Commit Graph

17 Commits

Author SHA1 Message Date
Dmitriy Novozhilov 5dfe100ae5 Allow use reference to reified type parameters in contracts since 1.4 2019-12-26 17:43:28 +03:00
Dmitriy Novozhilov f083edfac2 Allow contracts on final non-override members since 1.4 2019-12-26 17:43:27 +03:00
Dmitriy Novozhilov 2536fa0cd5 [FIR-TEST] Add new testdata generated after changes in previous commit 2019-12-12 16:11:46 +03:00
Dmitriy Novozhilov 9129210cfe Move changes from dfb379d under new language feature (#KT-28672)
Changes from that commit are breaking, and should be pass throw LC
2019-02-07 12:12:34 +03:00
Dmitriy Novozhilov 0501a108c0 Fix tests in DiagnosticsTestWithStdLibGenerated with NI 2019-02-01 11:40:20 +03:00
Dmitriy Novozhilov dfb379d999 Expand Effect System with receiver values (#KT-28672)
Add some classes to hierarchy of `ESValue`:
- `ESReceiver` -- representation of receiver
  of lambda function
- `ESDataFlow` and `ESDataFlowReceiver` -- little refactoring of entities
  that holds information for DFA (description in comments in code).

Also add kdoc to classes of `ESValue` hierarchy
2019-01-31 17:25:13 +03:00
Dmitriy Novozhilov 932e0234e7 Add testdata for issue #KT-28672 2019-01-31 17:25:13 +03:00
Dmitry Savvinov e483f83666 Make 'is'-operator more stupid
Consider following expression: 'call() is Foo'. Suppose that we know
something about the 'call()', e.g. 'returns(foo) -> <condition>'

Previously, we've tried to re-use knowledge about 'call()', constructing
some smart clause, like 'returns(true) -> foo is Foo && <condition>'.

The conceptual error here is that *we can't* argue that <condition>
holds. Imagine that 'call()' actually has unspecified 'returns(foo2) ->
<!condition>', and 'foo2 is Foo' also holds. Then we would get
'returns(true) -> foo2 is Foo && <condition>' <=> 'returns(true) ->
<condition>' for the whole call, which is not correct.

More concrete example would be something like:
'if (!x.isNullOrEmpty() is Boolean)'

^KT-27241 Fixed
2019-01-17 12:47:27 +03:00
Dmitry Savvinov 04ff2a3ee7 Add test on isNullOrEmpty-like contract
Currently, behavior is undesired, see KT-27241
2019-01-17 12:47:27 +03:00
Dmitry Savvinov 051ad0f79e Fix unsound smartcast due to returns(null) in safecall
^KT-26382 Fixed
2018-09-05 16:27:48 +03:00
Dmitry Savvinov 18338f4ea3 Add test on safecall of function with returns(null) contract
Currently with non-desired testdata, fix incoming in the next commit.
2018-09-05 16:27:47 +03:00
Dmitry Savvinov ee8702d21e Load of testdata change due to contracts publishing
See changes in e2606b72bdbec2fea567d4127197707869eb801e
2018-08-30 16:19:55 +03:00
Dmitry Savvinov 57bafcf9e9 Do not lose information about contracts if call uses default value
It is safe to treat DefaultValueArgument as UNKNOWN_COMPUTATION, because
default arguments can't break smartcasts.

Possibly, they can add new ones, but it can be supported later.

^KT-25278 Fixed
2018-08-10 13:19:32 +03:00
Denis Zharkov 375f3c0769 Fix type-intersection-related test data
The order of types enumeration has been changed recently:
previously it was collected at first from a child and then from
its parent, but now it's being collected in order of appearance
2018-08-10 15:40:50 +07:00
Dmitry Savvinov b29a6e48fb Refactor language features, which control effect system
- Introduce new language feature 'ReadDeserializedContracts', which
allows to deserialize contracts from metadata.

- Introduce new language feature 'AllowContractsForCustomFunctions',
which allows reading contracts from sources.

- Use new features instead of combination 'CallsInPlaceEffect ||
ReturnsEffect'

- Rename 'CallsInPlaceEffect' -> 'UseCallsInPlaceEffect',
'ReturnsEffect' -> 'UseReturnsEffect'. As names suggest, they control
if it is allowed to use corresponding effect in analysis.

We have to introduce separate 'ReadDeserializedContracts' to enable
contracts only in some modules of the project, because libraries are
read with project-wide settings (see KT-20692).
2018-01-26 11:30:44 +03:00
Dmitry Savvinov 816d89e393 [NI] Improved testdata after changes in applicabilities
This commits introduces testdata changes, where NI behaviour strictly
improved, after several previous fixes.

For some tests, just WITH_NEW_INFERENCE directive was added. It
indicates, that some of previous commits first introduced error in that
test, and then some other commit fixed it (netting no overall testdata
change). It is preferrably to keep those annotations until we will
migrate to NI completely, to prevent unexpected regressions.
2017-12-07 12:49:56 +03:00
Dmitry Savvinov 4434db4d69 Effects: add diagnostic tests on contracts
- Make AbstractDiagnosticsTest dump function contracts
- Add diagnostics tests on parsing contracts
- Add diagnostics tests on smartcats in presence of functions with
contracts
- Add diagnostics tests on initialization and flow in presence of
in-place called lambdas

==========
Introduction of EffectSystem: 16/18
2017-10-12 11:55:26 +03:00