Quickfix batch: clone() arguments (#498), print_table keywords (#499), and 56 orphaned test files rejoined to CI (#504)

[lmoresi]

Fixes #498, fixes #499, fixes #504

Three small independent fixes, one commit each.

#498 — EnhancedMeshVariable.clone() drops its arguments. The enhanced wrapper declared clone(self) and forwarded nothing, while the base MeshVariable.clone requires (name, varsymbol) — so clone was uncallable in both directions and six shipped examples aborted on it. clone(name, varsymbol) now mirrors the base signature and returns an EnhancedMeshVariable (same mesh/shape/type/degree/continuity, independent data), so chained clones (v1 = v0.clone(...), as in Ex_Stokes_Cartesian_SolC) keep the enhanced type. New regression suite tests/test_0508_enhanced_variable_clone.py (4 tests) covers each documented argument, clone-of-clone, data independence and discontinuous spaces; all four failed with the reported TypeError: clone() takes 1 positional argument but 3 were given before the fix. Spot-run: Ex_Darcy_1D_benchmark.py now runs past its clone calls, both Darcy solves and rendering (it still trips much later on a pre-existing, unrelated matplotlib shape issue — evaluate output shaped (100,1,1) fed to plt.plot).

#499 — timing.print_table dropped display_fraction/group_by/output_file. The PETSc-log rewrite cut the signature to (filename, format); 18 examples still pass the legacy keywords and aborted with TypeError at their first timing report. output_file is restored as a working alias for filename (identical behaviour — the table is written to the file; passing both raises TypeError). The PETSc backend cannot honour display_fraction (no row culling) or group_by (events, not lines/routines), so those are accepted with a visible FutureWarning rather than raising or being silently swallowed — per the issue's own analysis that silent ignoring of behaviour-changing keywords would be worse than raising, while a hard error keeps the 18 examples broken. New regression suite tests/test_0055_timing_print_table.py (6 tests) covers the exact example call patterns (including the Kramer three-keyword form), that output_file really writes, the conflict case, and that the current (filename, format) interface is unchanged; four failed with the reported TypeError pre-fix.

#504 — test_1018_ in no scripts/test.sh batch, plus the orphan audit.* test_101*py tests/test_102*py now runs as a serial batch, so the rotated free-slip suite (19 tests, incl. the #497 guard) is live in CI; verified locally with the exact test.sh invocation: pytest --config-file=tests/pytest.ini tests/test_1018*py → 19 passed in 52 s.

One-off audit of every tests/test_*.py against the batch globs found 56 orphaned files (not counting the deliberately disabled test_0050* auditor and test_06xx regression batches). Added to batches after verifying each batch passes locally:

• test_005[1-9]* + test_006[0-1]* (mesh/vtk/deform/callback utilities, 43 tests, 4 s)
• test_02* (solver smoke/warm-start/wallclock, 27 tests, 16 s)
• test_101* + test_102* (stokes DMPlex/nullspace/MG, rotated free-slip, boundary flux, FMG lockout, MG bundle — 91 tests, ~2.5 min total)
• test_105* (VE/VEP, ddt, yield smoother/homotopy/anchor, solve report — 78 tests + the existing test_1050, 5.3 min)
• test_1120* + test_1450* (vector SL, vector projection, 6 tests, 17 s)
• named files: test_docstring_utils, test_jit_cache, test_jit_deterministic_ordering, test_multicomponent_projection, test_snes_vector_asymmetric_jacobian, test_symbol_disambiguation_prototype (74 passed, 1 skipped, 21 s)
  Also fixed in passing: line 77 read $PYTEST tests/test_1110*py # Annulus version || status=1 — the inline comment swallowed the || status=1, so annulus adv-diff failures could never fail the script. Same gap class as CI: the rotated free-slip suite (test_1018_*) is in no scripts/test.sh batch — its regression tests never run #504.

Left as listed findings (not added):

• test_106* (Nitsche/constrained free-slip, jacobian layout, Newton cold start) and test_107* (free surface plume / moving mesh / spherical): these carry level_2/level_3 + slow + tier_b/tier_c markers (test_1064 is level_3 + slow + tier_c), and a probe run of the 106x batch was still grinding after 20 min wall / 12 min CPU — not a "trivially safe" add. They need a triage/deselect decision before gating CI; noted with a comment in test.sh.
• tests/parallel/ has its own orphans: the parallel batch runs only tests/parallel/test_075*py, so test_0005_xdmf_viz_topology_mpi, test_0700_*, test_076x–test_079x, test_0855_*, test_1017_*, and test_106x_*_parallel there never run via test.sh. Verifying those needs mpirun time and belongs in its own pass.

Gate: pixi run -e amr-dev pytest tests -m "level_1 and tier_a" -q -p no:cacheprovider --ignore=tests/test_0050_utils.py → 590 passed, 0 failed, 17 skipped (need --with-mpi), 1 xfailed, in 7:21.

Underworld development team with AI support from Claude Code

[Copilot]

Pull request overview

This PR batches three small compatibility/CI fixes: it restores the intended EnhancedMeshVariable.clone(name, varsymbol) interface, reintroduces legacy uw.timing.print_table() keywords for example compatibility (with deprecation warnings), and updates scripts/test.sh so previously orphaned test modules are collected in CI.

Changes:

• Fix EnhancedMeshVariable.clone() to accept and apply (name, varsymbol) and return an enhanced variable instance with independent data.
• Extend uw.timing.print_table() to accept legacy keywords (display_fraction, group_by, output_file) while preserving the current (filename, format) interface.
• Expand scripts/test.sh batching globs to include previously unrun serial tests (notably test_1018* and other audited orphans), and fix an inline-comment bug that prevented failures from propagating.

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
tests/test_0508_enhanced_variable_clone.py	Adds regression coverage for the corrected enhanced-variable clone(name, varsymbol) behavior.
tests/test_0055_timing_print_table.py	Adds regression coverage for print_table() legacy keywords and conflict handling.
src/underworld3/timing.py	Restores legacy print_table() keyword acceptance via compatibility warnings and an output_file alias.
src/underworld3/discretisation/enhanced_variables.py	Implements the corrected EnhancedMeshVariable.clone(name, varsymbol) to return an enhanced clone.
scripts/test.sh	Rejoins orphaned serial test files to CI batches and fixes a comment-related `

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.