[WIP] - Dimension handling

[Balinus]

This is a WIP for handling dimension in a more general way.

The idea is to first try to identify X, Y, Z and T dimensions using the axis attribute when present (optional in CF files). If not, the fallback is on the standard_name longitude, latitude and time (which is required in CF files).

Then, we should probably add units detection, etc.

I'll also add detection on long_names but this seems to be optional just like axis in CF files.

This is pretty rough right now, but it works with all my test files (CMIP, CORDEX and ERAINT files).

[codecov-io]

Codecov Report

Merging #137 into master will decrease coverage by 0.33%.
The diff coverage is 73.75%.

@@            Coverage Diff             @@
##           master     #137      +/-   ##
==========================================
- Coverage   88.94%   88.61%   -0.34%     
==========================================
  Files          11       12       +1     
  Lines        1475     1519      +44     
==========================================
+ Hits         1312     1346      +34     
- Misses        163      173      +10

Impacted Files	Coverage Δ
src/ClimateTools.jl	100% <ø> (ø)	⬆️
src/cf_conventions.jl	53.84% <53.84%> (ø)
src/extract.jl	81.44% <83.33%> (+1.63%)	⬆️

---

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 b3ac6f9...8e9e63c. Read the comment docs.

[Balinus]

ok, I think this is good to go. Now, detection is done by using axis, standard_name, long_name and units. Works on my test files but this would need another pair of eyes and some more testing.

[Balinus]

Just noticed it needs some love for 4D fields

[Balinus]

Decided to merge this PR as soon as possible due to some operational needs.