reposurgeon enables risky operations that version-control systems don’t want to let you do, such as (a) editing past comments and metadata, (b) excising commits, (c) coalescing commits, and (d) removing files and subtrees from repo history. The original motivation for reposurgeon was to clean up artifacts created by repository conversions.

reposurgeon is also useful for scripting very high-quality conversions from Subversion. It is better than git-svn at tag lifting, automatically cleaning up cvs2svn conversion artifacts, dealing with nonstandard repository layouts, recognizing branch merges, handling mixed-branch commits, and generally at coping with Subversion’s many odd corner cases. Normally Subversion repos should be analyzed at a rate of upwards of ten thousand commits per minute, though that rate can fall significantly on extremely large repositories.

An auxiliary program, repotool, performs various useful operations such as checkouts and tag listing in a VCS-independent manner. Yet another, repomapper, assists in automatically preparing contributor maps of CVS and SVN repositories.

The repocutter program is available for some specialized operations on Subversion dumpfiles; it may be useful in extracting portions of particularly gnarly Subversion repositories for conversion witth reposurgeon.

This distribution supports a generic conversion workflow using these tools, and includes the DVCS Migration Guide that describes how to use it.

The file reposurgeon-git-aliases` can be appended to your `~/.gitconfig to support working directly with action stamps in git.

Finally, an Emacs Lisp mode with useful functions for editing large comment message-boxes is included.

The hashbang lines in these tools invoke PyPy, an alternative Python implementation that performs much better than CPython on this code. You can use python2 or python3 to run these scripts, but it is best to have pypy installed. The performance difference is dramatic on large repositories.

We are in the process of moving this suite of tools from Python to Go for large performance increases. The subdirectory src contains the Go source. Do your builds through the Makefile, which sets the GOPATH variable to work around it not being embedded in a normal Go workspace.

To build the Go tools, simply run "make" in the toplevel directory. This also builds the suite documentation. Before running it the first time, run "make gosetup" to install package dependencies.

There is an extensive regression-test suite in the test/ directory. To test the correctness of this software, ensure that pypy and pylint and golang are installed and then type make check. For a portability check that includes Python 2 and Python 3 as well, do "make portcheck".

See reporting-bugs.asc for advice on how to troubleshoot problems with reposurgeon and report bugs.

The main reposurgeon website along with the documentation in HTML files lives at

The files Dockerfile, .dockerignore, .gitlab-ci.yml, and the contents of ci/ are not distributed; they are configuration for test builds on GitLab’s CI machinery.