зеркало из https://github.com/mozilla/gecko-dev.git
0e4eda282f
One of the big downsides to |mach try fuzzy| is that if you use the interactive finder, there's no way to repeat your last action. If you want to run the same push again, you have to manually re-select all the same tasks a second time. It is possible to save presets, but this is fairly heavy-weight and (more) permanent. Sometimes you just want to re-run a push a few times and forget about it. It's also possible to craft the query on the command line with -q, but then you don't get the immediate visual feedback, so you can't be sure that you typed out the right things without actually pushing. With |mach try again|, everytime you generate a try_task_config.json via 'fuzzy', 'empty' or any other subcommands that may exist in the future, it'll get stored in a history file under ~/.mozbuild. Then running |mach try again| will simply re-run the most recent try_task_config.json. You'll also be able to view the whole history via |mach try again --list| and select a specific try_task_config.json (i.e not the most recent one) via |mach try again --index <index>|. Example usage will be: $ ./mach try fuzzy <select a bunch of tasks> $ ./mach try again <re-pushes exact same set of tasks> MozReview-Commit-ID: 3EZjVCy08uq Depends on D1808. Differential Revision: https://phabricator.services.mozilla.com/D1867 --HG-- extra : moz-landing-system : lando |
||
|---|---|---|
| .. | ||
| conftest.py | ||
| cram.ini | ||
| python.ini | ||
| setup.sh | ||
| test_again.py | ||
| test_empty.t | ||
| test_fuzzy.py | ||
| test_fuzzy.t | ||
| test_message.t | ||
| test_preset.t | ||
| test_templates.py | ||