BLI: Adjust interpolation to support integers, other tweaks
ClosedPublic
Actions

Authored by Hans Goudey (HooglyBoogly) on Mar 25 2022, 5:00 AM.

Details

Reviewers

Jacques Lucke (JacquesLucke)
Clément Foucault (fclem)

Commits

rB378022c7973d: BLI: Adjust interpolation to support integers, other tweaks

Summary

In order to allow interpolation of integers with a float, add a separate
template parameter for the factor and multiplication types.
Also move some helper constexpr variables to the "base" header
(reversing the dependency to "base" -> "vector").

This also adds a distance function for scalar types, which is
helpful to allow sharing code between vectors and basic types.

Diff Detail

Repository: rB Blender

Event Timeline

Hans Goudey (HooglyBoogly) requested review of this revision.Mar 25 2022, 5:00 AM

Hans Goudey (HooglyBoogly) created this revision.

Hans Goudey (HooglyBoogly) added inline comments.

source/blender/blenlib/BLI_math_vec_types.hh
329	This same change could be applied to other operations here, but maybe it's better to just do it where necessary?

Jacques Lucke (JacquesLucke) added inline comments.Mar 25 2022, 12:31 PM

source/blender/blenlib/BLI_math_base.hh
95	Doesn't seem to be used.
source/blender/blenlib/BLI_math_vec_types.hh
15	Including that here seems unnecessary. You might have to add it in `BLI_math_vector.hh` and `BLI_float4x4.hh` though.
329	Doing it when necessary is fine. Looks like this should replace the function above though.

Jacques Lucke (JacquesLucke) requested changes to this revision.Mar 25 2022, 12:31 PM

This revision now requires changes to proceed.Mar 25 2022, 12:31 PM

Clément Foucault (fclem) added inline comments.Mar 25 2022, 1:16 PM

source/blender/blenlib/BLI_math_base.hh
95	Also I would expect `distance` to return a scalar distance between two points. Maybe there is instances where this is useful for scalar types?

Hans Goudey (HooglyBoogly) added inline comments.Mar 25 2022, 2:35 PM

source/blender/blenlib/BLI_math_base.hh
95	Ah, I forgot to mention that in the patch description. I'm using this in D14382 so the same code can work for vector and scalar types. I know it makes less sense to have a distance function for scalars, but the consistency is very helpful.