Recover an old version of a file

Git tracks the content of your files and keeps their history. Accidents happen when developing code, and sometimes you break it. In such cases it’s convenient to know how to recover an old version of your file. We’ll show you how through a hands-on example.

Hands-on example

First, let’s create a file with some text. Run

echo "The original file. " > my_file.txt

Now we’ll add, commit and push it to the remote:

git add my_file.txt && git commit -m "added my_file.txt to learn about recovery of files" && git push

Now, we’ll modify the file, commit the changes and push it to the remote. Run

echo "Modified the file. " > my_file.txt

and

git commit -am "modified the file" && git push

Feel free to check that the content of the file actually changed.

Now we want to recover the old version of the file. We need to look up the commit hash of the previous version to recover it. This is easy – just run

git log

and look for the commit with the correct commit message. Copy the commit hash (the string of numbers and letters) of this commit. Now we can recover the file by

git checkout COMMIT_HASH_YOU_COPIED my_file.txt

This will revert the content of the file back to what it was before you modified it. You can check this by running

cat my_file.txt

Git automatically adds the changes, so you can simply go ahead and commit and push it to the remote:

git commit -m "recovered the old version of my_file.txt" && git push

Note

If you want to keep the file you currently have as well, you need to change its name. Otherwise you’ll overwrite it in the recovery process.