Git Feature Branch Workflow
The core idea behind the Feature Branch Workflow is that all feature development should take place in a dedicated branch instead of the
master branch. This encapsulation makes it easy for multiple developers to work on a particular feature without disturbing the main codebase. It also means the
master branch will never contain broken code, which is a huge advantage for continuous integration environments.
Encapsulating feature development also makes it possible to leverage pull requests, which are a way to initiate discussions around a branch. They give other developers the opportunity to sign off on a feature before it gets integrated into the official project. Or, if you get stuck in the middle of a feature, you can open a pull request asking for suggestions from your colleagues. The point is, pull requests make it incredibly easy for your team to comment on each other’s work.
The Git Feature Branch Workflow is a composable workflow that can be leveraged by other high-level Git workflows. We discussed other Git workflows on the Git workflow overview page. Git Feature Branch Workflow is branching model focused, meaning that it is a guiding framework for managing and creating branches. Other workflows are more repo focused. The Git Feature Branch Workflow can be incorporated into other workflows. The Gitflow, and Git Forking Workflows traditionally use a Git Feature Branch Workflow in regards to their branching models.
The Feature Branch Workflow assumes a central repository, and
master represents the official project history. Instead of committing directly on their local
master branch, developers create a new branch every time they start work on a new feature. Feature branches should have descriptive names, like animated-menu-items or issue-#1061. The idea is to give a clear, highly-focused purpose to each branch. Git makes no technical distinction between the
master branch and feature branches, so developers can edit, stage, and commit changes to a feature branch.
In addition, feature branches can (and should) be pushed to the central repository. This makes it possible to share a feature with other developers without touching any official code. Since
master is the only “special” branch, storing several feature branches on the central repository doesn’t pose any problems. Of course, this is also a convenient way to back up everybody’s local commits. The following is a walk-through of the life-cycle of a feature branch.
Start with the master branch
All feature branches are created off the latest code state of a project. This guide assumes this is maintained and updated in the
git checkout master git fetch origin git reset --hard origin/master
This switches the repo to the
master branch, pulls the latest commits and resets the repo's local copy of
master to match the latest version.
Create a new-branch
Use a separate branch for each feature or issue you work on. After creating a branch, check it out locally so that any changes you make will be on that branch.
git checkout -b new-feature
This checks out a branch called new-feature based on
master, and the -b flag tells Git to create the branch if it doesn’t already exist.
Update, add, commit, and push changes
On this branch, edit, stage, and commit changes in the usual fashion, building up the feature with as many commits as necessary. Work on the feature and make commits like you would any time you use Git. When ready, push your commits, updating the feature branch on Bitbucket.
git status git add <fichier> git commit
Push feature branch to remote
It’s a good idea to push the feature branch up to the central repository. This serves as a convenient backup, when collaborating with other developers, this would give them access to view commits to the new branch.
git push -u origin new-feature
This command pushes new-feature to the central repository (origin), and the -u flag adds it as a remote tracking branch. After setting up the tracking branch,
git push can be invoked without any parameters to automatically push the new-feature branch to the central repository. To get feedback on the new feature branch, create a pull request in a repository management solution like Bitbucket Cloud or Bitbucket Server. From there, you can add reviewers and make sure everything is good to go before merging.
Now teammates comment and approve the pushed commits. Resolve their comments locally, commit, and push the suggested changes to Bitbucket. Your updates appear in the pull request.
Merge your pull request
Before you merge, you may have to resolve merge conflicts if others have made changes to the repo. When your pull request is approved and conflict-free, you can add your code to the
master branch. Merge from the pull request in Bitbucket.
Aside from isolating feature development, branches make it possible to discuss changes via pull requests. Once someone completes a feature, they don’t immediately merge it into
master. Instead, they push the feature branch to the central server and file a pull request asking to merge their additions into
master. This gives other developers an opportunity to review the changes before they become a part of the main codebase.
La revue du code est un avantage majeur des pull requests. En réalité, celles-ci sont conçues pour offrir un moyen générique de parler du code. On peut considérer les pull requests comme un espace de discussion dédié à une branche particulière. Elles peuvent donc aussi être utilisées beaucoup plus tôt dans le process de développement. Par exemple, si un développeur a besoin d'aide pour une fonctionnalité spécifique, il lui suffit de faire une pull request. Les parties intéressées seront automatiquement informées et elles pourront voir la question en regard des commits en question.
Once a pull request is accepted, the actual act of publishing a feature is much the same as in the Centralized Workflow. First, you need to make sure your local
master is synchronized with the upstream
master. Then, you merge the feature branch into
master and push the updated
master back to the central repository.
Pull requests can be facilitated by product repository management solutions like Bitbucket Cloud or Bitbucket Server. View the Bitbucket Server pull requests documentation for an example.
The following is an example of the type of scenario in which a feature branching workflow is used. The scenario is that of a team doing code review around on a new feature pull request. This is one example of the many purposes this model can be used for.
Marie commence une nouvelle fonctionnalité
Before she starts developing a feature, Mary needs an isolated branch to work on. She can request a new branch with the following command:
git checkout -b marys-feature master
This checks out a branch called
marys-feature based on
master, and the -b flag tells Git to create the branch if it doesn’t already exist. On this branch, Mary edits, stages, and commits changes in the usual fashion, building up her feature with as many commits as necessary:
git status git add <fichier> git commit
Marie va déjeuner
Mary adds a few commits to her feature over the course of the morning. Before she leaves for lunch, it’s a good idea to push her feature branch up to the central repository. This serves as a convenient backup, but if Mary was collaborating with other developers, this would also give them access to her initial commits.
git push -u origin marys-feature
This command pushes
marys-feature to the central repository (origin), and the -u flag adds it as a remote tracking branch. After setting up the tracking branch, Mary can call
git push without any parameters to push her feature.
Marie termine sa fonctionnalité
When Mary gets back from lunch, she completes her feature. Before merging it into
master, she needs to file a pull request letting the rest of the team know she's done. But first, she should make sure the central repository has her most recent commits:
Then, she files the pull request in her Git GUI asking to merge
master, and team members will be notified automatically. The great thing about pull requests is that they show comments right next to their related commits, so it's easy to ask questions about specific changesets.
Guillaume reçoit la pull request
Bill gets the pull request and takes a look at
marys-feature. He decides he wants to make a few changes before integrating it into the official project, and he and Mary have some back-and-forth via the pull request.
Marie apporte les changements
Pour apporter les changements, Marie utilise le même process que celui utilisé pour créer la première itération de sa fonctionnalité. Elle édite, stage, commite et pushe les mises à jour vers le dépôt centralisé. Toutes ses activités apparaissent dans la pull request, et Guillaume peut ajouter des commentaires à tout moment.
If he wanted, Bill could pull
marys-feature into his local repository and work on it on his own. Any commits he added would also show up in the pull request.
Marie publie sa fonctionnalité
Lorsque Guillaume est prêt à accepter la pull request, un utilisateur doit faire un merge de la fonctionnalité dans le projet stable (cette opération peut être exécutée par Marie ou Guillaume lui-même) :
git checkout master git pull git pull origin marys-feature git push
This process often results in a merge commit. Some developers like this because it’s like a symbolic joining of the feature with the rest of the code base. But, if you’re partial to a linear history, it’s possible to rebase the feature onto the tip of
master before executing the merge, resulting in a fast-forward merge.
Some GUI’s will automate the pull request acceptance process by running all of these commands just by clicking an “Accept” button. If yours doesn’t, it should at least be able to automatically close the pull request when the feature branch gets merged into
Dans le même temps, Jean fait exactement la même chose
While Mary and Bill are working on marys-feature and discussing it in her pull request, John is doing the exact same thing with his own feature branch. By isolating features into separate branches, everybody can work independently, yet it’s still trivial to share changes with other developers when necessary.
In this document, we discussed the Git Feature Branch Workflow. This workflow helps organize and track branches that are focused on business domain feature sets. Other Git workflows like the Git Forking Workflow and the Gitflow Workflow are repo focused and can leverage the Git Feature Branch Workflow to manage their branching models. This document demonstrated a high-level code example and fictional example for implementing the Git Feature Branch Workflow. Some key associations to make with the Feature Branch Workflow are:
- focused on branching patterns
- can be leveraged by other repo oriented workflows
- promotes collaboration with team members through pull requests and merge reviews
Utilizing git rebase during the review and merge stages of a feature branch will create enforce a cohesive Git history of feature merges. A feature branching model is a great tool to promote collaboration within a team environment.
Go one click deeper into Git workflows by reading our comprehensive tutorial of the Gitflow Workflow.