You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Now that I've set up Coveralls, it's easier to tell where the test coverage gaps are. Resyntax currently has about 94% coverage but there's several specific files that have much worse than that. Let's see how close to 100% we can get.
Current state as of master (4e4ffb5): 95.36% on Coveralls, 824 missed lines out of 17,755. The misses are concentrated in tool plumbing rather than refactoring rules — twelve files account for ~640 of the 824 lines, and most rule modules are already at 95–100%. Filed one issue per gap, each sized for a single small PR:
#815, #816, and #817 are mechanical (~170 lines of pure code and error branches). #818 is the biggest single win. #815–#819 together would put the project at roughly 97%. A realistic ceiling is around 98% — the long tail is defensive raises and debug-log branches with poor value per test.
Activity
jackfirth commentedon Jul 14, 2026
Current state as of master (4e4ffb5): 95.36% on Coveralls, 824 missed lines out of 17,755. The misses are concentrated in tool plumbing rather than refactoring rules — twelve files account for ~640 of the 824 lines, and most rule modules are already at 95–100%. Filed one issue per gap, each sized for a single small PR:
cli.rkt(run/output paths)main.rkt(mostly covered via #818/#819)grimoire/syntax-path.rkt(error branches)private/refactoring-result.rkt(pure functions)test/private/rackunit.rkt(failure reporting)private/github.rkt(env vars block testing)test.rkt(syntax error paths, part of #821)private/run-command.rktprivate/git.rkt#815, #816, and #817 are mechanical (~170 lines of pure code and error branches). #818 is the biggest single win. #815–#819 together would put the project at roughly 97%. A realistic ceiling is around 98% — the long tail is defensive raises and debug-log branches with poor value per test.