Configure
git config --global user.name "First Last"
git config --global user.email "first.last@emaildomain"
git config --global push.default simple
git config --global http.sslverify false
git config --global core.autocrlf false
git config --global alias.pull "pull --ff-only"
git config --global core.pager "less -F -X"
git config --global alias.logall "log --graph --decorate --pretty=oneline --abbrev-commit --all"List All System Properties
git config --list --globalStage the deletion of a file
git rm XGet status of the git repo
git status -sView the list of commits
git log --graph --onelineAdd an item to the staging area
git add XRemove an item from the staging area but keep on disk
git rm --cached XRemove an item from the staging area and remove from disk
git rm -f XUpdate last commit with contents of stage, if stage is empty then update the comment of the last commit.
git commit --amendUndo the last commit
git reset HEAD~1Move file from stage back to working area
git reset HEAD XUndo unstaged changes to file X
git checkout -- XRemove modified and untracked files
git cleanReset branch to BRANCH from origin
git fetch --all
git reset --hard origin/BRANCH
or
git fetch --all
git reset --hard FETCH_HEADView all branches
git branch -aCheckout and create a branch locally
git checkout -b feature/remove-rce-guiPush a branch that has been created locally and does not exist in the remote
git push --set-upstream origin feature/remove-rce-guiCreate a tracking branch
git branch --track bugfix/fixit origin/bugfix/fixitUpdate your subbranch with changes made to master - this is done before creating a Pull Request
# From your feature/bugfix branch run the following
git fetch origin master
git rebase origin/master
# Fix and merge issues and use git rebase --continue to complete the rebase
git push -fView all remotes
git remote -v View all tags
git tagView all branches
git branch -a -vDelete a remote
git remote rmDelete a branch
git branch -dRevert a commit by creating a commit that is the opposite of the given SHA
git revert <SHA>Edit last commit message
git commit --amend
or
git commit --amend -m "New Commit Message"Undo local changes that have not been committed yet
git checkout -- <edited file name>Reset local changes that have been committed
git reset <last good SHA>Reset local changes that have been committed and those local edits that have not been committed
git reset --hard <last good SHA>Use git reflog to get the history of HEAD label
$ git reflog
0ff0b0d (HEAD -> master) HEAD@{0}: commit: second one edited again
9b20920 HEAD@{1}: commit: third one
d0cd17c HEAD@{2}: commit: second one
f92cbc1 HEAD@{3}: commit (initial): first oneUse git reset to move the head back to the commit that you want
git reset d0cd17c$ git init .
$ tree .git/objects
.git/objects
├── info
└── pack
2 directories, 0 files
$Let's create an empty file and add it
$ touch file.txt
$ git add file.txt
$ tree .git/objects
.git/objects
├── e6
│ └── 9de29bb2d1d6434b8b29ae775ad8c2e48c5391
├── info
└── pack
3 directories, 1 file
$The content of the file has a header added to it to indicate what kind of content it is and then the result is zlib compressed and hashed using SHA-1. Subdirectories that start with the first two bytes of the hash are used to store the content, so the hash for the above is e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
The has will always be the same for the same content. So an empty file will always give the above hash which is efficient for sharing. Any empty files in git, regardless of location and filename will be represented by the above object.
Let's commit the file and see how many other objects are created.
$ tree .git/objects
.git/objects
├── b9
│ └── c6159978df9993fa252ba5bb54e989b57cd977
├── bd
│ └── d68b0120ca91384c1606468b4ca81b8f67c728
├── e6
│ └── 9de29bb2d1d6434b8b29ae775ad8c2e48c5391
├── info
└── pack
5 directories, 3 files
$There are now three objects per the above list. The empty file, the tree and the commit object itself.
Let's add some content to the file
$ echo "hello" >> file.txt
$ git add file.txt
$ tree .git/objects
.git/objects
├── b9
│ └── c6159978df9993fa252ba5bb54e989b57cd977
├── bd
│ └── d68b0120ca91384c1606468b4ca81b8f67c728
├── ce
│ └── 013625030ba8dba906f756967f9e9ca394464a
├── e6
│ └── 9de29bb2d1d6434b8b29ae775ad8c2e48c5391
├── info
└── pack
6 directories, 4 files
$Git will not pick up new content unless you add it. Note that we now have a new blob because we have new content.
We can use git cat-file -p to examine the contents of each blob.
$ git cat-file -p b9c6
tree bdd68b0120ca91384c1606468b4ca81b8f67c728
author thewaterwalker <[email protected]> 1572031531 +0000
committer thewaterwalker <[email protected]> 1572031531 +0000
first commit
$b9c6 is a commit object. It points to a tree and can have a parent as well.
$ git cat-file -p bdd6
100644 blob e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 file.txt
saj@querencia:~/test_git$bdd6 is the tree as referenced by the commit object. The tree object when printed shows the file permissions/type/SHA-1 hash. Note that they type could also be a tree which would represent other directories.
What this means is that objects are shared across different trees and different commits.
A ref is a pointer to an object. A branch is a ref to a commit. Tags are refs to commits but do not move automatically unlike branches when a new commit is added.
$ git tag a-tag -m "A tag"
$ git branch a-branch
$ tree .git/refs
.git/refs
├── heads
│ ├── a-branch
│ └── master
└── tags
└── a-tag
2 directories, 3 files
$ cat .git/refs/heads/a-branch
b9c6159978df9993fa252ba5bb54e989b57cd977
$master is always created by default. If you cat the content of a ref, it has a hash in it which refers to an object.
Reset just moves a ref to somewhere else. git records what you do to refs using the reflog. Reflog is in your local repository only and has just the last 90 days' worth of changes in it.
Over time, commits are merged into pack files which is also how git sends data across networks.