We're releasing Slipcover v0.1.8, with: - CPython 3.11 support; - tentative PyPy support, contributed by Carl Friedrich Bolz-Tereick; - a tweak in the missing lines output;

Coverage․py 6.4.3: a half-dozen small contributed fixes. Thanks everyone! pypi.org/project/covera… coverage.readthedocs.io/en/6.4.3/chang…

Coverage․py 6.4.4, with 3.11 wheels! pypi.org/project/covera… Now is the time to run your test suites on 3.11.0rc1. Python 3.11 is shipping in October, so make sure your code is ready for it!

Using GitHub actions to create a badge with your coverage total: nedbatchelder.com/blog/202209/ma…

Anyone want to help with coverage․py on Pyodide? github.com/nedbat/coverag…

This is the lowest-tech version of coverage measurement ever! 😅

Coverage․py 6.5.0: small changes, but packaging is slightly different. JSON report now has branch details. Let me know how it goes! coverage.readthedocs.io/en/6.5.0/chang… pypi.org/project/covera…

Looking for opinions: when combining data files, should “*/foo” match “foo/bar”? Should relative_files=True mean that we always match file paths by prefixes? github.com/nedbat/coverag…

today's video is about "code coverage" (and uses Coverage.py as an example and shows how to get started with it) -- check it out! youtu.be/eQM3TOjsM_Q

I don't use pytest-cov -- here's what I do instead -- youtu.be/sPgvHGkmd0U

“The idea that I can decide my open-source purpose feels a bit like a genie granting me a wish, and then I have to deal with the unanticipated plot twists that result.” nedbatchelder.com/blog/202210/id…

Coverage․py 6.6.0 beta 1: improvements to file patterns and relative_files that mean you MIGHT have to tweak your settings. Please test this pre-release version and let us know if something isn’t working! Install: pypi.org/project/covera… Changes: coverage.readthedocs.io/en/6.6.0b1/cha…

Should coverage․py report on excluded lines that were actually executed? How would you want that to work? github.com/nedbat/coverag…

Does PyCharm really use its own --omit switch on the command line? Should coverage․py combine the command line and the config? Or should the command line win like it does now? github.com/nedbat/coverag…

When running coverage, lines that should have run but didn't are called "unexecuted." What's a word for the opposite? Lines we said wouldn't run, but did: Sneaks? Surprises? Gremlins? Escapees? We need ideas!
github.com/nedbat/coverag… (also need opinions about how to report...)

New beta! Coverage․py 7.0.0b1 is out: pypi.org/project/covera… Many small improvements to path handling and combining data, plus markdown tables. Things might be broken! Let me know, please?

I’m posting on Mastodon these days, including this announcement of a beta of coverage․py 7.0.0: hachyderm.io/@nedbat/109451… Please try it and let me know if it works or if it doesn’t!

Coverage․py 7.0 is out, but you might need to update your settings! Let me know how it goes… (Also at hachyderm.io/@nedbat/109539…)

(From hachyderm.io/@nedbat/109761…) Have you wanted to measure coverage, but exclude certain data points because the lines ran "when it didn't count"? Help me figure out how to make that possible: github.com/nedbat/coverag…