fix: ignore hidden directories during server discovery

[Copilot]

Summary

plugboard server discover could fail when the target directory contained a hidden virtualenv such as .venv, because discovery attempted to recurse into and import files from hidden paths. This change makes discovery skip hidden directories while preserving normal module discovery.

Changes

• Discovery traversal

  • Keep the recursive import walk based on os.walk.
  • Prune hidden directories in place via dirs[:] so discovery does not descend into .venv and similar paths.
  • Simplify module name construction using Path(...).relative_to(...).with_suffix("").

• Regression coverage

  • Consolidate the discovery coverage into a single parametrized CLI test.
  • Cover package and non-package projects both with and without hidden directories.
  • Verify that visible modules are still discovered and discovery completes without raising when hidden directories are present.

[toby-coleman]

Parametrize the existing test to avoid duplication

[Copilot]

Updated in fe09f85 by folding the duplicated coverage into the existing discovery test as a parametrized set of cases.

[toby-coleman]

Actually keep the inplace update to dirs as it is less code @copilot

[Copilot]

Updated in 58f2032 to restore the os.walk + dirs[:] pruning approach, while keeping the module-name handling a bit cleaner with Path.relative_to(...).with_suffix("").

[github-actions]

Benchmark comparison for 97996e58 (base) vs 58f20321 (PR)

------------------------------------------------------------------------------------------------------------------ benchmark: 2 tests -----------------------------------------------------------------------------------------------------------------
Name (time in ms)                                                                         Min                 Max                Mean            StdDev              Median               IQR            Outliers     OPS            Rounds  Iterations
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_benchmark_process_run (pr/.benchmarks/Linux-CPython-3.14-64bit/0001_pr)         420.9185 (1.0)      423.7608 (1.0)      422.0182 (1.0)      1.1293 (1.0)      421.4850 (1.0)      1.5078 (1.0)           1;0  2.3696 (1.0)           5           1
test_benchmark_process_run (main/.benchmarks/Linux-CPython-3.14-64bit/0001_base)     422.6454 (1.00)     434.5691 (1.03)     426.7634 (1.01)     4.7228 (4.18)     425.5294 (1.01)     5.7463 (3.81)          1;0  2.3432 (0.99)          5           1
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Legend:
  Outliers: 1 Standard Deviation from Mean; 1.5 IQR (InterQuartile Range) from 1st Quartile and 3rd Quartile.
  OPS: Operations Per Second, computed as 1 / Mean

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!