Our most recent change to the script (58309b02a9) broke it because the
anchored regex's no longer match the beginning of the subject. Fix this by
omitting the hash, which is a bit unfortunate but probably acceptable (I rarely
look at the output of the script anyway).
Git's --grep option uses Basic Regular Expression syntax by default, which means
that the `[^\n]*` didn't do what was intended; it means "any character except
`\` or the literal letter `n`" — not "any character except newline."
Besides, `^` matched any line start, not only the start of the entire message,
so "any character except newline" would have been wrong anyway. Given this, the
script matched commits that have WIP or DROPME in the body, which is not what we
want. (The last commit of this branch is an example for that.)
Fix this by listing only the subject lines and grepping them outside of git;
this also lets us use a slightly simpler regex (we want to match WIP anywhere in
the subject).
This script is failing currently on
https://github.com/jesseduffield/lazygit/pull/3631 because that fork's
master branch is 300 commits behind our own, but the feature branch is
up to date.
The thing is, we don't actually need to involve the master branch. All
we care about is the feature branch's own commits, so this commit simply
fetches those commits and checks them.
Also check for squash! and amend! (these are all anchored to the beginning of
the subject), and WIP and DROPME for good measure (but only if they occur in the
first line, otherwise it wouldn't let me merge this very commit :)
My local git is not compiled with PCRE support, so using -E makes it easier for
me to test the script locally. And -E is good enough for the simple matching we
want to do here.
Stefan Haller
Jesse Duffield