10 KiB
| title | description | isPublished | publishedOn |
|---|---|---|---|
| 🗓️ Git version control: Ultimate Guide | What is `git`, what are the most used commands, best practices, and tips and tricks. The Ultimate guide to master `git` in your daily workflow. | true | 2022-10-27T14:33:07.465Z |
Hello! 👋
Welcome to the Ultimate Guide to master git in your daily workflow, we will see what are the most used commands, what are the best practices, and tips and tricks.
This guide is a summary of the most important things to know when working with git, and in general, will link to the official documentation of git or other resources for more details, it is on purpose to not go in depth in each topic, it allows to summarize git and vocabulary about it (you can use it as a git cheatsheet).
Note: Sources used to write this blog post are available at the end of this post.
Introduction
Git is a free and open-source distributed version control system for keeping track of changes across a set of files.
Git was originally authored by Linus Torvalds in 2005 for the development of the Linux kernel.
Git allows:
- to work with several people on the same codebase.
- track changes to know who did what and when.
- revert changes.
Git is decentralized, which means that every developer has a full copy of the repository and the complete history of the project.
Get started with git and .gitconfig config file
The first thing you should do when you install Git is to set your user name and email address.
git config --global user.name "Username"
git config --global user.email "email@example.com"
These configurations are stored in the .gitconfig file in your home directory (e.g: ~/.gitconfig) with this format:
[user]
name = Username
email = email@example.com
You can find more information and useful git configurations in the official documentation.
How git works?
Each git project is called a repository (or repo for short) and it contains all the files and folders for a project, as well as each file's revision history (commits) stored in the .git folder.
The history of a repository is represented by a graph.
Each node is called commit and contains:
- an instantaneous view (snapshot) of the state of the repository at a specific moment
- metadata: message, author, creation date, etc.
Commits are snapshots (not diffs on each file) of the project at specific moments in time.
There are several areas where the files in your project will live in Git:
- Working directory: the files that you see in your computer's file system.
- Staging area: the files that will go into your next commit (files added with
git add <filename>command). - Local repository: the
.gitdirectory, which contains all of your project's commits, branches, etc. (files added withgit commit -m "message"command). - Remote repository: the
.gitdirectory in a remote server (files added withgit pushcommand).
Commands cheatsheet
You can find the official documentation of git commands at git-scm.com/docs.
# Initialize a new git repository
git init
# Clone a repository
git clone <url>
# Add all the files to staging area
git add .
# Add specific file to staging area
git add <file>
# Commit changes
git commit -m "chore: initial commit"
# Add remote repository
git remote add <remote> <url>
# The main <remote> is often called `origin`
# Add forked repository
git remote add <remote> <url>
# The forked <remote> is often called `upstream`
# List all the remotes
git remote
# Sync forked repository
git fetch <remote>
git merge <remote>/<branch>
# Push changes to remote repository
git push <remote>
# Pull changes from remote repository
git pull <remote>
# Show the status of the working tree
git status
# Show the commit history
git log
# Create a new branch
git checkout -b <branch>
# Switch to a branch (or tag or commit)
git checkout <branch>
# Merge a branch into the current branch
git merge <branch>
# Note: Merge creates a "Merge commit" when the base branch and the branch to merge have diverged (they have different commits).
# To avoid creating a "Merge commit", we can use rebase instead of merge.
git rebase --interactive <branch-to-rebase-on>
# Combine multiple commits of a branch into one for a merge
git merge --squash <branch>
# Change several past commits (interactive rebase)
# HEAD points to the current consulted commit.
git rebase --interactive HEAD~<number-of-commits>
# Delete a branch
git branch --delete <branch>
git push <remote> --delete <branch>
# Fetch branches from remote repository and prune
git fetch --prune
# Revert a commit
git revert <commit>
# Reset the current branch, delete all commits since <branch> (without removing the changes)
git reset --soft <branch>
# Apply the changes introduced by some existing commits
# (by first being on the branch where you want to apply the commit)
git cherry-pick <commit>
# To list all commits that differ between two branches
git log <branch1>..<branch2> # commits in branch2 that are not in branch1 (branch2 ahead of branch1, branch2 behind branch1)
git log <branch2>..<branch1> # commits in branch1 that are not in branch2 (branch1 ahead of branch2, branch1 behind branch2)
# Summary of commit authors across all branches, excluding merge commits.
git shortlog --summary --numbered --all --no-merges
.gitignore file
The .gitignore file is a text file that tells git which files (or patterns) it should ignore.
The .gitignore file is usually placed in the root directory of the repository.
We usually ignore files that are generated by the build process or files that contain sensitive information.
Example of .gitignore file:
.env
build
*.exe
.gitkeep file
The .gitkeep file is a file that is used to keep an empty directory in a Git repository.
This is useful when you want to keep an empty directory in your repository but you don't want to commit any file inside it.
Git remote repositories (GitHub/GitLab)
Once you are ready to share your code over the internet, you will need to create a remote repository on a service like GitHub or GitLab.
There are many other services, you can also self-host your own Git server.
SSH vs HTTPS authentication
Once you have created a remote repository, you will need to authenticate to push and pull changes.
There are two main ways to authenticate:
- SSH: you will need to generate an SSH key pair and add the public key to your remote repository.
- HTTPS: you will need to provide your username and password each time you push or pull changes.
SSH authentication is the recommended way to authenticate to a remote repository.
You can find more information about SSH authentication in the official documentation.
Sign git commits with gpg
As we have seen in the Get started with git and .gitconfig config file section, we can configure git with a name and email address with a value of our choice.
That means that anyone can create a commit with any name and email address and claim to be whoever they want when they create a commit.
To avoid this, you can sign your commits with a GNU Privacy Guard (gpg) key.
You can find more information about signing commits in the official documentation.
Continous Integration/Continuous Delivery (CI/CD)
Once you have your code in a remote repository, everyone (with access) can potentially start contributing to the project. This is great, but it also means that you need to have a way to ensure that your code is working as expected for each change in the project.
You could do it manually, depending on the size and the complexity of the project, but it could be a tedious task.
Instead, you can use a Continuous Integration (CI) service to automate the process of testing your code, running linting, unit tests, e2e tests, etc.
There are many CI services, but the most popular ones are GitHub Actions, GitLab CI, CircleCI, Travis CI, and many others...
Then, once your code is ready, tested and working as expected, you can use a Continuous Delivery (CD) service to automate the process of deploying your code.
CI/CD services are usually integrated with remote repositories, so you can configure them to run automatically when you push changes to the remote repository.
Best practices and git workflows
Commit messages are very important, they are a way to easily know what has changed in the project.
There are many conventions for commit messages, but the most popular one is the Conventional Commits specification.
Then, we can use the commit messages to automatically determine a semantic version for the next release of the project.
When multiple developers are working on the same project, it is important to organize the work in a way that everyone can work on different features without conflicts (changes in the same files).
There are many ways to organize the work, but the most popular ones are:
They are called Git workflows, or Git branching strategies.
Conclusion
git is the tool that every programmer should know to do collaborative work (not only, git is also very powerful even when working alone) and keep track of changes across a set of files.