Scope for incremental compilation refers to binaries from previous step
of IC. It is used not only in IC context itself, but also it is
subtracted from original libraries scope. Before previous commit there
was such scheme:
1. create incremental compilation context for files of specific session
2. subtract IC scope from main libraries scope
3. use updated libraries scope to create library session
4. create all needed source session(s)
So here was a side effect of creating new IC context, which
1. is smelling code, because it increases mind complexity
2. hard to implement with new session utilities
So to fix this problem this commit changes the scheme above:
1. create IC scope and modify libraries scope
2. create libraries session
3. create source session(s) and IC context for them
These declarations should not be visible to users (and therefore are not added to FIR),
but plugin itself can reference them in already compiled serializable classes,
and therefore they should be available in metadata:
- synthetic deserialization constructor
- static write$Self function
See also:
^KT-55885
- KT-56505 occurred because `source.getChild(KtNodeTypes.MODIFIER_LIST)`
returns any modifier list in the subtree of the source element, not
necessarily the modifier list belonging to the checked element.
`depth = 1` restricts the search to the modifier list belonging to the
checked element itself.
- For example, given `f1` from KT-56505, `getChild` would return the
modifier list of `public var foo = 0`. Because it contains a
visibility modifier, `f1` wasn't marked with
`NO_EXPLICIT_VISIBILITY_IN_API_MODE`.
^KT-56505 fixed
- In explicit API mode, the `public` visibility modifier is not
redundant unless a declaration is hidden by a container. The
`REDUNDANT_VISIBILITY_MODIFIER` diagnostic is now not reported in such
cases.
^KTIJ-24485 fixed
- `REDUNDANT_MODIFIER_FOR_TARGET` is already reported for `open`
interfaces, but for consistency and IDE support, the compiler now
reports `REDUNDANT_MODALITY_MODIFIER` as well via the extended
checker.
- `REDUNDANT_MODIFIER_FOR_TARGET` cannot be disabled for this case
because it's reported via a basic checker while
`REDUNDANT_MODALITY_MODIFIER` is reported via an extended checker.
- Rename `implicitModality` to `redundantModalities` and return a set of
modalities. The idea of a *single* implicit/redundant modality doesn't
stand up to scrutiny. For example, for interfaces with `ABSTRACT`
implicit modality, `OPEN` is also a redundant modifier. But this is
not necessarily the case for all kinds of declarations.
Hypothetically, if it was possible to declare a class that is abstract
by default, adding an `open` modifier to that class would not be
redundant, as it would make the class instantiable.
If you need to mute a test for Gradle IC, the only way is to add exclude pattern and regenerate tests aka remove unnecessary tests. But the filter is absolute, so if you have the same test name in different subfolders (e.g. `pureKotlin/classRemoved/` and `classHierarchyAffected/classRemoved`) you cannot disable only one of them. This commit adds logic to specify which name pattern in which subfolder should be excluded
Construct compiler plugins classpath based on original
compiler arguments
Support of plugins in full pipeline test allows us to test projects
where some common compiler plugins are used without excluding modules
^KT-56075
Now testData contains info about original compiler arguments, which can
be reused to support plugins and various other arguments
Support original arguments in modularized tests
- Properly support JDK9 modules
- Opt-ins from original arguments
- jdkHome from module data or arguments, as well as noJdk option
- Fix friendPaths not configured
- jvmDefault now handled properly
^KT-56075
^KT-55879
In fact, test-data classpath contains references to stdlib and reflect
already
In order to avoid duplicating libraries on classpath set flags
accordingly
^KT-56075
JS IR BE incremental compilation infrastructure uses
LanguageVersionSettings::toString method to detect if any
compiler features or flags were enabled or disabled.
It is important that the features and flags order are stable
in the result string.
^KT-56580 Fixed
If a synthetic prop clashes with a real property (e.g. @JvmField
property from parent Kotlin class), don't generate fake override for the
synthetic property. This fixes a CONFLICTING_INHERITED_JVM_DECLARATIONS
error in a mixed hierarchy.
^KT-56538 Fixed
It's been introduced in the previous commit
("K2: Simplify handling mixed smartcast vs. original candidates")
Because previously, it was assumed wrongly that each next level of
ConeCallConflictResolver filter out the candidates that are 100% less
applicable/specific, but the main one (ConeOverloadConflictResolver)
either leaves the single candidate or the whole same set, thus at
FilteringOutOriginalInPresenceOfSmartCastConeCallConflictResolver
we've got 4 candidates only two of which we might filter out.
There's no real need in binding overrides for smartcast/original members
after FilteringOutOriginalInPresenceOfSmartCastConeCallConflictResolver
Instead, we just create candidates for all of them and filter out after
the ones that came from original type (if needed)
Dmitriy Novozhilov