nits from JET

[thchr]

Some minor nits found with JET.

The point of changes in src/SVector.jl and src/SArray.jl is that the if-statement otherwise cannot know that the call to .head is meaningful (it still thinks that ex.args[1] isa Any). This change just splits it up and type-asserts so the knowledge that ex isa Expr can propagate. The code duplication is a little annoying, but I couldn't think of a good way to avoid it.

Since it's a macro it doesn't really matter - but I figured it might be nice anyway in order to reduce noise in reports by JET in the future.

[mateuszbaran]

It would be nice to have coverage for these error messages but other than that it looks fine.

[thchr]

Would be happy to add some tests. Only... I don't know how to (if it is possible?) to create a :comprehension whose first argument is not a :generator? I.e., I'm not sure what would actually fail this test. Any ideas?

[mateuszbaran]

[1 for i in 1:10 for j in 1:10] has :flatten inside :comprehension so it would work as an example:

julia> dump(:([1 for i in 1:10 for j in 1:10]))
Expr
  head: Symbol comprehension
  args: Array{Any}((1,))
    1: Expr
      head: Symbol flatten
      args: Array{Any}((1,))
        1: Expr
          head: Symbol generator
          args: Array{Any}((2,))
            1: Expr
              head: Symbol generator
              args: Array{Any}((2,))
                1: Int64 1
                2: Expr
            2: Expr
              head: Symbol =
              args: Array{Any}((2,))
                1: Symbol i
                2: Expr

[thchr]

Fantastic, thanks - that'll do it! Added tests in ead793e.