Add expanded assembly subcomponent support for pxrUsdReferenceAssembly

[dj-mcg]

Fix double-transform rendering and add variant selection editing for subcomponent proxy shapes created when a pxrUsdReferenceAssembly is expanded.

Double-transform fix:

• Add virtual isRootPrimTransformInDagPath() to MayaUsdProxyShapeBase so subclasses can signal that the Maya DAG already contains the root prim's world transform.
• pxrUsdProxyShape implements this via a new hidden boolean attribute "skipRootPrimTransform", set by the assembly translator on expand.
• proxyDrawOverride, usdProxyShapeAdapter, and proxyRenderDelegate check the flag before applying the root prim transform.
• pxrUsdProxyShape::boundingBox() uses ComputeUntransformedBound when the flag is set, fixing picking and framing.

Variant selection support:

• Extend UsdMayaEditUtil to parse variant selection edits (usdVariantSet_* attributes) alongside transform edits.
• Add setDependentsDirty/compute overrides to pxrUsdProxyShape that apply variant selections from dynamic attributes onto the stage session layer.
• Replace AEpxrUsdProxyShapeTemplate.mel with a Python version that builds variant set UI dynamically from the USD prim.
• Register the AE template Python module in init.py.

Tests:

• Add testUsdReferenceAssemblyExpandedTransforms verifying correct transforms and bounding boxes for expanded sub-proxies.

[AramAzhari-adsk]

Just some minor adjustments

[AramAzhari-adsk]

@dj-mcg one of the unit tests seem to fail on OSX platforms. Does that also occur for you?

======================================================================
13:17:23      FAIL: testClipboardCutPaste (testClipboard.ClipboardHandlerTestCase.testClipboardCutPaste)
13:17:23      Basic test for the Clipboard cut/paste support.
13:17:23      ----------------------------------------------------------------------
13:17:23      Traceback (most recent call last):
13:17:23        File ".../ecg-maya-usd/maya-usd/test/lib/ufe/testClipboard.py", line 190, in testClipboardCutPaste
13:17:23          self.assertTrue(ch.hasItemsToPaste_())
13:17:23          ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^
13:17:23      AssertionError: False is not true
13:17:23
13:17:23      ----------------------------------------------------------------------
13:17:23      Ran 9 tests in 0.701s
13:17:23
13:17:23      FAILED (failures=1)

[seando-adsk]

@AramAzhari-adsk / @dj-mcg I think that was just a random glitch test failure. I re-ran the OSX 2027 build and it passed the 2nd time. Aram I can also see that you re-ran the entire preflight and OSX 2027 also passed there, but then there was a Linux build with bunch of interactive tests failures - I think that was just a network glitch. I'm marking this as ready for merge.

Sean

[AramAzhari-adsk]

@AramAzhari-adsk / @dj-mcg I think that was just a random glitch test failure. I re-ran the OSX 2027 build and it passed the 2nd time. Aram I can also see that you re-ran the entire preflight and OSX 2027 also passed there, but then there was a Linux build with bunch of interactive tests failures - I think that was just a network glitch. I'm marking this as ready for merge.

Sean

if the test failures changed to something else then I think we're good to check in.

Thanks!