Interrogations about "servless CI with git" bash script

Question

I am in reference to the excellent article by David Gageot: Serverless CI with git.

Let me include David's script here:

#!/bin/bash
if [ 0 -eq `git remote -v | grep -c push` ]; then
  REMOTE_REPO=`git remote -v | sed 's/origin//'`
else
  REMOTE_REPO=`git remote -v | grep "(push)" | sed 's/origin//' | sed 's/(push)//'`
fi

if [ ! -z "$1" ]; then
  git add .
  git commit -a -m "$1"
fi

git pull

if [ ! -d ".privatebuild" ]; then
  git clone . .privatebuild
fi

cd .privatebuild
git clean -df
git pull

if [ -e "pom.xml" ]; then
  mvn clean install

  if [ $? -eq 0 ]; then
    echo "Publishing to: $REMOTE_REPO"
    git push $REMOTE_REPO master
  else
    echo "Unable to build"
    exit $?
  fi
fi

If I understand correctly this script, it will clone the initial git repository to a second hidden git repository where unit tests will be run.

If unit tests pass, the second hidden repository is pushed to the initial working repository.

My questions are as follows:

• How is one supposed to use this script? By invoking it instead of the git commit command? As a pre-commit hook?
• What happens if changes are made to the initial git repository when units test are running?

Answer 1

The script is a little outdated but here's how it worked: You'd use the script either to push local changes or commit AND push local changes. The latter behaviour I got rid of in newer versions of the script because it should have a single responsibility. Before running the build, it would pull from remote, clone the whole repo, run the build from this clone then push. If changes are pushed by someone else in the meantime, the push will simply fail. If you make changes in your repo in the meantime, theses changes are unknown to the clone.

Here's the version I use everyday now https://github.com/dgageot/dotfiles/blob/master/bin/git-build

Hope this helps.