are still implementations that do not make sense, for
example in those subclasses of VariableDescriptor which are not also
subclasses of CallableMemberDescriptor (e.g. ValueParameterDescriptor).
Those can be removed by making CallableMemberDescriptor (instead of
CallableDescriptor) inherit from Substitutable. However, that would
require more changes in the compiler because CallableDescriptor is used
rather often in places where in fact only CallableMemberDescriptor
instances can appear.
Explicit return types and casts are required in some places now because
there's no single non-trivial supertype for
ClassifierDescriptorWithTypeParameters and CallableDescriptor.
Previously it was DeclarationDescriptorWithVisibility, now it's both
that and Substitutable<...>
that was moved to the script-runtime, that it a row caused the incompatibility
with old GSK implementations. Now the legacy template resolving related code
is restored, marked obsolete and wrapped for compatibility with the new code.
There's no need to recalculate results if all of the
previous instructions' results remain unchanged.
It's a crucial optimizations because otherwise algorithm
can degrate to O(n * m) [n - instructions number, m - variables number]
even in a linear code like this:
var a_1 = 1
var a_2 = 1
...
var a_m = 1
We perform analysis iteration once, then we repeat it to be sure
that no changes happened and here for each instruction we're
starting to check if recomputed value is the same, that basically
compares maps of all the variables, so it works O(n * m)
After this change on the second iteration we won't recompute new values
as none of predecessor has been changed
Note that there are some other problems, for example:
`val a = if (true) 1 as Number else 2`, here we'll get useless cast
#KT-9551 Fixed
#KT-9645 Fixed
Forbid underscore-only (_, __, ___, ...) names as callees and as types.
If CHECK_TYPE directive is on, filter out UNDERSCORE_USAGE_WITHOUT_BACKTICKS messages.
Types of the corresponding parameters (or type parameter bounds, types
in supertypes, etc) are now compatible not only if they're equal, but
also if values of those types are mutually assignable (if "a" is subtype
of "b" and "b" is subtype of "a")
#KT-17005 Fixed
Dmitry Petrov