fix: memoized decorator memory leak

[dpgaspar]

SUMMARY

This is a followup/based on PR from: #17319
Fixes: #15132

Currently memoized data structure is not thread safe, can grow without bounds and is not safe on impure functions.

This PR uses python's lru_cache https://docs.python.org/3/library/functools.html#functools.lru_cache.
The goal is to make it clear how cache is implemented and we should use it only on functions that respect the following on python's docs:

In general, the LRU cache should only be used when you want to reuse previously computed values. Accordingly, it doesn’t make sense to cache functions with side-effects, functions that need to create distinct mutable objects on each call, or impure functions such as time() or random().

BEFORE/AFTER SCREENSHOTS OR ANIMATED GIF

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

[codecov]

Codecov Report

Merging #23139 (e961933) into master (eb8386e) will increase coverage by 1.57%.
The diff coverage is 60.59%.

❗ Current head e961933 differs from pull request most recent head 6ea9af6. Consider uploading reports for the commit 6ea9af6 to get more accurate results

@@            Coverage Diff             @@
##           master   #23139      +/-   ##
==========================================
+ Coverage   65.81%   67.39%   +1.57%     
==========================================
  Files        1881     1897      +16     
  Lines       72398    73030     +632     
  Branches     7882     7945      +63     
==========================================
+ Hits        47647    49216    +1569     
+ Misses      22729    21778     -951     
- Partials     2022     2036      +14     

Flag	Coverage Δ
hive	?
mysql	78.46% <79.87%> (+0.09%)	⬆️
postgres	78.53% <79.87%> (?)
presto	52.71% <56.70%> (?)
python	82.12% <79.87%> (+3.46%)	⬆️
sqlite	77.00% <79.67%> (+0.10%)	⬆️
unit	52.55% <56.09%> (?)

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

Impacted Files	Coverage Δ
...ackages/superset-ui-core/src/utils/featureFlags.ts	100.00% <ø> (ø)
...rontend/src/views/CRUD/allentities/AllEntities.tsx	0.00% <0.00%> (ø)
...nd/src/views/CRUD/allentities/AllEntitiesTable.tsx	0.00% <0.00%> (ø)
superset-frontend/src/views/CRUD/tags/TagCard.tsx	0.00% <0.00%> (ø)
superset-frontend/src/views/CRUD/tags/TagList.tsx	0.00% <0.00%> (ø)
superset/models/datasource_access_request.py	51.85% <ø> (-1.72%)	⬇️
superset-frontend/src/tags.ts	13.72% <13.72%> (ø)
superset-frontend/src/components/Tags/utils.tsx	21.05% <21.05%> (ø)
...d/src/explore/components/PropertiesModal/index.tsx	50.53% <25.80%> (-12.37%)	⬇️
superset-frontend/src/views/routes.tsx	51.11% <28.57%> (-4.16%)	⬇️
... and 135 more

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

[EugeneTorap]

Thanks for this PR!

[villebro]

Thanks for cleaning this up. This has been a big potential memory leak, so this should help guard against OOMs on big deployments. If this does introduce a perf regression then we'll at least know where we need additional caching.