Rename circle modifier for symbols that are circled to "o"

[Enivex]

This is in conflict with #34 . However, circled symbols are extremely common, both in terms of the number of characters (we're still missing a bunch of symbols) and in usage. Therefore, I believe that it deserves its own short and cute modifier, which also has the benefit of resolving the ambiguous conflicting uses of "circle". There is also precedence, in the form of "oo" to mean infinity.

Notes:

• ast.circle was really a circled variant of ast.op, so I used ast.op.o.
• circle.nested was really a circled variant of circle.stroked.tiny, so I used circle.stroked.tiny.o
• The aliases compose.o and convolve.o were missing. I added them

I initially wanted to add additional symbols in this PR, but in the process of looking into this I discovered a number of issues that should ideally be resolved first. I therefore prioritized just getting the ball rolling.

[MDLC01]

If circle.stroked.tiny.o is the same as compose.o, maybe we can only keep compose.o. I feel like the .o modifier is less clear in meaning when it comes to geometrical shapes, compared to mathematical operators.

[Enivex]

If circle.stroked.tiny.o is the same as compose.o, maybe we can only keep compose.o. I feel like the .o modifier is less clear in meaning when it comes to geometrical shapes, compared to mathematical operators.

I'd be okay with that

[Enivex]

I did not add a deprecation to slash.circle, since it will likely never see release.

[T0mstone]

If circle.stroked.tiny.o is the same as compose.o, maybe we can only keep compose.o. I feel like the .o modifier is less clear in meaning when it comes to geometrical shapes, compared to mathematical operators.

I'm a bit bugged by the inconsistency of having compose have a unique variant instead of being a strict alias. Can't we keep circle.nested somehow? Or maybe go the other way and deprecate circle.stroked.tiny, given that its purpose in Unicode is strictly as a mathematical operator?

[Enivex]

I'd rather go the second route. There's a number of symbols that are currently named with little regard for what they are actually used for.

That being said, I'm not sure if that change should be made in this PR or not.

[mkorje]

That being said, I'm not sure if that change should be made in this PR or not.

I think it'd be better to do in another PR then, where we could review other symbols' names.

[MDLC01]

Yes there is definitely some work to do around geometrical shapes, in the same vein as #28.

[MDLC01]

With #86 merged, this can now use the dedicated deprecation syntax for variants.

[Enivex]

With #86 merged, this can now use the dedicated deprecation syntax for variants.

Done. I also readded circle .stroked.tiny.o as deprecated

[Enivex]

What do people feel about renaming dash.o to minus.o.alt while we're already deprecating dash.circled? Despite the official unicode name, which I think comes from the fact that the horizontal bar has approximately en-dash length, it's a binary operator and seemingly used for the same things as minus.o. It doesn't feel at home among the punctuation.

It would also aid discoverability.

[T0mstone]

I also readded circle .stroked.tiny.o as deprecated

It never existed in the firstplace, I don't think there's a point to adding it now when it'd be deprecated anyway.

[Enivex]

I also readded circle .stroked.tiny.o as deprecated

It never existed in the firstplace, I don't think there's a point to adding it now when it'd be deprecated anyway.

Oops, brain fart