Revisions to How do I undo 'git add' before commit?

Remove unnecessary spaces

Source Link

edited Jan 15 at 15:03

5.5k
4
18
37

Git has commands for every action imaginable, but it needs extensive knowledge to get things right and because of that it is counter-intuitive at best...

What you did before:

Changed a file and used git add ., or git add <file>.

What you want:

Remove the file from the index, but keep it versioned and left with uncommitted changes in working copy:
```
 git reset HEAD <file>
```

Reset the file to the last state from HEAD, undoing changes and removing them from the index:

 # Think `svn revert <file>` IIRC.
 git reset HEAD <file>
 git checkout <file>

 # If you have a `<branch>` named like `<file>`, use:
 git checkout -- <file>

This is needed since git reset --hard HEAD won't work with single files.

Remove <file> from index and versioning, keeping the un-versioned file with changes in working copy:
```
 git rm --cached <file>
```
Remove <file> from working copy and versioning completely:
```
 git rm <file>
```

Git has commands for every action imaginable, but it needs extensive knowledge to get things right and because of that it is counter-intuitive at best...

What you did before:

Changed a file and used git add ., or git add <file>.

What you want:

Remove the file from the index, but keep it versioned and left with uncommitted changes in working copy:
```
 git reset HEAD <file>
```

Reset the file to the last state from HEAD, undoing changes and removing them from the index:

 # Think `svn revert <file>` IIRC.
 git reset HEAD <file>
 git checkout <file>

 # If you have a `<branch>` named like `<file>`, use:
 git checkout -- <file>

This is needed since git reset --hard HEAD won't work with single files.

Remove <file> from index and versioning, keeping the un-versioned file with changes in working copy:
```
 git rm --cached <file>
```
Remove <file> from working copy and versioning completely:
```
 git rm <file>
```

Git has commands for every action imaginable, but it needs extensive knowledge to get things right and because of that it is counter-intuitive at best...

What you did before:

Changed a file and used git add ., or git add <file>.

What you want:

Remove the file from the index, but keep it versioned and left with uncommitted changes in working copy:
```
git reset HEAD <file>
```
Reset the file to the last state from HEAD, undoing changes and removing them from the index:
```
# Think `svn revert <file>` IIRC.
git reset HEAD <file>
git checkout <file>

# If you have a `<branch>` named like `<file>`, use:
git checkout -- <file>
```
This is needed since git reset --hard HEAD won't work with single files.
Remove <file> from index and versioning, keeping the un-versioned file with changes in working copy:
```
git rm --cached <file>
```
Remove <file> from working copy and versioning completely:
```
git rm <file>
```

edited body

Source Link

edited Nov 27, 2020 at 14:10

sjas

19.2k
14
92
94

Git has commands for every action imaginable, but it needs extensive knowledge to get things right and because of that it is counter-intuitive at best...

What you did before:

Changed a file and used git add ., or git add <file>.

What you want:

Remove the file from the index, but keep it versioned and left with uncommitted changes in working copy:
```
 git reset headHEAD <file>
```
Reset the file to the last state from HEAD, undoing changes and removing them from the index:
```
 # Think `svn revert <file>` IIRC.
 git reset HEAD <file>
 git checkout <file>

 # If you have a `<branch>` named like `<file>`, use:
 git checkout -- <file>
```
This is needed since git reset --hard HEAD won't work with single files.
Remove <file> from index and versioning, keeping the un-versioned file with changes in working copy:
```
 git rm --cached <file>
```
Remove <file> from working copy and versioning completely:
```
 git rm <file>
```

Git has commands for every action imaginable, but it needs extensive knowledge to get things right and because of that it is counter-intuitive at best...

What you did before:

Changed a file and used git add ., or git add <file>.

What you want:

Remove the file from the index, but keep it versioned and left with uncommitted changes in working copy:
```
 git reset head <file>
```
Reset the file to the last state from HEAD, undoing changes and removing them from the index:
```
 # Think `svn revert <file>` IIRC.
 git reset HEAD <file>
 git checkout <file>

 # If you have a `<branch>` named like `<file>`, use:
 git checkout -- <file>
```
This is needed since git reset --hard HEAD won't work with single files.
Remove <file> from index and versioning, keeping the un-versioned file with changes in working copy:
```
 git rm --cached <file>
```
Remove <file> from working copy and versioning completely:
```
 git rm <file>
```

Git has commands for every action imaginable, but it needs extensive knowledge to get things right and because of that it is counter-intuitive at best...

What you did before:

Changed a file and used git add ., or git add <file>.

What you want:

Remove the file from the index, but keep it versioned and left with uncommitted changes in working copy:
```
 git reset HEAD <file>
```
Reset the file to the last state from HEAD, undoing changes and removing them from the index:
```
 # Think `svn revert <file>` IIRC.
 git reset HEAD <file>
 git checkout <file>

 # If you have a `<branch>` named like `<file>`, use:
 git checkout -- <file>
```
This is needed since git reset --hard HEAD won't work with single files.
Remove <file> from index and versioning, keeping the un-versioned file with changes in working copy:
```
 git rm --cached <file>
```
Remove <file> from working copy and versioning completely:
```
 git rm <file>
```

Active reading.

Source Link

edited Nov 3, 2019 at 13:30

Peter Mortensen

31.3k
22
109
132

Git has commands for every action imaginable, but it needs extensive knowledge to get things right and because of that it is counter-intuitive at best...

What you did before:

Changed a file and used git add ., or git add <file>.

What you want:

Remove the file from the index, but keep it versioned and left with uncommitted changes in working copy:
```
 git reset head <file>
```
Reset the file to the last state from HEAD, undoing changes and removing them from the index:
```
 # Think `svn revert <file>` IIRC.
 git reset HEAD <file>
 git checkout <file>

 # If you have a `<branch>` named like `<file>`, use:
 git checkout -- <file>
```
This is needed since git reset --hard HEAD won't work with single files.
Remove <file> from index and versioning, keeping the un-versioned file with changes in working copy:
```
 git rm --cached <file>
```
Remove <file> from working copy and versioning completely:
```
 git rm <file>
```

Git has commands for every action imaginable, but needs extensive knowledge to get things right and because of that it is counter-intuitive at best...

What you did before:

Changed a file and used git add ., or git add <file>.

What you want:

Remove the file from the index, but keep it versioned and left with uncommitted changes in working copy:
```
 git reset head <file>
```
Reset the file to the last state from HEAD, undoing changes and removing them from the index:
```
 # Think `svn revert <file>` IIRC.
 git reset HEAD <file>
 git checkout <file>

 # If you have a `<branch>` named like `<file>`, use:
 git checkout -- <file>
```
This is needed since git reset --hard HEAD won't work with single files.
Remove <file> from index and versioning, keeping the un-versioned file with changes in working copy:
```
 git rm --cached <file>
```
Remove <file> from working copy and versioning completely:
```
 git rm <file>
```

Git has commands for every action imaginable, but it needs extensive knowledge to get things right and because of that it is counter-intuitive at best...

What you did before:

Changed a file and used git add ., or git add <file>.

What you want:

Remove the file from the index, but keep it versioned and left with uncommitted changes in working copy:
```
 git reset head <file>
```
Reset the file to the last state from HEAD, undoing changes and removing them from the index:
```
 # Think `svn revert <file>` IIRC.
 git reset HEAD <file>
 git checkout <file>

 # If you have a `<branch>` named like `<file>`, use:
 git checkout -- <file>
```
This is needed since git reset --hard HEAD won't work with single files.
Remove <file> from index and versioning, keeping the un-versioned file with changes in working copy:
```
 git rm --cached <file>
```
Remove <file> from working copy and versioning completely:
```
 git rm <file>
```

formatting fix

Source Link

edited Oct 30, 2015 at 1:37

Mark Amery

150.3k
87
415
465

Loading

Cleaned up formatting, now it's a little easier to read.

Source Link

edited Apr 5, 2014 at 5:42

user456814

Loading

added 2 characters in body

Source Link

edited Nov 8, 2013 at 7:37

sjas

19.2k
14
92
94

Loading

added 56 characters in body

Source Link

edited Jul 14, 2013 at 22:34

sjas

19.2k
14
92
94

Loading

edited body

Source Link

edited Mar 29, 2013 at 11:20

sjas

19.2k
14
92
94

Loading

Source Link

created Mar 29, 2013 at 11:14

sjas

19.2k
14
92
94

Loading

Collectives™ on Stack Overflow

Return to Answer