Feature extract dialogue

[local-minimum]

This is fully integrated with existing API and replaces the functionality of the /feature_extract

Button only visible on completed scan-jobs. Though this doesn't necessarily mean you can do feature extraction as there has to have been a compilation and an analysis too. I think this quirk is fine for now, for this iteration.

[codecov-io]

Codecov Report

Merging #364 into master will increase coverage by 0.07%. The diff coverage is 96.15%.

@@            Coverage Diff             @@
##           master     #364      +/-   ##
==========================================
+ Coverage   40.29%   40.37%   +0.07%     
==========================================
  Files         195      198       +3     
  Lines       16729    16752      +23     
  Branches     2917     2919       +2     
==========================================
+ Hits         6741     6763      +22     
+ Misses       9573     9555      -18     
- Partials      415      434      +19

Impacted Files	Coverage Δ
...er_data/js/src/components/ScanningJobPanelBody.jsx	100% <ø> (ø)	:arrow_up:
scanomatic/ui_server_data/js/src/prop-types.js	100% <100%> (ø)	:arrow_up:
...server_data/js/src/components/ScanningJobPanel.jsx	100% <100%> (ø)	:arrow_up:
...c/components/ScanningJobFeatureExtractDialogue.jsx	100% <100%> (ø)
.../ui_server_data/js/src/components/ScanningRoot.jsx	100% <100%> (ø)	:arrow_up:
...r_data/js/src/containers/ScanningRootContainer.jsx	100% <100%> (ø)	:arrow_up:
scanomatic/ui_server_data/js/src/api.js	84.37% <100%> (+0.16%)	:arrow_up:
...er_data/js/src/containers/ScanningJobContainer.jsx	92% <92%> (ø)
scanomatic/ui_server/general.py	22.41% <0%> (ø)	:arrow_up:
scanomatic/data_processing/phases/features.py	17.23% <0%> (ø)	:arrow_up:
... and 9 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 5af537a...0a37275. Read the comment docs.

[local-minimum]

I think it is quite feasible to make system tests for this given some short cuts, specifically that we don't really want to check that a feature extraction gets started and is running. The latter would require an experiment directory and a full analysis directory in it. That fixture would need to be quite a few MB. To be able to run the system test though in any reasonable manner you have to prep the system with a completed experiment in the db and log in to the container and add an analysis folder ot that experiments' folder.

[gdetrez]

I think for this feature, it's enough to check that you get the "... is enqueued" message in the UI (and maybe the corresponding message on errors).