Closed diabelb closed 2 years ago
Thanks for this PR. I've made a PR superseding this one (with an alternative implementation and tests for avoiding regressions).