Add unit tests for conductor redfish module

[berendt]

Closes #2225.

Adds tests/unit/tasks/conductor/test_redfish.py, covering every
function in osism/tasks/conductor/redfish.py. Part of Tier 3 (#2199).

Change set (single commit)

Add unit tests for conductor redfish module — a new test module that
mirrors the conventions already used in tests/unit/tasks/conductor/
(SPDX header, the shared loguru_logs fixture from tests/conftest.py,
a local _has_log helper, MagicMock/patch, section separators).

What is covered

• _normalize_redfish_data (pure function, built from inline dicts):
  None removal, dict/list → json.dumps, bool → "true"/
  "false", int/float → str(...), plain str pass-through, empty
  dict, and a mixed input combining all of the above.
• get_resources: parametrized delegation to _get_ethernet_interfaces,
  _get_network_adapters, _get_network_device_functions (asserting the
  other two helpers are not called), plus the unsupported-resource-type
  branch that returns [] and logs a warning.
• _get_ethernet_interfaces: no connection, top-level exception,
  system without the ethernet_interfaces attribute, ethernet_interfaces = None, two systems × two interfaces (4 normalized entries), full
  optional-attribute exposure, mac_address=None stripping, and
  per-interface exception isolation.
• _get_network_adapters: no connection, top-level exception, chassis
  without network_adapters, two normalized adapter entries, and
  per-adapter exception isolation.
• _get_network_device_functions: no connection, top-level exception,
  adapter missing network_device_functions (adapter-level except, then
  processing continues with the next adapter), per-device-function
  exception isolation, MAC extraction from device_func.ethernet, missing
  ethernet (MACs stripped by normalization), and adapter_id/
  adapter_name correlation on every entry.

Mocking notes

The Redfish connection chain (get_system_collection() /
get_chassis_collection() → get_members() → members) is faked with
MagicMock. del mock.attr drives the hasattr(...) is-False cases.

The per-item except branches log member.identity, so the failing
member must keep identity readable while a different attribute raises.
A MagicMock cannot raise on attribute read without leaking the
descriptor to every other mock, so a small explicit _RaisesOn shim is
used for those cases (its __getattr__ raises on one named attribute and
returns AttributeError — i.e. getattr(..., None) default — for the
rest). This is a deliberate, minor divergence from the issue's
"MagicMock is sufficient" hint; the observable behaviour the issue
specifies (item skipped, warning logged, others still returned) is
unchanged.

Verification

• black --check — clean.
• flake8 — clean.
• pytest / coverage: not run locally (project convention is to let the
  python-osism-unit-tests Zuul job run the suite on the PR). Line
  coverage of the target module was reviewed by hand and every line is
  exercised; please rely on the CI job for the 100 % figure.

🤖 Generated with Claude Code

[sourcery-ai]

Hey - I've left some high level feedback:

• Consider replacing the custom _RaisesOn shim with MagicMock plus side_effect/PropertyMock on the specific attributes that should raise, which would reduce bespoke test-only helpers while keeping the same behavior.

Prompt for AI Agents

Please address the comments from this code review:

## Overall Comments
- Consider replacing the custom `_RaisesOn` shim with `MagicMock` plus `side_effect`/`PropertyMock` on the specific attributes that should raise, which would reduce bespoke test-only helpers while keeping the same behavior.

---

Sourcery is free for open source - if you like our reviews please consider sharing them ✨

• X
• Mastodon
• LinkedIn
• Facebook

_{Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.}