v1.1.2

1.1.2 - 2026-05-23

• TAG: v1.1.2
• COVERAGE: 96.06% -- 195/203 lines in 11 files
• BRANCH COVERAGE: 80.00% -- 40/50 branches in 11 files
• 18.46% documented

Added

• CLEAN_RESULTSET / K_SOUP_COV_CLEAN_RESULTSET: deletes coverage/.resultset.json
  before SimpleCov starts to prevent stale entries from prior runs (e.g. after a refactor
  that shifts line/branch IDs) from producing phantom uncovered branches. Defaults to
  false on CI (each job already starts clean) and true locally (where developers
  re-run tests frequently). Override via K_SOUP_COV_CLEAN_RESULTSET=true/false or by
  setting the constant directly in Ruby. Kettle::Soup::Cover.clean_resultset! is also
  exposed as a public method for callers who want to trigger cleanup explicitly.
  Note: set K_SOUP_COV_CLEAN_RESULTSET=false inside .simplecov_spawn.rb so
  spawned subprocesses do not wipe the resultset being accumulated by the main process.
• Added turbo_tests:setup and turbo_tests:cleanup rake hooks for
  turbo_tests2 coverage collection.
• Added K_SOUP_COV_TURBO_TESTS and K_SOUP_COV_TURBO_TESTS_DIR support so
  parallel workers write coverage under per-worker directories and the cleanup
  hook can collate the resultsets.

Changed

• Updated documentation on hostile takeover of RubyGems
  • https://dev.to/galtzo/hostile-takeover-of-rubygems-my-thoughts-5hlo

Fixed

• Deferred hard coverage minimum enforcement for turbo_tests2 workers to the
  final collated coverage report, so individual workers with partial coverage do
  not fail before merged coverage is available.

Official Discord 👉️

Many paths lead to being a sponsor or a backer of this project. Are you on such a path?