Skip to content

Latest commit

 

History

History
480 lines (335 loc) · 22.5 KB

README.md

File metadata and controls

480 lines (335 loc) · 22.5 KB

Git Workflow Introduction

So what is this repository?

This repository serves as an introduction to the basic git workflow that walks through setting up a git repository that will be used to create your own GitHub Pages website. While this should provide a good introduction to git and its various uses, it is strongly recommended to follow up with the Pro Git Book written by Scott Chacon and Ben Straub from which parts of this guide are derived. The git reference documentation is also another great place to find tips, tricks, and other guides to getting the most out of git.

What is git?

git (not to be confused with GitHub) is a distributed version control system that is designed to allow multiple collaborators to share and keep track of project files and versions across all of their computers. It was originally created in 2005 as a tool to help the Linux kernel development community keep track of changes made through their ever growing non-linear development process. Though it was originally created for software development it can be used in a variety of different ways for any project that requires versioning of files across multiple people/computers. Examples of this are standards like OWASP ASVS or proposals such as Git Law.

Getting Started

All changes that git tracks are stored in a repository which acts as a versioned warehouse for all of your project's files. The following command will initialize a brand new git repository inside the current directory.

$ git init

Once this command completes, you have your first git repository! You can inspect its current status using the status command which is very useful when you get into trouble. We will be using this one a lot throughout the rest of this guide to show some of the various states that a git repository can be in.

$ git status

Below you can see a sample of the output of the status command which will show you the branch you are on (in this case master, but more on that later), your current commit status (also more on that later), and some help text to guide you onto potential next steps.

On branch master

No commits yet

nothing to commit (create/copy files and use "git add" to track)

If you inspect your local file structure at this point, you will also see that git has created a hidden .git folder inside of your current directory.

$ ls -la
total 0
drwxr-xr-x   3 starr  staff   96 Apr 18 18:09 .
drwxr-xr-x  17 starr  staff  544 Apr 18 18:09 ..
drwxr-xr-x   9 starr  staff  288 Apr 18 18:09 .git

And if you navigate into that directory, you will see the various files and folders that git uses to track changes and maintain its configuration. If you are curious what these files do, and what they are for, most of it is explained in the Pro Git Book mentioned above.

$ ls -la .git
total 24
drwxr-xr-x   9 starr  staff  288 Apr 18 18:09 .
drwxr-xr-x   3 starr  staff   96 Apr 18 18:09 ..
-rw-r--r--   1 starr  staff   23 Apr 18 18:09 HEAD
-rw-r--r--   1 starr  staff  137 Apr 18 18:09 config
-rw-r--r--   1 starr  staff   73 Apr 18 18:09 description
drwxr-xr-x  13 starr  staff  416 Apr 18 18:09 hooks
drwxr-xr-x   3 starr  staff   96 Apr 18 18:09 info
drwxr-xr-x   4 starr  staff  128 Apr 18 18:09 objects
drwxr-xr-x   4 starr  staff  128 Apr 18 18:09 refs

First Commit

Commits are the core way that git records changes in the repository, and they represent a snapshot of files in time. Commits are made whenever you reach a state in your project that you would like to record, say if you were working on rewording a document or refactoring some code, but wanted to try a different path without deleting your current progress. In order to create a commit, we need some changes first so let's run the following command to create an index.html file in the current directory.

$ touch index.html

Once this completes we can use the status command to see what effect this had on the git repository.

$ git status
On branch master

No commits yet

Untracked files:
  (use "git add <file>..." to include in what will be committed)

	index.html

nothing added to commit but untracked files present (use "git add" to track)

As you can see we still don't have any commits yet, but git has picked up the new Untracked file that we created above. Files can be in one of four states (Untracked, Unmodified, Modified, and Staged) depending on where they are in the git process, but will always initially be Untracked when they are created (more on this in the Pro Git Book). You can also see that the status command is trying to be helpful and is pointing us to the next command we want to run to create a commit, which is add. The add command will move any Untracked or Modified file to the Staged state which will tell git to stage it to get ready for the next commit. This allows you to be selective about which files go into each of your commits. The add command can be ran using the following command.

$ git add index.html

Once this completes we can again use the status command to see what effect this had on the git repository.

$ git status
On branch master

No commits yet

Changes to be committed:
  (use "git rm --cached <file>..." to unstage)

	new file:   index.html

Now status shows that our index.html file is Staged and is ready to be committed along with showing us the rm command which will allow us to unstage the file if we would like. Now that we are ready to commit, we can simply run the following command.

$ git commit

This will open a vi window by default in which you can enter a message to describe your commit (you can bypass this and enter a message from the command line by using the -m option). Git commit messages could be a topic in and of themselves (see this page for things to keep in mind when writing them), but for now you can simply use the message My first commit. Once you save the message you've successfully created your first commit!

NOTE: If this is your absolute first time creating a commit, git may have prompted you to enter a global name and email address to associate with your commit so that other collaborators can contact you if they need to. For this you can simply follow the info that git provides in that message and proceed above as normal. You will only be asked to set this once.

Now if we run the status command again, we will see the output for a normal, Unmodified git repository.

$ git status
On branch master
nothing to commit, working tree clean

And if we want to actually see the commit history we have created thus far, we can use the log command.

$ git log
commit dfbd01f915e535b6936693b43e297236a095f4bb (HEAD -> master)
Author: Your Name <[email protected]>
Date:   Thu Apr 18 20:15:23 2019 -0700

    My first commit

This will show us the unique hash for each commit (which can be used to look up the commit later), as well as the Author, Date, and Commit Message. It also shows which commit is currently set as the local HEAD. For this, think of a tape playing through a cassette player. The head of that cassette can be at any point in the music track and can be rewound, or fast-forwarded to move around through the music.

Making Changes

Now that we have one commit, let's add some content to our index.html page to spice it up a bit. For this you can use nano, vi, emacs or any text editor of your choosing.

$ nano index.html

Once you have this open, paste the contents of the index.html page from this repository to act as a base for your index.html page. When you are done, save it, and run the status command.

$ git status
On branch master
Changes not staged for commit:
  (use "git add <file>..." to update what will be committed)
  (use "git checkout -- <file>..." to discard changes in working directory)

	modified:   index.html

no changes added to commit (use "git add" and/or "git commit -a")

As you can see , our index.html page is now in the Modified state. To stage the file, and commit, simply run the same commands as before.

NOTE: The . selects all of the files in the current directory

$ git add .
$ git commit

From here you can again run the status and log commands to check in on what git did.

$ git status
On branch master
nothing to commit, working tree clean
$ git log
commit 4ad7c47199abb9df0613a16828a0cdae0560628a (HEAD -> master)
Author: Your Name <[email protected]>
Date:   Thu Apr 18 21:05:51 2019 -0700

    Adds content to index.html page

commit dfbd01f915e535b6936693b43e297236a095f4bb
Author: Your Name <[email protected]>
Date:   Thu Apr 18 20:15:23 2019 -0700

    My first commit

As you can see you now have a second commit on top of your previous commit which has become our new HEAD as well as now being at the new top of master. New commits will always automatically advance the HEAD, but let's say we wanted to get back to our previous, blank index.html page. In order to do this, we will use the checkout command followed by your first commit's hash.

$ git checkout dfbd01f915e535b6936693b43e297236a095f4bb
Note: checking out 'dfbd01f915e535b6936693b43e297236a095f4bb'.

You are in 'detached HEAD' state. You can look around, make experimental
changes and commit them, and you can discard any commits you make in this
state without impacting any branches by performing another checkout.

If you want to create a new branch to retain commits you create, you may
do so (now or later) by using -b with the checkout command again. Example:

  git checkout -b <new-branch-name>

HEAD is now at dfbd01f My first commit

From the above, you can see that git is warning us about the state we are in because if we make changes at this point we will likely lose our changes (unless we create a branch off of this commit as described in the message, and as expanded on later on). If we were to run status and log at this point we would see some warnings about our HEAD being detached, and see that we are actually no longer on a defined branch.

$ git status
HEAD detached at dfbd01f
nothing to commit, working tree clean
$ git log
commit dfbd01f915e535b6936693b43e297236a095f4bb (HEAD)
Author: Your Name <[email protected]>
Date:   Thu Apr 18 20:15:23 2019 -0700

    My first commit

In order to get out of this state after looking around, we can simply run the checkout command, but specify our branch master instead of a git hash.

$ git checkout master
Previous HEAD position was dfbd01f My first commit
Switched to branch 'master'

Branching Out

Branches in git can be thought of in many ways as branches of a tree, they can both branch out from one another, as well as merge back together again. Another way to think of them is as named commits, and since every commit builds off of the previous commits they retain those commits in that branch's log. Git commits can also exist across branches and in each other's history, again since every commit builds off of those prior. Git repositories can have thousands of branches with thousands of parallel tracks of work all existing simultaneously. If you wanted to list the branches in a repository you can run the following command.

git branch
* master

As you can see we only have one branch, which happens to be the default master branch. Normally this branch is treated as a special branch that contains the canonical source of whatever you are creating, and normally you would not want to push breaking or draft changes directly to the master branch.

If you went ahead and took a look at our index.html page, you might notice that it looks slightly bland, but it is functional for the most part. Let's add some styling to it, but create a branch so as to avoid making breaking changes to our page as we modify and commit changes to it. In order to create a new branch, you can run the following command:

$ git branch css-test

Now you have a css-test branch (you can see it with git branch), but have not checked it out yet. In order to do this, run the checkout command for the new css-test branch as you did with master before.

NOTE: A shortcut to both create a branch and check it out is to use the -b switch on the checkout command as in git checkout -b css-test

$ git checkout css-test
Switched to branch 'css-test'

Now if we run the status command, we will see that we are on the css-test branch.

$ git status
On branch css-test
nothing to commit, working tree clean

Now that we are on the right branch, we can bring over the cover.css file from this repository into yours as we did with the index.html file earlier. Feel free to edit the file as you like, and then run the following commands to stage and commit the file:

$ git add .
$ git commit

If we were to run the log command at this point, we would see that our new commit is at the top of our css-test branch, and that our HEAD is set to the same, but that our master branch is now behind by one commit.

$ git log
commit 44066ba020e78edcaa8e740db0692ada548e5ab6 (HEAD -> css-test)
Author: Your Name <[email protected]>
Date:   Thu Apr 18 21:50:19 2019 -0700

    Adds styling to the cover page

commit 4ad7c47199abb9df0613a16828a0cdae0560628a (master)
Author: Your Name <[email protected]>
Date:   Thu Apr 18 21:05:51 2019 -0700

    Adds content to index.html page

commit dfbd01f915e535b6936693b43e297236a095f4bb
Author: Your Name <[email protected]>
Date:   Thu Apr 18 20:15:23 2019 -0700

    My first commit

Once we are done with our changes, we will need to merge our css-test branch back into master in order to bring it up to speed. To do this, we will need to checkout the master branch again, and use the merge command to bring in the changes from our css-test branch.

$ git checkout master
Switched to branch 'master'
$ git merge css-test
Updating 4ad7c47..44066ba
Fast-forward
 cover.css | 110 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 110 insertions(+)
 create mode 100644 cover.css

As you can see, git fast-forwarded our changes from css-test onto the masterbranch. It was able to do this without any user interaction because the changes on each branch did not conflict, and git was able to figure out how to splice the changes together on its own, if this was not the case, git would show a merge conflict which will be explained in the next section. If we run log at this point, we will see that master and css-test are set to the same commit.

$ git log
commit 44066ba020e78edcaa8e740db0692ada548e5ab6 (HEAD -> master, css-test)
Author: Your Name <[email protected]>
Date:   Thu Apr 18 21:50:19 2019 -0700

    Adds styling to the cover page

commit 4ad7c47199abb9df0613a16828a0cdae0560628a
Author: Your Name <[email protected]>
Date:   Thu Apr 18 21:05:51 2019 -0700

    Adds content to index.html page

commit dfbd01f915e535b6936693b43e297236a095f4bb
Author: Your Name <[email protected]>
Date:   Thu Apr 18 20:15:23 2019 -0700

    My first commit

Merge Conflicts

Merge conflicts arise when two branches are merged together that have changes that cover the same file(s) and line number(s). Let's create a merge conflict so that we can resolve it and get back into a good state. First we'll create a new branch to change the top nav bar to show our name instead of the generic Cover.

$ git checkout -b name-test

Once on the new branch, change the index.html page on line 27 to use your name instead of Cover. Once you save the file, you can stage and commit to the name-test branch.

$ git add .
$ git commit

After that, let's checkout master again so we can edit the same line in the same file on a different branch.

$ git checkout master

Now edit your index.html page on line 27 to be something different than Cover, and different from the name you put in before.

NOTE: git can be smart sometimes if the exact same change is made in the exact same way, so make sure it is different than before.

Once saved, stage and commit the file as before.

$ git add .
$ git commit

NOTE: If we run the log command at this point we will see that our name-test commit is not visible because it is not part of the history of master

We are now ready to create our merge conflict. Since we are already on the master branch, run the following command to merge name-test into master:

$ git merge name-test
Auto-merging index.html
CONFLICT (content): Merge conflict in index.html
Automatic merge failed; fix conflicts and then commit the result.

Oh no, a merge conflict! If we run the status command now, we will see what we would expect, that both branches edited the index.html page as well as some helpful tips to get out of the current state that we are in.

On branch master
You have unmerged paths.
  (fix conflicts and run "git commit")
  (use "git merge --abort" to abort the merge)

Unmerged paths:
  (use "git add <file>..." to mark resolution)

	both modified:   index.html

no changes added to commit (use "git add" and/or "git commit -a")

And if we look at our index.html file, we will see that git has helpfully segmented the changes for our review so that we can manually merge the two changes together.

<div class="inner">
<<<<<<< HEAD
<h3 class="masthead-brand">My Dog</h3>
=======
<h3 class="masthead-brand">Your Name</h3>
>>>>>>> name-test
<nav class="nav nav-masthead justify-content-center">

In this case, we can simply delete the <<<<<<< HEAD, =======, and >>>>>>> name-test lines and choose which <h3> tag to accept, either the one from our current HEAD or from the name-test branch. Once these edits are made, we can save the file, but aren't out of the woods yet. In order to fully resolve the conflicts and finalize our changes, we need to stage and commit the index.html file to the current branch.

$ git add .
$ git commit

Once this is done, we can run status again, and see that we are on a clean master branch again.

$ git status
On branch master
nothing to commit, working tree clean

Using Remotes

NOTE: From here on out, this guide assumes that you have a GitHub account. GitLab, and other git remote hosts provide similar features but may differ slightly in their implementation.

Now that you have learned the basics of using git locally, it's time to add a remote to your project. A remote is simply a named server that you push your commits to so that others can pull them back down. For this we will create a GitHub repository by creating a GitHub account, and selecting the + button in the top right, then New Repository, and filling out the form with the information it requests. Once this is done, we can copy the remote's .git URL and proceed with adding a remote on the command line.

To add a remote for your GitHub repository named origin (the default name for most single-remote git projects), run the following command:

NOTE: For simplicity's sake it may be easier to use an https remote URL instead of an ssh URL. If you are unfamiliar with creating SSH keys, I would recommend using the https version.

$ git remote add origin <your-git-remote-url>

Once this is done, you simply need to use the push command to push your local commits to the remote.

$ git push -u origin master
Enumerating objects: 16, done.
Counting objects: 100% (16/16), done.
Delta compression using up to 8 threads
Compressing objects: 100% (13/13), done.
Writing objects: 100% (16/16), 3.22 KiB | 3.22 MiB/s, done.
Total 16 (delta 2), reused 0 (delta 0)
remote: Resolving deltas: 100% (2/2), done.
To github.com:<your-username>/<your-repo>
 * [new branch]      master -> master
Branch 'master' set up to track remote branch 'master' from 'origin'.

NOTE: The -u in the command sets up the remote to use for a given branch. Any new branches will use the origin remote by default, and after running this command master now will as well. Because of this, from here on out you can simply use the git push command without any switches.

One of the great things about remotes is that changes can be made on other machines or by other people and be synced to your local machine. To test this out let's add a README.md file through the GitHub page for your repository by clicking the "Create new file" button and naming the file README.md. Place whatever text you want in the file, and when you have committed that from the web UI, run the following command to pull it down to your local machine:

$ git pull
remote: Enumerating objects: 4, done.
remote: Counting objects: 100% (4/4), done.
remote: Compressing objects: 100% (3/3), done.
remote: Total 3 (delta 0), reused 0 (delta 0), pack-reused 0
Unpacking objects: 100% (3/3), done.
From github.com:<your-username>/<your-repo>
   b28e0de..422c12a  master     -> origin/master
Updating b28e0de..422c12a
Fast-forward
 README | 1 +
 1 file changed, 1 insertion(+)
 create mode 100644 README

Once this runs, you should be able to see the file you created populate in your local filesystem.

GitHub Pages

Another great thing about remotes are the custom extensions that some of them have built around the git infrastructure, such as issue tracking, collaboration management, and even website hosting. GitHub supports a feature called GitHub Pages that we can use to get the website we have been building hosted on GitHub! To enable GitHub pages, go to your project's settings in GitHub, scroll to the GitHub Pages section, and set the Source to your master branch.

Once that's done, your site is published, at the URL provided! From here feel free to edit the site as you wish, and customize it to make it your own, all within your new found git workflow! You can also add collaborators to your site if you wish, and they can get an exactcopy of the repository on their devices using the clone command:

$ git clone <your-git-remote-url>

This will create a new folder with the same name as your git repository for them to collaborate with you on!