[jakduch]

What

• expose the snippet flag in partition table show, create, and update responses
• verify that creating and updating a partition table persists the flag and returns it through the API

Why

The API already accepts the snippet parameter, but omits it from partition table responses. API clients therefore cannot read or confirm the stored state.

Testing

• ruby -c test/controllers/api/v2/ptables_controller_test.rb
• git diff --check
• The focused Rails test requires the Foreman bundle, which is not installed in this clean worktree; it is left to CI.

AI assistance disclosure:

• Assisted by Codex 5.6 Sol High.

[jakduch]

Updated the API update test to clear os_family when converting the fixture to a snippet, since snippets require it to be blank. New commit: ca942075a.

[jakduch]

The CI failures are unrelated to the partition-table API change. Both JavaScript jobs hit the same time-dependent DateTimePicker test. The remaining failures are separate existing flakes: HostgroupJSTest timed out after all retries and Katello hit an unhandled VCR request in AlternateContentSourceRefreshTest; the Test suite failures are only the aggregate result. I am preparing a separate fix for the DateTimePicker test.

[jakduch]

The JavaScript matrix is failing on the shared DateTimePicker midnight-boundary test, unrelated to this PR. I fixed that deterministic failure in #11280. The Ruby/integration failures are separate upstream test failures and are being tracked independently.

[jakduch]

The matrix failures are the unrelated DateTimePicker future-only midnight-boundary failure fixed by #11280 (Redmine #39792). The Ruby 3.3 integration failure is the shared HostgroupJSTest timeout fixed by #11291 (Redmine #39803). The remaining Katello failure is an unrelated VCR cassette mismatch in AlternateContentSourceRefreshTest; this PR only exposes the partition-table snippet field in the API.

[jakduch]

The independent Katello VCR failure is now resolved upstream by Katello/katello#11867, merged as a4a5ed2. That change fixed the new Pulp binding request-body defaults that broke body_json cassette matching, and its Ruby 3.0/3.3 Katello suites passed. The remaining Foreman-side failures are covered by #11280 and #11291.

[lzap]

Great, thanks.

Can you also update the index action so the flag is exposed there?

[ofedoren]

I guess it won't hurt to add this to the hammer part as well: https://github.com/theforeman/hammer-cli-foreman/blob/master/lib/hammer_cli_foreman/partition_table.rb

[jakduch]

Thanks, snippet is now exposed by the shared partition table view, so it is included in both index and show responses. I added index coverage and rebased onto current develop. I will handle the Hammer exposure in a separate hammer-cli-foreman PR.

[jakduch]

The Hammer follow-up is ready in theforeman/hammer-cli-foreman#663. It exposes snippet in both partition table list and info output.

[jakduch]

The functional failure was in the new assertion: the index response wraps records in results, so the test was checking an entry from the metadata hash instead of a partition table. Fixed that and squashed the branch back to one commit.

[jakduch]

Updated: the snippet flag is now exposed from the shared main response used by the index action, with an index assertion in the controller test. The full CI is green.

[lzap]

LGTM