- 23 Jun, 2022 2 commits
-
-
Antoine Beaupré authored
-
Antoine Beaupré authored
-
- 09 Apr, 2022 1 commit
-
-
Antoine Beaupré authored
-
- 30 Oct, 2021 2 commits
-
-
Antoine Beaupré authored
-
Antoine Beaupré authored
-
- 22 Oct, 2021 1 commit
-
-
Antoine Beaupré authored
Some of this comes from the research and tests I've done in feed2exec.
-
- 21 Sep, 2021 4 commits
-
-
Antoine Beaupré authored
-
Antoine Beaupré authored
-
Antoine Beaupré authored
This was surprisingly hard, but I think I got it about right. The actions do get fired twice, however, which is not great. In particular, this will be slow on large config files. It might be possible to avoid parsing the config file twice by disabling the config action somehow when it's first parsed, but that seems like a needless optimization for now. This should actually fix undertime#19
-
Antoine Beaupré authored
-
- 30 Aug, 2021 1 commit
-
-
Antoine Beaupré authored
-
- 27 Apr, 2021 1 commit
-
-
Antoine Beaupré authored
-
- 23 Apr, 2021 3 commits
-
-
Antoine Beaupré authored
With this, it's possible to check if a flag is set, e.g. do: if args.verbose: pass Before this, the parameter was always set to None. Now it's set to the `const` set by the caller or whatever value is passed as an argument (e.g. for a --loglevel FOO argument).
-
Antoine Beaupré authored
-
Antoine Beaupré authored
-
- 09 Feb, 2021 1 commit
-
-
Antoine Beaupré authored
-
- 17 Nov, 2020 2 commits
-
-
Antoine Beaupré authored
-
Antoine Beaupré authored
-
- 26 Oct, 2020 6 commits
-
-
Antoine Beaupré authored
It's unclear where those errors come from: E _pytest.pathlib.ImportPathMismatchError: ('ecdysis._version', '/home/anarcat/src/ecdysis/.tox/py/lib/python3.7/site-packages/ecdysis/_version.py', PosixPath('/home/anarcat/src/ecdysis/ecdysis/_version.py')) They do not occur in feed2exec or other projects I have, and were not happening before the "matrix" switch. It might be a relative import thing, but I don't have time to diagnose this so for now just use the workaround found here: https://stackoverflow.com/questions/44067609/getting-error-importmismatcherror-while-running-py-test
-
Antoine Beaupré authored
-
Antoine Beaupré authored
-
Antoine Beaupré authored
-
Antoine Beaupré authored
-
Antoine Beaupré authored
-
- 19 Oct, 2020 1 commit
-
-
Antoine Beaupré authored
-
- 25 May, 2020 1 commit
-
-
Antoine Beaupré authored
-
- 25 Apr, 2020 2 commits
-
-
Antoine Beaupré authored
-
Antoine Beaupré authored
-
- 21 Apr, 2020 9 commits
-
-
Antoine Beaupré authored
The rationale behind this is that we want minimal entrypoints, and they way to do that is to tell pip and other deployment tools that we want the "wheel" dist, which is more clever about that. I am not sure yet how this relates to setup.py.
-
Antoine Beaupré authored
-
Antoine Beaupré authored
-
Antoine Beaupré authored
-
Antoine Beaupré authored
-
Antoine Beaupré authored
fix a typo in CONTRIBUTING.rst See merge request !1
-
Antoine Beaupré authored
-
Antoine Beaupré authored
-
Antoine Beaupré authored
-
- 20 Apr, 2020 1 commit
-
-
Antoine Beaupré authored
-
- 09 Mar, 2020 1 commit
-
-
Antoine Beaupré authored
The previous configuration showed usage like this: stressant [--disk] ... Which seems to imply that, by default, no disk job is run. But the usage actually said: --disk, --no-disk run disk tests (default: True) which is actually correct, but ambiguous. Now instead, we show this in the short usage: stressant [--no-disk] ... And this in the long version: --no-disk, --disk run disk tests (default: True) Which is much clearer. Actual commandline parsing is not affected, only usage output.
-
- 08 Jan, 2020 1 commit
-
-
Antoine Beaupré authored
-