Correct developer guide's function template

[gmou3]

Add spaces in tuples.

P.S. Should the default values 1 and 2 have single or double backticks?

[github-actions]

Documentation preview for this PR (built with commit a3a1a9f; changes) is ready! 🎉
This preview will update shortly after each push to this PR.

[saraedum]

The 1 and 2 should have no backticks at all. They could also have two but I don't think that LaTeX typeset makes sense here.

I think the 1 and 2 should have two backticks here. No backticks is also fine, but it's easier to teach people to use `` here.

[gmou3]

I simplified a bit the examples, and changed the note. Previously, the note was more like an algorithm block for which we have a good example.

[saraedum]

Sorry, I misunderstood where this goes in the entire documentation. I think the original OUTPUT was actually fine.

Then again, we have many examples of OUTPUT: tuple in the code base. So it's a good template to follow that.

[gmou3]

Thank you.

[gmou3]

@kwankyu I agree with these recommendations. It is clearer if one thinks of something more complicated as the default value, a field for example. We would like to show some nice math instead of some code I presume (the code would appear further up in the arguments).

I think I need to do a follow-up PR.

[kwankyu]

If you agree, I may do it in #38117.

[saraedum]

I'm confused. Should the defaults be typeset with single backticks and not double-backticks anymore?

I find 218 single backticks for defaults and 6567 double backticks for defaults in our codebase.

[gmou3]

Something from Python like None, or True, should have double backticks.
Something mathematical should have single backticks.

Let us know of any counter-argument.

[saraedum]

Something from Python like None, or True, should have double backticks. Something mathematical should have single backticks.

Let us know of any counter-argument.

My main argument is that the code base uses more double-backticks than single-backticks. I believe that this document should not set new standards but report what the code base is already doing.

Also, 1 is the actual default value as a Python int, it's not really a "mathematical object" arguably. I believe that the defaults should report the actual default of a function, imho this can never be a mathematical object. If it's None and None means the ring of integers, then INPUT: should say so instead of claiming that ring (default: `\ZZ`) simply because `\ZZ` is not something you can type into SageMath. I have no strong opinion on that latter part though.

There's of course the caveat in my argument that 1 is the Python int and not the SageMath preparsed int but we tend to gloss over this oddity almost everywhere in the documentation (and our code tries to treat both of them equally.)

[gmou3]

I think the part after the "--" should be treated the same as the text body preceding the "INPUT". One can see how to type the default value from the function header.

I am unsure about what the rule is/was. The template had (and still has) single backticks however. We can determine here what is desirable based on some logic, and everything else can be adapted to comply.