cebe67d90c
Fix default value generation after suspension point
...
Replace coercion from VOID with call 'pushDefaultValueOnStack'
It's necessary because coercion of VOID to java/lang/Object ends with Unit instance on stack
that makes variables spilling algorithm thinking that variable is Unit
#KT-13409 Fixed
2016-09-01 15:03:31 +03:00
1226d8fc2c
Refine 'handleResult' calls generation within coroutines
...
Before this change everything works just fine for 'handleResult' methods
accepting non-Unit parameters
For other cases the same coercion-to-unit strategy is in plain lambdas:
- if last statement is not Unit type, execute it, pop from the stack, then put Unit instance
- for 'return@label' (no expression) just put Unit on the stack
#KT-13531 Fixed
2016-09-01 15:03:31 +03:00
cd717467f0
Nested classes are no more allowed inside any anonymous objects or local classes #KT-13510 Fixed
...
(cherry picked from commit 5c3ec6c)
2016-08-31 19:28:13 +03:00
f8039249c6
CFA: additional jumps to catch / finally generated in the end of try / before exits from try #KT-5469 Fixed
...
Also #KT-13612 Fixed
(cherry picked from commit 7c188b3)
2016-08-31 19:28:09 +03:00
6a9d058db4
UNUSED_VALUE is now reported on expression only if this expression is unused in all (e.g. finally) branches #KT-9825 Fixed
...
(cherry picked from commit aac8e94)
2016-08-31 19:28:00 +03:00
3298649bd7
Support "-no-reflect" in "kotlin" command
...
#KT-13491 Fixed
2016-08-26 15:17:00 +03:00
25eefdb6ec
Pseudocode label now cannot be bound to a non-owning pseudocode #KT-13555 Fixed
...
Also EA-77641 Fixed
(cherry picked from commit 07643ce)
2016-08-25 10:32:04 +03:00
d08d8af407
USELESS_ELVIS_RIGHT_IS_NULL : build fix (add to different files in DiagnosticsTestGenerated), do not report if USELESS_ELVIS already exists
...
(cherry picked from commit c71b656)
2016-08-23 16:26:54 +03:00
a3f131d6c6
KT-13521 Warning right part of "expression ?: null" is useless
...
#KT-13521 Fixed
2016-08-23 01:10:10 +09:00
ac3dfd93bd
More correct handling of compile time constant types #KT-13421 Fixed
2016-08-18 14:42:16 +03:00
ae856e834a
Store map (Call to KotlinType) for SMARTCAST slice #KT-13426 Fixed
...
Also EA-86848 Fixed
2016-08-17 16:09:45 +03:00
cbcef67d82
KT-13426: store map from ImplicitReceiver into KotlinType for implicit receiver smart casts
2016-08-17 16:09:34 +03:00
b69661df8f
Pseudocode and CFG changed after 'let' is resolved correctly.
2016-08-17 12:40:50 +03:00
58a010464a
Exception typealiases rendering in test output (CFG)
2016-08-17 12:40:50 +03:00
0ebfba7a1a
Test data: do not use exceptions from java.lang, as they are not available in JS
2016-08-17 12:40:50 +03:00
d943079557
Exception typealiases rendering in test output
2016-08-17 12:40:50 +03:00
e0d525b72a
Support bound function reference inlining
2016-08-17 11:26:45 +03:00
7325baa06a
Fix for KT-13374: CompilationException: Inline function call with anonymous object implementing an interface by delegation
...
#KT-13374 Fixed
2016-08-16 10:47:01 +03:00
7b8ee5fba1
Add test case for obsolete issue
...
#KT-13172 Fixed
2016-08-15 19:41:47 +03:00
bd88b0941d
Fix incorrect unsupported error on synthetic extension call on LHS of ::
...
#KT-13271 Fixed
2016-08-15 19:41:47 +03:00
5b1ee13db8
Make KClasses for primitives equal to KClasses for wrapper types
...
Both primitive int and wrapper type java.lang.Integer are represented by the
single type kotlin.Int in Kotlin, so inequality between the corresponding
KClasses was confusing here. To keep the old behavior, one may call 'k1.java ==
k2.java' instead of `k1 == k2`
#KT-13462 Fixed
2016-08-15 19:41:47 +03:00
3efa738bc0
Fix NPE on .companionObjectInstance for built-in classes
...
#KT-13370 Fixed
2016-08-15 19:41:47 +03:00
85a951db52
Fixed exception AssertionError: Illegal resolved call to variable with invoke
...
#KT-13330 Fixed
#KT-13349 Fixed
2016-08-15 19:24:17 +03:00
1c5c6506ce
Report of UNINITIALIZED_ENUM_COMPANION also for implicit receivers #KT-11769 Fixed
2016-08-12 15:06:43 +03:00
04f71bccf2
KT-11769 case with companion function call: error is now detected if companion receiver is stated explicitly
2016-08-12 15:06:38 +03:00
6cf90cfc4e
Fix for SOE in VarianceChecker #KT-13401 Fixed
2016-08-11 13:04:58 +03:00
46a7c706e1
Fixes and refactorings after review
2016-08-11 09:11:11 +02:00
e99697e3a2
Fix testdata for cli help tests
2016-08-11 09:11:11 +02:00
d21061ff60
CAST_NEVER_SUCCEEDS: some type alias tests (1.1 only!)
2016-08-10 12:23:09 +03:00
687698da9a
Do not report USELESS_CAST when casting null to nullable (special case)
2016-08-10 12:23:02 +03:00
098520de64
CAST_NEVER_SUCCEEDS: do not report when casting nullable to nullable #KT-260 Fixed
2016-08-10 12:22:53 +03:00
1df9724c0c
Refine stack frames markup calculation
...
It fixes VerifyError with coroutines on Dalvik happening because of
variables spilling before/after suspension point
BasicInterpreter from ASM does not distinct 'int' types from other
int-like types like 'byte' or 'boolean', neither do HotSpot and JVM spec.
But it seems like Dalvik does not follow it, and spilling
boolean value into an 'int' field fails with VerifyError on Android 4,
so it's necessary to distinct int types for variables spilling
#KT-13289 Fixed
2016-08-09 16:36:52 +03:00
0d26087040
CLI, Ant: add kotlin-reflect.jar to classpath by default, support "-no-reflect"
...
Note that now "-no-stdlib" implies "-no-reflect".
#KT-13237 Fixed
2016-08-09 11:24:24 +03:00
30fd22499b
Test data fix (after cherry-pick from 1.0.4)
2016-08-09 10:32:15 +03:00
b668206aa7
Minor: test comment, relevant to KT-13322
...
(cherry picked from commit 7606b1d)
2016-08-09 10:28:49 +03:00
729de536de
Rewrite slice error removed for LEAKING_THIS #KT-13371 Fixed
...
Also EA-86478 fixed
(cherry picked from commit 5ac3126)
2016-08-09 10:28:40 +03:00
13e64c18d9
Correct CFA order for enums: first own members, then entries, and at last companion object members #KT-6054 Fixed
...
(cherry picked from commit 94d3b4c)
2016-08-09 10:28:27 +03:00
1bc7b4e363
Fix test data
2016-08-08 21:48:10 +03:00
7f142253bf
Reflection: simplify KTypeProjection
2016-08-08 17:54:11 +03:00
a7f4037206
Reflection: change order of arguments of inner generic type
...
As in KClassifier.createType and everywhere in the compiler, specify arguments
for the innermost type first. This is more convenient to use because generally
the construction/introspection of such type starts from the innermost class
anyway (i.e. something like generateSequence can be used, without the need to
call .reverse() in the end)
2016-08-08 17:54:11 +03:00
89d69bc7eb
Reflection: add KClass.createInstance
2016-08-08 17:54:11 +03:00
0525b1e6c1
Reflection: add KClass.declaredMembers
2016-08-08 17:54:11 +03:00
30b7334e48
Reflection: add utilities to filter parameters by kind and name
2016-08-08 17:54:11 +03:00
d78988a12a
Reflection: add KVisibility, KClass.visibility, KCallable.visibility
2016-08-08 17:54:11 +03:00
7e317f7a7c
Reflection: add API for declaration modifiers
...
#KT-10447 Fixed
2016-08-08 17:54:11 +03:00
ada81923dc
Reflection: add KClass.isInstance, KClass.cast, KClass.safeCast
...
#KT-11284 Fixed
2016-08-08 17:54:11 +03:00
3702f4a7a7
Reflection: add KClass.{isSubclassOf,isSuperclassOf}, KType.{isSubtypeOf,isSupertypeOf}
...
The behavior in primitives.kt is likely to be reconsidered
#KT-8998 Fixed
2016-08-08 17:54:11 +03:00
4cd252d9d5
Reflection: add KClass.{supertypes,superclasses,allSupertypes,allSuperclasses}
...
Some things are not implemented yet for allSupertypes, such as KType->Type
mapping of types and generic substitution of built-in types
2016-08-08 17:54:11 +03:00
486ea62c72
Reflection: add KType.withNullability
2016-08-08 17:54:11 +03:00
f958483e56
Reflection: deprecate KClass.defaultType
...
Its semantics were unclear and its javaType was implemented incorrectly (it
should have returned ParameterizedType with star projections for generic type)
2016-08-08 17:54:11 +03:00
Denis Zharkov