ChatGPT解决这个技术问题 Extra ChatGPT

How do I delete a file from a Git repository?

How do I delete "file1.txt" from my repository?

git rm is the right answer, but remember that the file will still be there in history. If you want to remove a file because it had sensitive information, you'll need to do something more drastic. (Changing history, especially for content you've already pushed, is a drastic action, and should be avoided if possible.)
Note: on GitHub, you now can directly delete a file from the web interface (without having to even clone the repo). See my answer below.
@KeithThompson what steps might that be if I desperately want to do that?

j
johannchopin

Use git rm.

If you want to remove the file from the Git repository and the filesystem, use:

git rm file1.txt
git commit -m "remove file1.txt"

But if you want to remove the file only from the Git repository and not remove it from the filesystem, use:

git rm --cached file1.txt
git commit -m "remove file1.txt"

And to push changes to remote repo

git push origin branch_name

Also handy: git rm -r directory // To remove directory and content
That's not going to get rid of that file in other commits though.
@SaulOrtega: That's correct. To remove a file from previous commits (changing past history), see GitHub's help page on Remove sensitive data.
Take note this will delete the file locally too. If you only want to delete it from the repo do: git rm --cached file1.txt
It's worth noting that if that file contained sensitive information (e.g. credentials) you should change those credentials immediately. To quote GitHub "Once you have pushed a commit to GitHub, you should consider any data it contains to be compromised. If you committed a password, change it! If you committed a key, generate a new one."
C
Community

git rm file.txt removes the file from the repo but also deletes it from the local file system.

To remove the file from the repo and not delete it from the local file system use:
git rm --cached file.txt

The below exact situation is where I use git to maintain version control for my business's website, but the "mickey" directory was a tmp folder to share private content with a CAD developer. When he needed HUGE files, I made a private, unlinked directory and ftpd the files there for him to fetch via browser. Forgetting I did this, I later performed a git add -A from the website's base directory. Subsequently, git status showed the new files needing committing. Now I needed to delete them from git's tracking and version control...

Sample output below is from what just happened to me, where I unintentionally deleted the .003 file. Thankfully, I don't care what happened to the local copy to .003, but some of the other currently changed files were updates I just made to the website and would be epic to have been deleted on the local file system! "Local file system" = the live website (not a great practice, but is reality).

[~/www]$ git rm shop/mickey/mtt_flange_SCN.7z.003
error: 'shop/mickey/mtt_flange_SCN.7z.003' has local modifications
(use --cached to keep the file, or -f to force removal)
[~/www]$ git rm -f shop/mickey/mtt_flange_SCN.7z.003
rm 'shop/mickey/mtt_flange_SCN.7z.003'
[~/www]$ 
[~/www]$ git status
# On branch master
# Changes to be committed:
#   (use "git reset HEAD <file>..." to unstage)
#
#   deleted:    shop/mickey/mtt_flange_SCN.7z.003
#
# Changed but not updated:
#   (use "git add <file>..." to update what will be committed)
#   (use "git checkout -- <file>..." to discard changes in working directory)
#
#   modified:   shop/mickey/mtt_flange_SCN.7z.001
#   modified:   shop/mickey/mtt_flange_SCN.7z.002
[~/www]$ ls shop/mickey/mtt_flange_S*
shop/mickey/mtt_flange_SCN.7z.001  shop/mickey/mtt_flange_SCN.7z.002
[~/www]$ 
[~/www]$ 
[~/www]$ git rm --cached shop/mickey/mtt_flange_SCN.7z.002
rm 'shop/mickey/mtt_flange_SCN.7z.002'
[~/www]$ ls shop/mickey/mtt_flange_S*
shop/mickey/mtt_flange_SCN.7z.001  shop/mickey/mtt_flange_SCN.7z.002
[~/www]$ 
[~/www]$ 
[~/www]$ git status
# On branch master
# Changes to be committed:
#   (use "git reset HEAD <file>..." to unstage)
#
#   deleted:    shop/mickey/mtt_flange_SCN.7z.002
#   deleted:    shop/mickey/mtt_flange_SCN.7z.003
#
# Changed but not updated:
#   modified:   shop/mickey/mtt_flange_SCN.7z.001
[~/www]$

Update: This answer is getting some traffic, so I thought I'd mention my other Git answer shares a couple of great resources: This page has a graphic that help demystify Git for me. The "Pro Git" book is online and helps me a lot.


You'd then need to commit this with something like git commit -m "Just removing file1.txt" and then, if you have a remote repository, push the commit with something like git push origin master.
C
Community

First, if you are using git rm, especially for multiple files, consider any wildcard will be resolved by the shell, not by the git command.

git rm -- *.anExtension
git commit -m "remove multiple files"

But, if your file is already on GitHub, you can (since July 2013) directly delete it from the web GUI!

Simply view any file in your repository, click the trash can icon at the top, and commit the removal just like any other web-based edit.

Then "git pull" on your local repo, and that will delete the file locally too.
Which makes this answer a (roundabout) way to delete a file from git repo?
(Not to mention that a file on GitHub is in a "git repo")

https://i.stack.imgur.com/259au.jpg

(the commit will reflect the deletion of that file):

https://i.stack.imgur.com/fgXe9.png

And just like that, it’s gone. For help with these features, be sure to read our help articles on creating, moving, renaming, and deleting files. Note: Since it’s a version control system, Git always has your back if you need to recover the file later.

The last sentence means that the deleted file is still part of the history, and you can restore it easily enough (but not yet through the GitHub web interface):

See "Restore a deleted file in a Git repo".


You have to scroll to the bottom and hit "Commit changes" to commit the delete!
This does not work for large text files. The server tries to render the file as in the screenshot and then fails with a 500 error.
This answer is refering to github, the question is about a git repo. Not how to use the github interface.
y
yehonatan yehezkel

This is the only option that worked for me.

git filter-branch -f --index-filter 'git rm --cached --ignore-unmatch *.sql'

Note: Replace *.sql with your file name or file type. Be very careful because this will go through every commit and rip this file type out.

EDIT: pay attention - after this command you will not be able to push or pull - you will see the reject of 'unrelated history' you can use 'git push --force -u origin master' to push or pull


This is a great answer, but note that in contrast to the other answers, this command rewrites the commit history and removes all trace of the files matching the pattern. This is particularly useful if you found you accidentally committed some sensitive information (and, if needed can successfully use --force to rewrite history of any remote repos this has been pushed to).
That was exactly the problem that I had and it took some trial and error to get it right. A guy that I was working with had committed several large database dumps into a repo going back through many many commits. Which is why I used .sql in my example. My repo got so bloated and large that github wouldn't accept pushes anymore.
If you are using a remote master you need to force a commit and push it up with git push --force -u origin master
@AlbertMarkovski Sorry I wasn't specific. Github actually wouldn't accept the push because the repo was beyond their size limit. At least several GBs. They cut me off and emailed me about it. So, unfortunately, no amount of forcing a push would help at that point.
@JasonGlisson I've also faced that issue where a video got added to a repository by accident.
J
Jazi

Additionally, if it's a folder to be removed and it's subsequent child folders or files, use:

git rm -r foldername

This remove the folder from the local as well! Use git rm --cached -r <foldername>
J
James Polley

More generally, git help will help with at least simple questions like this:

zhasper@berens:/media/Kindle/documents$ git help
usage: git [--version] [--exec-path[=GIT_EXEC_PATH]] [--html-path] [-p|--paginate|--no-pager] [--bare] [--git-dir=GIT_DIR] [--work-tree=GIT_WORK_TREE] [--help] COMMAND [ARGS]

The most commonly used git commands are:
   add        Add file contents to the index
   :
   rm         Remove files from the working tree and from the index

For more complex things, I find git help confusing, but it's good for the simple things :)
It's not actually that great for git noobs because "index" is not a concept git noobs are familiar with. I speak from personal experience of being a git noob :) Also, I feel it's much less confusing to say that rm stages a file for deletion rather than removes from the index (though it's still meaningless for noobs).
Agreeing with romkyns: The description "Remove files from the working tree and from the index" doesn't sound to me like it's a way to remove the file from the repo. Maybe if I understood git better it would. Instead it sounds like you're just unstaging files (and optionally removing them from the working tree -- would that necessarily affect the repo?)
佚名

If you want to delete the file from the repo, but leave it in the the file system (will be untracked):

bykov@gitserver:~/temp> git rm --cached file1.txt
bykov@gitserver:~/temp> git commit -m "remove file1.txt from the repo"

If you want to delete the file from the repo and from the file system then there are two options:

If the file has no changes staged in the index: bykov@gitserver:~/temp> git rm file1.txt bykov@gitserver:~/temp> git commit -m "remove file1.txt" If the file has changes staged in the index: bykov@gitserver:~/temp> git rm -f file1.txt bykov@gitserver:~/temp> git commit -m "remove file1.txt"


n
nachoparker

git rm will only remove the file on this branch from now on, but it remains in history and git will remember it.

The right way to do it is with git filter-branch, as others have mentioned here. It will rewrite every commit in the history of the branch to delete that file.

But, even after doing that, git can remember it because there can be references to it in reflog, remotes, tags and such.

If you want to completely obliterate it in one step, I recommend you to use git forget-blob

https://ownyourbits.com/2017/01/18/completely-remove-a-file-from-a-git-repository-with-git-forget-blob/

It is easy, just do git forget-blob file1.txt.

This will remove every reference, do git filter-branch, and finally run the git garbage collector git gc to completely get rid of this file in your repo.


In 2021, I tried this and it didn't seem to work. I could still see the file history. Instead I ended up renaming the file with git mv, cleared all contents from the file committed it.
B
Blasanka

Note: if you want to delete file only from git use below:

git rm --cached file1.txt

If you want to delete also from hard disk:

git rm file1.txt

If you want to remove a folder(the folder may contain few files) so, you should remove using recursive command, as below:

git rm -r foldername

If you want to remove a folder inside another folder

git rm -r parentFolder/childFolder

Then, you can commit and push as usual. However, if you want to recover deleted folder, you can follow this: recover deleted files from git is possible.

From doc: git rm [-f | --force] [-n] [-r] [--cached] [--ignore-unmatch] [--quiet] [--] …​ OPTIONS … Files to remove. Fileglobs (e.g. *.c) can be given to remove all matching files. If you want Git to expand file glob characters, you may need to shell-escape them. A leading directory name (e.g. dir to remove dir/file1 and dir/file2) can be given to remove all files in the directory, and recursively all sub-directories, but this requires the -r option to be explicitly given. -f --force Override the up-to-date check. -n --dry-run Don’t actually remove any file(s). Instead, just show if they exist in the index and would otherwise be removed by the command. -r Allow recursive removal when a leading directory name is given. -- This option can be used to separate command-line options from the list of files, (useful when filenames might be mistaken for command-line options). --cached Use this option to unstage and remove paths only from the index. Working tree files, whether modified or not, will be left alone. --ignore-unmatch Exit with a zero status even if no files matched. -q --quiet git rm normally outputs one line (in the form of an rm command) for each file removed. This option suppresses that output.

Read more on official doc.


w
wat er

The answer by Greg Hewgill, that was edited by Johannchopin helped me, as I did not care about removing the file from the history completely. In my case, it was a directory, so the only change I did was using:

git rm -r --cached myDirectoryName

instead of "git rm --cached file1.txt" ..followed by:

git commit -m "deleted myDirectoryName from git"
git push origin branch_name

Thanks Greg Hewgill and Johannchopin!


S
Suhaib

Another way if you want to delete the file from your local folder using rm command and then push the changes to the remote server.

rm file1.txt

git commit -a -m "Deleting files"

git push origin master

T
Tomas Baran

According to the documentation.

git rm --cached file1.txt

When it comes to sensitive data—better not say that you removed the file but rather just include it in the last known commit:

0. Amend last commit

git commit --amend -CHEAD

If you want to delete the file from all git history, according to the documentation you should do the following:

1. Remove it from your local history

git filter-branch --force --index-filter \ "git rm --cached --ignore-unmatch PATH-TO-YOUR-FILE-WITH-SENSITIVE-DATA" \  --prune-empty --tag-name-filter cat -- --all
# Replace PATH-TO-YOUR-FILE-WITH-SENSITIVE-DATA with the path to the file you want to remove, not just its filename

Don't forget to include this file in .gitignore (If it's a file you never want to share (such as passwords...):

echo "YOUR-FILE-WITH-SENSITIVE-DATA" >> .gitignore
git add .gitignore
git commit -m "Add YOUR-FILE-WITH-SENSITIVE-DATA to .gitignore"

3. If you need to remove from the remote

git push origin --force --all

4. If you also need to remove it from tag releases:

git push origin --force --tags

O
OlegV

In my case I tried to remove file on github after few commits but save on computer

git filter-branch -f --index-filter 'git rm --cached --ignore-unmatch file_name_with_path' HEAD
git push --force -u origin master

and later this file was ignored


S
Sitesh

To delete a specific file

git rm filename

To clean all the untracked files from a directory recursively in single shot

git clean -fdx


git clean -fdx <-- Beware as it will remove everything that is not tracked.
@Philip By To clean all the untracked files from a directory recursively I meant exactly same.
Can be clarified in a better way imho.
j
jushusted

If you have the GitHub for Windows application, you can delete a file in 5 easy steps:

Click Sync.

Click on the directory where the file is located and select your latest version of the file.

Click on tools and select "Open a shell here."

In the shell, type: "rm {filename}" and hit enter.

Commit the change and resync.


A
Ajay Kumar

First,Remove files from local repository. git rm -r File-Name or, remove files only from local repository but from filesystem git rm --cached File-Name Secondly, Commit changes into local repository. git commit -m "unwanted files or some inline comments" Finally, update/push local changes into remote repository. git push


would this remove everything related with the accidentally added and pushed file?
@bapors no it will remain in git history
r
rojen

go to your project dir and type:

git filter-branch --tree-filter 'rm -f <deleted-file>' HEAD

after that push --force for delete file from all commits.

git push origin --force --all

A
Adam Burke

For the case where git rm doesn't suffice and the file needs to be removed from history: As the git filter-branch manual page now itself suggests using git-filter-repo, and I had to do this today, here's an example using that tool. It uses the example repo https://example/eguser/eg.git

Clone the repository into a new directory git clone https://example/eguser/eg.git Keep everything except the unwanted file. git-filter-repo --path file1.txt --invert-paths Add the remote repository origin back : git remote add origin https://example/eguser/eg.git. The git-filter-repo tool removes remote remote info by design and suggests a new remote repo (see point 4). This makes sense for big shared repos but might be overkill for getting rid a single newly added file as in this example. When happy with the contents of local, replace remote with it. git push --force -u origin master. Forcing is required due to the changed history.

Also note the useful --dry-run option and a good discussion in the linked manual on team and project dynamics before charging in and changing repository history.


j
jacanterbury

I tried a lot of the suggested options and none appeared to work (I won't list the various problems). What I ended up doing, which worked, was simple and intuitive (to me) was:

move the whole local repo elsewhere clone the repo again from master to your local drive copy back the files/folder from your original copy in #1 back into the new clone from #2 make sure that the problem large file is either not there or excluded in the .gitignore file do the usual git add/git commit/git push


g
glisu

After you have removed the file from the repo with git rm you can use BFG Repo-Cleaner to completely and easily obliterate the file from the repo history.


n
nitesh mishra

Just by going on the file in your github repository you can see the delete icon beside Raw|Blame and don't forget to click on commit changes button. And you can see that your file has been deleted.


N
Nick.McDermaid

I have obj and bin files that accidentally made it into the repo that I don't want polluting my 'changed files' list

After I noticed they went to the remote, I ignored them by adding this to .gitignore

/*/obj
/*/bin

Problem is they are already in the remote, and when they get changed, they pop up as changed and pollute the changed file list.

To stop seeing them, you need to delete the whole folder from the remote repository.

In a command prompt:

CD to the repo folder (i.e. C:\repos\MyRepo) I want to delete SSIS\obj. It seems you can only delete at the top level, so you now need to CD into SSIS: (i.e. C:\repos\MyRepo\SSIS) Now type the magic incantation git rm -r -f obj rm=remove -r = recursively remove -f = means force, cause you really mean it obj is the folder Now run git commit -m "remove obj folder"

I got an alarming message saying 13 files changed 315222 deletions

Then because I didn't want to have to look up the CMD line, I went into Visual Sstudio and did a Sync to apply it to the remote


E
Eugene

if your file is sensitive (for example, settings or keys you accidently added and committed) then you can remove it from all versions.

To remove from all versions use the command below (warning: careful because you won't be able to restore the removed file in the repo if do not have a copy):

Using Git

git filter-branch --index-filter 'git rm -rf --cached --ignore-unmatch file.ext' HEAD

Using git-filter-repo (get it here, new recommended way by Git)

git filter-repo --path file.ext

S
Sergi Ramón

If you need to remove files from a determined extension (for example, compiled files) you could do the following to remove them all at once:

git remove -f *.pyc

P
Poseidon

Incase if you don't file in your local repo but in git repo, then simply open file in git repo through web interface and find Delete button at right corner in interface. Click Here, To view interface Delete Option