test: follow up FileSystemSeedReader coverage cleanup

[eric-tramel]

Summary

• consolidate the shared line-fanout FileSystemSeedReader test helper so engine and interface tests use the same implementation
• collapse duplicate engine-only hydration test readers and reuse a shared alpha/beta fixture to trim repeated setup
• add the missing follow-up coverage for None invalid returns and the full-output all-empty-fanout guard

Context

• follow-up to the non-blocking review comments on feat: support 1-to-many FileSystemSeedReader hydration #424

Testing

• uv run ruff check packages/data-designer-engine/src/data_designer/engine/testing/seed_readers.py packages/data-designer-engine/tests/engine/resources/test_seed_reader.py packages/data-designer/tests/interface/test_data_designer.py
• uv run pytest packages/data-designer-engine/tests/engine/resources/test_seed_reader.py packages/data-designer/tests/interface/test_data_designer.py

[greptile-apps]

Greptile Summary

This PR is a clean follow-up cleanup to #424 that consolidates duplicate FanoutDirectorySeedReader-style test helpers into a single, shared LineFanoutDirectorySeedReader living in data_designer.engine.testing, and merges the InvalidHydrationReturnSeedReader / SchemaMismatchFanoutSeedReader pair into a single ConfigurableHydrationDirectorySeedReader. It also adds two previously missing coverage cases: the None return path in test_filesystem_seed_reader_rejects_invalid_hydrate_row_returns and the new test_filesystem_seed_reader_full_output_raises_when_all_manifest_rows_fan_out_to_empty guard test.

• seed_readers.py is a new shared testing helper module; LineFanoutDirectorySeedReader is correctly re-exported from __init__.py, resolving the inconsistency raised in the prior review.
• Column ordering with include_file_name=True (relative_path → file_name → line_index → line) exactly matches the old FanoutDirectorySeedReader class attribute, so no behaviour change is introduced.
• include_file_name=False (default) preserves the FanoutCustomDirectorySeedReader column schema used in the interface tests.
• The new write_alpha_beta_text_files pytest fixture neatly reduces per-test file setup boilerplate across four fanout-related engine tests.
• All changes are test-scoped; no production code is modified.

Confidence Score: 5/5

• This PR is safe to merge — it is a pure test-layer refactoring with no production code changes.
• All changes are confined to test utilities and test files. The consolidated LineFanoutDirectorySeedReader correctly preserves both the include_file_name=True and include_file_name=False column schemas, and the ConfigurableHydrationDirectorySeedReader is a straightforward union of two previously separate reader stubs. The new None-return test case and the all-empty-fanout guard test exercise real code paths in the engine. No regressions are expected.
• No files require special attention.

Important Files Changed

Filename	Overview
packages/data-designer-engine/src/data_designer/engine/testing/seed_readers.py	New file introducing the shared LineFanoutDirectorySeedReader helper, consolidating two previously duplicate implementations from the engine and interface test files. Logic is correct; instance-level output_columns setup works at runtime.
packages/data-designer-engine/src/data_designer/engine/testing/init.py	Adds the LineFanoutDirectorySeedReader import and re-export to address the previous review comment about inconsistency with other test helpers.
packages/data-designer-engine/tests/engine/resources/test_seed_reader.py	Removes FanoutDirectorySeedReader, InvalidHydrationReturnSeedReader, and SchemaMismatchFanoutSeedReader in favour of the shared LineFanoutDirectorySeedReader and a new ConfigurableHydrationDirectorySeedReader. Adds write_alpha_beta_text_files fixture, None-return parametrize case, and the all-empty-fanout full-output guard test.
packages/data-designer/tests/interface/test_data_designer.py	Replaces the local FanoutCustomDirectorySeedReader with the canonical LineFanoutDirectorySeedReader from the testing package; no behavioural changes, column schema is identical.

Flowchart

%%{init: {'theme': 'neutral'}}%%
flowchart TD
    A["LineFanoutDirectorySeedReader\n(engine/testing/seed_readers.py)"]
    B["include_file_name=True\noutput_columns: relative_path, file_name, line_index, line"]
    C["include_file_name=False\noutput_columns: relative_path, line_index, line"]

    A --> B
    A --> C

    B --> D["test_seed_reader.py\n(engine tests)"]
    C --> E["test_data_designer.py\n(interface tests)"]

    F["ConfigurableHydrationDirectorySeedReader"]
    F --> G["hydrated_return=None → NoneType error"]
    F --> H["hydrated_return=scalar → scalar type error"]
    F --> I["hydrated_return=list with wrong schema → column mismatch error"]
    F --> D

Loading

_{Last reviewed commit: "test: trim FileSyste..."}