1.+im returns a Complex{Int}

[mauro3]

julia> typeof(1.+im)
Complex{Int64}

julia> typeof(1.)
Float64

julia> typeof(1+1.im)
Complex{Float64}

should be a Float64

[StefanKarpinski]

This is parsed as 1 .+ im rather than 1. + im. So the title is a bit misleading.

[mauro3]

I see the problem. Feel free to close if it cannot be resolved.

[StefanKarpinski]

I wouldn't mind this being a warning or something.

[mbauman]

Dup of #5246, which is nearly impossible to search for. But I'll leave this open for now.

Parser warnings are annoying since the REPL parses your statements twice, so they print twice. But maybe that can be fixed, too.

[JeffBezanson]

I don't really like giving a warning for this. It means you have to change your code if you want to remove the annoying message, so it might as well be an error if anything.

[kmsquire]

Is it easy for the parser to error and give both possible parses for
situations like this?

On Tue, Apr 21, 2015 at 10:18 AM, Jeff Bezanson [email protected]
wrote:

I don't really like giving a warning for this. It means you have to change
your code if you want to remove the annoying message, so it might as well
be an error if anything.

—
Reply to this email directly or view it on GitHub
#10920 (comment).

[JeffBezanson]

Yes I think it would be pretty easy to give an error.

[jiahao]

+1 to parser error

Old crufty me is used to writing 2./x in C and Fortran to mean 2.0e0/x and this occasionally trips me up.

[stevengj]

+1 for parser error

[jakebolewski]

See comments in #11529.