Chapter 26 Introduction to Git

26.1 Introduction

26.1.1 Useful shell commands

See location / working directory: pwd
See what is in current directory: ls (list files)
Changing directory: cd file_place
Editing a file: nano file_name
- Delete, add, change contents of a file
- Save changes: Ctrl + O
- Exit the text editor: Ctrl + X
Create or edit a file: echo
- Create a new file
  
  echo "Review for duplicate records" > todo.txt
- Add content to existing file
  
  echo "Review for duplicate records" >> todo.txt
Checking Git version: git --version

26.1.2 Saving files

Git workflow

Modify a file

nano , echo
Save the draft
- Adding a single file: git add file_name
- Adding all modified files: git add .
Commit the updated file
- git commit -m "..."
Repeat

Check the status of files

git status

26.1.3 Comparing files

Compare an unstaged file with the last committed version:
- git diff filename
Compare a staged file with the last committed version:
- git diff -r HEAD filename
Compare all staged files with the last committed versions:
- git diff -r HEAD

26.2 Making changes

26.2.1 Storing data with Git

The commit structure

Git commits have three parts:

Commit
- contains the metadata
- Git hash
  - allow data sharing between repos
  - If two files are the same, then their hashes are the same
    
    eg., the last summary_statistics.csv hash is 3f5003f
Tree
- tracks the names and locations in the repo
Blob
- binary large object
- may contain data of any kind
- compressed snapshot of a file’s contents

Viewing a repository’s history

git log

Show more recent commits: press space
Quit the log and return to the terminal: press q

Finding a particular commit

git show c27fa856

Only need the first 6-8 characters of the hash
Useful for viewing changes made in a particular commit

(vs git diff compare changes between commits)

26.2.2 Viewing changes

The HEAD shortcut

Compares staged files to the version in the last commit
Use a tilde ~ to pick a specific commit to compare versions

Changes per document by line

git annotate file_name

Summary

Command	Function
`git show HEAD~1`	Show what changed in the second most recent commit
`git diff 35f4b4d 186398f`	Show changes between two commits
`git diff HEAD~1 HEAD~2`	Show changes between two commits
`git annotate file`	Show line-by-line changes and associated metadata

26.2.3 Undoing changes before committing

Staged files

Unstaging a single file
- git reset HEAD file_name
Unstaging all files
- git reset HEAD

Unstaged files

Undo changes to an unstaged file
- git checkout -- file_name
  - checkout means switching to a different version, defaults to the last commit
  - losing all changes made to the unstaged file forever
Undo changes to all unstaged files
- git checkout .
  - This command must be run in the main directory

26.2.4 Restoring and reverting

Customizing the log output

By restrict the number with -

git log -3

shows the three most recent commits

git log -3 file_name

shows the three most recent commits of one file

By restrict with date

git log --since='Apr 2 2022'

since particular date

git log --since='Apr 2 2022' --until='Apr 11 2022

between two dates

Cleaning a repository

See what files are not being tracked
- git clean -n
Delete those files
- git clean -f

26.3 Git workflows

26.3.1 Configuring Git

Levels of settings

git config --list : view the list of all customizable settings
Git has three levels of settings:

1. --local : settings for one specific project

2. --global : settings for all of our projects

3. --system : settings for every users on this computer

Changing our settings

git config --global setting value

Change email address to johnsmith@datacamp.com:

git config --global user.email johnsmith@datacamp.com
Change username to John Smith:

git config --global user.name 'John Smith'

Creating a custom alias

Set up an alias through global settings
Typically used to shorten a command

eg., To create an alias for committing files by executing ci :

git config --global alias.ci 'commit -m'

We can now commit files by executing: git ci
Tracking aliases: git config --global --list

Ignoring specific files

nano .gitignore

26.3.2 Branches

There’re 3 branches, 2 merges in the picture.

Source and destination

When merging two branches:

the commits are called parent commits
source : the branch we want to merge from
destination : the branch we want to merge into

eg., When merging Analysis into Main,

Analysis = source

Main = destination

Identifying branches

git branch

* = current branch

Creating a new branch

git checkout -b branch_name

The difference between branches

git diff branch_1 branch_2

26.3.3 Working with branches

Switch branches

git checkout branch_name

Why do we merge branches?

main = ground truth
Each branch should be for a specific task
Once the task is complete we should merge our changes into main
- to keep it up to date and accurate

Merging branches

git merge source destination

eg., To merge summary-statistics into main

git merge summary-statistics main

26.3.4 Handling conflict

A conflict occurs when a file in different branches has different contents that prevent them from automatically merging into a single version.

Git conflicts

nano todo.txt

reserve only c) line and delete others lines

Another example to see how to delete lines:

26.4 Collaborating with Git

26.4.1 Creating repos

Benefits of repos
- Systematically track versions
- Collaborate with colleagues
- Git stores everything!
Don’t create a nested repos

Creating a new repo

git init repo_name

Converting a project

git init

26.4.2 Working with remotes

Benefits of remote repos

Everything is backed up
Collaboration, regardless of location
git clone is a very useful command for copying other repos onto your local computer, whether from another local directory or remote storage such as GitHub.

Cloning locally

git clone path-to-project-directory

git clone /home/john/repo
git clone /home/john/repo new_repo_name

Cloning a remote

Remote repos are stored in an online hosting service e.g., GitHub, Bitbucket, or Gitlab
We can clone a remote repo on to our local computer
- git clone [URL]
  
  ed., git clone https://github.com/datacamp/project

Identifying a remote

git remote

Git stores a remote tag in the new repo’s configuration
Getting more information
- git remote -v

Creating a remote

git remote add name URL

Defining remote names is useful for merging branches

git remote add george https://github.com/george_datacamp/repo

26.4.3 Pulling from a remote

Two ways to Synchronize local and remote repos

fetch and merge
1. Fetching from a remote
  
  git fetch remote_name local_branch
2. Synchronizing content
  
  git merge remote_name local_branch
pull
- Short cut of above 2 steps process
  
  git pull remote_name local_branch

Important to save locally before pulling from a remote

26.4.4 Pushing to a remote

git push

Save changes locally first
Push into remote_name from local_branch

git push remote_name local_branch

Resolving a conflict

git pull remote_name local_branch

Git will automatically open the nano text editor and ask us to add a message for the merge
Leave a message that we are pulling the latest report from the remote