This project is not completely done yet. The README describes ideas for the future API.
Fetch a hook from the local filesystem (see below for directory format):
$ githooks fetch ~/projects/hooks/ctags
>> Copying hook to ~/.githooks/post-merge/ctags...
Fetch a hook from a remote github repo (NOT DONE):
$ githooks fetch github.com/AndrewRadev/my-magical-hooks/ctags
>> Copying hook to ~/.githooks/post-merge/ctags...
List all hooks, installed in the project (NOT DONE):
$ githooks list
post-merge/rails-migrations :: github.com/AndrewRadev/githooks-storage/rails-migrations
Runs `rake db:migrate` if any new migrations were added.
post-merge/ruby-bundler :: github.com/AndrewRadev/githooks-storage/ruby-bundler
Runs `bundle install` if Gemfile.lock was changed.
<hook-type>/<hook-name> :: <remote-source>
<description>
Install a new hook in the project from the githooks storage in ~/.githooks
(this will eventually automatically fetch & install whenever possible):
$ githooks install post-merge/ruby-bundler
$ githooks install <hook-type>/<hook-name>
Run a hook manually (it would need some arguments to work, see man githooks):
$ githooks run post-merge/rails-migrations <args>
The format of the source of a githooks hook is a directory with the following structure:
hook-name/
meta.yml
run
After running githooks fetch hook-name, the run file and the metadata in
meta.yml will be processed and stored in ~/.githooks. The run file is the
actual script to run and it can be written any way you like. The meta.yml
file contains metadata and should have the form:
---
type: <hook-type>
description: <description of the hook's effect>Fetching from a remote repo is still not implemented.
Each hook is written to the relevant .git/hooks/* file. For example, with the
abovementioned post-merge/ruby-bundler and post-merge/rails-migrations
would result in the .git/hooks/post-merge file looking like this:
## Start of githooks scripts
githooks run post-merge/ruby-bundler
githooks run post-merge/rails-migrations
## End of githooks scripts