fix nibabel get_data deprecation

[xgui3783]

per https://nipy.org/nibabel/reference/nibabel.dataobj_images.html#nibabel.dataobj_images.DataobjImage.get_data nibabel.get_data will raise Exception as of >=5.0

Per suggestion, this PR now uses numpy.asanyarray(img.dataobj)

another option is img.get_fdata(img.header.get_data_dtype())

[codecov]

Codecov Report

Merging #28 (4fd3f02) into master (b1baf42) will not change coverage.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##           master      #28   +/-   ##
=======================================
  Coverage   92.45%   92.45%           
=======================================
  Files          25       25           
  Lines        1485     1485           
  Branches      219      219           
=======================================
  Hits         1373     1373           
  Misses         63       63           
  Partials       49       49           

Impacted Files	Coverage Δ
src/neuroglancer_scripts/volume_reader.py	82.69% <100.00%> (ø)

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[xgui3783]

probably supercedes #27

[xgui3783]

@ylep can you take a look at this PR, but also a number of other possible fixes (#27) or using img.get_fdata(img.header.get_data_dtype()) instead of numpy.asanyarray(img.dataobj) (this PR) ?

[ylep]

It seems to me that this is actually the correct replacement. Using get_fdata here would be incorrect, because we are handling the scaling in neuroglancer-scripts instead of having it applied by nibabel.

[blowekamp]

Thank you for a proper fix.

Any idea when the next published release will occur?

[xgui3783]

Thank you for a proper fix.

Any idea when the next published release will occur?

already happened, I think: https://pypi.org/project/neuroglancer-scripts/#history