UsdMtlxSplitStringArray Parses Spaces Incorrectly

[feldstj]

Description of Issue

UsdMtlxSplitStringArray splits strings currently based on commas and spaces. Certain enums such as ND_lama_sss's sssMode assume that the spaces in the enum labels will not be used to split the enum labels. The fix is to only split on commas and to remove leading and trailing whitespace from the results.

Steps to Reproduce

from pxr import Sdr
r = Sdr.Registry()
n = r.GetShaderNodeByIdentifier("ND_lama_sss")
t = n.GetShaderInput("sssMode")
print(t.GetOptions())

Result: [('Path-traced', ''), ('Davis', ''), ('Path-traced', ''), ('exponential', ''), ('Diffusion', ''), ('Burley', ''), ('Diffusion', ''), ('Burley', ''), ('(mean', ''), ('free', ''), ('path)', '')]

Expected: [('Path-traced Davis', '0'), ('Path-traced exponential', '1'), ('Diffusion Burley', '2'), ('Diffusion Burley (mean free path)', '3')]

System Information (OS, Hardware)

Package Versions

Build Flags

Enable MaterialX.

[tallytalwar]

Filed as internal issue #USD-8006