Parsing problem with nested splatting named tuple constructor in a function call

[bkamins]

In the code below definitions f1 and f2 should be equivalent I think.
However, f1 is a syntax error, while f2 works as expected.

julia> h(x; k=1) = (x, k)
h (generic function with 1 method)

julia> f1(c) = h((;c...), k=true)
ERROR: syntax: "..." expression outside call
Stacktrace:
 [1] top-level scope at REPL[8]:1

julia> f2(c) = (nt=(;c...); h(nt, k=true))
f2 (generic function with 1 method)

julia> f2(Dict(:a=>1, :b=>2))
((a = 1, b = 2), true)

It seems that a parser does not correctly recognize that ... in this case is used in a nested NamedTuple constructor call within a call to h.

The problem seems to be present only if the call involves keyword arguments, so e.g.

julia> f1(c) = h((;c...), true)
f1 (generic function with 1 method)

parses correctly.

Also

julia> f1(c) = h(;c..., k=true)
f1 (generic function with 1 method)

parses, but it gives a different definition of course.