Commit Graph

134 Commits

Author SHA1 Message Date
Mikhail Glukhikh ed8ccdc15a Take into account data flow info changes for special call (if/when/elvis/!!) arguments #KT-10824 Fixed
Smart casts on complex expressions look as no more possible
2016-01-28 19:30:20 +03:00
Mikhail Glukhikh 5062ce2804 Regression: correct smart cast from Type? to Type! 2016-01-28 17:08:15 +03:00
Stanislav Erokhin 298a075381 Minor. removed functions which declared in Standard.kt from diagnostic testdata.
Removed declaration for functions: TODO, run, with, apply, let.
2016-01-25 19:13:31 +03:00
Stanislav Erokhin d7438d9f2b Add mock runtime to diagnostic tests.
Mock runtime contains function from file Standard.kt
2016-01-25 19:13:30 +03:00
Mikhail Glukhikh e9af4d25d0 Handling of try / catch / finally with 'Nothing' in all catch blocks #KT-10735 Fixed 2016-01-22 16:37:22 +03:00
Mikhail Glukhikh 366ab508e1 Data flow analysis: assignment of null makes variable effectively of type 'Nothing?' 2016-01-22 16:37:10 +03:00
Mikhail Glukhikh 859059d714 Bug fix: jump out of a loop inside if condition 2016-01-22 16:37:07 +03:00
Mikhail Glukhikh 3106a8508a Front-end and back-end test for KT-7801 2016-01-22 16:37:00 +03:00
Mikhail Glukhikh 7ac55cefcb Extra test with smart cast to 'T' 2016-01-22 16:36:57 +03:00
Dmitry Petrov f371e67ce8 PatternMatchingTypingVisitor:
rewrite type inference for 'when' using special constructs.
This fixes several type inference issues for 'when':
KT-9929, KT-9972, KT-10439, KT-10463
along with some other diagnostics-related issues.
2016-01-22 10:41:55 +03:00
Ilya Gorbunov f4822cd757 Fix testData in compiler: add collections and ranges package to fq-names. 2016-01-22 05:54:38 +03:00
Alexander Udalov 078a53e6a0 Minor, add regression test for KT-10444
#KT-10444 Obsolete
2016-01-17 16:28:22 +03:00
Mikhail Glukhikh 0f80df7b2e initialDataFlowInfoForArguments introduced in CallExpressionResolver methods #KT-10175 Fixed 2016-01-15 17:22:42 +03:00
Stanislav Erokhin 3d44471659 KT-10483 Missing smart cast on variable in variable as function call
#KT-10483 Fixed
2015-12-25 20:13:08 +03:00
Denis Zharkov b4bb92d136 Fix overload resolution ambiguity for types intersection
There are two different forms of types intestion:
1. Type parameters with multiple bounds
2. Smart casts

The problem was that when member scope of type intersection contained
effective duplicates and that lead to overload resolution ambiguity in
strange cases like `x.hashCode()`

For first type we do effectively the same thing as when building member
scope for class extending several interfaces: group all descriptors by
both-way-overridability relation and then choose most-specific in each
group.

For smart casts we do basically the same thing but with special
treatments:
1. From all descriptors that _equal_ to most specific we choose
   the one that works without smartcast if possible (i.e. we choose first from candidates list)
2. If smart-cast value seems to be unstable we use only member scope
   of receiver type + all descriptors from smart cast possible types
   that has incompatible signature. If we'd include all of them and
   choose one as more specific, and it would lead to false
   SMART_CAST_IMPOSIBLE (see test unstableSmartCast.kt)

 #KT-3996 Fixed
 #KT-10315 Fixed
2015-12-15 16:18:31 +03:00
Mikhail Glukhikh 6fb871a088 A set of additional tests for complex expressions DFA 2015-12-11 15:17:37 +03:00
Mikhail Glukhikh d024045638 Data flow values for Elvis / parenthesized expressions, smart casts on them 2015-12-11 15:17:29 +03:00
Mikhail Glukhikh 745a3aeeac Accurate representation of smart casts on complex expressions 2015-12-11 15:17:25 +03:00
Mikhail Glukhikh 7cd867b936 DFA: take into account if / else and when expression assignments #KT-10002 Fixed
Also #KT-7479 Fixed
2015-12-11 15:17:22 +03:00
Mikhail Glukhikh c73f01927a 'sealed' is now not 'final' by default + related code changes #KT-10266 Fixed 2015-12-09 19:19:04 +03:00
Dmitry Jemerov e97e82d119 use terms "lambda expression" and "anonymous function" instead of "function literal" and "function expression" 2015-12-09 13:55:40 +01:00
Mikhail Glukhikh 57b3bc0496 Regression: getPredictableNullability for unnecessary safe call check #KT-10276 Fixed 2015-12-08 11:23:35 +03:00
Mikhail Glukhikh 5c9e55f3fb DFA bug fix: assignment of unstable to stable uses predictable type set (see test) 2015-12-08 10:13:38 +03:00
Mikhail Glukhikh 2e73bcb3a6 DFA bug fix: assignment of unstable to stable uses predictable nullability (see test) 2015-12-08 10:13:34 +03:00
Mikhail Glukhikh 5729855e4f Type parameter as result type: more precise expected type definition #KT-5385 Fixed
Tests for KT-5385 KT-10223 KT-7800 KT-7817 + something related
2015-12-04 19:00:24 +03:00
Mikhail Glukhikh fe04cc513b Original type is taken into account when intersecting possible types in resolve #KT-10232 Fixed 2015-11-30 19:09:44 +03:00
Mikhail Glukhikh 8788d8e2d5 Call chains: receiver data flow info is dropped for extensions with nullable receiver #KT-10056 Fixed 2015-11-27 20:23:25 +03:00
Mikhail Glukhikh 9a6d4c32c5 More accurate handling of "always null" for receivers 2015-11-25 18:26:15 +03:00
Mikhail Glukhikh 89e56093a2 Limit "always null" scope: only for !!, is and dot; senseless comparison rolled back; "smart constant" information for nulls #KT-10029 Fixed 2015-11-25 18:26:08 +03:00
Dmitry Petrov da90c21284 Warn about comma-separated conditions in when without argument.
See KT-5143.
2015-11-25 15:39:23 +03:00
Mikhail Glukhikh 3a4f6d8226 Smart cast impossible on when subject is no more recorded #KT-10061 Fixed 2015-11-18 18:02:59 +03:00
Mikhail Glukhikh 811ba8110f Implicit receiver smart casts implementation and highlighting 2015-11-17 10:26:42 +03:00
Mikhail Glukhikh b468d5b0c6 No smart casts to Nothing? 2015-11-09 16:36:37 +03:00
Mikhail Glukhikh 2d9fbf5696 Always null detection 2015-11-09 16:36:34 +03:00
Mikhail Glukhikh b556037915 Smart casts to nothing after comparison with null 2015-11-06 10:57:06 +03:00
Mikhail Glukhikh 5e9aa38950 More precise type handling for equality #KT-2311 Fixed 2015-11-05 17:29:34 +03:00
Mikhail Glukhikh 41ebfd025e More precise diagnostics of smart cast impossible #KT-7240 Fixed 2015-11-03 13:38:51 +03:00
Mikhail Glukhikh 03aaddd379 More correct handling of possible type set during assignment #KT-8229 Fixed 2015-11-03 12:49:21 +03:00
Mikhail Glukhikh 0a5a5a2e36 More precise diagnostics is added for smart cast impossible #KT-8810 Fixed 2015-11-03 11:15:58 +03:00
Mikhail Glukhikh 182cde1e02 Standard data flow analysis results are taken into account during condition analysis #KT-7933 Fixed 2015-10-30 16:14:34 +03:00
Mikhail Glukhikh dd1196ae6b Condition analysis: left part of and (true condition) / or (false condition) is used for right part analysis #KT-8780 Fixed 2015-10-30 16:14:33 +03:00
Mikhail Glukhikh 12103d19d2 Smart casts inside property chains like a?.b?.c.foo(a, b) are now handled correctly #KT-7290 Fixed 2015-10-30 16:14:32 +03:00
Mikhail Glukhikh d2c65a5336 Iterative call chain handling to get rid of JetTypeInfoInsideCallChain and other nasty things #KT-9754 Fixed 2015-10-28 16:00:30 +03:00
Alexey Tsvetkov 1e2256207c Update test data 2015-10-19 20:45:02 +03:00
Alexey Tsvetkov dc27363452 Remove @annotation from test data 2015-10-19 20:45:01 +03:00
Alexey Tsvetkov 1973573a89 Remove @data from test data 2015-10-19 20:45:00 +03:00
Svetlana Isakova bb9f7094e0 Report 'TYPE_INFERENCE_EXPECTED_TYPE_MISMATCH'
on call element, not on callee expression
2015-10-19 14:15:05 +03:00
Mikhail Glukhikh 4b35e3b135 Preliminary declaration visitor for estimating local variable's predictability for smart casts
Predictability estimation algorithm is completely new, but backward compatibility should present.
A large set of tests. Some updated tests.
Smart casts allowed for captured variables if they are not modified in closure #KT-9051 Fixed
Also #KT-8643 Fixed
Also #KT-7976 Fixed
Correct handling of lambda arguments in functions #KT-9143 Fixed
2015-10-16 20:47:47 +03:00
Denis Zharkov bd37a10677 Transform *Array.size to properties 2015-10-16 18:16:02 +03:00
Alexander Udalov ec1b4776fe Deprecate PropertyMetadata, use KProperty<*> for delegated properties instead 2015-10-14 20:45:25 +03:00