Commit Graph

438 Commits

Author SHA1 Message Date
Doug Gregor
0d568a93d4 [SE-0289] Update diagnostics & many other strings to "result builders" 2020-10-20 21:44:09 -07:00
Nathan Hawes
15f5222bbd [CodeCompletion][Sema] Allow missing args when solving if the completion location indicates the user may intend to write them later.
func foo(a: Int, b: Int) {}
func foo(a: String) {}

// Int and String should both be valid, despite the missing argument for the
// first overload since the second arg may just have not been written yet.
foo(a: <complete here>

func bar(a: (Int) -> ()) {}
func bar(a: (String, Int) -> ()) {}

// $0 being of type String should be valid, rather than just Int, since $1 may
// just have not been written yet.
bar { $0.<complete here> }
2020-10-19 12:16:19 -07:00
Pavel Yaskevich
461eafff54 [ConstraintSystem] NFC: Move ConstraintSystem.h to include/swift/Sema 2020-10-08 10:45:47 -07:00
Pavel Yaskevich
4954763524 [ConstraintSystem] NFC: Move CSFix.h to include/swift/Sema 2020-10-08 10:43:18 -07:00
Pavel Yaskevich
f2614dec4a [ConstraintSystem] NFC: Move ConstraintLocator.h to include/swift/Sema 2020-10-08 10:42:39 -07:00
Pavel Yaskevich
087906a013 [ConstraintSystem] NFC: Move OverloadChoice.h to include/swift/Sema 2020-10-08 10:42:24 -07:00
Pavel Yaskevich
1fb69a7d43 [Diagnostics] Diagnose cases when it's impossible to infer type for nil
Detect and diagnose situations when it's invalid to use `nil`
literal without more context e.g. `_ = nil`, `nil as? Int`, or
`_ = nil!`.
2020-09-25 14:59:54 -07:00
Pavel Yaskevich
8c6098a3de [CSFix] Add a fix to detect when type of couldn't be inferred 2020-09-24 15:47:07 -07:00
Holly Borla
3332df6f89 Merge pull request #33777 from hborla/rename-storage-wrapper
[NFC][Property Wrappers] Rename "storage wrapper var" to "projection var/projected value"
2020-09-03 18:29:04 -07:00
Holly Borla
3fd882cadf [NFC] Rename "storage wrapper var" to "projection var" or "projected value"
in property wrapper-related code.
2020-09-02 18:07:40 -07:00
Pavel Yaskevich
c17bdaf715 [AST] Adjust Diagnostic transaction to only track pending errors 2020-09-01 13:57:03 -07:00
Pavel Yaskevich
1bfb0b0747 [ConstraintSystem] Extend invalid function body fix to cover constraint generation failures
Ignore function builder body if it emits at least one diagnostic
during constraint generation.

Resolves: rdar://problem/65983237
2020-09-01 12:23:50 -07:00
Pavel Yaskevich
da6bb3657f [CSFix] Adjust IgnoreInvalidFunctionBuilderBody to return true only if errors are emitted 2020-08-17 12:39:29 -07:00
Pavel Yaskevich
bc319779d5 [TypeChecker] Replace invalid refs with ErrorExpr only when explicitly allowed
Add a flag to `ConstraintSystem::preCheckExpression` and subsequently
to `TypeChecker::resolveDeclRefExpr` to indicate whether it's allowed
to replace invalid member refs with `ErrorExpr`.

It is useful for diagnostics and code completion to preserve AST
in it's original state otherwise it's impossible to diagnose errors
post factum or extract `CodeCompletionExpr` when it's a child of an
invalid reference.
2020-08-17 12:39:29 -07:00
Pavel Yaskevich
3ad8ea86bb [CSFix] Add a fix to ignore function builder body which fails pre-check
Fix is anchored on a `BraceStmt` associated with an invalid closure.
It's diagnosing the problem by calling `ConstraintSystem::preCheckExpression`
for all expressions in the body without suppressing errors.
2020-08-17 12:39:29 -07:00
Luciano Almeida
b1eccb594d [Sema] Do not diagnose contextual type mismatches for malformed key path expressions (#33230)
* [AST] Adding hasSingleInvalidComponent to key path expression

* [Sema] Adding a new fix and failure to diagnose missing key path component

* [Sema] Recording new fix for key path missing components and remove diagnose from pre-check

* [tests] Adjusting key path missing component contextual tests

* [Sema] Renaming missing component key path fix and failure

* [Sema] Correcting comments typos
2020-08-16 12:27:21 -03:00
Pavel Yaskevich
44d686f71d [ConstraintSystem] Use new "specify label" fix/diagnostic to warn about backward scan 2020-08-07 11:50:29 -07:00
Pavel Yaskevich
adbbc00181 [CSFix] Add a warning fix for situations when trailing closure is matched via backward scan 2020-08-06 16:55:37 -07:00
Doug Gregor
7f2d4e00fc [Diagnostics] Remove the now-unused AllowInvalidUseOfTrailingClosure
Diagnosis for invalid uses of trailing closures has been folded in
with argument-matching diagnostics, so remove all of the machinery
around the syntactic "mismatched trailing closure" logic.
2020-07-24 08:10:54 -07:00
Luciano Almeida
9924719ced [CSFix] Creating fix for diagnose missing unwrap of optional base type on key path application 2020-07-15 08:55:31 -03:00
Luciano Almeida
efc46bd635 [TypeChecker] Not warning element cast to unrelated for known collection types (Set,Array and Dictionary) 2020-07-05 22:54:51 -03:00
Luciano Almeida
43fd786ae0 [SR-5688] [Sema] Handle key path component base type on MemberAccessOnOptionalBaseFailure (#32376)
* [CSDiagnostics] Adjusting MemberAccessOnOptionalBaseFailure to be able to handle key path component member base types

* [tests] Adding regression tests for SR-5688

* [CSDiagnostics] Adjusting source range to diagnose/insert the fixes in correct location

* [tests] Adjusting regression tests to handle the fixits

* [AST] Creating an helper getSourceRange function for KeyPathExpr::Component

* [ConstraintSystem] Store member base type when recording UnwrapOptionalBase fix

* [AST] Creating a new diagnostic note for removing optional from written type

* [CSDiagnostics] Adjusting logic around MemberAccessOnOptionalBaseFailure to emit the correct diagnostics and fixes

* [tests] Adjusting regression tests to add subscript and key path root cases with respective diagnostics

* [Diagnostics] Adjusting message to mention base type

* [CSDiagnostics] Better naming for method/variable that represents base source range

* [CSDiagnostics] Adjusting to use the stored base member only when member is a key path component.

* [Diagnostics] Adjusting minor typos and code

* [AST] Adjusting keypath root diagnostic note message for use unwrapped type

* [CSDiagnostics] Adjusments in MemberAccessOnOptionalBaseFailure diagnostics as per suggestion

* [tests] Adding more test cases for SR-5688

* [CSDiagnostics] Adjusting fixits for key path root and range for diagnostics

* [CSSimplify] Attempt to diagnose InsertExplicitCall for optional function return types when possible on simplifyOptionalObjectConstraint

* [tests] Adding TODO to improve the diagnostics refering to key path root infered as optional types

* [CSDiagnostics] Adjusting comments

* [CSSimplify] Adjusting logic on simplifyOptionalObjectConstraint to attempt InsertCall fix before remove unwrap

* [CSDiagnostics] Adjust the logic to use resolveType on MemberAccessOnOptionalBaseFailure construction
2020-06-25 20:07:03 -03:00
Pavel Yaskevich
73667cdc7e [Diagnostics] Use Id_MatchOperator to check whether given operator is pattern match 2020-06-15 11:57:48 -07:00
Pavel Yaskevich
2c0b8ca76a [Diagnostics] Add a tailored diagnostic for function mismatch in argument positions
If parameter has a function type and no argument overload choices
match let's diagnose that specifically.

Resolves: [SR-12689](https://bugs.swift.org/browse/SR-12689)
Resolves: rdar://problem/62481592
2020-06-12 15:02:53 -07:00
Pavel Yaskevich
d23b938331 [Diagnostics] Implement IgnoreAssignmentDestinationType::diagnoseForAmbiguity
For cases like this:

```swift
struct X {}
struct Y {}

func overloaded<T>(_ value: T) -> T { value }
func overloaded<T>(_ value: T) -> Int { 0 }

func test(x: inout X, y: Y) {
  x = overloaded(y)
}
```

Solver would record a `IgnoreAssignmentDestinationType` fix per overload,
`diagnoseForAmbiguity` could be used to properly diagnose ambiguity cases
like that.
2020-06-12 11:47:04 -07:00
Pavel Yaskevich
e2e57ada1b [ConstraintSystem] Turn 'omit ' into a contextual mismatch it is 2020-06-12 11:47:04 -07:00
Pavel Yaskevich
b0070f5739 [Diagnostics] Check whether all contextual mismatches has the same type before diagnosing ambiguity
Instead of requiring sub-classes of `ContextualMismatch` to implement
`diagnoseForAmbiguity` let's implement it directly on `ContextualMismatch`
itself and check whether all of the aggregated fixes have same types on
both sides and if so, diagnose as-if it was a single fix.
2020-06-12 11:47:04 -07:00
Robert Widmann
ca728917e1 Merge pull request #32138 from AnthonyLatsis/eliminate-typeloc
Sema: Remove TypeLoc from ExplicitCastExpr (via TypeExpr)
2020-06-10 11:28:50 -07:00
Greg Titus
c73b144ad8 Perform the unintended generic detection in simplifyMemberConstraint and define a ConstraintFix to produce the new note. 2020-06-09 09:02:42 -07:00
Anthony Latsis
4b4634141f Sema: Remove TypeLoc from ExplicitCast 2020-06-09 15:20:25 +03:00
Pavel Yaskevich
b938f960df [Diagnostics] NFC: Rename a fix/diagnostic for missing .rawValue reference 2020-06-03 08:48:33 -07:00
Pavel Yaskevich
734f5b660c [Diagnostics] NFC: Fix typo(s) raw representative -> representable 2020-06-02 12:24:42 -07:00
Pavel Yaskevich
9805b0ab89 [ConstraintSystem] Do more checking before suggesting to use of .rawValue
Introduce `repairByUsingRawValueOfRawRepresentableType` which is used for
assignment, argument-to-parameter conversions and contextual mismatches.

It checks whether `from` side is a raw representable type and tries
to match `to` side to its `rawValue`.

Also extract logic common for `repairByUsingRawValueOfRawRepresentableType`
and `repairByExplicitRawRepresentativeUse` into `isValueOfRawRepresentable`.
2020-06-01 14:10:24 -07:00
Pavel Yaskevich
80c4aa6e80 [Diagnostics] NFC: Rename ValueType to ExpectedType in raw representable fix/diagnostics 2020-06-01 12:18:03 -07:00
Pavel Yaskevich
942d967297 [CSFix] Move type verification out of UseValueTypeOfRawRepresentative 2020-05-29 12:36:34 -07:00
Pavel Yaskevich
959c6d5606 [CSFix] Move type verification out of ExplicitlyConstructRawRepresentable
Do the verification during constraint repair and provide fix itself
with raw representative type and its raw value to be used for diagnostics.
2020-05-27 17:52:52 -07:00
Pavel Yaskevich
b90e579927 Merge pull request #32001 from LucianoPAlmeida/SR-12869-tuple-mismatch-crash
[SR-12869] [Diagnostics] Fixes assertion hit on TupleContextualFailure involving optional tuple type
2020-05-26 09:36:24 -07:00
Luciano Almeida
e43cd9d784 [CSFix] Ajusting contextual type purpose for subscript destination for assign expr in getStructuralTypeContext 2020-05-25 20:18:12 -03:00
Luciano Almeida
f13b00fc6d [CSFix] Adjusting tuple mismatch to handle optionals 2020-05-24 10:39:48 -03:00
Suyash Srijan
f026c61dce [ConstraintSystem] Disallow use of enum case as a key path component (#31969) 2020-05-22 20:41:45 +01:00
Pavel Yaskevich
5db1c70cd0 [CSFix] Extend locator with ArgumentAttribute for escaping/inout fixes
If either `MarkExplicitlyEscaping` or `TreatRValueAsLValue` fix
points to an argument-to-parameter conversion, let's add more
more clarifying argument to the locator to pinpoint the problem.
2020-05-18 13:14:01 -07:00
Pavel Yaskevich
d111f119d8 [ConstraintSystem] Detect and diagnose inability to infer type of closure parameter(s)
Detect situation when it's impossible to determine types for
closure parameters used in the body from the context. E.g.
when a call closure is associated with refers to a missing
member.

```swift
struct S {
}

S.foo { a, b in } // `S` doesn't have static member `foo`

let _ = { v in } // not enough context to infer type of `v`

_ = .foo { v in } // base type for `.foo` couldn't be determined
```

Resolves: [SR-12815](https://bugs.swift.org/browse/SR-12815)
Resolves: rdar://problem/63230293
2020-05-15 01:14:30 -07:00
Pavel Yaskevich
bf4c513f98 [CSFix] Add a fix to use when type of closure parameter can't be inferred 2020-05-15 01:13:06 -07:00
Luciano Almeida
02c454c976 [Diagnostics] Diagnose that we cannot infer the key path type when binding to a hole 2020-05-11 18:08:40 -03:00
Artem Chikin
7aed042817 Add fix on Keypath -> Function with multiple arguments type mismatch.
When simplifying a keypath constraint with a function type binding, single-parameter functions have the parameter type and the return type matched against the keypath root and value; whereas multiple-parameter functions cause an ambiguous failure (in `simplifyKeyPathConstraint`).

Resolves rdar://problem/57930643
2020-05-06 19:15:08 -07:00
Pavel Yaskevich
bd09d0c776 [ConstraintSystem] Switch ConstraintFix to use ASTNode instead of TypedNode 2020-04-29 17:03:45 -07:00
Joe Groff
b4c7a7e85d Merge pull request #31224 from AnthonyLatsis/rename-getfullname-2
AST: Rename getFullName -> getName on ValueDecl & MissingMemberDecl
2020-04-24 12:51:28 -07:00
Pavel Yaskevich
12a63ce907 [ConstraintSystem] Switch getContextualType and its variants to use TypedNode 2020-04-23 01:13:13 -07:00
Pavel Yaskevich
7669e0d059 [CSFix] Switch ConstraintFix::getAnchor to return TypedNode 2020-04-23 01:13:13 -07:00
Anthony Latsis
74252028ca AST: Rename getFullName -> getName on ValueDecl & MissingMemberDecl 2020-04-23 05:16:55 +03:00