git-filter-branch - Rewrite branches
git filter-branch [--setup <command>] [--subdirectory-filter <directory>]
[--env-filter <command>] [--tree-filter <command>]
[--index-filter <command>] [--parent-filter <command>]
[--msg-filter <command>] [--commit-filter <command>]
[--tag-name-filter <command>] [--prune-empty]
[--original <namespace>] [-d <directory>] [-f | --force]
[--state-branch <branch>] [--] [<rev-list-options>...]
git filter-branch has a plethora of pitfalls that can
produce non-obvious manglings of the intended history rewrite (and can leave
you with little time to investigate such problems since it has such abysmal
performance). These safety and performance issues cannot be backward
compatibly fixed and as such, its use is not recommended. Please use an
alternative history filtering tool such as git filter-repo[1]. If you
still need to use git filter-branch, please carefully read the
section called “SAFETY” (and the section called
“PERFORMANCE”) to learn about the land mines of filter-branch,
and then vigilantly avoid as many of the hazards listed there as reasonably
possible.
Lets you rewrite Git revision history by rewriting the branches
mentioned in the <rev-list-options>, applying custom filters on each
revision. Those filters can modify each tree (e.g. removing a file or
running a perl rewrite on all files) or information about each commit.
Otherwise, all information (including original commit times or merge
information) will be preserved.
The command will only rewrite the positive refs mentioned
in the command line (e.g. if you pass a..b, only b will be
rewritten). If you specify no filters, the commits will be recommitted
without any changes, which would normally have no effect. Nevertheless, this
may be useful in the future for compensating for some Git bugs or such,
therefore such a usage is permitted.
NOTE: This command honors .git/info/grafts file and
refs in the refs/replace/ namespace. If you have any grafts or
replacement refs defined, running this command will make them permanent.
WARNING! The rewritten history will have different object
names for all the objects and will not converge with the original branch.
You will not be able to easily push and distribute the rewritten branch on
top of the original branch. Please do not use this command if you do not
know the full implications, and avoid using it anyway, if a simple single
commit would suffice to fix your problem. (See the "RECOVERING FROM
UPSTREAM REBASE" section in git-rebase(1) for further
information about rewriting published history.)
Always verify that the rewritten version is correct: The original
refs, if different from the rewritten ones, will be stored in the namespace
refs/original/.
Note that since this operation is very I/O expensive, it might be
a good idea to redirect the temporary directory off-disk with the -d
option, e.g. on tmpfs. Reportedly the speedup is very noticeable.
The filters are applied in the order as listed below. The
<command> argument is always evaluated in the shell context using the
eval command (with the notable exception of the commit filter, for
technical reasons). Prior to that, the $GIT_COMMIT environment
variable will be set to contain the id of the commit being rewritten. Also,
GIT_AUTHOR_NAME, GIT_AUTHOR_EMAIL, GIT_AUTHOR_DATE, GIT_COMMITTER_NAME,
GIT_COMMITTER_EMAIL, and GIT_COMMITTER_DATE are taken from the current
commit and exported to the environment, in order to affect the author and
committer identities of the replacement commit created by
git-commit-tree(1) after the filters have run.
If any evaluation of <command> returns a non-zero exit
status, the whole operation will be aborted.
A map function is available that takes an "original
sha1 id" argument and outputs a "rewritten sha1 id" if the
commit has been already rewritten, and "original sha1 id"
otherwise; the map function can return several ids on separate lines
if your commit filter emitted multiple commits.
--setup <command>
This is not a real filter executed for each commit but a
one time setup just before the loop. Therefore no commit-specific variables
are defined yet. Functions or variables defined here can be used or modified
in the following filter steps except the commit filter, for technical
reasons.
--subdirectory-filter <directory>
Only look at the history which touches the given
subdirectory. The result will contain that directory (and only that) as its
project root. Implies the section called “Remap to
ancestor”.
--env-filter <command>
This filter may be used if you only need to modify the
environment in which the commit will be performed. Specifically, you might
want to rewrite the author/committer name/email/time environment variables
(see
git-commit-tree(1) for details).
--tree-filter <command>
This is the filter for rewriting the tree and its
contents. The argument is evaluated in shell with the working directory set to
the root of the checked out tree. The new tree is then used as-is (new files
are auto-added, disappeared files are auto-removed - neither .gitignore files
nor any other ignore rules HAVE ANY EFFECT!).
--index-filter <command>
This is the filter for rewriting the index. It is similar
to the tree filter but does not check out the tree, which makes it much
faster. Frequently used with
git rm --cached --ignore-unmatch ..., see
EXAMPLES below. For hairy cases, see
git-update-index(1).
--parent-filter <command>
This is the filter for rewriting the commit’s
parent list. It will receive the parent string on stdin and shall output the
new parent string on stdout. The parent string is in the format described in
git-commit-tree(1): empty for the initial commit, "-p parent"
for a normal commit and "-p parent1 -p parent2 -p parent3 ..." for a
merge commit.
--msg-filter <command>
This is the filter for rewriting the commit messages. The
argument is evaluated in the shell with the original commit message on
standard input; its standard output is used as the new commit message.
--commit-filter <command>
This is the filter for performing the commit. If this
filter is specified, it will be called instead of the
git commit-tree
command, with arguments of the form "<TREE_ID> [(-p
<PARENT_COMMIT_ID>)...]" and the log message on stdin. The commit
id is expected on stdout.
As a special extension, the commit filter may emit multiple commit
ids; in that case, the rewritten children of the original commit will have
all of them as parents.
You can use the map convenience function in this filter,
and other convenience functions, too. For example, calling skip_commit
"$@" will leave out the current commit (but not its changes!
If you want that, use git rebase instead).
You can also use the git_commit_non_empty_tree
"$@" instead of git commit-tree "$@" if you
don’t wish to keep commits with a single parent and that makes no
change to the tree.
--tag-name-filter <command>
This is the filter for rewriting tag names. When passed,
it will be called for every tag ref that points to a rewritten object (or to a
tag object which points to a rewritten object). The original tag name is
passed via standard input, and the new tag name is expected on standard
output.
The original tags are not deleted, but can be overwritten; use
"--tag-name-filter cat" to simply update the tags. In this case,
be very careful and make sure you have the old tags backed up in case the
conversion has run afoul.
Nearly proper rewriting of tag objects is supported. If the tag
has a message attached, a new tag object will be created with the same
message, author, and timestamp. If the tag has a signature attached, the
signature will be stripped. It is by definition impossible to preserve
signatures. The reason this is "nearly" proper, is because ideally
if the tag did not change (points to the same object, has the same name,
etc.) it should retain any signature. That is not the case, signatures will
always be removed, buyer beware. There is also no support for changing the
author or timestamp (or the tag message for that matter). Tags which point
to other tags will be rewritten to point to the underlying commit.
--prune-empty
Some filters will generate empty commits that leave the
tree untouched. This option instructs git-filter-branch to remove such commits
if they have exactly one or zero non-pruned parents; merge commits will
therefore remain intact. This option cannot be used together with
--commit-filter, though the same effect can be achieved by using the
provided git_commit_non_empty_tree function in a commit filter.
--original <namespace>
Use this option to set the namespace where the original
commits will be stored. The default value is refs/original.
-d <directory>
Use this option to set the path to the temporary
directory used for rewriting. When applying a tree filter, the command needs
to temporarily check out the tree to some directory, which may consume
considerable space in case of large projects. By default it does this in the
.git-rewrite/ directory but you can override that choice by this
parameter.
-f, --force
git filter-branch refuses to start with an
existing temporary directory or when there are already refs starting with
refs/original/, unless forced.
--state-branch <branch>
This option will cause the mapping from old to new
objects to be loaded from named branch upon startup and saved as a new commit
to that branch upon exit, enabling incremental of large trees. If
<branch> does not exist it will be created.
<rev-list options>...
Arguments for git rev-list. All positive refs
included by these options are rewritten. You may also specify options such as
--all, but you must use -- to separate them from the git
filter-branch options. Implies the section called “Remap to
ancestor”.
By using git-rev-list(1) arguments, e.g., path limiters,
you can limit the set of revisions which get rewritten. However, positive
refs on the command line are distinguished: we don’t let them be
excluded by such limiters. For this purpose, they are instead rewritten to
point at the nearest ancestor that was not excluded.
On success, the exit status is 0. If the filter
can’t find any commits to rewrite, the exit status is 2. On
any other error, the exit status may be any other non-zero value.
Suppose you want to remove a file (containing confidential
information or copyright violation) from all commits:
git filter-branch --tree-filter 'rm filename' HEAD
However, if the file is absent from the tree of some commit, a
simple rm filename will fail for that tree and commit. Thus you may
instead want to use rm -f filename as the script.
Using --index-filter with git rm yields a
significantly faster version. Like with using rm filename, git rm
--cached filename will fail if the file is absent from the tree of a
commit. If you want to "completely forget" a file, it does not
matter when it entered history, so we also add --ignore-unmatch:
git filter-branch --index-filter 'git rm --cached --ignore-unmatch filename' HEAD
Now, you will get the rewritten history saved in HEAD.
To rewrite the repository to look as if foodir/ had been
its project root, and discard all other history:
git filter-branch --subdirectory-filter foodir -- --all
Thus you can, e.g., turn a library subdirectory into a repository
of its own. Note the -- that separates filter-branch options
from revision options, and the --all to rewrite all branches and
tags.
To set a commit (which typically is at the tip of another history)
to be the parent of the current initial commit, in order to paste the other
history behind the current history:
git filter-branch --parent-filter 'sed "s/^\$/-p <graft-id>/"' HEAD
(if the parent string is empty - which happens when we are dealing
with the initial commit - add graftcommit as a parent). Note that this
assumes history with a single root (that is, no merge without common
ancestors happened). If this is not the case, use:
git filter-branch --parent-filter \
'test $GIT_COMMIT = <commit-id> && echo "-p <graft-id>" || cat' HEAD
or even simpler:
git replace --graft $commit-id $graft-id
git filter-branch $graft-id..HEAD
To remove commits authored by "Darl McBribe" from the
history:
git filter-branch --commit-filter '
if [ "$GIT_AUTHOR_NAME" = "Darl McBribe" ];
then
skip_commit "$@";
else
git commit-tree "$@";
fi' HEAD
The function skip_commit is defined as follows:
skip_commit()
{
shift;
while [ -n "$1" ];
do
shift;
map "$1";
shift;
done;
}
The shift magic first throws away the tree id and then the -p
parameters. Note that this handles merges properly! In case Darl committed a
merge between P1 and P2, it will be propagated properly and all children of
the merge will become merge commits with P1,P2 as their parents instead of
the merge commit.
NOTE the changes introduced by the commits, and which are
not reverted by subsequent commits, will still be in the rewritten branch.
If you want to throw out changes together with the commits, you
should use the interactive mode of git rebase.
You can rewrite the commit log messages using --msg-filter.
For example, git svn-id strings in a repository created by git
svn can be removed this way:
git filter-branch --msg-filter '
sed -e "/^git-svn-id:/d"
'
If you need to add Acked-by lines to, say, the last 10
commits (none of which is a merge), use this command:
git filter-branch --msg-filter '
cat &&
echo "Acked-by: Bugs Bunny <bunny@bugzilla.org>"
' HEAD~10..HEAD
The --env-filter option can be used to modify committer
and/or author identity. For example, if you found out that your commits have
the wrong identity due to a misconfigured user.email, you can make a
correction, before publishing the project, like this:
git filter-branch --env-filter '
if test "$GIT_AUTHOR_EMAIL" = "root@localhost"
then
GIT_AUTHOR_EMAIL=john@example.com
fi
if test "$GIT_COMMITTER_EMAIL" = "root@localhost"
then
GIT_COMMITTER_EMAIL=john@example.com
fi
' -- --all
To restrict rewriting to only part of the history, specify a
revision range in addition to the new branch name. The new branch name will
point to the top-most revision that a git rev-list of this range will
print.
Consider this history:
D--E--F--G--H
/ /
A--B-----C
To rewrite only commits D,E,F,G,H, but leave A, B and C alone,
use:
git filter-branch ... C..H
To rewrite commits E,F,G,H, use one of these:
git filter-branch ... C..H --not D
git filter-branch ... D..H --not C
To move the whole tree into a subdirectory, or remove it from
there:
git filter-branch --index-filter \
'git ls-files -s | sed "s-\t\"*-&newsubdir/-" |
GIT_INDEX_FILE=$GIT_INDEX_FILE.new \
git update-index --index-info &&
mv "$GIT_INDEX_FILE.new" "$GIT_INDEX_FILE"' HEAD
git-filter-branch can be used to get rid of a subset of files,
usually with some combination of --index-filter and
--subdirectory-filter. People expect the resulting repository to be
smaller than the original, but you need a few more steps to actually make it
smaller, because Git tries hard not to lose your objects until you tell it
to. First make sure that:
•You really removed all variants of a filename, if
a blob was moved over its lifetime. git log --name-only --follow --all --
filename can help you find renames.
•You really filtered all refs: use
--tag-name-filter cat -- --all when calling git-filter-branch.
Then there are two ways to get a smaller repository. A safer way
is to clone, that keeps your original intact.
•Clone it with
git clone
file:///path/to/repo. The clone will not have the removed objects. See
git-clone(1). (Note that cloning with a plain path just hardlinks
everything!)
If you really don’t want to clone it, for whatever reasons,
check the following points instead (in this order). This is a very
destructive approach, so make a backup or go back to cloning it. You
have been warned.
•Remove the original refs backed up by
git-filter-branch: say git for-each-ref --format="%(refname)"
refs/original/ | xargs -n 1 git update-ref -d.
•Expire all reflogs with git reflog expire
--expire=now --all.
•Garbage collect all unreferenced objects with
git gc --prune=now (or if your git-gc is not new enough to support
arguments to --prune, use git repack -ad; git prune
instead).
The performance of git-filter-branch is glacially slow; its design
makes it impossible for a backward-compatible implementation to ever be
fast:
•In editing files, git-filter-branch by design
checks out each and every commit as it existed in the original repo. If your
repo has 10^5 files and 10^5 commits, but each commit only
modifies five files, then git-filter-branch will make you do 10^10
modifications, despite only having (at most) 5*10^5 unique blobs.
•If you try and cheat and try to make
git-filter-branch only work on files modified in a commit, then two things
happen
•you run into problems with deletions whenever the
user is simply trying to rename files (because attempting to delete files that
don’t exist looks like a no-op; it takes some chicanery to remap
deletes across file renames when the renames happen via arbitrary
user-provided shell)
•even if you succeed at the
map-deletes-for-renames chicanery, you still technically violate backward
compatibility because users are allowed to filter files in ways that depend
upon topology of commits instead of filtering solely based on file contents or
names (though this has not been observed in the wild).
•Even if you don’t need to edit files but
only want to e.g. rename or remove some and thus can avoid checking out each
file (i.e. you can use --index-filter), you still are passing shell snippets
for your filters. This means that for every commit, you have to have a
prepared git repo where those filters can be run. That’s a significant
setup.
•Further, several additional files are created or
updated per commit by git-filter-branch. Some of these are for supporting the
convenience functions provided by git-filter-branch (such as map()), while
others are for keeping track of internal state (but could have also been
accessed by user filters; one of git-filter-branch’s regression tests
does so). This essentially amounts to using the filesystem as an IPC mechanism
between git-filter-branch and the user-provided filters. Disks tend to be a
slow IPC mechanism, and writing these files also effectively represents a
forced synchronization point between separate processes that we hit with every
commit.
•The user-provided shell commands will likely
involve a pipeline of commands, resulting in the creation of many processes
per commit. Creating and running another process takes a widely varying amount
of time between operating systems, but on any platform it is very slow
relative to invoking a function.
•git-filter-branch itself is written in shell,
which is kind of slow. This is the one performance issue that could be
backward-compatibly fixed, but compared to the above problems that are
intrinsic to the design of git-filter-branch, the language of the tool itself
is a relatively minor issue.
•Side note: Unfortunately, people tend to fixate
on the written-in-shell aspect and periodically ask if git-filter-branch could
be rewritten in another language to fix the performance issues. Not only does
that ignore the bigger intrinsic problems with the design, it’d help
less than you’d expect: if git-filter-branch itself were not shell,
then the convenience functions (map(), skip_commit(), etc) and the
--setup argument could no longer be executed once at the beginning of
the program but would instead need to be prepended to every user filter (and
thus re-executed with every commit).
The git filter-repo[1] tool is an alternative to
git-filter-branch which does not suffer from these performance problems or
the safety problems (mentioned below). For those with existing tooling which
relies upon git-filter-branch, git filter-repo also provides
filter-lamely[2], a drop-in git-filter-branch replacement (with a few
caveats). While filter-lamely suffers from all the same safety issues as
git-filter-branch, it at least ameliorates the performance issues a
little.
git-filter-branch is riddled with gotchas resulting in various
ways to easily corrupt repos or end up with a mess worse than what you
started with:
•Someone can have a set of "working and
tested filters" which they document or provide to a coworker, who then
runs them on a different OS where the same commands are not working/tested
(some examples in the git-filter-branch manpage are also affected by this).
BSD vs. GNU userland differences can really bite. If lucky, error messages are
spewed. But just as likely, the commands either don’t do the filtering
requested, or silently corrupt by making some unwanted change. The unwanted
change may only affect a few commits, so it’s not necessarily obvious
either. (The fact that problems won’t necessarily be obvious means they
are likely to go unnoticed until the rewritten history is in use for quite a
while, at which point it’s really hard to justify another flag-day for
another rewrite.)
•Filenames with spaces are often mishandled by
shell snippets since they cause problems for shell pipelines. Not everyone is
familiar with find -print0, xargs -0, git-ls-files -z, etc. Even people who
are familiar with these may assume such flags are not relevant because someone
else renamed any such files in their repo back before the person doing the
filtering joined the project. And often, even those familiar with handling
arguments with spaces may not do so just because they aren’t in the
mindset of thinking about everything that could possibly go wrong.
•Non-ascii filenames can be silently removed
despite being in a desired directory. Keeping only wanted paths is often done
using pipelines like git ls-files | grep -v ^WANTED_DIR/ | xargs git
rm. ls-files will only quote filenames if needed, so folks may not notice
that one of the files didn’t match the regex (at least not until
it’s much too late). Yes, someone who knows about core.quotePath can
avoid this (unless they have other special characters like \t, \n, or "),
and people who use ls-files -z with something other than grep can avoid this,
but that doesn’t mean they will.
•Similarly, when moving files around, one can find
that filenames with non-ascii or special characters end up in a different
directory, one that includes a double quote character. (This is technically
the same issue as above with quoting, but perhaps an interesting different way
that it can and has manifested as a problem.)
•It’s far too easy to accidentally mix up
old and new history. It’s still possible with any tool, but
git-filter-branch almost invites it. If lucky, the only downside is users
getting frustrated that they don’t know how to shrink their repo and
remove the old stuff. If unlucky, they merge old and new history and end up
with multiple "copies" of each commit, some of which have unwanted
or sensitive files and others which don’t. This comes about in multiple
different ways:
•the default to only doing a partial history
rewrite (--all is not the default and few examples show it)
•the fact that there’s no automatic
post-run cleanup
•the fact that --tag-name-filter (when used to
rename tags) doesn’t remove the old tags but just adds new ones with
the new name
•the fact that little educational information is
provided to inform users of the ramifications of a rewrite and how to avoid
mixing old and new history. For example, this man page discusses how users
need to understand that they need to rebase their changes for all their
branches on top of new history (or delete and reclone), but that’s only
one of multiple concerns to consider. See the "DISCUSSION" section
of the git filter-repo manual page for more details.
•Annotated tags can be accidentally converted to
lightweight tags, due to either of two issues:
•Someone can do a history rewrite, realize they
messed up, restore from the backups in refs/original/, and then redo their
git-filter-branch command. (The backup in refs/original/ is not a real backup;
it dereferences tags first.)
•Running git-filter-branch with either --tags or
--all in your <rev-list-options>. In order to retain annotated tags as
annotated, you must use --tag-name-filter (and must not have restored from
refs/original/ in a previously botched rewrite).
•Any commit messages that specify an encoding will
become corrupted by the rewrite; git-filter-branch ignores the encoding, takes
the original bytes, and feeds it to commit-tree without telling it the proper
encoding. (This happens whether or not --msg-filter is used.)
•Commit messages (even if they are all UTF-8) by
default become corrupted due to not being updated — any references to
other commit hashes in commit messages will now refer to no-longer-extant
commits.
•There are no facilities for helping users find
what unwanted crud they should delete, which means they are much more likely
to have incomplete or partial cleanups that sometimes result in confusion and
people wasting time trying to understand. (For example, folks tend to just
look for big files to delete instead of big directories or extensions, and
once they do so, then sometime later folks using the new repository who are
going through history will notice a build artifact directory that has some
files but not others, or a cache of dependencies (node_modules or similar)
which couldn’t have ever been functional since it’s missing some
files.)
•If --prune-empty isn’t specified, then the
filtering process can create hoards of confusing empty commits
•If --prune-empty is specified, then intentionally
placed empty commits from before the filtering operation are also pruned
instead of just pruning commits that became empty due to filtering
rules.
•If --prune-empty is specified, sometimes empty
commits are missed and left around anyway (a somewhat rare bug, but it
happens...)
•A minor issue, but users who have a goal to
update all names and emails in a repository may be led to --env-filter which
will only update authors and committers, missing taggers.
•If the user provides a --tag-name-filter that
maps multiple tags to the same name, no warning or error is provided;
git-filter-branch simply overwrites each tag in some undocumented pre-defined
order resulting in only one tag at the end. (A git-filter-branch regression
test requires this surprising behavior.)
Also, the poor performance of git-filter-branch often leads to
safety issues:
•Coming up with the correct shell snippet to do
the filtering you want is sometimes difficult unless you’re just doing
a trivial modification such as deleting a couple files. Unfortunately, people
often learn if the snippet is right or wrong by trying it out, but the
rightness or wrongness can vary depending on special circumstances (spaces in
filenames, non-ascii filenames, funny author names or emails, invalid
timezones, presence of grafts or replace objects, etc.), meaning they may have
to wait a long time, hit an error, then restart. The performance of
git-filter-branch is so bad that this cycle is painful, reducing the time
available to carefully re-check (to say nothing about what it does to the
patience of the person doing the rewrite even if they do technically have more
time available). This problem is extra compounded because errors from broken
filters may not be shown for a long time and/or get lost in a sea of output.
Even worse, broken filters often just result in silent incorrect
rewrites.
•To top it all off, even when users finally find
working commands, they naturally want to share them. But they may be unaware
that their repo didn’t have some special cases that someone
else’s does. So, when someone else with a different repository runs the
same commands, they get hit by the problems above. Or, the user just runs
commands that really were vetted for special cases, but they run it on a
different OS where it doesn’t work, as noted above.
- 1.
- git filter-repo
- 2.
- filter-lamely