ENH: Add cache to SpatialReferences

[mgxd]

• adds cached property to allow checkpointing spaces.
  • this can be leveraged to differentiate user requested spaces
• make SpatialReferences standard spaces class attribute private

[pull-assistant]

Best reviewed: commit by commit

---

Optimal code review plan

     enh: add cache for spaces

     Apply suggestions from code review

     Update niworkflows/utils/spaces.py

Powered by Pull Assistant. Last update 8ea3df6 ... d8c584c. Read the comment docs.

[mgxd]

I'll leave this here @oesteban

[codecov]

Codecov Report

Merging #453 into master will increase coverage by 11.4%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master     #453      +/-   ##
==========================================
+ Coverage   45.98%   57.38%   +11.4%     
==========================================
  Files          61       61              
  Lines        6048     6113      +65     
  Branches        0      710     +710     
==========================================
+ Hits         2781     3508     +727     
+ Misses       3267     2547     -720     
- Partials        0       58      +58

Flag	Coverage Δ
#reportlettests	100% <ø> (?)
#unittests	46.06% <100%> (+0.08%)	⬆️

Impacted Files	Coverage Δ
niworkflows/utils/spaces.py	94.3% <100%> (+0.34%)	⬆️
niworkflows/utils/misc.py	80% <0%> (+0.2%)	⬆️
niworkflows/interfaces/registration.py	52.15% <0%> (+3.44%)	⬆️
niworkflows/func/util.py	28% <0%> (+4%)	⬆️
niworkflows/interfaces/plotting.py	84.74% <0%> (+10.16%)	⬆️
niworkflows/interfaces/mni.py	42.28% <0%> (+16.57%)	⬆️
niworkflows/interfaces/masks.py	62.87% <0%> (+18.56%)	⬆️
niworkflows/interfaces/report_base.py	68.85% <0%> (+21.31%)	⬆️
niworkflows/viz/plots.py	46.31% <0%> (+39.69%)	⬆️
niworkflows/viz/utils.py	74.09% <0%> (+64.5%)	⬆️
... and 1 more

---

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 0421e6f...d8c584c. Read the comment docs.

[oesteban]

I wonder whether we should accept several checkpoints, instead of raising error. (and provide access to earlier timepoints).