diff options
| author | Pjotr Prins <pjotr.public01@thebird.nl> | 2026-10-05 11:12:15 +0000 |
|---|---|---|
| committer | Pjotr Prins <pjotr.public01@thebird.nl> | 2026-10-07 08:26:56 +0200 |
| commit | 15a5c8abeb73584eeae6cae1d92176da42fa6bc3 (patch) | |
| tree | f5926ef740ef3e59b298cd99b32f093709c1e913 /test/python/test_version.py | |
| parent | 81662183f5cad305d7c011fc53b6bb750b3b0833 (diff) | |
| download | genenetwork4-15a5c8abeb73584eeae6cae1d92176da42fa6bc3.tar.gz | |
Add tests for denying a genuinely <1-year-old non-public dataset
All four 403-deny tests so far used UTHSC_BXD_Cochlea_0225, which was a real non-public, non-embargoed dataset when first used as a fixture but is now ~20 months old -- still correctly denied (under the 5-year embargo threshold), but no longer actually tests "freshly added, private" data specifically, only "not yet old enough to auto-expire." Found UTHSC_BXD_2wk_Cochlea_0926 (public=1, confidentiality=1, created 2026-09-01, the most recently added ProbeSetFreeze in the database as of this writing) and added a parallel test alongside each existing UTHSC_BXD_Cochlea_0225 test -- dataset_probeset_info, its /values sibling, the dataset_probesets listing, and the unified /traits endpoint, in both Rust and Python -- confirming the deny path holds for data that's actually less than a year old, not just data that merely hasn't aged past 5 years yet. Verified live against the real database before adding (is_publicly_readable's two conditions -- flags and age -- both still correctly deny for this dataset; values/probeset data exists for it, so the test exercises the authorization check rather than a 404). Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01XDDpXvZBQat9uSgDeEMJ8d
Diffstat (limited to 'test/python/test_version.py')
0 files changed, 0 insertions, 0 deletions
