Snapshot tests

In short: A snapshot test stores the output of a piece of the program once as a reference and compares it with the new result on later runs.

In more detail: Instead of checking every value individually, for example the rendered HTML of a component is saved as a file. If the output changes, the test fails and you decide: unintended bug or intended change (then update the snapshot).

In Depth

• Good for: Large, stable outputs such as error messages, serialized data or reports.
• Risk: Huge snapshots get confirmed unread during updates and lose their value.
• Keep small: Capture targeted excerpts rather than whole pages.
• Unstable values: Replace timestamps and random IDs with fixed placeholders beforehand.
• Visual variant: Screenshot comparison detects layout bugs but is sensitive to fonts and rendering differences.

See also: Property-based testing