Make sure skewed dataset is cast to bool properly

[paulromano]

Description

I ran into a bug loading a thermal scattering file from HDF5 recently. Namely, the skewed dataset for discrete incoherent inelastic angle-energy distribution is not read and converted to a bool properly. Right now, we're applying bool(...) on the dataset object itself, which is always true because it's not an empty object. We should be applying bool(...) on the value of the dataset. Below is a demonstration of the bug from a Python prompt (pay attention to the last two lines with bool). Note that when passing iwt=1 when generating the data, skewed should be false:

In [3]: lch4 = openmc.data.ThermalScattering.from_njoy(
   ...:     '/opt/data/endf/endf-b-viii.0/neutrons/n-001_H_001.endf',
   ...:     '/opt/data/endf/endf-b-viii.0/thermal_scatt/tsl-l-CH4.endf',
   ...:     iwt=1, stdout=True)

 njoy 2016.68  30Sep22                                       05/14/24 12:42:03
 *****************************************************************************

 reconr...                                                                0.0s

 broadr...                                                                0.0s

 thermr...                                                                0.1s

 wrote thermal data for temp = 1.0000E+02                                 2.5s

 thermr...                                                                2.5s

 ***warning***maximum value of beta limits the allowed energy transfer
 the sct approx. will be used for transfers larger than  1.389 ev.

 wrote thermal data for temp = 1.0000E+02                                13.8s

 acer...                                                                 13.8s
                                                                         14.3s
 *****************************************************************************

In [4]: lch4.export_to_hdf5('lch4_iwt1.h5')

In [5]: f = h5py.File('lch4_iwt1.h5', 'r')

In [6]: skewed_dataset = f['c_H_in_CH4_liquid/100K/inelastic/distribution/skewed']

In [7]: bool(skewed_dataset)
Out[7]: True

In [8]: bool(skewed_dataset[()])
Out[8]: False

Unfortunately we don't have any thermal scattering data in our test cross section set that triggers this problem and I'm hesitant to add another test that runs NJOY because of the time required, so for now I haven't added a test.

Checklist

• I have performed a self-review of my own code
• I have run clang-format (version 15) on any C++ source files (if applicable)
• I have followed the style guidelines for Python source files (if applicable)
• I have made corresponding changes to the documentation (if applicable)
• I have added tests that prove my fix is effective or that my feature works (if applicable)

[eepeterson]

Good catch @paulromano!