azalea/kotlin-fork
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
96,470 Commits 1 Branch 0 Tags
62b4c7e52aa0f7517d6f9611bf089030022a5010
Commit Graph

41554 Commits

Author SHA1 Message Date
Dmitrii Gridin a539203e73 [SLC] drop KtAnalysisSession from SymbolLightClassBase 
^KT-54051
 2022-11-29 13:33:33 +00:00
Dmitriy Novozhilov 46928eaa0b [FIR] Unconditionally enable BooleanElvisBoundSmartCasts feature in K2 
^KT-54694 Fixed
^KT-26357
 2022-11-29 12:25:45 +00:00
Dmitriy Novozhilov ac7fddaad5 [FE] Always infer upper types to intersection types in K2... 
...and always use old algorithm for K1

^KT-51221
 2022-11-29 12:25:44 +00:00
Sergej Jaskiewicz d50c072af0 [klib] Validate paths when unpacking zip archives 2022-11-29 12:21:04 +00:00
Mads Ager 30002e29d1 [KAPT+JVM_IR] Do not generate initializers for delegated properties. 
^KT-54870 Fixed
 2022-11-28 22:27:09 +01:00
Alexander Korepanov 9dab8637a8 [JS Gradle] Pass incremental cache root dir to K2JS compiler 2022-11-28 20:46:40 +00:00
Alexander Korepanov 693258ae91 [JS IR] Invalidate all klib dependencies after removing it 
Without the invalidation, broken JS code
 (with broken cross-module references) may appear.

 ^KT-54911 Fixed
 2022-11-28 20:46:40 +00:00
Alexander Korepanov ca19d71a00 [JS IR] Commit cache header after lowering 
Commiting cache header before lowering may
 break caches in case of lowering errors.

^KT-54912 Fixed
 2022-11-28 20:46:40 +00:00
Alexander Korepanov 8cde0a81bc [JS IR] Use topological order for loading klibs in JS IR IC 
Generally, the library order doesn't matter.

 However, after lowering we need to post process
 lowered klibs in topological order:
  - commit incremental cache artifacts,
  - produce JS AST.

 The patch uses an ability of the Kotlin library resolver to load klibs
 in topological order and drops its reimplementation from IC infrastructure.

 There is one side effect:
   klibs which are not reachable from the main module, will produce their JS.
   This should be more correct than just ignoring them.
 2022-11-28 20:46:39 +00:00
Alexander Korepanov ec6a7094e6 [JS IR] Manage cache root dir from IC cache updater 
- pass cache root directory instead of dir list
 - manage klib cache dirs from cache updater
 2022-11-28 20:46:39 +00:00
Alexander Korepanov c7589245c4 [JS IR] Minor IC refactoring 
Move metadata classes from cache updater to separate file.
 2022-11-28 20:46:38 +00:00
Mikhail Glukhikh 8ffabddf75 FIR2IR: cleanup field static fake override caching 2022-11-28 17:27:51 +00:00
Mikhail Glukhikh 91a3c78844 Fir2IrDeclarationStorage: drop redundant 'with' + fix formatting 2022-11-28 17:27:51 +00:00
Mikhail Glukhikh 237f357016 FIR: don't copy initializer for field override, unwrap it instead 2022-11-28 17:27:50 +00:00
Mikhail Glukhikh c698d060c6 FIR2IR: support static fake overrides for fields (related to KT-53441) 
#KT-54921 Fixed
 2022-11-28 17:27:50 +00:00
Sergej Jaskiewicz aa1b18b0c8 [IR] Prevent infinite recursion when rendering bound symbol references 
Refactor the renderer, make BoundSymbolReferenceRenderer a static class
to prevent calling RenderIrElementVisitor's methods from it to avoid
infinite recursion in the future.

^KT-52677 Fixed
 2022-11-28 16:43:53 +00:00
Vladimir Dolzhenko af08e4121d Fix fragment element types class hierarchy. 
They must not extend `IStubFileElementType`.

#KT-55160 Fixed
 2022-11-28 15:44:37 +00:00
Vladimir Dolzhenko e751b11b5e Add extra checks 
#EA-721785
 2022-11-28 15:20:29 +00:00
Ilya Chernikov e4a361a4b4 K2 Scripting: disable scripts in source roots by default 
fixes ultimate FP test
 2022-11-28 15:10:55 +00:00
Pavel Kunyavskiy 09518a457a Better unbound symbols support for dumpKotlinLike 
If there is an unbound symbols in IrTree near backend exception,
compiler failed again during processing an error.

It led to hard-understandable error. Now at least it would show
problematic part of code, which is better for making workarounds.

^KT-55088
 2022-11-28 14:22:46 +00:00
Ilya Chernikov 4af32b8d36 K2 Scripting, IR: non-class based scripts support in IR 2022-11-26 18:01:49 +00:00
Ilya Chernikov cc9a7480a2 K2 Scripting: add FIR2IR conversion for FirScript 2022-11-26 18:01:49 +00:00
Ilya Chernikov 47448d779c K2 Scripting: enable script diagnostic tests for FIR 
also add script scopes test
 2022-11-26 18:01:49 +00:00
Ilya Chernikov a3a1550933 K2 scripting: add initial scripting support to K2 frontend 2022-11-26 18:01:48 +00:00
Evgeniy.Zhelenskiy 002f6bd34a [FIR] Fix JvmName annotation diagnostics 
Signed-off-by: Evgeniy.Zhelenskiy <Evgeniy.Zhelenskiy@jetbrains.com>

#KT-55143
 2022-11-25 16:30:22 +01:00
Zalim Bashorov 7f8f7aa050 [Wasm] Revert ref cast and test instructions to be compatible with Binaryen 2022-11-25 15:31:48 +01:00
Zalim Bashorov 5f140dac21 [Wasm] Remove wasm_ref_cast and use wasm_ref_cast_null inside the compiler 2022-11-25 15:31:47 +01:00
Zalim Bashorov e7d0e451e7 [Wasm] Fix immediate type for recently (in M6) introduced instructions 
STRUCT_TYPE_IDX -> HEAP_TYPE
WasmImmediate.TypeIdx -> WasmImmediate.HeapType
 2022-11-25 15:31:47 +01:00
Anna Kozlova abb45a0728 [AA] don't treat incomplete function as anonymous 
^KTIJ-23672
sync anonymous function predicate with RawFirBuilder/DeclarationsConverter
 2022-11-25 14:27:03 +00:00
Egor Kulikov d585f068dd [FIR] Make anonymous initializer lazy in RawFirBuilder 
Fourth step for KT-52615

Merge-request: KT-MR-7798
Merged-by: Egor Kulikov <Egor.Kulikov@jetbrains.com>
 2022-11-25 11:41:44 +00:00
Nikolay Lunyak a454d42e65 [FIR] KT-54587: Report the missing diagnostics 
^KT-54587 Fixed

Merge-request: KT-MR-7791
Merged-by: Nikolay Lunyak <Nikolay.Lunyak@jetbrains.com>
 2022-11-25 10:19:58 +00:00
Nikolay Lunyak 7147b7d17b [FIR] KT-55033: Split runTransaction into smaller functions 
^KT-55033 Fixed

Merge-request: KT-MR-7779
Merged-by: Nikolay Lunyak <Nikolay.Lunyak@jetbrains.com>
 2022-11-25 09:12:15 +00:00
vladislav.grechko 50e92d2238 Add null checks in constructors taking value class types 
^KT-53492: Fixed
 2022-11-24 20:12:05 +00:00
Vladislav Grechko cd6e865fb3 Improve support of 'lateinit' modifier 
- Allow 'lateinit' for inline classes which underlying type
is suitable for 'lateinit'

- K2: report all problems related to 'lateinit' modifier

^KT-55052: Fixed
 2022-11-24 19:47:21 +00:00
vladislav.grechko e0c13e5276 Fix addAll & putAll invocations on inline mutable collections 
^KT-54950: Fixed
 2022-11-24 19:27:42 +00:00
Dmitrii Gridin 3bc9299b3b [FIR] SignatureEnhancement: drop redundant 'resolvePhase' for type parameter 
^KT-54826
 2022-11-24 18:32:45 +00:00
Dmitrii Gridin ec613e57ef [FIR] add Enhancement wrapping for type parameters 
^KT-55095 Fixed
 2022-11-24 18:32:43 +00:00
Dmitrii Gridin 2155a23e4e [AA FIR] implement FirCallableSignature to simplify search by signature 
^KT-54826 Fixed
 2022-11-24 18:32:43 +00:00
Denis.Zharkov ca12cfb90d K2: Do not fix variables that has yet unprocessed constraints in forks 
Otherwise, exception from org.jetbrains.kotlin.resolve.calls.inference.components.ConstraintInjector.TypeCheckerStateForConstraintInjector.fixedTypeVariable
might happen during forks resolution

The test data is extracted from intelliJ FP test

^KT-43296 Fixed
 2022-11-24 17:29:30 +00:00
Denis.Zharkov b73acd7a3a K2: Apply constraints from forks at UNTIL_FIRST_LAMBDA completion phase 
Otherwise, OVERLOAD_RESOLUTION_AMBIGUITY is reported for the calls to
the functions annotated as @OverloadResolutionByLambdaReturnType

^KT-43296 In Progress
 2022-11-24 17:29:29 +00:00
Denis.Zharkov 10d63cc52a K2: Postpone inference forks resolution until FULL completion 
It helps to use expected type information when choosing the branch
even in the nested calls

^KT-43296 In Progress
 2022-11-24 17:29:29 +00:00
Denis.Zharkov c958c79362 Minor cleanup at ConstraintInjector.kt 2022-11-24 17:29:29 +00:00
Denis.Zharkov 02b2927921 Clarify naming around inference fork points 
^KT-43296 In Progress
 2022-11-24 17:29:28 +00:00
Denis.Zharkov 715a73c8fb K2: Support inference case with a mix of smart-cast and expected type 
The idea is that we should not fix (i.e. choose any of the fork branches)
on the stage of candidate processing before completion, but it's enough
just to check that current state can be converged to success.

And when completion starts, and we add expected type to the system,
we've got more information to choose the correct fork branch.

NB: The old `processForkConstraints` is being called just
at the beginning of the completion phase.

^KT-43296 In Progress
 2022-11-24 17:29:28 +00:00
Sergej Jaskiewicz 77e197d46a [IR] Add documentation comments for IdSignatures 2022-11-24 17:13:31 +00:00
Sebastian Sellmair 7870c82221 [Gradle][Native] K2NativeCompilerArguments.refinesPaths: Use Array<String> instead of String 
^KT-55071 Verification Pending
 2022-11-24 14:42:11 +00:00
Sebastian Sellmair 752ff04245 [Gradle][Native] Setup refinesPaths as expectedBy module dependencies 
Previously, dependsOn dependencies were just passed as friend modules
to shared native compilations.

^KT-55071 Verification Pending
 2022-11-24 14:42:10 +00:00
Ivan Kylchik a1477983ba Properly handle vararg interpretation for classic frontend 
#KT-55108 Fixed
 2022-11-24 11:07:07 +00:00
Denis.Zharkov f7d8fd54ca K2: Fix deserialization of flexible type based on type parameter 
It's anyway safe to use avoidComprehensiveCheck = true because
during deserialization we're sure that we need DNN type because
it's been serialized as such.
 2022-11-24 09:39:58 +00:00
Artem Kobzar ad78f27cce [K/JS] Fix problem with dynamic return type inside coroutine 2022-11-23 21:33:56 +00:00