Git remove file from not last commit

Crossposted on my blog
You can read here

CAUTION

Before continue we need to understand that this will break the Git history.

This means that after the fact, in case the commit in question was already pushed you will need to overwrite it with git push --force, and every repo co-worker will need to git pull --force.

Git and Ethereum, somewhat similar

Git has a chain of commits, called history.
Ethereum/Bitcoin has a chain of blocks, called blockchain.

Every commit (Git) or block (Ethereum) has a parent commit/block, so when you edit a block/commit you will change also its children.

Only what precedes the changed commit/block remains identical, all that follows will be a new thing.

This it's why they say "Git rebase rewrites history".

Note that in practice is not possibile to edit a commit, you can only overwriting it with a different version of it.

Git History and blockchain are both immutable.

Going back to our problem

My problem is that i've scraped some date on the Web and saved this data on a .json file, this file size is 220MB, and Github let us upload files with a limit of 100MB per file.

remote: error: File XXX is 220.00 MB; this exceeds GitHub's file size limit of 100 MB

But i became aware of that only when i tried to push the repo.
So i need to remove that file from my Git history.

Git History

This is the Git History of this example.

e6cfcb4 (HEAD -> main) build: update docs about docker adminer # OK
d817a6c feat(api): get-XXX-by-id now use internal data         # OK
9176b01 feat: seed database with scraped data  # 👈 WRONG ONE
c1301d1 build: install ts-node for running scripts             # OK
8c26fd5 feat: add XXX model to db                              # OK
e51acdb refactor: folder structure                             # OK
e0f1c7f build: add env.example                                 # OK
a7d7e34 feat: vote persists in db                              # OK
f32d6c4 build: imported shadow-database docker instance        # OK
987a245 build: implement prisma with first migration           # OK
...

We need to :

• keep all commit previous to 9176b01
• edit 9176b01 commit , removing the unwanted file
• commit a new version of 9176b01 with changes
• recreate all untouched commits after 9176b01

Git rebase interactive

I could be wrong if i say that this is the only way to resolve this problem, I am only a regular developer that stackoverflows when need some help, like all of us 😎.

If you know other solutions or has suggestions write them in the comments !

Anyway, let's start.

Start

As a recap, the wrong commit is 9176b01,
and last commit before the wrong one is c1301d1.

# Start rebasing from last correct commit
git rebase -i c1301d1

Git will ask instruction on the rebase process.

pick 9176b01 feat: seed database with scarped data
pick d817a6c feat(api): get-XXX-by-id now use internal data
pick e6cfcb4 build: update docs about docker adminer

# Rebase c1301d1..e6cfcb4 onto c1301d1 (3 commands)
#
# Commands:
# p, pick <commit> = use commit
# r, reword <commit> = use commit, but edit the commit message
# e, edit <commit> = use commit, but stop for amending
# s, squash <commit> = use commit, but meld into previous commit
# f, fixup [-C | -c] <commit> = like "squash" but keep only the previous
#                    commit's log message, unless -C is used, in which case
#                    keep only this commit's message; -c is same as -C but
#                    opens the editor
# x, exec <command> = run command (the rest of the line) using shell
# b, break = stop here (continue rebase later with 'git rebase --continue')
# d, drop <commit> = remove commit
# l, label <label> = label current HEAD with a name
# t, reset <label> = reset HEAD to a label

We must tell Git which commit we want to preserve and which ones we want to edit.
So we tell Git we want to edit 9176b01 by swapping pick with edit.

edit 9176b01 feat: seed database with scarped data
pick d817a6c feat(api): get-XXX-by-id now use internal data
pick e6cfcb4 build: update docs about docker adminer

# Rebase c1301d1..e6cfcb4 onto c1301d1 (3 commands)
#
# Commands:
# ...

Save with :wq and Git tell us

Stopped at 9176b01...  feat: seed database with scraped data
You can amend the commit now, with

  git commit --amend 

Once you are satisfied with your changes, run

  git rebase --continue

Now if we git log --oneline, we see that our last commit is the wrong one, we are back at that point.

Let's remove the unwanted file, by

rm path/to/unwanted-file

Tip:

If you want to keep the file but undo changes made by this wrong commit use git show path/to/changed-file | git apply -R instead of rm path/to/file

In this example both command produces that same effect because the file didn't exists before this commit, so undo is same as delete.

Add changes to staging area with

git add path/to/unwanted-file

and

git commit --amend

Edit the commit message if you want and confirm with :wq.

Then complete the rebase

git rebase --continue

Voilà !

Now we have rewritten the Git history from that point.

Conclusion

I write articles mainly to help future myself or to help the growth of tools I use in my work.

If this article was helpful to you leave a like.

Would you like me to talk about a particular topic ?

Tell me in the comments !