GoldenTests

GoldenTests, also known as golden file testing or snapshot testing, is a software testing technique in which a component's or function's output is compared against a stored baseline known as a golden file. The golden file serves as a reference representing the correct or expected result.

In a typical GoldenTests workflow, a test runs code to produce an output—such as a rendered user interface, an API response, or a generated file—and then loads the corresponding golden file. If the outputs match exactly, the test passes; if not, the test fails and highlights the differences. In some setups, the golden file can be automatically updated when a change is intentional and approved, effectively evolving the baseline.

GoldenTests are widely used for detecting regressions where visual appearance, text, or data structure changes would be undesirable. They are common in UI development for catching visual regressions, as well as in data transformation, code generation, and compiler output scenarios where outputs are expected to be stable and deterministic.

There are several forms of GoldenTests, including text-based golden files for textual outputs and image-based goldens for pixel-perfect UI verification. Some toolchains support binary or serialized object goldens as well.

Advantages of GoldenTests include rapid regression checks, a clear reference baseline, and support for safe refactoring when updates to goldens are properly reviewed. Limitations include brittleness with nondeterministic outputs, maintenance burden from large numbers of goldens, and the need for governance around when and how goldens are updated. Best practices emphasize deterministic outputs, isolating dependencies, storing goldens in version control, and using visual or textual diff tools to review changes.