Skip to content

git.md

Latest commit

 

History

History
271 lines (182 loc) · 6.98 KB

git.md

File metadata and controls

271 lines (182 loc) · 6.98 KB

Git

Beginner guides

Undo changes

Undo changes to a local file

This is for undoing changes on an untracked file.

git checkout -- <file>

There is also the new command git restore.

git restore <file>

Type man git restore for more about this command, especially the --staged and --worktree options.

Undo a git add

git reset <file>

See here for more.

Undo an Unpublished Commit

git reset HEAD~

git reset is the command responsible for the undo. It will undo your last commit while leaving your working tree (the state of your files on disk) untouched. You'll need to add them again before you can commit them again.

More explanations here.

Hard reset

This will destroy any local modifications. Don't do it if you have uncommitted work you want to keep.

git reset --hard 0d1d7fc3 # Replace the commit hash by the one you want to rollback to

Alternatively, if there's work to keep:

git stash
git reset --hard 0d1d7fc3 # Replace the commit hash by the one you want to rollback to
git stash pop

This saves the modifications, then reapplies that patch after resetting. You could get merge conflicts, if you've modified things which were changed since the commit you reset to.

Source: SO.

Revert a Git repository to a previous commit

git revert --no-commit 0d1d7fc3..HEAD # Replace the commit hash by the one you want to rollback to
git commit

This will revert everything from the HEAD back to the commit hash (excluded), meaning it will recreate that commit state in the working tree as if every commit after 0d1d7fc3 had been walked back. You can then commit the current tree, and it will create a brand new commit essentially equivalent to the commit you "reverted" to.

The --no-commit flag lets git revert all the commits at once - otherwise you'll be prompted for a message for each commit in the range, littering your history with unnecessary new commits.

See this SO answer for detailled explanations.

IMPORTANT

Note that this only works if there was no merge commit since the commit we want to rollback to. See this answer if merge commits exist.

Push only specific commits

Sometimes there are a few commits pending to be pushed but you don't want to push all of them for some reason, e.g. partial deployment, and so you want to push them only up to a certain commit.

Then you can do this:

git push <remote> <commit hash>:<branch>

For example, you have 5 commits A > B > C > D > E pending (for simplicity, ABCDE are commit hashes), and you want to push up to commit "C". The following will then push A, B, and C to origin/main.

git push origin C:main

Sources

Copy file from another commit or branch

There are two options to perform this: using git checkout or git restore.

Source: StackOverflow.

Using git restore

git restore --source otherbranch path/to/myfile.txt

The difference with git checkout (see below) is that git checkout copies the file to the working directory (your files on disk) but also to the staging area (much like a git add). git checkout by default changes only the working directory.

Using git checkout

Run this from the branch where you want the file to end up:

git checkout otherbranch myfile.txt

General formulas:

git checkout <commit_hash> <relative_path_to_file_or_dir>
git checkout <remote_name>/<branch_name> <file_or_dir>

Some notes (from comments in the SO thread):

  • Using the commit hash, you can pull files from any commit.
  • This works for files and directories.
  • Overwrites the file myfile.txt and mydir.
  • Wildcards don't work, but relative paths do.
  • Multiple paths can be specified.

Fetch a remote branch

If you need to fetch a remote branch that is not in your local repository, use git switch.

git switch name-of-remote-branch

You will get an output similar to this:

branch 'foo-bar-branch' set up to track 'origin/foo-bar-branch'.
Switched to a new branch 'foo-bar-branch'

Reset a branch based on another

To reset a branch named my-feature-branch based on the main branch for example:

git switch main && git pull   # get the latest changes from remote
git switch my-feature-branch  # switch back to the feature branch
git reset --hard main

Using the stash

A simple and concise tutorial is available here.

The following sections are taken from the article.

What is the stash

Git features the stash, which is as much as a good place to store uncommitted changes. When you stash you changes, they will be stored, and your working copy will be reverted to HEAD (the last commit revision) of your code.

When you restore your stash, you changes are reapplied and you continue working on your code.

Stash your current changes

$ git stash save <optional message for later reference>

Example

git stash save 'a dummy stash operation'

You will get an output like:

Saved working directory and index state On main: a dummy stash operation

List current stashes

It is possible to have more than one stash. The stash works like a stack. Every time a new stash is saved, it's put on top of the stack.

git stash list
stash@{0}: On main: a dummy stash operation

The stash@{0} is the stash ID, it will be used to restore it later.

The stash ID changes with every stash made. stash@{0} refers to the last stash made.

Apply a stash

git stash apply stash@{0}

You may notice the stash is still there after you have applied it. You can drop it if you don't need it any more.

git stash drop stash@{0}

Apply and remove the last stash saved

git stash pop

Wipe all the stashes away

git stash clear

Search in commit messages

If you need to search for a commit which messages contains a given text, use:

git log --grep='pattern'

Note: This will only search in the history of the active branch.