Database Testing with pytest

on YouTube ↗27m

" summary

Shows a reproducible pytest setup that tests real Microsoft SQL Server objects using pyodbc connections, fixtures placed in conftest.py, and transaction rollbacks to reset state. Demonstrates unit tests for tables, a view and a stored procedure (including a 4,000‑line SP), finds a miscalculated cost (test saw $100 instead of expected $200) and a bad join that returned 16 rows instead of 8, and fixes both so tests pass. Uses fixture scopes (scope='module') and yield for setup/teardown.

AI-written summary · about

" chapters (16)

1. 0:00Introduction
2. 0:39Accidental DBA journey
3. 2:11Unit testing basics
4. 3:24Why pytest
5. 4:47Testing approach & setup
6. 6:15Pytest fixtures
7. 7:43Writing test cases
8. 8:49Helpful pytest options
9. 10:21Building fixtures
10. 12:33Managing scope & files
11. 13:57Populating test data
12. 15:06Example test suite
13. 16:32Handling test failures
14. 17:41Verification & success
15. 20:14Pros and cons
16. 22:47Q&A and conclusion

" related talks

• 39mTest-Drive Your DatabaseAlastair Smith · TestCon Vilnius 2017
• 48mTesting the Database Layer: typical problems & best practicesRandy Syring · PyOhio 2015
• 37mYes, You Can Test SQLPaul Zuradzki · PyCon US 2026
• 19mEffortless Low-Code Data TestingTomas Tucek, Petr Pařík · SQLBits 2019
• 33mUnit Testing with PytestSofía Denner · PyCon US 2023
• 20mEffortless Low-Code Data TestingTomas Tucek, Petr Pařík · SQLBits 2019
• 19mEffortless Low-Code Data TestingTomas Tucek, Petr Pařík · SQLBits 2022
• 28mAdvanced Testing TechniquesAlan Braithwaite · GopherCon UK 2019