Repository navigation
Delete CLUSTER_AUDIT_GPU_C.md (epic #2424 hard gate) - #2476
Merged
Merged
Conversation
brendancol
added a commit
that referenced
this pull request
May 27, 2026
…2440) Final pre-merge gate. The cluster-16 audit doc was a working artefact used to capture per-section moves, helper renames, and citation rewires while building the consolidation. Every prior cluster issue deleted its audit doc in a dedicated final commit before merge (#2474, #2476). Closing cluster follows the same convention.
brendancol
added a commit
that referenced
this pull request
May 27, 2026
…pus / tail (#2440) (#2480) * geotiff tests: consolidate closing cluster 16 — features / golden-corpus / tail (#2440) Final cluster of long-tail epic #2424. Folds 26 top-level test files into their natural homes and lands the final audit pass. Sub-PR A — features / tiers (4 files) → release_gates/test_features.py: - test_features.py (relocated + retained as the spine of the merged file) - test_supported_features_shape_2348.py (section header) - test_supported_features_tiers_2137.py (section header) - test_unsupported_features_2349.py (section header) - test_vrt_stable_only_2443.py (release-contract gate, folded too) Sub-PR B — golden-corpus relocation (13 files) → golden_corpus/: Moved with rename (drop _1930 suffix). Rewires the relative ._helpers import in test_http.py to ..\_helpers after the move. Sub-PR C — uncategorised tail (11 files) → individual homes per the issue's placements. test_remaining_fail_closed_1987.py lands in unit/test_metadata.py alongside the existing _1987 framework tests. Out-of-scope cleanup: five numbered files left at the top level by prior clusters folded in (test_inconsistent_geokeys_2417, test_overview_kernels_2413, test_invalid_int_nodata_rejection_2441, test_compound_crs_reject_2418), plus the three remaining numbered top-level files renamed to drop the suffix (test_fuzz_hypothesis_1661, test_namespace_no_leak_1708, test_polish_1488). The two near-duplicate round-trip modules consolidated into one test_round_trip.py. Release-gate citations in docs/source/reference/release_gate_geotiff.rst, docs/source/reference/geotiff.rst, golden_corpus_baselines.rst, the package docstring, and a stray code comment in _header.py updated to the new paths. The release-gate citation meta-gate now passes. Definition-of-done audit (epic #2424): - find xrspatial/geotiff/tests -name 'test_*.py' | wc -l → 72 (target 60–80) - find xrspatial/geotiff/tests -maxdepth 1 -name 'test_*.py' | wc -l → 5 (target ≤5) - filenames matching test_*_[0-9]{4,}.py → 0 - pytest xrspatial/geotiff/tests/ -q → green Tests-only. Closes #2425 Closes #2434 Closes #2438 Closes #2440 Closes #2424 * Address review: rename property-test helpers, fix stale citation (#2440) - test_round_trip.py: rename second-section _assert_fixed_point / _LOCKED_ATTRS to _assert_property_fixed_point / _PROPERTY_LOCKED_ATTRS so the invariant tests at lines 149-682 keep their original byte-equivalent tolerance (np.testing.assert_array_equal + pytest.approx 1e-15) and the Hypothesis property tests use their intended math.isclose tolerance. Both helpers now coexist instead of one shadowing the other on module load. - test_round_trip.py: drop the duplicate numpy/pytest/xarray imports in the Hypothesis section since they're already imported at the top. - read/test_georef.py: update stale citation pointing at the deleted test_remaining_fail_closed_1987.py to point at the new home in unit/test_metadata.py. - release_gates/test_features.py: add a docstring note that the release_gate markers are localised to the contract sections so pytest -m release_gate selects the right subset. * Delete CLUSTER_AUDIT_CLOSE.md (epic #2424 hard gate for cluster 16 / #2440) Final pre-merge gate. The cluster-16 audit doc was a working artefact used to capture per-section moves, helper renames, and citation rewires while building the consolidation. Every prior cluster issue deleted its audit doc in a dedicated final commit before merge (#2474, #2476). Closing cluster follows the same convention. * Fix CI: gate slow corpus fixtures in dask backends + fold #2449 test (#2440) Two follow-ups after the consolidation surfaced fresh issues: 1. CI fast-lane failure (ImportError: lerc) in ``test_dask_numpy_parity[compression_lerc_float32]``. The original top-level ``test_golden_corpus_dask_numpy_1930.py`` was never reached by the ``pytest xrspatial/geotiff/tests/golden_corpus/`` workflow target, so a latent bug in its ``_build_param`` (no call to ``fast_slow_marks_for``) was invisible. The Sub-PR B relocation moved the file into ``golden_corpus/`` per the issue spec, and the fast-lane CI lane started running it without the ``slow`` filter applying. Wire ``fast_slow_marks_for`` into the three dask / GPU backend modules (``test_dask_numpy``, ``test_dask_gpu``, ``test_gpu``) so the manifest's tagged-slow ``compression_*`` fixtures (including the optional-dep ``lerc`` cell) drop out of ``pytest -m "not slow"`` the same way they do in the eager backend module. Nightly / push-to-main lanes still exercise the full set. 2. The merge from origin/main pulled in ``test_file_source_context_2449.py`` (PR #2479's coverage for the new ``_FileSource`` context-manager protocol). That landed at the top level with a numeric suffix, pushing the Definition-of-Done counts off-target (top-level=6, suffixed=1). Folded into ``read/test_basic.py`` alongside the other input-source variants so the counts return to 5 top-level / 0 suffixed. Verified locally: - ``pytest xrspatial/geotiff/tests/golden_corpus/ -m "not slow" -q`` green with 30 deselected (the slow compression cells, including ``lerc``). - ``pytest xrspatial/geotiff/tests/golden_corpus/ -q`` green (full lane: 339 passed, 25 skipped). - ``pytest xrspatial/geotiff/tests/ -q`` green (5941 passed, 81 skipped, 2 xfailed).
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Follow-up cleanup for Sub-PR C of #2438. The original PR #2472 was
merged before the final pre-merge audit-delete commit could land.
CLUSTER_AUDIT_GPU_C.mdis a per-sub-PR review aid used during theconsolidation; it is not part of the test surface or the release-gate
registry. Epic #2424 requires the audit file to be deleted before any
sub-PR row is considered closed.
This PR removes only that single markdown file.
Refs #2438, #2424.