| Message ID | 20191221113846.169538-2-dev+git@drbeat.li (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | None | expand |
On Sat, Dec 21, 2019 at 12:38:46PM +0100, Beat Bolli wrote: > After committing, I often want to return to the place of the latest > change to continue my work. Add the new mode "commit" which does exactly > this. That's one of my primary uses for git-jump, too. But you can already do that by jumping to the diff of HEAD^. Which has the additional advantage that it's a diff against the working tree. So if you did a partial commit, the diff will include any leftover changes. So I'm not opposed to this patch per se, given that it's not very many lines. But I'm not sure I see much advantage over "git jump diff HEAD^". It's slightly less typing, but I already alias "git jump diff" since it's so long. > Optional arguments are given to the "git show" call. So it's possible to > jump to changes of other commits than HEAD. This can also be done with "git jump diff $commit^ $commit". However, I've found that jumping based on older diffs is mostly useless, because the line numbers at $commit and those in the working tree don't always match up (and inherently you're always jumping in the working tree copy). -Peff
On 21.12.19 20:23, Jeff King wrote: > On Sat, Dec 21, 2019 at 12:38:46PM +0100, Beat Bolli wrote: > >> After committing, I often want to return to the place of the latest >> change to continue my work. Add the new mode "commit" which does exactly >> this. > > That's one of my primary uses for git-jump, too. But you can already do > that by jumping to the diff of HEAD^. Which has the additional advantage > that it's a diff against the working tree. So if you did a partial > commit, the diff will include any leftover changes. > > So I'm not opposed to this patch per se, given that it's not very many > lines. But I'm not sure I see much advantage over "git jump diff HEAD^". > It's slightly less typing, but I already alias "git jump diff" since > it's so long. > >> Optional arguments are given to the "git show" call. So it's possible to >> jump to changes of other commits than HEAD. > > This can also be done with "git jump diff $commit^ $commit". However, > I've found that jumping based on older diffs is mostly useless, because > the line numbers at $commit and those in the working tree don't always > match up (and inherently you're always jumping in the working tree > copy). Thanks, I didn't realize that git diff could just as well be used to generate the diff of an arbitrary commit. I have added this new shortcut to my bash aliases: gjc() { git jump diff ${1:-HEAD}^ ${1:-HEAD}; } Cheers, Beat
diff --git a/contrib/git-jump/git-jump b/contrib/git-jump/git-jump index 776fa90f7f..e7192073c6 100644 --- a/contrib/git-jump/git-jump +++ b/contrib/git-jump/git-jump @@ -9,6 +9,9 @@ The <mode> parameter is one of: diff: elements are diff hunks. Arguments are given to diff. +commit: element are the hunks of a commit (default HEAD). Arguments are + given to git show. + merge: elements are merge conflicts. Arguments are ignored. grep: elements are grep hits. Arguments are given to git grep or, if @@ -27,6 +30,10 @@ mode_diff() { git diff --no-prefix --relative "$@" | diff_to_quickfix } +mode_commit() { + git show --no-prefix --relative "$@" | diff_to_quickfix +} + diff_to_quickfix() { perl -ne ' if (m{^\+\+\+ (.*)}) { $file = $1; next }
After committing, I often want to return to the place of the latest change to continue my work. Add the new mode "commit" which does exactly this. Optional arguments are given to the "git show" call. So it's possible to jump to changes of other commits than HEAD. Signed-off-by: Beat Bolli <dev+git@drbeat.li> --- contrib/git-jump/git-jump | 7 +++++++ 1 file changed, 7 insertions(+)