Repository navigation
scripts/test.sh: eleven test files match no glob, so the integration-point and swarm suites have never run in CI #721
Description
What we found
scripts/test.sh selects serial tests with numeric batch globs. One of them,
$PYTEST tests/test_005[1-9]*py tests/test_006[0-1]*py || status=1
stopped covering its own band as soon as test_0062 was added. Eleven test
files currently match no glob in the script and have never run in CI:
tests/test_0062_rotated_constraint_exclusivity.py
tests/test_0063_mpi_hang_supervisor.py
tests/test_0064_quadrature_point_fe.py <- #703
tests/test_0065_integration_point_variable.py <- #703
tests/test_0066_integration_point_slcn.py <- #703
tests/test_0067_integration_point_proxy.py <- #707
tests/test_0068_swarm_repopulation.py <- #713
tests/test_0069_swarm_midtime_velocity.py <- #713
tests/test_0070_ve_stress_history_on_particles.py <- #713
tests/test_0301_meshvariable_clone.py
tests/test_1030_analytic_zhong2008.py
That is the entire integration-point suite delivered by #703 and #707, and the
repopulation / mid-time-velocity / VE-stress-on-particles tests delivered by
#713. Each of those PRs shipped green.
This is separate from test_106*/test_107*, which are deferred on purpose
and documented in the script against #504.
Why it is the shape of the problem, not one glob
A numeric range has to be widened by hand every time the band grows, and
nothing says when it wasn't. This is the #570 class. Widening
test_006[0-1] to test_00[6-9] fixes today's instance and leaves the next
one to chance.
Fix
Widen the stale globs, and add scripts/check_test_coverage.py, which fails
the run if any tests/test_*.py matches no glob. Files deferred by decision
go in a DEFERRED list carrying the reason and issue, on the same terms as
the allowlist in check_deprecated_patterns.py: it may only shrink.
Underworld development team with AI support from Claude Code
Activity
Nitsche free-slip tests: a wiring check instead of an accuracy claim,…
Nitsche free-slip tests: a wiring check, a real assert, and the compa…
A guard for the test globs, and 22 session scripts out of the test tr…
Review (#723): the guard listed a covered file as deferred, and never…
Merge pull request #723 from underworldcode/docs/audit-2026-09
lmoresi commentedon Oct 5, 2026
Already fixed. Probed against
development(1391e4aa):The checker that this issue motivated now enforces it, so a file falling outside the globs fails rather than going dark. Closed under the issue-adjudication rule in #819.
Merge development: the two conflicts are the branch's own additions
Merge development: the two conflicts are the branch's own additions
Merge development: the two conflicts are the branch's own additions
Merge development: two mechanical conflicts and one rename
Merge development into the stack root
lmoresi commentedon Oct 12, 2026
Reopened and labelled
fixed-in-development, under the policy now in CLAUDE.md:The fix stands and the evidence above is unchanged — this is a bookkeeping correction, not a reversal. Closing it on merge was wrong because
developmenthas not been released, so anyone running the release still has the defect; the label says where the fix is and the release sweep closes it.CI reaches the SUPG predictor-corrector tests