Commit Graph

8769 Commits

Author SHA1 Message Date
Mikhail Zarechenskiy 82d43c23b6 [NI] Perform full type checking in NI if there is no contradiction
In NI type checking actually performs in constraint system. To be conservative, now we still perform full type cheking but only if there is no contradiction in constraint system
2017-11-13 16:33:20 +03:00
Mikhail Zarechenskiy f51f7a0fe4 [NI] Report diagnostic about receiver as super expression 2017-11-13 16:33:08 +03:00
Mikhail Zarechenskiy 1d6e9c244b [NI] Report diagnostic about spread argument to non vararg parameter 2017-11-13 16:32:58 +03:00
Mikhail Zarechenskiy d45f4e71de [NI] Don't try to search methods on undefined integer type 2017-11-13 16:31:22 +03:00
Mikhail Zarechenskiy 4ad885afb0 [NI] Correctly propagate base constraint system of invoke call
Consider the following example:

class A {
  operator fun <T> invoke(): Foo<T> = throw Exception()
}

fun foo(f: Foo<Int>) {}

fun test(a: A) {
  foo(a()) // after resolve of `invoke`, it has non-fixed type variable
}
2017-11-13 16:31:04 +03:00
Mikhail Zarechenskiy fbbe396881 Add configuration key to enable new inference for compiler tests 2017-11-13 16:27:18 +03:00
Mikhail Zarechenskiy 6486c8dccc Add ability to reuse one diagnostic test file for old and new inference 2017-11-13 16:23:05 +03:00
Mikhail Zarechenskiy 73b4e24d54 Introduce language feature for NI, remove old setting 2017-11-13 16:23:04 +03:00
Alexander Udalov 6f276c623b Do not use IntelliJ extensions for uninitialized variable suppressor in JS 2017-11-13 13:04:54 +01:00
Alexander Udalov 6f90a0545c Do not use IntelliJ extensions for declaration without body suppressor in JS 2017-11-13 13:04:54 +01:00
Alexander Udalov 7f81601123 Do not use IntelliJ extensions for unused parameter suppressor in JS 2017-11-13 13:04:54 +01:00
Alexander Udalov d782c8a644 Use ServiceLoader instead of IntelliJ extensions for ScriptHelper 2017-11-13 13:04:54 +01:00
Alexander Udalov 02981038f3 Split core into descriptors, descriptors.jvm, deserialization, descriptors.runtime 2017-11-10 18:58:31 +01:00
Denis Zharkov b4499d52ee Introduce OpenAddressLinearProbingHashTable to be used in binding context
See comments in the implementation for clarification
2017-11-10 13:01:13 +03:00
Nikolay Krasko 329322e268 Increase stub version because of change in overload sorting
See commit f197f36e239e3500baf4af87d9b719a123fe2f26
2017-11-07 18:00:23 +03:00
Alexander Udalov 3049b5d92f Remove logging in KotlinSuppressCache.isSuppressedByAnnotated
It produces lots of useless output in stderr when tests are run with
Gradle in the terminal
2017-11-02 17:15:59 +01:00
Simon Ogorodnik 339d06d040 Improve diagnosticMissingPackageFragment reporting
Single execution path to report missing package fragment problems

Split failures on PluginDeclarationProviderFactory site by
known reasons to improve exception analysis
2017-11-02 16:31:43 +03:00
Ilmir Usmanov 817f79520a Implement coroutineContext intrinsic
This intrinsic allows coroutine to access its context without suspention
and, furthermore, disabling tail-call optimization.
KT-17609: Fixed
2017-11-02 12:45:24 +03:00
Dmitry Savvinov e17610f378 Don't throw on recursion when computing member scope
When recursion is detected while computing
`ClassResolutionScopesSupport.scopeForMemberDeclarationResolution`,
create 'ThrowingLexicalScope' (as with other scopes), instead of
throwing ISE immediately. That allows to report error properly in cases
like in KT-18514

#KT-18514 Fixed
2017-11-01 15:57:56 +03:00
Denis Zharkov 2df4d9496a Fix incorrect UNINITIALIZED_ENUM_COMPANION diagnostic
#KT-20959 Fixed
2017-10-27 09:36:00 +03:00
Simon Ogorodnik 98310d3bdb KT-20884: Support receiver reference in KDoc comments
#KT-20884 fixed
2017-10-25 17:54:47 +03:00
Denis Zharkov be630f93dd Postpone full resolution for a candidate known to be failed
It's necessary for performance, because there are some resolution
parts that actually can be omitted and at the same time they aren't
very cheap (like inference or value arguments types checking)
2017-10-25 16:33:04 +03:00
Dmitry Petrov 57602c420d Sanitize declaration return types
On JVM, strip @EnhancedNullability annotation from inferred types for
functions, properties, and local variables, so that these annotations
do not "escape" from Kotlin declarations.
2017-10-25 12:30:32 +03:00
Alexander Udalov 1a8be635b9 Use ServiceLoader instead of IntelliJ extensions for DefaultErrorMessages
#KT-10473 Fixed
2017-10-24 18:25:38 +02:00
Alexander Udalov 1c6dce3674 Do not use DefaultErrorMessages.Extension as IntelliJ extension in plugins 2017-10-24 18:22:04 +02:00
Alexander Udalov f2be34ca1c Support KClass<*> annotation arguments in serialization/deserialization
#KT-11586 Fixed
2017-10-23 19:16:44 +02:00
Mikhail Glukhikh 4b8c04a3d6 Minor: fix typo 2017-10-20 15:16:03 +03:00
Dmitry Savvinov 5b9b8778b5 Minor: rename 'mapToPlatformClasses' -> 'mapToPlatformIndependentClasses' 2017-10-19 18:21:56 +03:00
Dmitry Savvinov 58b815d842 Fix false USELESS_CAST in function with expression body
Note that current behaviour is made similar to the case with
properties initializers/accessors, which means that more complex
cases are not covered yet (see KT-20801) #KT-20802 fixed.
2017-10-19 18:21:56 +03:00
Dmitry Savvinov a1778a7da8 Fix CAST_NEVER_SUCCEEDS when casting Nothing to some type 2017-10-19 18:21:56 +03:00
Mikhail Glukhikh 3fb0354676 Minor: add / fix comments (related to KT-20752) 2017-10-16 20:00:47 +03:00
Stanislav Erokhin b9fa8d4d96 Support parenthesized left expression for ExpectedTypeFromCast 2017-10-16 18:28:57 +03:00
Mikhail Glukhikh a55c6f0c95 Don't register safe cast type info for unstable values
Important: to be removed in 1.3
So #KT-20752 Fixed
2017-10-16 16:21:45 +03:00
Alexander Udalov 0510c553c3 Mark class files as pre-release if language version > LATEST_STABLE
#KT-20547 Fixed
2017-10-16 15:09:36 +02:00
Alexander Udalov 3665255a2b Extract skipPreReleaseCheck out of DeserializationConfiguration.skipMetadataVersionCheck 2017-10-16 15:09:36 +02:00
Alexander Udalov 8a0dcca957 Do not consider 'expect' class type constructors final
Because even a final expected class can be actualized with an open
actual class and thus have subtypes in the platform code
2017-10-16 15:08:50 +02:00
Alexander Udalov 0a861fd4ed Minor, refine return type of AbstractClassTypeConstructor.getDeclarationDescriptor 2017-10-16 15:08:50 +02:00
Alexander Udalov 2682837fd7 Support 'expect final' -> 'actual open' for callables
#KT-17944
2017-10-16 15:08:50 +02:00
Mikhail Zarechenskiy 61d1963ddb Mark incompatibilities related to the declaration signature as STRONG 2017-10-13 16:14:43 +03:00
Mikhail Zarechenskiy b2dc82f17e Improve diagnostic for non-actual declaration with weak incompatibility 2017-10-13 16:14:17 +03:00
Mikhail Zarechenskiy d0cbd73307 Don't report incompatibility expect/actual errors on usual overloads
Divide incompatibility on two groups: strong and weak. Strong incompatibility means that if declaration with such incompatibility has no `actual` modifier then it's considered as usual overload and we'll not report any error on it.

 #KT-20540 Fixed
 #KT-20680 Fixed
2017-10-13 16:14:14 +03:00
Mikhail Zarechenskiy eb828e3497 Report errors about incompatible constructors of actual class
#KT-20540 In Progress
 #KT-20680 In Progress
2017-10-13 16:13:40 +03:00
Stanislav Erokhin 4932fa1ddd Support expected type from explicit cast
This commit support the following case.
Suppose we have such declaration:
  fun <T> foo(): T { ... }

Then in code we want to use it like this: `foo() as String`.
But in LV <= 1.1 we have type inference error: "Not enough
information for type parameter `T`". This error happened because we
do not use type from cast as expected type for call.

In this commit we fix this problem and use this type as expected type
in following cases:
 - our function has only one type parameter (this can be relaxed later)
 - function parameter types and extension receiver type not contains `T`

Also this fix problem with `findViewById`.
Already signature was: `fun findViewById(...): View`
and was used like: `findViewById() as MyView`.
New signature is `fun <T : View> findViewById(...): T`
and old usage was broken because of problem described above
2017-10-11 19:23:52 +03:00
Alexander Udalov a4f378d04d Decrease resolution priority of declarations from kotlin-stdlib-jre7/8
In order to allow "overload resolution ambiguity" when both
kotlin-stdlib-jre7/8 and kotlin-stdlib-jdk7/8 are in the dependencies
2017-10-11 19:23:38 +03:00
Alexander Udalov ecfea9e340 Forbid private 'expect' declarations
#KT-19170 Fixed
2017-10-11 17:21:55 +02:00
Mikhail Zarechenskiy 88595e1a58 Don't require to override abstract methods in expect class
#KT-16099 Fixed
2017-10-11 15:27:50 +03:00
Mikhail Zarechenskiy 5695e76a00 Prohibit inheritance by delegation in expect classes
#KT-20431 Fixed
2017-10-11 15:27:46 +03:00
Dmitry Savvinov eae9923dbe Add allowKotlinPackage AnalysisFlag
This flag is used internally by EffectSystem as a sign of compiling
stdlib. If this flag is present, then EffectSystem will read contracts
on functions and serialize them into metadata even if corresponding
LanguageFeatures are turned off. This is done solely for building
1.2-runtime with contracts in it without the need to turn on
LanguageFeatures manually.

======
This commit finishes a first series of commits related to effect
system. After it, compiler is ready to work with contracts, but it is
impossible to actually annotate anything, because there are no
contracts DSL in stdlib yet.
2017-10-11 13:40:16 +03:00
Dmitry Savvinov bdbb161cfa Effects: support called-in-place lambdas
Support initialization of local variables in capture, if that capture is
anonymous lambda that was passed to function which explicitly expressed
its intent to call closure "here and now"

Changes:
- LocalFunctionDeclarationInstruction made open

- Introduce InlinedLocalFunctionDeclarationInstruction, which is subtype
of LocalFunctionDeclarationInstruction with additional semantic: it
is statically known that this function will be called in-place and maybe
with some definite amount of invocations.

- Next-instruction of InlinedLocalFunctionDeclarationInstruction depends
on whether flow can exit its body normally, i.e. on wheter EXIT is
reachable. If flow can reach EXIT, then .next is just straight next
instruction, otherwise it's SINK.

- Take non-local instructions into consideration when analyzing
reachability. We didn't it before because no one ever needed proper
analysis. Now we have InlinedLocalFunctionDeclarationInstruction which
cares about proper reachability of EXIT.

- Pull control-flow information from
InlinedLocalFunctionDeclarationInstruction's EXIT into parent's
pseudocode, thus allowing it to participate in initialization/use
analysis.

- Change logic of 'isCapturedWrite' to not report
"CAPTURED_VAL_INITIALIZATION" if value is captured by the inline-lambda.

==========
Introduction of EffectSystem: 12/18
2017-10-11 13:34:54 +03:00
Dmitry Savvinov 0dd6d1f4c7 Effects: support assert-like contracts
- Support functions that explicitly express relation between successfull
return and passed arguments (e.g., 'check')
- Note that we have to correct resulting data flow for arguments *after*
resoling (because we get contracts only after resolving). To do so, we
have to extend MutableDataFlowInfoForArguments interface with
'updateResultInfo' method.

==========
Introduction of EffectSystem: 11/18
2017-10-11 13:34:54 +03:00