This repo holds captures of broot instances in text format for comparison.
To run the test, execute ./test_screen_captures.sh
.
These tests have been only tested on one machine.
But the tests themselves are robust as long as the testing infrastructure stays the same.
(Testing infra staying same is no longer a requirement since the bug
referred here https://unix.stackexchange.com/a/698013 has been fixed in the latest version)
If the instrastructure changes, one glance at the diffs might be enough
Checkout and cat
master/initial.md
and other files inside `master.
- Do a
diff -side-by-side
on the capture and master corresponding to the failed case. - If everything looks alright, use vi to diff.
vi -d <PID>.BROOT.CAPTURES/<CASE>.capture master/<CASE>
. Install AnsiEsc from https://github.com/powerman/vim-plugin-AnsiEsc. And issue:AnsiEsc
- Estimate if this is a regression or expected.
- If expected, copy the
.capture
file to become the new.master
.
- Rename (run in powershell)
(Get-Item *.capture).name |% { Rename-Item $_ $_.replace( "capture", "master" ) }
- Manual diff
for i in ls master/*
do
f=$(basename $i)
diff /tmp/<PID OF TEST>.BROOT.CAPTURES/$f.capture $i
done