Fixing sort issue with area chart and adding tests

[michellethomas]

There's an issue where the area (stacked line) chart doesn't sort correctly. At the end of get_data for NVD3 charts the data gets sorted. I asked Beto about this line added here and it looks like it's necessary for time shift coloring. I believe the correct way to handle this is to see if the chart sorts the series (with self.sort_series flag), but I'm not sure if I'm missing any edge cases.

I've added tests to make sure the area chart is sorted correctly, and make sure the line chart with time shift has the correct colors.

@betodealmeida @john-bodley @graceguo-supercat

[codecov-io]

Codecov Report

Merging #6358 into master will increase coverage by <.01%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master    #6358      +/-   ##
==========================================
+ Coverage   56.07%   56.07%   +<.01%     
==========================================
  Files         525      525              
  Lines       23212    23214       +2     
  Branches     2776     2776              
==========================================
+ Hits        13016    13018       +2     
  Misses       9787     9787              
  Partials      409      409

Impacted Files	Coverage Δ
superset/viz.py	71.69% <100%> (+0.03%)	⬆️

---

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 3ae7d32...2f041b6. Read the comment docs.

[michellethomas]

This PR is a bit old but still relevant, so I rebased and will merge it in. I don't think the line chart_data = sorted(chart_data, key=lambda x: tuple(x['key'])) is necessary anymore for determining the color for time shift lines because we have code that assigns a color based on the label, but I'm leaving it in because it keeps the legend sorted. The self.sort_series happens before this line so we don't want to be resorting in that case.

[kristw]

LGTM