fix(sds): hide field options that come back as an empty list

[albert-tb]

What

An SDS field the tenant has not configured no longer shows up as a question. Viscosity - SDS Output, pH - SDS Output, and the other Section 9 fields that come back as a bare empty list are now hidden.

Why

Those entities are not in the TEN1 or TEN5 field configuration. The field-options endpoint returns [] for them, with no display flag. get_field_options stored every list as displayed, so Ask Albert rendered them as free-text inputs. EX-2811.

How

A bare empty list means the field is not configured, so display is false. A non-empty list is still the option list and stays displayed. An object that already says display: true is unchanged, including a free-text field whose data is empty.

Testing

• uv run ruff format src/albert/collections/sds.py tests/unit/collections/test_sds.py
• uv run ruff check src/albert/collections/sds.py tests/unit/collections/test_sds.py
• uv run pytest tests/unit/collections/test_sds.py -v

SDK Changes

sds.get_field_options reports display=False when the field-options response is an empty list.

[claude]

Code Review

The fix is correct: display=bool(payload) maps a bare empty list to display=False while preserving the non-empty and explicit-object cases. No bugs or correctness issues.

Two minor pattern notes (not blocking):

• src/albert/collections/sds.py:194 Stray run of spaces in the docstring (after flash_point) — collapses fine in rendered markdown but reads oddly in source; worth tidying.
• tests/unit/collections/test_sds.py OPINIONS.md says don't unit-test public collection methods (extract branching into a pure helper and test that). Here it's transport-faked (responses) and asserts on how the SDK reacts to each response shape, so it's defensible for a one-line branch, but it deviates from the stated convention.

Summary: 0 bugs found. Fix is correct and well-tested; only minor style/convention nits.