الاختبار المدفوع باللقطات والمعاينات

Some bugs are not crashes; they are layout slips you only see. Snapshot tests catch these by comparing images of your views. 👀

A snapshot test renders a view to an image once, saves it, then fails later if a new render no longer matches.

The first run records a trusted reference image. You review it by eye, and from then on it becomes the source of truth.

SwiftUI previews already show a view in many states. Designing those states first makes your snapshots easy to write.

Add previews for empty, loading, and error states. Each one becomes a snapshot, so regressions in any state are caught.

Render each view in light and dark mode. A color that vanishes in dark mode is exactly what a snapshot will reveal. 🌗

Snapshot a view at the largest Dynamic Type size too. Text that overflows or clips will fail the comparison immediately.

Always snapshot on a fixed device or frame. Different screen sizes produce different images and cause false failures.

When a snapshot fails, the tool shows old, new, and a diff. You decide: a real bug, or an intended change to re-record.

Store reference images in version control. Teammates then share the same baseline, and CI can catch visual changes in pull requests.

Changed the design intentionally? Switch to record mode once to refresh the baselines, then turn it off again.

What does a snapshot test actually compare?

You learned snapshot testing: record a reference, cover many preview states, test light and dark, and commit baselines so visual bugs never sneak in. 🎉