GIT_BASICS.md

Git Tutorial

This tutorial will help you understand how git works and how to use git to submit your commit on GitHub.

note

This tutorial is about using Git in bash/cmd, which we highly recommend, : as it's cleaner. GitHub is a totally different thing, it is the web interface or app.

How to install Git

First step is installing Git. Supposing you are on a Debian-based distribution, this will do:

$ sudo apt-get install git

For installing Git on a Mac OS system, you can use the homebrew package manager as follows:

$ brew install git

Alternatively, Git comes pre-installed with XCode (or its Command Line Tools) on a Mac OS system. You can directly install the Command Line Tools as follows:

$ xcode-select --install

Getting Started with NITSkm / Algorithms

First of all, you have to fork the repository you are going to contribute to. This will basically give you a clone of the repository to your own repository. You can do this by opening this to fork the NITSkmOS/Algorithms repository and then clicking 'Fork' in the upper right corner.

Grabbing NITSkm / Algorithms on your local machine

Now you should clone the repository to your local machine so that you can have access to all the code locally and start fixing issues! To do this, you can use these to clone the NITSkm / Algorithms repository:

$ git clone -o upstream https://github.com/NITSkmOS/Algorithms

note

-o upstream sets the remote name of the original NITSkmOS/Algorithms repositories as upstream.

upstream is just a name we used for simplicity. You can name it however you want.

Don't worry if you're not familiar with what remotes are. The following section will explain more about remotes.

Now you have all your code on your local machine!

Getting to work

First let's talk about remotes. To communicate with the outside world, git uses what are called remotes. These are repositories other than the one on your local disk which you can push your changes into (so that other people can see them) or pull from (so that you can get others changes). Now you should add a remote of your fork to your local machine so that you can pull and push your commits. This can be simply done by using the command:

$ git remote add myfork <your_fork_link>

note

myfork is just a name we used for simplicity. You can name it however you want.

The next thing you should know before getting down to work is how to check on the changes you've made to your project, and your current branch. The ability to check your current branch is also extremely important, as you'll see in the next section. The command to check this information is:

$ git status

Before we move onto the next section, you need to know about a very important branch called master. Master is the default branch that git checkouts for you when you clone a repository. It's our policy here at NITSkmOS to never develop on your fork's master branch. This is why we create new branches, which leads us to the next section.

Creating a new branch

To start working on an issue, you first need to create a new branch where you will work. Do not change files when you are on your fork's master branch. If you submit a Pull Request from your fork's master branch, maintainers will assume that you didn't read this guide. The NITSkmOS developers may even reject your work (even if it is a good patch), because you are showing you haven't checked our documentation. The reason why you should never develop on your master branch is because your fork's master branch should always be synchronized with the main repository's master branch, which is much more challenging if it has new commits on it. This is why we create our own branch:

$ git checkout -b branchname

note

checkout will switch to the newly created branch.

-b will create a new branch if the branch doesn't already exist.

Some sample naming conventions for branches: + issueXXX + patchXXX + gh-XXX + A short form of the issue name (Where XXX is your issue number.)

We also recommend naming your first branch "my-first-good-pull-request", for the purpose of this guide.

Checking your work

After the issue is fixed and you have tested it (tests are very important! never submit a change that isn't tested), you should check your progress. Type:

$ git status

It will give you an idea about what files are currently modified and which branch you're developing on.

note

Tip: If there's something you don't find, you can always use:

$ git grep "syntax"

This will search through the whole repository and show you the files that contain the syntax.

Run coala

Now you can check if your commit messages and code formattings conform with the community guidelines. If something goes wrong, coala will let you know. The continuous integration (CI) will fail if coala reports errors which means that we cannot proceed with merging your fix/pull request.

$ coala

Adding the files and commiting

First, make sure you're on the correct branch and not developing on master! If you've been following this guide, and this is your first pull request, you should be developing on the "my-first-good-pull-request" branch. You can check your branch with:

$ git status

Now you can add your files/folders to the current commit:

$ git add <file/folder_name>

Do this until you have added all the files needed for your commit. Then type:

$ git commit

This will lead you to a text editor. Now you need to write your commit message. We are very strict about writing commit messages as they help us maintain the repository clean and stable. Commit messages usually consists of three main parts. They should have a newline between them.

• The header

  The header should have the name of the file that you have made the change on, followed by ":", a space, and then a short title that explains the change made.

  Example: .gitignore: Add a new Constants variable

• The body

  The body should have a short paragraph that briefly describes the change that was made, and the reason why this change was needed in imperative. Its maximum length is 72 characters.

• The issue that is being fixed

  This part will usually have "Fixes <issue_link>", so the issue gets referenced on GitHub.

Now that your message is written, you will have to save the file. Press escape to exit insert mode, and save the file (in Vim that is being done by pressing shift + Z twice).

Pushing the commit

Before you push the commit, ensure that you are not developing on master again by running:

$ git status

Now you will need to push the commit to the fork. All you have to do is:

$ git push myfork

It will most likely ask for your login credentials from GitHub. Type them in, and your commit will be pushed online.

Creating a Pull Request

If you've made it this far, and you're still using your 'master' branch, then we're definitely going to be able to tell you have not been reading this documentation. Naughty, naughty, but there is still a way to fix your changes if you have already commited. You can run the following command, which will take you to a new branch containing all of your commited changes (Note: Some sample naming conventions can be found under the "Creating a branch" section). Then, to set your fork's master branch back to a pristine state, check the commands in our Common Git Issues section

$ git checkout -b <branchname>

Now you would like to get your commit into the actual master branch. Making your changes available to all future users of the project. For this, you will have to create a Pull Request. To do this, you will have to go on GitHub, on your fork page. You should change the branch to the one you have worked on and submitted the commit on. Now you can create a Pull Request by clicking the New Pull Request button in the pull request tab.

Congratulations! You have just created your first Pull Request! You are awesome!

note

If you see any error like 1 commit ahead of the master branch you need : to sync your local fork with the remote repository before sending a pull request.

More information regarding syncing can be found here.

Follow-up

Now after you have created the Pull Request, there are two possibilities:

• your PR will get accepted, and your commit will get merged into the master branch - sadly, this rarely happens on the first Pull Request.

• your PR will be rejected. There are 2 cases when a PR is rejected:

  • Test fails
  • Reviewer wants something changed (This also causes bot to fail)

It's highly unlikely that your PR will be accepted on the first attempt

• but don't worry that's just how it works. It helps us maintain the repository clean and stable.

Now if you need to modify your code, you can simply edit it again, add it and commit it using

$ git commit -a --amend

This will edit your last commit message. If your commit message was considered fine by our reviewers, you can simply send it again like this. If not, edit it and send it. Now you have successfully edited your last commit!

If you need to rebase, or want to edit an older commit from your branch, we have an amazing tutorial that you can watch to understand how it works.

Rebasing

As people work on NITSkmOS/Algorithms, new commits will be added. This will result in your local fork going out of sync with the remote repository. To sync your changes with the remote repository run the following commands in the desired branch:

note

This assumes that the remote upstream is the original NITSkmOS/Algorithms repository at https://github.com/NITSkmOS/Algorithms, not your fork.

If you have followed the steps outlined in this guide and cloned the original NITSkmOS/Algorithms repository, upstream should refer to it. You can proceed to the following section without worry.

If you're unsure about this, run git remote -v to check which remote points to the original repository and use that instead of upstream in the following section.

$ git fetch upstream
$ git rebase upstream/master

This will fetch the commits from the remote repository and will merge it into the branch where you are currently working, and move all of the local commits that are ahead of the rebased branch to the top of the history on that branch.

note

After following these instructions when you try to push to remote you may get fast-forwarding error. If that is the case, then you will have to force push since you are attempting to rewrite the git commit history. To do that append the --force argument in the push command:

$ git push myfork --force

Warning: Never force-push on the master branch, or any branch not owned by you.

To verify whether you have rebased correctly, go to the web page of the branch in your fork. If it says your branch is n commits behind Algorithms:master, then you haven't correctly rebased yet. Otherwise, you're good to go!

Squashing your commits

It's possible that you have more than one commit and you want them to be squashed into a single commit. You can take your series of commits and squash them down into a single commit with the interactive rebasing tool. To squash your commits run the following command:

$ git rebase -i master

note

master is the SHA1 hash of the commit before which you want to squash all the commits and make sure that rebase is done onto master branch.

An editor will be fired up with all the commits in your current branch (ignoring merge commits), which come after the given commit. Keep the first one as "pick" and on the second and subsequent commits with "squash". After saving, another editor will be fired up with all the messages of commits which you want to squash. Clean up all the messages and add a new message to be displayed for the single commit.

Common Git Issues

Sometimes, you use git add -A and add files you didn't want to your push (often after rebasing) and push it to the remote. Here, is a short outline of, how can you remove (or revert changes in) particular files from your commit even after pushing to remote.

In your local repo, to revert the file to the state before the previous commit run the following:

$ git checkout HEAD^ /path/to/file

Now , after reverting the file(s) update your last commit, by running :

$ git commit -a --amend

To apply these changes to the remote you need to force update the branch :

$ git push -f myfork

note

The procedure outlined above helps roll back changes by one commit only. 'myfork' mentioned above is your forked repository, where you push your commits.

The git checkout <revision sha> path/to/file command offers you more flexibility in reverting the changes in a file, done even from earlier than the last commit. By replacing the HEAD^ by the revision number of the particular HEAD commit, you can refer to the required revision of the file.

Might sound a little intimidating, but don't worry, an example has been provided for you. First you can check the commit's revision number, where the file was revised by running the following command:

$ git log /path/to/file

The revision number might look like 3cdc61015724f9965575ba954c8cd4232c8b42e4 Now, to revert the file to that revision, run the command:

$ git checkout 3cdc61015724f9965575ba954c8cd4232c8b42e4 /path/to/file.txt

Now, after the file gets reverted back to the required revision, commit the changes and (force) push to the remote.

While rebasing, you may come across mid-rebase conflicts. For information regarding how to resolve mid-rebase conflicts, please check this tutorial.

http://ohshitgit.com/ contains helpful Git snippets for recovering from various common Git issues. It is a great resource to check out when something has gone wrong.

If at any stage you are confused, or have an issue, do not close your Pull Request. Instead, contact us on gitter so that we can help you resolve your problem.

Useful Git commands

This section will briefly explain some other Git commands you will most likely use and will really make your work easier.

$ git config

The git config command lets you configure your Git installation (or an individual repository) from the command line. This command can define everything from user info to preferences to the behavior of a repository.

$ git log

The git log command displays committed snapshots. It lets you list the project history, filter it, and search for specific changes. While git status lets you inspect the working directory and the staging area, git log only operates on the committed history.

$ git push --force myfork

While we normally use git push myfork to push your commit to your fork, after further editing and work on your commit, you will need to use the --force parameter to your push to automatically update your Pull Request.

$ git reset --hard

Reset the staging area and the working directory to match the most recent commit. In addition to unstaging changes, the --hard flag tells Git to overwrite all changes in the working directory, too. Put another way: this obliterates all uncommitted changes, so make sure you really want to throw away your local developments before using it.

$ git clean

The git clean command removes untracked files from your working directory. This is really more of a convenience command, since it’s trivial to see which files are untracked with git status and remove them manually. Like an ordinary rm command, git clean is not undoable, so make sure you really want to delete the untracked files before you run it.

$ git checkout <branch>

The git checkout command is used to switch to another branch in the repository. Here <branch> is the name of the branch you want to switch to.

$ git rebase

Rebasing is the process of moving a branch to a new base commit. From a content perspective, rebasing really is just moving a branch from one commit to another. But internally, Git accomplishes this by creating new commits and applying them to the specified base—it’s literally rewriting your project history. It’s very important to understand that, even though the branch looks the same, it’s composed of entirely new commits.

$ git rebase -i

Running git rebase with the -i flag begins an interactive rebasing session. Instead of blindly moving all of the commits to the new base, interactive rebasing gives you the opportunity to alter individual commits in the process. This lets you clean up history by removing, splitting, and altering an existing series of commits. It’s like git commit --amend on steroids. Usage is $ git rebase -i <base>. Rebase the current branch onto <base>, but use an interactive rebasing session. This opens an editor where you can enter commands (described below) for each commit to be rebased. These commands determine how individual commits will be transferred to the new base. You can also reorder the commit listing to change the order of the commits themselves.

If you would like more information/commands, please use your favourite search engine to look for it. Git is widely used throughout the world and there are many good tutorials and git related Q&A threads out there.

Attribution

This Git Basics is adapted from the coala's Git Basics Guide, available at http://api.coala.io/en/latest/Developers/Git_Basics.html.