Github tutorial

First steps

• Make sure git is installed

  • SSH keys for those who want them

• Git config

  • git config --global user.name "Name"

  • git config --global user.email "[email protected]"

  • Editor

    • $EDITOR or $VISUAL [System Wide]
      • e.g. export EDITOR=vim in ~/.bashrc
    • git config --global core.editor vim/emacs/etc [Git only]

Most important commands to remember

• git status -- shows current status, including which files are tracked/untracked and which are staged
• git log -- Shows a log of all commits.
  • By default, this is hard to read. I personally have alias lg="git log --all --decorate --oneline --graph" specified in my .bashrc, so all I need to type is lg for a more readable git log
• git add -- add (stages) file
• git commit -- Commits all staged files
• git push -- pushes all commits to remote repository
  • git pull -- pulls commits from remote repository

Create your first repository

• Starting from command line:
  • git init -- Sets up git repository in current folder
  • Add remote repository
    • git remote add <name> <url>
    • <name> can be any name, typically the main remote branch is called origin
    • <url> is the URL of the git repo. This looks like https://github.com/<username>/<repo_name> if using HTTPS and [email protected]:<username>/<repo_name>.git if using SSH.
• Starting from web
  • Create repository on GitHub.com (or equivalent)
  • Clone repository
    • git clone https://github.com/<username>/<repo_name> -- HTTPS/no SSH
    • git clone [email protected]:<username>/<repo_name>.git -- SSH
  • Remote repository will automatically be added with this method, with the default name origin.

Adding and committing files

• Check the status of your repository with git status

  • Will list untracked files, tracked files with unstaged changes, and staged files
    • Staging = telling git to add the file to the next commit

• git add <file> -- Stages file to be committed

  • git add . -- Stages all files (even those in folders)
  • git add *.py -- Only stage .py files
  • Did you accidentally stage a file that you didn't want to?
    • git restore --staged <filename>
  • If the file has been updated since the last git add but before a commit, you will need to git add the file again.
    • I generally only stage files immediately before committing so that I know they're all at the latest version

Note: Git will continue to stage the file, even if it's deleted. If a staged file is deleted, you will need to git add <filename> it again to make git aware of its deletion

• Commit the files
  • git commit -- Opens a text editor (set via git config) to write commit message on the top line. Save the file and quit the editor to continue the commit process.
    • Alternatively, git commit -m "<message>" bypasses the editor entirely

Branches

This section may be incomplete in time for the Git workshop on Feb 25th.

Branches are very powerful, but may not be necessary if you're just using Git to keep your code backed up. A good usecase for branches are

1. if you're making a change that would potentially break master, and you want to keep master "clean" as the working version or
2. if you're contributing to someone elses repository

Creating branches

• git checkout -b <branch_name>
  • This is short for git branch <branch_name> and git checkout <branch_name>, so it creates and checks out a new branch
    • By default, this branch is identical to wherever HEAD is (HEAD is git-speak for what is currently checked out. This could be an old commit, a branch, etc). Unless you have specified otherwise, HEAD is at the latest commit of the current branch.
• Now there are two branches, one master (by default) and the second <branch_name>.
  • Switch between the two with git checkout master, git checkout <branch_name>
• The two branches are completely isolated from each other. Changes to one branch do not affect the other

Merging

Often, you're going to eventually want to merge the contents of one branch into the other.

• git merge <branch_name> will merge all changes from <branch_name> into the current branch.
  • Git always merges the specified branch into the current checked-out branch.
• Git will automatically detect any conflicts between the two versions. Often, it's able to reconcile the differences on its own but sometimes it requires some manual intervention.

Diff

Much like the command diff, git diff can provide information on the differences between files in different commits.

• By default, git diff alone will show the changes between all unstaged and the last commit.
  • git diff --staged shows changes between staged and last commit
• You can also see the differences between two versions of a file, using the commit hash found by running git log
  • e.g. git diff db821bc 5104f19 my_file.txt where the two commit hashes are the shorter versions given by git log --oneline
• git diff <branch1> <branch2> see diff between branches

Working with remote repositories

This section may be incomplete in time for the Git workshop on Feb 25th.

Adding and committing files works even if you're not syncing with a remote repository, allowing for a local "track changes" of your code (which can be useful!). However, the power of Git comes from being able to publish your code online and work collaboratively.

• If you don't have a remote repository set up yet, make sure you go back to that section and do that first
• Once you have made commit, you can push it online. This can be done with git push origin master. This pushes the current commit to the master branch of the origin remote repository.
  • You can specify -u (git push -u origin master) to let git know that this is the "upstream" repository, meaning in the future you only need to git push without specifying the branch or repository name.
• Similarly, if there was a change in the remote that you want to pull into your current repository, you can so git pull origin master (or git pull if origin master is set to be the upstream branch).

External Links

• Git Online Ebook -- Great resource!
• Cheat Sheet
• Group Folder