[pavlushkin]

The GP CCMP population tests need practitioner and PractitionerRole launch context, but published @aehrc/questionnaire-test-toolkit@0.1.0 accepts only patient context. This change adds the missing inputs so the suite can run from the IG repository using the published toolkit.

Changes

• Add optional user and fhirContext wrapper props; forward them to population, expose the practitioner to renderer expressions, and repopulate when either input changes.
• Make selectTab leave an already expanded accordion open while preserving normal clicks for other tabs.
• Prepare toolkit version 0.1.1, keeping the existing renderer and SDC dependency baseline and dual ESM/CJS package exports.
• Add eight persistent regression tests against the built public package. These cover independent user/context rerenders, context isolation, tab behavior, validation, terminology URL configuration and extraction.
• Run the toolkit tests and typechecking in the shared Build-Test-Lint workflow on Node 24. The existing release workflow also calls this workflow.

The context rerender test keeps the same request-definition array across renders and changes only fhirContext. The fixtures have no external ValueSets, so these regression tests do not depend on Ontoserver.

Validation

• Toolkit regression suite: 8/8 passed across two files.
• Toolkit and regression-test typechecking: passed.
• Three separate temporary mutations were rejected by the corresponding tests: removing the user dependency, removing the fhirContext dependency, and removing the expanded-tab guard. The original build was restored and the normal suite passed again.
• ESLint on applicable added files, Prettier, workflow YAML validation and diff whitespace checks: passed.
• npm package dry run: nine files; ESM/CJS and both declaration formats remain present, and test sources/configuration are excluded.
• During the underlying context implementation, packed-consumer acceptance passed and the existing Aboriginal extraction suite passed 26/26. The new committed tests additionally correct and preserve the isolated context-rerender check.

Release dependency

This is the toolkit prerequisite for aehrc/gpccmp-fhir-ig#8. Merging this PR does not publish a package: a maintainer must release questionnaire-test-toolkit@0.1.1 through the existing release workflow. After publication, the IG lockfile must be regenerated from npm and the actual registry artifact, clean installation, 41-test suite and CI verified. The earlier packed candidate integrity must not be assumed to match the final published tarball.

[leoniedickson]

Thanks @pavlushkin - could we start a CHANGELOG.md for this package?

[pavlushkin]

Hi Leonie, Sure, I will all soon.

…

On Mon, 26 Oct 2026 at 15:23, leoniedickson ***@***.***> wrote: *leoniedickson* left a comment (aehrc/smart-forms#2193) <#2193 (comment)> Thanks @pavlushkin <https://github.com/pavlushkin> - could we start a CHANGELOG.md for this package? — Reply to this email directly, view it on GitHub <#2193?email_source=notifications&email_token=AHSGO26WJDWQ6H3CD24C43L5V3UX3A5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTMNBQHE3DKNJXGYYKM4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJLDGN5XXIZLSL5RWY2LDNM#issuecomment-6409655760>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AHSGO2Y6YLMI77UJOVLTZO35V3UX3AVCNFSNUABFKJSXA33TNF2G64TZHM2DONZUHA3DONRZHNEXG43VMU5TMMBTGM4TMMRXGI32C5QC> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[pavlushkin]

Follow-up toolkit PR (saved responses, comparison extraction, strict interaction guards, population callback seam, repeat helpers and Health Check acceptance) opened against this branch: beda-software#27. It will be retargeted to main after this PR merges.

[pavlushkin]

Follow-up toolkit PR (saved responses, comparison extraction, strict interaction guards, population callback seam, repeat helpers and Health Check acceptance): #2194 — built on top of this branch and will be rebased onto main after this PR merges.