tests.test_manager_yarn module

Yarn Classic query tests.

These tests replay captured yarn --silent --json streams through YarnClassic.search and YarnClassic.outdated. They do not invoke yarn.

tests.test_manager_yarn.finished(status)

The --verbose record yarn prints when the registry answers.

Return type:

str

tests.test_manager_yarn.yarn(monkeypatch)

tests.test_manager_yarn.replay(monkeypatch, manager, stdout, stderr)

Stand in for run_cli, recording the run the way run() does.

Return type:

None

tests.test_manager_yarn.spawn(monkeypatch, manager, code, stdout, stderr)

Stand in for the child process, so the failure gate of run() runs.

Return type:

None

tests.test_manager_yarn.test_outdated_reads_exit_1_as_updates(yarn, monkeypatch, caplog)

Yarn exits 1 when updates exist, its warnings on <stderr>.

tests.test_manager_yarn.test_outdated_registry_failure(yarn, monkeypatch, caplog)

A failed registry request still fails, relaying only its error record.

tests.test_manager_yarn.test_search_found(yarn, monkeypatch)

tests.test_manager_yarn.test_search_missing_package(yarn, monkeypatch, caplog, status)

Yarn resolves these answers to “no such package”: an empty result.

tests.test_manager_yarn.test_search_registry_failure(yarn, monkeypatch, caplog, stdout, diagnosis)

Any other answer, or none, is a failure yarn hid behind a zero exit.

tests.test_manager_yarn.test_search_ignores_an_earlier_run(yarn, stub_run_cli)

A failure recorded by an earlier call never leaks into this one.