feat: add MotherDuck DB engine spec

[betodealmeida]

SUMMARY

Add a separate DB engine spec for MotherDuck, a cloud version of DuckDB. While adding the spec I did a few QoL improvements:

• Fixed the "Refer to the SQLAlchemy docs for more information on how to structure your URI." text, which said only "Refer to the for more information on how to structure your URI."
• Display the sqlalchemy_uri_placeholder as the placeholder for the SQLAlchemy URI.
• Add the optional duckdb dependency to Superset, with the library needed to run DuckDB and connect to MotherDuck.

BEFORE/AFTER SCREENSHOTS OR ANIMATED GIF

Before, using the DuckDB driver, note that we show the generic placeholder and that the text is missing the link to the docs:

After, we show the correct placeholder, and the text is fixed:

TESTING INSTRUCTIONS

ADDITIONAL INFORMATION

• Has associated issue:
• Required feature flags:
• 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

[betodealmeida]

SupersetText is either a module or {}, and both are truthy.

superset/superset-frontend/src/utils/textUtils.ts

Lines 18 to 25 in ec9e9a4

	const loadModule = () => {
	try {
	// eslint-disable-next-line global-require, import/no-unresolved
	return require('../../../superset_text') || {};
	} catch (e) {
	return {};
	}
	};

[betodealmeida]

Moving this from the "dynamic form" mixin to the base DB engine spec.

[codecov]

Codecov Report

Merging #24934 (23af5eb) into master (ec9e9a4) will increase coverage by 0.00%.
Report is 4 commits behind head on master.
The diff coverage is 73.43%.

❗ Current head 23af5eb differs from pull request most recent head de7d069. Consider uploading reports for the commit de7d069 to get more accurate results

@@           Coverage Diff           @@
##           master   #24934   +/-   ##
=======================================
  Coverage   69.00%   69.00%           
=======================================
  Files        1906     1906           
  Lines       74133    74149   +16     
  Branches     8208     8211    +3     
=======================================
+ Hits        51153    51169   +16     
+ Misses      20857    20856    -1     
- Partials     2123     2124    +1     

Flag	Coverage Δ
hive	54.17% <24.13%> (-0.01%)	⬇️
javascript	55.84% <100.00%> (+<0.01%)	⬆️
mysql	79.21% <70.68%> (+<0.01%)	⬆️
postgres	79.32% <70.68%> (+<0.01%)	⬆️
presto	54.07% <24.13%> (-0.01%)	⬇️
python	83.37% <70.68%> (+<0.01%)	⬆️
sqlite	77.89% <70.68%> (+<0.01%)	⬆️
unit	55.05% <24.13%> (-0.01%)	⬇️

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

Files Changed	Coverage Δ
...rset-frontend/src/components/ErrorMessage/types.ts	100.00% <ø> (ø)
superset-frontend/src/dashboard/actions/hydrate.js	2.04% <ø> (ø)
...et-frontend/src/dashboard/components/SaveModal.tsx	39.28% <ø> (ø)
superset/cli/native_filters.py	0.00% <0.00%> (ø)
superset/databases/api.py	89.95% <ø> (ø)
...uperset/utils/dashboard_filter_scopes_converter.py	15.38% <0.00%> (-0.41%)	⬇️
superset/security/manager.py	94.43% <94.44%> (ø)
...set-frontend/src/dashboard/util/permissionUtils.ts	88.23% <100.00%> (+1.56%)	⬆️
...eatures/databases/DatabaseModal/SqlAlchemyForm.tsx	63.63% <100.00%> (+2.09%)	⬆️
...end/src/features/databases/DatabaseModal/index.tsx	43.66% <100.00%> (ø)
... and 7 more

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

[john-bodley]

@betodealmeida should this be >=0.8.0,<1.0.0?

[betodealmeida]

@john-bodley could be, but in general I prefer to pin 0.x versions instead of using a range, since in theory the API is not sable and 0.8.1 could introduce breaking changes.

[rusackas]

On that note, 0.8.1 is out, and seems to follow semver
https://github.com/duckdb/duckdb/releases/

[rusackas]

I could go either way on this... trying to rationalize a preference.

If we pin the version, and there's an issue (potentially security) then we might have a fire drill on our hands.

If we use a range, and something breaks, we might have a bug on our hands.

By the "lesser evil" metric, I think this might put the range idea ahead in my mind by about the width of a frog hair, as long as they stick to semver ;)

[betodealmeida]

Yeah, I know they were planning to change how databases are returned by their dialect, so maybe we keep this frozen until they hit 1.0?

Also, this is an optional dependency, people can always install a newer version if they want.

[rusackas]

Not sure if we want to pin or not... and not sure if we want to bump the pinned version or not. But neither issue will keep me up at night and the rest looks good to me!