postgresql-stored-program-tests

Creating end-to-end tests for PostgreSQL stored programs involves analyzing function and procedure SQL files, docs, and schema definitions to produce Gauge specs and seed fixtures. The generated tests include cleanup steps and assertions on updated tables for deterministic validation.

Deterministic test fixtures for PostgreSQL functions use CSV files whenever possible for readability, or SQL files, placed under the fixtures directory. They seed required tables with test data and include cleanup steps to ensure consistent test execution.

Gauge specs can test both PostgreSQL procedures and functions by generating test scenarios under the specs directory. These specs leverage seed data and cleanup steps to validate stored program behavior deterministically.

PostgreSQL stored program testing applies to both functions and procedures, reading their SQL files and optional docs to infer test branches. It generates Gauge specs and fixtures tailored to each stored program's inputs and outputs.

Testing PostgreSQL stored programs end-to-end requires table schema definitions that the stored programs read or modify. The tool analyzes these schemas to generate seed data, expected fixtures, and table assertions for deterministic test validation.

CSV fixtures are chosen over SQL fixtures for PostgreSQL E2E testing to improve readability of seed and expected data. While both formats are supported, CSV files make deterministic test scenarios easier to review and maintain.