fix(api): serialize empty read results as [] instead of null

[YO-WHATS-UP2]

fix(api): serialize empty read results as [] instead of null

Fixes #588

dbreads declared its result slices with var x []T, so a query returning zero rows
yielded a nil slice — which encoding/json marshals as null. spec.yaml declares
these fields type: array and marks none of them nullable, so the API contradicted
its own contract, and callers compensated with || [].

Initialise the slices as empty at the sites whose values reach a JSON response. Two of
them were explicit ExtraFiles = nil assignments, so "extra_files": null was emitted
for every package without extra files — the common case.

Also drops the workaround in GetGroupVersionBreakdown, which returned []string{}
where the schema expects versionBreakdownEntry objects, and only worked because empty
arrays serialize identically regardless of element type.

Testing done

The new test asserts the marshalled bytes equal [] rather than using assert.Empty,
which passes for a nil slice too. Confirmed failing on the previous code.

• Changelog entries added in the respective changelog/ directory (user-facing change, bug fix, security fix, update)
• Inspected CI output for image differences: /boot and /usr size, packages, list files for any missing binaries, kernel modules, config files, kernel modules, etc.

[Copilot]

Pull request overview

This PR fixes JSON serialization of empty “list” results in the backend API to consistently emit [] (empty array) rather than null, bringing responses in line with the OpenAPI schema (non-nullable arrays) and removing a handler-level workaround that returned the wrong element type.

Changes:

• Initialize result slices to non-nil empty slices ([]T{}) in dbreads query methods so zero-row queries marshal as [].
• Ensure package extra_files is emitted as [] (including the common “no extra files” case) by initializing/assigning empty slices rather than nil.
• Add a regression test that asserts marshalled JSON is exactly [], and update an existing activity test to assert non-nil empty results.

Reviewed changes

Copilot reviewed 11 out of 11 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
backend/pkg/handler/packages.go	Ensures ExtraFiles is always a non-nil slice when building package objects for JSON responses.
backend/pkg/handler/groups.go	Removes the incorrect-type workaround for empty version breakdown responses; relies on proper empty-slice behavior instead.
backend/pkg/api/internal/dbreads/packages.go	Returns non-nil empty slices for empty package reads and ensures ExtraFiles is set to [] when absent.
backend/pkg/api/internal/dbreads/packages_floors.go	Returns a non-nil empty slice when a package has no floor channels.
backend/pkg/api/internal/dbreads/instances.go	Returns a non-nil empty slice for empty instance status history and instance list results.
backend/pkg/api/internal/dbreads/groups.go	Returns a non-nil empty slice for empty group lists and group version breakdown results.
backend/pkg/api/internal/dbreads/channels.go	Returns a non-nil empty slice for empty channel lists.
backend/pkg/api/internal/dbreads/applications.go	Returns a non-nil empty slice for empty application lists.
backend/pkg/api/internal/dbreads/activity.go	Returns a non-nil empty slice for empty activity results.
backend/pkg/api/empty_slice_json_test.go	Adds regression coverage asserting JSON marshals to [] (not null) for several empty-result endpoints/fields.
backend/pkg/api/activity_test.go	Updates expectations for non-existent teams to return an empty (but non-nil) slice.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 11 out of 11 changed files in this pull request and generated no new comments.

[Copilot]

Pull request overview

Copilot reviewed 11 out of 11 changed files in this pull request and generated no new comments.

[ervcz]

Sorry for the slow reply, and thanks for the contribution.

Nebraska is one of the projects proposed for CNCF/LFX mentorship this term, and since that went up we've had a lot of attention here at once. The queue grew much faster than we can review it, so replies are slower than usual. Your PR hasn't been forgotten and it will get a proper look.

Thanks for your patience.

[Copilot]

Pull request overview

Copilot reviewed 11 out of 11 changed files in this pull request and generated no new comments.

[Copilot]

Pull request overview

Copilot reviewed 11 out of 11 changed files in this pull request and generated no new comments.