feat: apply Time Grain to X-Axis column

[zhaoyongjie]

SUMMARY

Currently, the Time Grain control defines a couple of date truncate functions on the Time Column. Time Grain only be applied if the Time column and X-Axis are set to the same value. This PR is the first step to removing the entire time section.

After this PR, when enabling the GENERIC_CHART_AXES

1. Time Grain Control removes from Time Section.
2. Time Grain Control will show if i) applies a Date Column in the X-Axis. ii) applies a Adhoc Column
3. Time Grain Control will be displayed under X-Axis.

This change allows Time Column and Time Filter to use different columns, Time Series Chart will support different column to set Time Filter and X-Axis, and Time Grain will be applied to the X-Axis.

Supported Charts

• Mixed Timeseries Chart
• Area Chart
• Bar V2 Chart
• Line V2 Chart
• Scatter Chart
• SmoothLine Chart
• Step Chart

BEFORE/AFTER SCREENSHOTS OR ANIMATED GIF

SQL Pattern

Apply Categorical Column or Date Column to the X-Axis

Apply.Time.Grain.to.X-Axis.mov

Custom SQL

Custom.SQL.mov

TESTING INSTRUCTIONS

In order to test this PR, the multiple Date column should provide, so I prepare a CSV (superstore_orders.csv) file for that. Just load this CSV to test this PR.

1. Native filters should work as before
2. Time Grain should work as before
3. Time Grain should apply to X-Axis

ADDITIONAL INFORMATION

• Has associated issue:
• Required feature flags: GENERIC_CHART_AXES
• Changes UI
• Includes DB Migration (follow approval process in SIP-59)
  • Migration is atomic, supports rollback & is backwards-compatible
  • Confirm DB migration upgrade and downgrade tested
  • Runtime estimates and downtime expectations provided
• Introduces new feature or API
• Removes existing feature or API

[codecov]

Codecov Report

Merging #21163 (1de9492) into master (f71ee2e) will decrease coverage by 0.02%.
The diff coverage is 84.09%.

❗ Current head 1de9492 differs from pull request most recent head f6d31b3. Consider uploading reports for the commit f6d31b3 to get more accurate results

@@            Coverage Diff             @@
##           master   #21163      +/-   ##
==========================================
- Coverage   66.45%   66.42%   -0.03%     
==========================================
  Files        1789     1789              
  Lines       68291    68322      +31     
  Branches     7273     7273              
==========================================
+ Hits        45382    45384       +2     
- Misses      21034    21063      +29     
  Partials     1875     1875              

Flag	Coverage Δ
hive	53.04% <70.27%> (+0.02%)	⬆️
mysql	80.83% <81.08%> (+<0.01%)	⬆️
postgres	?
presto	52.94% <70.27%> (+0.02%)	⬆️
python	81.28% <81.08%> (-0.08%)	⬇️
sqlite	79.43% <81.08%> (+<0.01%)	⬆️
unit	50.74% <16.21%> (-0.04%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
...t-controls/src/sections/echartsTimeSeriesQuery.tsx	50.00% <ø> (ø)
...ackages/superset-ui-core/src/query/types/Column.ts	100.00% <ø> (ø)
...chart-echarts/src/MixedTimeseries/controlPanel.tsx	54.16% <ø> (ø)
...chart-echarts/src/Timeseries/Area/controlPanel.tsx	40.00% <ø> (ø)
...charts/src/Timeseries/Regular/Bar/controlPanel.tsx	31.25% <ø> (ø)
...harts/src/Timeseries/Regular/Line/controlPanel.tsx	33.33% <ø> (ø)
...ts/src/Timeseries/Regular/Scatter/controlPanel.tsx	40.00% <ø> (ø)
...src/Timeseries/Regular/SmoothLine/controlPanel.tsx	40.00% <ø> (ø)
...chart-echarts/src/Timeseries/Step/controlPanel.tsx	33.33% <ø> (ø)
...perset/utils/pandas_postprocessing/contribution.py	100.00% <ø> (ø)
... and 21 more

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[villebro]

Pretty cool stuff on the way here..! 👏 Left a few comments. Also, it's slightly difficult to give a good review in the absence of having seen designs or frontend POC code, so if there's anything like that out there I'd love to see it to get a more complete understanding of how this is meant to work eventually.

[zhaoyongjie]

Pretty cool stuff on the way here..! 👏 Left a few comments. Also, it's slightly difficult to give a good review in the absence of having seen designs or frontend POC code, so if there's anything like that out there I'd love to see it to get a more complete understanding of how this is meant to work eventually.

@villebro Thanks for the review! I will add a frontend part this week for convenience to review.

[michael-s-molina]

Thank you for PR @zhaoyongjie. I left some suggestions.

[zhaoyongjie]

@michael-s-molina Thanks for the review, I have addressed your comments.

[villebro]

Tested to work with the FF enabled and disabled with both legacy and new Timeseries charts. A few minor nits, but apart from that LGTM and hoping to see this merged soon! Also, we may want to consider adding a line in UPDATING.md to state that with the feature flag enabled, the time grain control will move below the x-axis control.

[michael-s-molina]

LGTM. Thanks for addressing the suggestions @zhaoyongjie. I agree with @villebro about adding something to UPDATING.md before merging this PR. Another suggestion would be to get approval from @jinghua-qa as well. Awesome work @zhaoyongjie!

[zhaoyongjie]

/testenv up FEATURE_GENERIC_CHART_AXES=true FEATURE_DRILL_TO_DETAIL=true FEATURE_DASHBOARD_CROSS_FILTERS=true

[github-actions]

@zhaoyongjie Ephemeral environment spinning up at http://35.91.72.50:8080. Credentials are admin/admin. Please allow several minutes for bootstrapping and startup.

[jinghua-qa]

Tested in ephemeral env, and the new feature LGTM.

However i have found 2 issues that also in master branch, i will file tickets to follow up.
1, default datetime is not autoselected as the x-axis when multiple column set is temporal

Screen.Recording.2022-09-07.at.1.07.16.AM.mov

2, when db=sqlite, and use custom sql in x-axis, time grain did not apply

Screen.Recording.2022-09-07.at.1.11.07.AM.mov

[zhaoyongjie]

@jinghua-qa Thanks for testing, I'm gonna follow up in a separate PR.

[github-actions]

Ephemeral environment shutdown and build artifacts deleted.

[john-bodley]

@zhaoyongjie would you mind expanding on this bullet point? From reading through the PR description et al. it seems that the biggest change is that the time grain is now decoupled from the time filter column. As it currently reads now "... will move below the X-Axis control." it sounds like it's purely a cosmetic change as opposed to a scoping change.

[zhaoyongjie]

@john-bodley update the document at PR