reverse(A; dims): Allow tuple for "dims" keyword

[AndiMD]

Hi,

I suggest allowing a tuple for the dims keyword to make it consistent with sum(A;dims=(1,2)) etc.
Also, size(A,dims) could support a dims keyword for consistency.

Thanks!
Andreas

[HDictus]

I like both these ideas, having a kwarg 'dims' only accept a single integer is bad naming anyway.

[HDictus]

Ohh, if I understand it correctly, reverse is implemented in lisp, this should be interesting.
Is anyone already working on this?

[StefanKarpinski]

if I understand it correctly, reverse is implemented in lisp

Huh? Where are you getting that?

[HDictus]

grepping through the files, the definition I found for reverse was in a lisp file.
I probably missed the actual one, I can't understand how dims ends up in that lisp definition.
I found a different one in 'base' , I guess that's the actual one? I'm a little confused that it references Base.reverse when it itself is in base, and I can't find the other definition. The lisp one is eager (as a comment in base.iterators.jl mentions about Base.reverse), so maybe that's it?

Sorry for getting confused, it's my first time taking a close look at the actual main body julia code.

[StefanKarpinski]

The Julia repo contains an entire (tiny) Lisp implementation which is used for the parser. Lisp has many common function names like reverse. No Julia functions are implemented in Lisp, only logic for parsing source files and converting them into Julia AST. If you're looking for the definition of a function, you generally want to restrict your search to the base directory or stdlib if it's defined in a standard library.

[HDictus]

Ah, ok,. thanks!

[HDictus]

#31716 should close this, only I'm having some trouble with the win32 build, which seems to have had some trouble cleaning and running tests on time, and the xcode tests, which can't seem to connect to github.
If anyone knows where to start wuth looking into this, please let me know.

[mcabbott]

Closed by #37367, I think.