Documentation for starters

Git Tutorial: part 1, part 2.

Git User’s manual. Everyday GIT With 20 Commands Or So. Git workflows.

Advanced documentation

Git Magic, with a number of translations.

Pro Git. The Book about git. Buy it at Amazon or download in PDF, mobi, or ePub form. It has translations to many different languages. Download Russian translation from GArik.

Git Wiki.

Git Buch (German).

Offline documentation

Git has builtin help: run git help $TOPIC. For example, run git help git or git help help.

Download and installation

Unix users: download and install using your package manager.

Microsoft Windows: download git-for-windows.

MacOS X: use git installed with XCode or download from MacPorts or git-osx-installer or install git with Homebrew: brew install git.

git-cola (repository) is a Git GUI written in Python and GPL licensed. Linux, Windows, MacOS X.

TortoiseGit is a Windows Shell Interface to Git based on TortoiseSVN; open source.

Initial configuration

This simple code is often appears in documentation, but it is important so let repeat it here. Git stores author and committer names/emails in every commit, so configure your real name and preferred email:

$ git config --global user.name "User Name"
$ git config --global user.email user.name@example.org

Updating local and remote-tracking branches

To update remote-tracking branches without updating local branches run git remote update [$REMOTE...]. For example:

$ git remote update
$ git remote update origin

$ git fetch $REMOTE $BRANCH

$ git fetch $REMOTE $BRANCH:$BRANCH

$ git pull $REMOTE $BRANCH

$ git fetch $REMOTE $BRANCH
$ git merge FETCH_HEAD

$ git fetch origin v1:v1  # Update v1
$ git pull --rebase origin master  # Update the current branch master
                                   # using rebase instead of merge
$ git merge v1

$ git checkout v1
$ git pull --rebase origin v1
$ git checkout master
$ git pull --rebase origin master
$ git merge v1

$ git pull origin

$ git pull

$ git push origin v1 master

$ git push origin v1:v1 master:master

$ git push origin

$ git push

$ git config push.default simple

$ git config push.default matching

$ git fetch origin tag $TAG1 tag $TAG2...

$ git fetch origin tag 1.4.2
$ git fetch origin v1:v1 tag 2.1.7

$ git push origin tag 1.4.2
$ git push origin v1 master tag 2.1.7

$ git push --tags origin

git checkout: restore file’s content

git checkout, for example, can be used to restore the content of file(s) to that one of a commit. Like this:

git checkout HEAD~ README

The commands restores the contents of README file to the last but one commit in the current branch. By default the commit ID is simply HEAD; i.e. git checkout README restores README to the latest commit.

(Do not use git checkout to view a content of a file in a commit, use git cat-file -p; e.g. git cat-file -p HEAD~:path/to/README).

git reset: remove (non-pushed) commits

git reset moves the head of the current branch. The head can be moved to point to any commit but it’s often used to remove a commit or a few (preferably, non-pushed ones) from the top of the branch - that is, to move the branch backward in order to undo a few (non-pushed) commits.

git reset has three modes of operation - soft, hard and mixed. Default is mixed. ProGit explains the difference very clearly. Bare repositories don’t have indices or working trees so in a bare repo only soft reset is possible.

git reflog: reference log

Removing commits with git reset or moving the head of a branch sounds dangerous and it is. But there is a way to undo: another reset back to the original commit. Git doesn’t remove commits immediately; unreferenced commits (in git terminology they are called “dangling commits”) stay in the database for some time (default is two weeks) so you can reset back to it or create a new branch pointing to the original commit.

For every move of a branch’s head - with git commit, git checkout, git fetch, git pull, git rebase, git reset and so on - git stores a reference log (reflog for short). For every move git stores where the head was. Command git reflog can be used to view (and manipulate) the log.

In addition to the moves of the head of every branch git stores the moves of the HEAD - a symbolic reference that (usually) names the current branch. HEAD is changed with git checkout $BRANCH.

By default git reflog shows the moves of the HEAD, i.e. the command is equivalent to git reflog HEAD. To show the moves of the head of a branch use the command git reflog $BRANCH.

So to undo a git reset lookup the original commit in git reflog, verify it with git show or git log and run git reset $COMMIT_ID. Git stores the move of the branch’s head in reflog, so you can undo that undo later again.

In a more complex situation you’d want to move some commits along with resetting the head of the branch. Cherry-pick them to the new branch. For example, if you want to reset the branch master back to the original commit but preserve two commits created in the current branch do something like:

$ git branch save-master  # create a new branch saving master
$ git reflog  # find the original place of master
$ git reset $COMMIT_ID
$ git cherry-pick save-master~ save-master
$ git branch -D save-master  # remove temporary branch

git revert: revert a commit

git revert reverts a commit or commits, that is, it creates a new commit or commits that revert(s) the effects of the given commits. It’s the only way to undo published commits (git commit --amend, git rebase and git reset change the branch in non-fast-forwardable ways so they should only be used for non-pushed commits.)

There is a problem with reverting a merge commit. git revert can undo the code created by the merge commit but it cannot undo the fact of merge. See the discussion How to revert a faulty merge.

One thing that cannot be undone

Whatever you undo, there is one thing that cannot be undone - overwritten uncommitted changes. Uncommitted changes don’t belong to git so git cannot help preserving them.

Most of the time git warns you when you’re going to execute a command that overwrites uncommitted changes. Git doesn’t allow you to switch branches with git checkout. It stops you when you’re going to rebase with non-clean working tree. It refuses to pull new commits over non-committed files.

But there are commands that do exactly that - overwrite files in the working tree. Commands like git checkout $PATHs or git reset --hard silently overwrite files including your uncommitted changes.

With that in mind you can understand the stance “commit early, commit often”. Commit as often as possible. Commit on every save in your editor or IDE. You can edit your commits before pushing - edit commit messages, change commits, reorder, combine, split, remove. But save your changes in git database, either commit changes or at least stash them with git stash.

Line endings

Git has builtin mechanisms to handle line endings between platforms with different end-of-line styles. To allow git to do CRLF conversion assign text attribute to files using .gitattributes. For files that have to have specific line endings assign eol attribute. For binary files the attribute is, naturally, binary.

For example:

$ cat .gitattributes
*.py text
*.txt text
*.png binary
/readme.txt eol=CRLF

To check what attributes git uses for files use git check-attr command. For example:

$ git check-attr -a -- \*.py

Useful assets

GitAlias (repository) is a big collection of aliases. A careful selection of aliases for frequently used commands could save you a lot of keystrokes!

GitIgnore and https://github.com/github/gitignore are collections of .gitignore files for all kinds of IDEs and programming languages. Python included!

pre-commit (repositories) is a framework for managing and maintaining multi-language pre-commit hooks. The framework is written in Python and has a lot of plugins for many programming languages.

Staging area

Staging area aka index aka cache is a distinguishing feature of git. Staging area is where git collects patches before committing them. Separation between collecting patches and commit phases provides a very useful feature of git: you can review collected patches before commit and even edit them - remove some hunks, add new hunks and review again.

To add files to the index use git add. Collecting patches before committing means you need to do that for every change, not only to add new (untracked) files. To simplify committing in case you just want to commit everything without reviewing run git commit --all (or just -a) - the command adds every changed tracked file to the index and then commit. To commit a file or files regardless of patches collected in the index run git commit [--only|-o] -- $FILE....

To add hunks of patches to the index use git add --patch (or just -p). To remove collected files from the index use git reset HEAD -- $FILE... To add/inspect/remove collected hunks use git add --interactive (-i).

To see the diff between the index and the last commit (i.e., collected patches) use git diff --cached. To see the diff between the working tree and the index (i.e., uncollected patches) use just git diff. To see the diff between the working tree and the last commit (i.e., both collected and uncollected patches) run git diff HEAD.

See WhatIsTheIndex and IndexCommandQuickref in Git Wiki.

Root

Git switches to the root (top-level directory of the project where .git subdirectory exists) before running any command. Git remembers though the directory that was current before the switch. Some programs take into account the current directory. E.g., git status shows file paths of changed and unknown files relative to the current directory; git grep searches below the current directory; git apply applies only those hunks from the patch that touch files below the current directory.

But most commands run from the root and ignore the current directory. Imagine, for example, that you have two work trees, one for the branch v1 and the other for master. If you want to merge v1 from a subdirectory inside the second work tree you must write commands as if you’re in the top-level dir. Let take two work trees, project-v1 and project, for example:

$ cd project/subdirectory
$ git fetch ../project-v1 v1:v1
$ git merge v1

Please note the path in git fetch ../project-v1 v1:v1 is ../project-v1 and not ../../project-v1 despite the fact that we run the commands from a subdirectory, not from the root.

ReReRe

Rerere is a mechanism that helps to resolve repeated merge conflicts. The most frequent source of recurring merge conflicts are topic branches that are merged into mainline and then the merge commits are removed; that’s often performed to test the topic branches and train rerere; merge commits are removed to have clean linear history and finish the topic branch with only one last merge commit.

Rerere works by remembering the states of tree before and after a successful commit. That way rerere can automatically resolve conflicts if they appear in the same files.

Rerere can be used manually with git rerere command but most often it’s used automatically. Enable rerere with these commands in a working tree:

$ git config rerere.enabled true
$ git config rerere.autoupdate true

You don’t need to turn rerere on globally - you don’t want rerere in bare repositories or single-branch repositories; you only need rerere in repos where you often perform merges and resolve merge conflicts.

See Rerere in The Book.

Database maintenance

Git object database and other files/directories under .git require periodic maintenance and cleanup. For example, commit editing left unreferenced objects (dangling objects, in git terminology) and these objects should be pruned to avoid collecting cruft in the DB. The command git gc is used for maintenance. Git automatically runs git gc --auto as a part of some commands to do quick maintenance. Users are recommended to run git gc --aggressive from time to time; git help gc recommends to run it every few hundred changesets; for more intensive projects it should be something like once a week and less frequently (biweekly or monthly) for lesser active projects.

git gc --aggressive not only removes dangling objects, it also repacks object database into indexed and better optimized pack(s); it also packs symbolic references (branches and tags). Another way to do it is to run git repack.

There is a well-known message from Linus Torvalds regarding “stupidity” of git gc --aggressive. The message can safely be ignored now. It is old and outdated, git gc --aggressive became much better since that time.

For those who still prefer git repack over git gc --aggressive the recommended parameters are git repack -a -d -f --depth=20 --window=250. See this detailed experiment for explanation of the effects of these parameters.

From time to time run git fsck [--strict] to verify integrity of the database. git fsck may produce a list of dangling objects; that’s not an error, just a reminder to perform regular maintenance.

Command-line options and arguments

git help cli recommends not to combine short options/flags. Most of the times combining works: git commit -av works perfectly, but there are situations when it doesn’t. E.g., git log -p -5 cannot be combined as git log -p5.

Some options have arguments, some even have default arguments. In that case the argument for such option must be spelled in a sticky way: -Oarg, never -O arg because for an option that has a default argument the latter means “use default value for option -O and pass arg further to the option parser”. For example, git grep has an option -O that passes a list of names of the found files to a program; default program for -O is a pager (usually less), but you can use your editor:

$ git grep -Ovim  # but not -O vim

BTW, if git is instructed to use less as the pager (i.e., if pager is not configured in git at all it uses less by default, or if it gets less from GIT_PAGER or PAGER environment variables, or if it was configured with git config [--global] core.pager less, or less is used in the command git grep -Oless) git grep passes +/$pattern option to less which is quite convenient. Unfortunately, git grep doesn’t pass the pattern if the pager is not exactly less, even if it’s less with parameters (something like git config [--global] core.pager less -FRSXgimq); fortunately, git grep -Oless always passes the pattern.

bash/zsh completion

It’s a bit hard to type git rebase --interactive --preserve-merges HEAD~5 manually even for those who are happy to use command-line, and this is where shell completion is of great help. Bash/zsh come with programmable completion, often automatically installed and enabled, so if you have bash/zsh and git installed, chances are you are already done - just go and use it at the command-line.

If you don’t have necessary bits installed, install and enable bash_completion package. If you want to upgrade your git completion to the latest and greatest download necessary file from git contrib.

Git-for-windows comes with git-bash for which bash completion is installed and enabled.

bash/zsh prompt

For command-line lovers shell prompt can carry a lot of useful information. To include git information in the prompt use git-prompt.sh. Read the detailed instructions in the file.

Search the Net for “git prompt” to find other prompt variants.

SSH connection sharing

SSH connection sharing is a feature of OpenSSH and perhaps derivatives like PuTTY. SSH connection sharing is a way to decrease ssh client startup time by establishing one connection and reusing it for all subsequent clients connecting to the same server. SSH connection sharing can be used to speedup a lot of short ssh sessions like scp, sftp, rsync and of course git over ssh. If you regularly fetch/pull/push from/to remote repositories accessible over ssh then using ssh connection sharing is recommended.

To turn on ssh connection sharing add something like this to your ~/.ssh/config:

Host *
ControlMaster auto
ControlPath ~/.ssh/mux-%r@%h:%p
ControlPersist 600

See OpenSSH wikibook and search for more information.

SSH connection sharing can be used at GitHub, GitLab and SourceForge repositories, but please be advised that BitBucket doesn’t allow it and forcibly closes master connection after a short inactivity period so you will see errors like this from ssh: “Connection to bitbucket.org closed by remote host.”