[CodeCompletion][Sema] Allow missing args when solving if the completion location indicates the user may intend to write them later. #34337
Conversation
|
@swift-ci please test |
lib/Sema/CSSimplify.cpp
Outdated
| @@ -1135,6 +1175,28 @@ class ArgumentFailureTracker : public MatchCallArgumentListener { | |||
| getExtraneousArguments() const { | |||
| return ExtraArguments; | |||
| } | |||
|
|
|||
| private: | |||
| bool areMissingArgsInvalid(ConstraintLocator *locator) { | |||
There was a problem hiding this comment.
Maybe we should convert this into a filter over MissingArguments in argument tracker instead and only record the ones which come before the token?
There was a problem hiding this comment.
Or even better - ignore missing argument recorded if it comes after the token that way destructor of the ArgumentFailureTracker could stay untouched?
lib/Sema/CSSimplify.cpp
Outdated
| SourceRange range = closure->getSourceRange(); | ||
| if (range.isValid() && SM.rangeContainsCodeCompletionLoc(range)) { | ||
| shouldFix = !closure->hasAnonymousClosureVars() && | ||
| (!args.empty() || closure->getInLoc().isValid()); |
There was a problem hiding this comment.
Maybe an early return here in reverse situation instead?
There was a problem hiding this comment.
Would missing the single tuple argument handling below (need to expand a few lines) be a problem?
Actually I guess these cases should be exclusive.
There was a problem hiding this comment.
Yeah, I think that should be okay.
nathawes
force-pushed
the
allow-missing-args-when-type-checking-for-completion
branch
from
36f89a0 to
b8a4af2
Compare
|
Build failed |
nathawes
force-pushed
the
allow-missing-args-when-type-checking-for-completion
branch
from
b8a4af2 to
cd6f200
Compare
| struct SynthesizedArg { | ||
| unsigned paramIdx; | ||
| AnyFunctionType::Param param; | ||
| }; | ||
|
|
There was a problem hiding this comment.
This is no longer necessary (in my original changes it also tracked the index to insert the missing arg at) but I left it in for the slight readability improvement over a std::pair (e.g. paramIdx vs first)
There was a problem hiding this comment.
That works for me, thank you!
|
Build failed |
|
@swift-ci please test |
…ion 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> }
nathawes
force-pushed
the
allow-missing-args-when-type-checking-for-completion
branch
from
cd6f200 to
15f5222
Compare
|
@swift-ci please test and merge |
|
Build failed |
|
Build failed |
IntandStringmembers should both be prioritized for the completion below despite the missing argument for the first overload since the second argument may just have not been written yet:Similarly, members of type
Stringshould also be prioritized for the completion below (rather than justInt) since the user may intend to use$1later in the expression.