Tiny suggestions for selections.

[pslacerda1]

I was thinking...

And if PyMOL selections supported like 1UYL.{D,B,DL,BL}.00 instead 1UYL.D.00 1UYL.B.00 1UYL.DL.00 1UYL.BL.00 which are the objects I want instead of 1UYL.*.00.

[JarrettSJohnson]

My gut reaction to this is that this is all but tiny. I'll look into this more, but the selection parser is already so complex as it is and trying to introduce another expansion mechanism sounds like it could get us into some trouble if it's not done carefully. You could probably introduce some sort of higher-level python function that manually performs the expansion for you that could be passed as a selection string. Probably not ideal, but at least reduces opportunities of regressions for other users.

Also, I feel like this could also get messy if you're writing a selection that's built off of an f-string. It wouldn't be clear what the syntax would then have to be.

cmd.select(f"1UYL.{D,B}.00") -> cmd.select(f"1UYL.{{D,B}}.00") # ?

Nested brace expansions also add yet another level of infeasibility.

{a,b{1,2},c}

Though even if we did decide to do something like this, one-level only would probably be hard-cap.

[pslacerda1]

I agree that f-strings would become cumbersome. Maybe it show be done with `[]` or `<>`. Maybe do it as a preprocessor would become more feasible because of the reduced complexity to do only context-free string manipulation. It would be different than `*` or name completion because there wouldn'd be no need to know the object names a priori. And yes, I agree with one level deep be the hard limit. Thank you for your quick and comprehensive response.