问题
I have a repo in which I want to push changes. I want to sync two repositories: ./pages.git into ./pages. My git setup on my development machine pushes into pages.git. But the web-application works with ./pages.
git clone pages.git pages
fatal: destination path 'pages' already exists and is not an empty directory.
Well now I delete the pages directory and git clone works. But that's not the clean solution, isn't it?
Is there any way to do that without having to delete the pages directory manually? I'd like to automate that process so that git automatically performs the actions whenever I push.
Best, Marius
回答1:
Why not just create a post-receive hook for the pages.git
repository that will do a git pull
from within the pages
repository to pull over the changes? git clone
is designed to be run only once per resulting clone (hence why it complains until you delete the previous clone), whereas git pull
is designed to be run repeatedly whenever you want to update a repo with changes from another.
I have a hook setup for something essentially the same as this with a particular repository, here's the gist of it:
cd /var/repos/path_to_mirror || exit
unset GIT_DIR
git pull origin master
- In this case,
path_to_mirror
is the equivalent ofpages
in your situation. - This script would be in the
pages.git/.git/hooks/post-receive
file.
回答2:
In my understanding you're working on local
copy, push to pages.git
and have a copy for webapp
?
If so, you can use post-receive hook on pages.git
that will call git pull
on pages
or make a simple copy to pages
.
Post-receive
hook works on remote repo after push from local one. It is executed after all refs have been updated due to push
.
来源:https://stackoverflow.com/questions/2355718/how-to-automate-git-to-automatically-clone-a-repo-on-the-server