Requre: Simplify your Integration&E2E tests
summary
Recording-based test harness called Requre records live API calls and function outputs to YAML files and replays them so tests run offline, without CI secrets, and without heavy mocking. Implemented in Python and used on the Augur project against GitHub GitLab Pagure Koji Copr and git libraries, it also archives files/directories, supports post-processing to scrub tokens/headers, and enforces consumption counts so tests fail if code calls more than recorded. Tradeoffs: large YAML blobs, frequent regeneration when external APIs or dependency outputs change, and library-specific capture plumbing.
AI-written summary · about
chapters (13)
- 0:00Intro
- 0:58Motivation Testing Packit project
- 3:32Benefits simplicity
- 5:55Benefits extension
- 6:59Cons | code stability
- 9:37Cons | complexity of libraries
- 11:54Cons | storage files/security
- 14:38basic storage
- 18:11file decorators
- 20:34request replacement
- 21:29python import wrapper
- 21:57very simple real life example
- 23:19Similar projects
related talks
46mRecord-Replay API Test Cases & Data Mocks without writing code
14mRecord-Replay API Test Cases with Data Mocks using Keploy
35mEvolution of a RESTful Testing Framework
16mAuto-Generating E2E Test-Cases and Mocks w/ Keploy
45mTest Automation Requirements Through Recordings
46mFrom Betamax to VCR: harden your API integration testing
36mAPI Testing Without Writing Test Cases and Mocks
39mHow to Generate Test-Cases and Data Mocks for Microservices at Kernel Using eBPF