search

enthought / mayavi

3D visualization of scientific data in Python
http://docs.enthought.com/mayavi/mayavi/
Other
1.3k stars 284 forks source link

Quick fix for failures with Traits 6.3 #1092

Closed mdickinson closed 2 years ago

mdickinson commented 2 years ago

This PR represents a quick fix for failures of the Mayavi test suite with the recently-released Traits 6.3.0.

Note that this is only a band-aid workaround to get the CI up and running again. A better fix would be to rewrite RevPrefixMap to inherit directly from TraitType (which is designed to be subclassed), and not to either subclass PrefixMap (which wasn't designed to be subclassed), or to depend on internal details of PrefixMap.

mdickinson commented 2 years ago

Tests pass on Linux, failed on macOS and were cancelled on Windows. The test failure on macOS appears to be an intermittent failure that's unrelated to this PR:

======================================================================
FAIL: Test if all the TVTK classes can be instantiated
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/runner/work/mayavi/mayavi/tvtk/tests/test_tvtk.py", line 818, in test_all_instantiable
    raise AssertionError(message.format(''.join(errors)))
AssertionError: Not all classes could be instantiated:
Traceback (most recent call last):
  File "/Users/runner/work/mayavi/mayavi/tvtk/tests/test_tvtk.py", line 813, in test_all_instantiable
    tvtk_klass()
  File "tvtk_classes/euclidean_cluster_extraction.py", line 76, in __init__
    tvtk_base.TVTKBase.__init__(self, vtk.vtkEuclideanClusterExtraction, obj, update, **traits)
  File "/Users/runner/work/mayavi/mayavi/tvtk/tvtk_base.py", line 386, in __init__
    self.update_traits()
  File "/Users/runner/work/mayavi/mayavi/tvtk/tvtk_base.py", line 539, in update_traits
    setattr(self, name, val)
  File "/Users/runner/hostedtoolcache/Python/3.6.15/x64/lib/python3.6/site-packages/traits/base_trait_handler.py", line 75, in error
    object, name, self.full_info(object, name, value), value
traits.trait_errors.TraitError: The 'radius' trait of an EuclideanClusterExtraction instance must be 0.0 <= a floating point number <= 9.999999680285692e+37, but a value of 3.105036184601418e+231 <class 'float'> was specified.

I'll restart the test run.

mdickinson commented 2 years ago

appears to be an intermittent failure

Indeed, it looks like a case of #861.

mdickinson commented 2 years ago

I've opened #1094 for the underlying issue. This PR should probably be updated to add a reference to that issue.

mdickinson commented 2 years ago

Closing: we've incorporated this fix into Traits 6.3.1 and made a Traits release. It's possible that Mayavi's dependencies should be updated to explicitly exclude Traits 6.3.0.