- To initialize empty repository in current working directory
git init - To stage (add) files before committing:
git add <files> - To commit:
git commit -m "first commit" - To stage (add) and commit at the same time:
git commit -m "commit message" -a
git add .
git reset .
- If you use
git reset --hard, you will unstage the files and delete them forever
git commit -m "Something terribly misguided"
git reset HEAD~
<< edit files as necessary >>
git add ...
- To create (if you haven't already) and change to branch specified if you haven't done so:
git checkout -b <branch-name> - To just checkout existing branch:
git checkout <branch-name>
You don't necessarily have to commit your files before switching to the master branch there's a another mechanism to save your work in progress if you're working on something that's half-finished or experimental.
You can use a command called git stash
git stash -u
This will save all of your current changes without committing them and then revert back to a clean working directory then later at some point in the future when you want to get back to work. You can either pop or apply the changes in that stash to your current working directory.
git stash pop
So kind of just like it sounds you're stashing away your current changes to be used at some later point so stashing is just a handy thing to know and an alternative to committing your code if you're not quite ready to do so.
A lot of times on a feature branch you'll create a lot of different commits and these commits are kind of irrelevant to what's going on in the master branch, e.g. you can see here that our feature branch is three commits ahead of our master branch and it has a bunch of comments about adding useless emojis to the code. Instead of merging all of these commits into the master branch you can use the squash flag to squash them down into a single commit. When you do so, your merge this will keep the change history nice and concise on the master branch but still preserve all of the original commits on the feature branch itself.
git merge <branch-name> squash
When you merge with the squash flag, it won't actually change the head commit on the master branch so you'll need to add an additional commit on your own that says something like merged in feature branch and that gives us a nice compact change history on the master (so if you checkout to another branch without committing this squash merge, when you check back out to master the changes wont be "saved")
git remote add origin git@github.com:hwlee96/<reponame>.git
git push -u origin <branch-name> (e.g. master or name of branch)
- Git & GitHub Tutorial for Beginners #11 - Collaborating on GitHub (really good)
- https://www.youtube.com/watch?v=MnUd31TvBoU
- Note that even if you delete a branch after merging the pull request on GitHub, the branch still exists on Git installed on your local system. If you want to delete this branch, you have to research how to do so, but should be straightforward.
- https://www.youtube.com/watch?v=rgbCcBNZcdQ
- Same as no forking generally
- http://rogerdudler.github.io/git-guide/ (I used to use this a lot)
- Pro Git Book (downloaded, but not verified if good and useful)
- Rebasing (watched before but can't rmb) - https://www.youtube.com/watch?v=f1wnYdLEpgI