bugfix: erroneous implementation of reading RGB nifti

[xgui3783]

bugfix: rgb nifti bug
tests: added tests for arbitary dimension rgb nii
tests: added testing values before and after volume_file_to_precomputed

previous implementation of reading rgb nii was errorneous.

[codecov]

Codecov Report

Merging #23 (43e5faa) into master (b63b2bd) will increase coverage by 0.02%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master      #23      +/-   ##
==========================================
+ Coverage   92.40%   92.43%   +0.02%     
==========================================
  Files          25       25              
  Lines        1475     1480       +5     
  Branches      211      230      +19     
==========================================
+ Hits         1363     1368       +5     
  Misses         63       63              
  Partials       49       49              

Impacted Files	Coverage Δ
...r_scripts/scripts/volume_to_precomputed_pyramid.py	89.58% <100.00%> (-0.22%)	⬇️
src/neuroglancer_scripts/volume_reader.py	83.33% <100.00%> (+0.64%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update b63b2bd...43e5faa. Read the comment docs.

[xgui3783]

Apologies for the spam. I noticed that there (still) were some error in the code (turns out, the byte order matters very much how ndarray.view collapses rgb nifti.

The PR should be ready for review now.

[xgui3783]

Apologies. There are still some issues that I am trying to sort out re this PR. I will mark ready for review & tag you when it's ready.

[ylep]

Hi @xgui3783, what is the status of this PR? Is it still relevant? If so, do you see any chance to get it ready before we release version 1.1.0?

[xgui3783]

@ylep , sorry, still relevant, but no progress.

It was brought to me attention that, since there are no unanimous format on how rgb is encoded in nifti, different softwares seems to use their own system to encode the rgb channel.

Perhaps I could close thie PR and add an issue instead.