Property-based testing library

A property-based testing library lets a test describe its inputs instead of listing them, and state what must hold for all of them. Each language has its own, so the one in use is found in the project rather than assumed.

Finding it

• The dependency manifest: the library among the test dependencies.
• Existing tests that use it: @given in Python, fc.assert in JavaScript, @Property in Java, proptest! in Rust.
• Its settings: the number of cases per property, and profiles for local runs and CI (Hypothesis's settings profiles, often in conftest.py).

It has no command of its own: property tests run with the project's test command.

Using it

• Reproduce a failure from its seed: each library prints the seed or the failing case, and Hypothesis also keeps failing examples in its .hypothesis folder.
• Keep the shrunk input as a fixed case (@example in Hypothesis, or a plain unit test), so it is checked on every run.
• More cases on a schedule than on every change (max_examples in Hypothesis, numRuns in fast-check).

Examples: Hypothesis (Python), fast-check (JavaScript), jqwik (Java), proptest (Rust), QuickCheck (Haskell).