d37bdbc07d
Saving and loading projection kind of array element type in bytecode.
2012-11-20 17:20:00 +04:00
21e97b408c
Loading java arrays as Array<out T> at method return type position, as well.
2012-11-20 16:16:47 +04:00
d2a308964d
Skipping java.util.Collection and similar classes in JdkAnnotationsSanityTest.
2012-11-20 16:16:47 +04:00
d1f217f2df
Minor. LinkedHashMap for easier output parsing.
2012-11-20 16:16:46 +04:00
62b89655b9
Simplified calculateTypeArgumentsFromSuper and added samples in comments.
2012-11-20 16:16:46 +04:00
f20c4dac15
Cleanup in tests.
2012-11-20 16:16:46 +04:00
5ff718b94c
Minor. Renamed variables where 'projection kind' term should be used instead of 'variance'.
2012-11-20 16:16:46 +04:00
4a37c130df
Minor. Better name for CollectionClassMapping class.
2012-11-20 16:16:46 +04:00
f287e2b55d
Minor. Various fixups after code review.
2012-11-20 16:16:46 +04:00
94cad16d1d
Moved all code relevant to signatures propagation to separate class.
...
#KT-2776 in progress
2012-11-20 16:16:46 +04:00
00bed4a2a4
Added tests with two superclasses.
...
#KT-2776 in progress
2012-11-20 16:16:46 +04:00
3b63e6e061
Supported variance propagation.
2012-11-20 16:16:45 +04:00
9177080d8b
Fixed loading variance from alternative signature.
2012-11-20 16:16:45 +04:00
c5e9deb906
Reverted modification in TypeTransformingVisitor in "Renamed & moved AllInjectorsGenerator".
...
It was not relevant to that commit: it was squashed accidentally during interactive rebasing.
2012-11-20 16:16:45 +04:00
044487cf37
Supported propagating read-onliness.
...
#KT-2776 in progress
2012-11-20 16:16:45 +04:00
89b413b7ad
Supported generic subclasses in submethods.
...
#KT-2776 in progress
2012-11-20 16:16:45 +04:00
e2d213f1fd
Loading functions from superclasses before current to avoid heisenbug.
...
Added test with two supertypes.
#KT-2776 in progress
2012-11-20 16:16:45 +04:00
494345ce10
Improved algorithm, added simple tests with generic types.
...
#KT-2776 in progress
2012-11-20 16:16:44 +04:00
1f4d994480
Introduced simple propagation of nullability on loading Java.
...
#KT-2776 in progress
2012-11-20 16:16:44 +04:00
961fde3c8a
Remove redundant abstract tests. Use AbstractCodegenTest instead
2012-11-20 14:35:58 +04:00
3fc2e5d20c
Attempt to split long or double on the stack exception
...
#KT-3042 Fixed
2012-11-20 13:35:30 +04:00
900788a8e0
Fix javaClass property for long and double
...
#KT-3046 Fixed
2012-11-19 21:19:48 +04:00
7bb7445c57
Fix for invoking abstract method in enum
2012-11-19 20:18:45 +04:00
6384b7340c
IDEA 122.813 is required now (API of the new make depends on the new build)
2012-11-19 18:15:20 +04:00
47e057d529
Migrate to IDEA 122.813
2012-11-19 17:04:18 +04:00
636deb6a09
KT-3016 Exception during analyze when some function has no name
...
#KT-3016 Fixed
2012-11-19 13:44:55 +04:00
be7989af36
Moved JVM-specific String.repeat method to StringsJVM.kt
2012-11-19 13:29:17 +04:00
7ccded5f6d
KT-2503 Added String.repeat method
...
#KT-2503 fixed
2012-11-17 14:43:06 +04:00
712fafddb6
DataFlowInfo.getPossibleTypes() now returns Set
...
This solves some performance problems related to DataFlowInfo (lots of nested
conditions etc.)
2012-11-16 17:56:31 +04:00
6c7b32dc03
Fix data flow info after inner loops with continue-outer
...
ControlStructureTypingVisitor.containsBreak() is refined and now also takes
into account continue to outer loops
2012-11-16 17:56:30 +04:00
d1de446ae0
Fix & optimize DataFlowInfo.or()
...
Get complete nullability & type info before creating a new data flow info.
Don't create unnecessary data flow infos
2012-11-16 17:56:29 +04:00
6b65c44d45
Optimize DataFlowInfo.and()
...
Instead of copying everything, we now only completely copy one data flow info
and use another as a delegation target
2012-11-16 17:56:29 +04:00
f5ed3ff6bd
Refactor DataFlowInfo.establishSubtyping()
...
Don't create useless copies of data flow infos.
Also get rid of vararg subject variable in type-checking 'when' expression
2012-11-16 17:56:28 +04:00
d52387a7eb
Delegating data flow info
...
A delegating data flow info can have a parent data flow info, which it queries
when is queried itself. This helps to avoid lots of copying of the same data
2012-11-16 17:56:27 +04:00
83d56213cd
Separate DataFlowInfo into interface & implementation
2012-11-16 17:56:27 +04:00
ed02c6c39e
Remove DataFlowInfo.CompositionOperator
...
It was used only in one place, which wouldn't suffer without it
2012-11-16 17:56:26 +04:00
4ec95ff4d6
Minor, replace Name creation with constants usage
2012-11-16 17:56:25 +04:00
57f18fca2e
Retain data flow info after array accesses
2012-11-16 17:56:24 +04:00
7c0ea67356
Fix erroneous check when saving EXPRESSION_DATA_FLOW_INFO
...
Before saving a data flow info, we checked if it's the same as that of context,
which was passed to us. This proves wrong when eventually somebody asks for a
data flow info of the same expression, but with a different context (i.e. with
a different data flow info in it) and we return this (different) context's
info, whereas we should have returned that of the original context.
This could've happened in CandidateResolver, for example.
Save data flow info after every expression now, only if it's not empty.
No new tests added, see next commit.
2012-11-16 17:56:23 +04:00
ac7ee9d3f5
Retain data flow info after when-expressions
...
This means the subject expression and the common data flow info of all entries'
conditions
2012-11-16 17:56:23 +04:00
d0a2ba5737
Retain data flow info after is-expression
2012-11-16 17:56:22 +04:00
ef3e38071c
Retain data flow info after try-finally
2012-11-16 17:56:21 +04:00
d629fe2d91
Retain data flow info after 'for' loop range expressions
2012-11-16 17:56:20 +04:00
0b1c87a812
Retain data flow info after conditions in do-while statements
2012-11-16 17:56:20 +04:00
b0e9c7aab7
Retain data flow info after conditions in while-statements
2012-11-16 17:56:19 +04:00
ae77bf0a4e
Retain data flow info after conditions in if-statements
2012-11-16 17:56:18 +04:00
2872a0799a
Retain data flow info after string template expressions
2012-11-16 17:56:17 +04:00
950f912154
Retain data flow info after multi-declaration
...
#KT-2825 In Progress
2012-11-16 17:56:17 +04:00
a54e34282a
Retain data flow info after assignment operations (+= etc)
...
#KT-2825 In Progress
2012-11-16 17:56:16 +04:00
f949d9e804
Retain data flow info after assignment statements
...
#KT-2825 In Progress
2012-11-16 17:56:15 +04:00
Evgeny Gerashchenko