Re: Picking up multiple cherries at one go.

From: Alex Riesen <raa.lkml@gmail.com>
Date: Mar 24, 2010, 09:21 UTC
Message-ID: <81b0412b1003240221o53196568r9c8f2e0b7a2f1fd7@mail.gmail.com>
In-Reply-To: <ffef8f9a1003240147o777d058jfacd79aa86908925@mail.gmail.com>

On Wed, Mar 24, 2010 at 09:47, Jozef Babjak <jozef.babjak@gmail.com> wrote:

Show 8 quoted lines

> However, last days I needed to split a long multi-purpose branch into
> several topic branches. The branch was created as a result of 'general
> cleanup and refactoring', but at the end of day I realized that some
> well-defined parts can be identified there and so I wanted to extract
> them to dedicated branches. Thus, cherry-pick and rebase were my
> friend for couple of days. Everything goes well, but there was a
> repetitive task - cherry picking multiple commits. Sometimes the
> process failed, because cherry picking led to conflicts. ...

Try git rebase -i (interactive) next time. It does not exactly what your script does, but ... You may like it.

> To simplify this task, I wrote the shell script attached at the end of this
> email.The script named git_cherries.sh reads commit IDs from stdin and
> cherry-picks those commits to current branch. Everything is done on an
> auxilliary branch, ...

You can commit without a branch (called "detached HEAD"). Try "git checkout HEAD^0", for example. No need to invent new branch names (and risk collisions).