Anonymous types are not approximated by frontend for private declarations.
Class IDs for such types are replaced by StringTable before being written
to metadata. JVM string table mangles such types and keeps them generic
for reflection. For other purposes the types are replaced with `Any`.
Type arguments of the replaced type should be ignored in the latter case.
Otherwise decompiled text builder crashes on an attempt to restore `Any`
type with non-zero number of type arguments.
The code pretending to replace a type with its first supertype was dropped
from ApproximatingStringTable for two reasons:
- the first type from `getAllSuperClassifiers` is the original type itself
which doens't provide a ClassId for anonymous type, so it was a noop
- tracking potential type arguments of the first anonymous type's supertype
would be a complication with almost no practical value (types in decompiled
text would be slightly closer to the real type of a private declaration).
^KT-46393 Fixed
This directive anyway does not make test run twice with OI, and with NI
It only once run the test with specific settings (// LANGUAGE)
and ignores irrelevant (OI or NI tags)
There is some behavior change regarding the new WrapWithSafeLetCall quickfix
1. it now works correctly on binary expressions by wrapping it with `()`
2. it now looks for a nullable position upward and do the modification there,
if possible. For example, consider the following code
```
fun bar(s: String): String = s
fun test(s: String?) {
bar(bar(bar(<caret>s)))
}
```
After applying this fix, FE1.0 yields
```
bar(bar(s?.let { bar(it) }))
```
while the new implementation yields
```
s?.let { bar(bar(bar(it))) }
```
This behavior aligns with FE1.0 if `bar` accepts nullable values.
(e.g., `nullable.a = b`), and use positioning strategies to locate the
dot in the LHS expression.
Without it, only the callee reference is reported on, which makes the
highlighting of the error and application of quickfixes incorrect in the
IDE.
Also fixed issue with annotated and/or labeled expressions on LHS of
assignment (e.g., `(@Ann label@ i) = 34`).
`iterator()` function that does NOT have `operator` modifier.
This is different from FE1.0. Adding `!!` will then surface the error
that `operator` modifier needs to be added (with corresponding fix).
UNSAFE_OPERATOR_CALL, UNSAFE_INFIX_CALL, ITERATOR_ON_NULLABLE,
ARGUMENT_TYPE_MISMATCH, RETURN_TYPE_MISMATCH.
TODO: Don't offer fix when target is known to be null (from data flow
analysis).
isAvailable and in invoke) by moving computation of element to modify
to before instantiation (i.e., to the factories or equivalent).
This lets us to move it to idea-frontend-independent and re-use it FIR.
Make it work just the same as the analogue from FE 1.0
This change is necessary since many tests start failing after
we began reporting diagnostics after call completion
Mark Punzalan