Aglet

Verify null handling in an API client

Null handling is verified by preserving meaning across the mapper, storage, and UI. Exercise every representation the operation allows, then check a consumer that acts on the field. Ensure unknown or filtered data does not become an invented success, identity, or permission state.

Check whether the outcome improved

  1. Define representation outcomes

    Write expected decoded, stored, displayed, and downstream states for value, null, absent, empty, filtered, and malformed fields. State which cases may replace existing data and which must remain pending or visible as unknown.

  2. Run mapper and storage checks

    Feed each fixture through the real client and compare raw presence, parsed value, default branch, persistence, and rendering. Confirm explicit null stays distinguishable from absence where the contract requires it. Inspect a resource update, not only an initial read.

  3. Exercise adjacent consumers

    Use a report, eligibility check, or workflow that reads the field. Repeat null and absent cases after refresh and process restart. Verify a permission-filtered value does not widen access or erase known-good state during a retry.

What to carry forward

Accept when representation differences remain intentional through decode, storage, and consumer behavior, with no silent privilege or state change. Keep verification partial for undocumented filtered responses. Record fixtures, field contract, client revision, and untested consumers.

Technical background: Project documentation on GitHub.

Keep the decision with the work.

Use a Work Item in Aglet to record the problem, the evidence you have, and the next decision. Add an owner and priority, then keep updates in the discussion so the next person can follow the reasoning.

Create an account See the product workflow