Support injecting additional CLI arguments to the commit step

[ohadlb]

This branch adds the new command-line flag -c that allows modifying the git commit step in the sync process. This is a flexible way to adapt git-sync to support different usecases

[ohadlb]

I'd like to add some context for this branch, because I'm not too sure about it myself.
As I alluded to in the README update, my main usecase here is to use git-sync on a company repo that has some strict requirements in the pre-commit hook, but given that I seem to be the first person to run into this problem in the years this project's been online I think it's safe to say this is a pretty marginal usecase that might not justify adding a dedicated flag. In addition, it's very likely that additional rare usecases will eventually appear that will require more and more for-one-user-only flags, so instead I opted to add this catch-all to just allow the user to set any additional arguments they find necessary.
However, I am aware that it is a problematic design. It requires the user to know the details of the commit commands the script uses and so on, while a dedicated abstract flag would allow the script to just do The Right Thing™.
Moreover I can point out that a user who truly wants to inject additional flags does not really need this new -c flag at all: via the magic of dynamic binding in the shell it's possible to change literally any aspect of the git invocation by creating a wrapper script for git-sync like this:

git () {
    case "$1" in
        commit)
            shift
            command git commit --no-verify "$@"
            ;;
        *)
            command git "$@"
            ;;
    esac
}
. git-sync "$@"

Note that the last line is . git-sync, sourcing the script rather than executing it.

Anyway, let me know what you think when you get around to reviewing this. I'm fine with any decision you make as project owner.