cvs2git - convert a cvs repository into a git repository
]... OUTPUT-OPTIONS CVS-REPOS-PATH
Create a new git repository based on the version history stored in a CVS
repository. Each CVS commit will be mirrored in the git repository, including
such information as date of commit and id of the committer.
The output of this program are a "blobfile" and a
"dumpfile", which together can be loaded into a git repository using
is the filesystem path of the part of the CVS repository
that you want to convert. This path doesn't have to be the top level directory
of a CVS repository; it can point at a project within a repository, in which
case only that project will be converted. This path or one of its parent
directories has to contain a subdirectory called CVSROOT (though the CVSROOT
directory can be empty).
It is not possible directly to convert a CVS repository to which you only have
remote access, but the FAQ describes tools that may be used to create a local
copy of a remote CVS repository.
- Read the conversion options from path instead of
from the command line. This option allows far more conversion flexibility
than can be achieved using the command-line alone. See the documentation
for more information. Only the following command-line options are allowed
in combination with --options: -h/--help,
--help-passes, --version, -v/--verbose,
-q/ --quiet, -p/--pass/--passes,
--dry-run, --profile, --trunk-only,
--encoding, and --fallback-encoding. Options are processed
in the order specified on the command line.
- Write the "blob" data (containing revision
contents) to path.
- Write the revision data (branches and commits) to
- Do not create any output; just print what would
- Enables writing cvsusername@domain styleeMail
addresses to the git commits.
- Convert only trunk commits, not tags nor branches.
- Use encoding as the encoding for filenames, log
messages, and author names in the CVS repos. This option may be specified
multiple times, in which case the encodings are tried in order until one
succeeds. Default: ascii. See
http://docs.python.org/lib/standard-encodings.html for a list of other
- If none of the encodings specified with --encoding
succeed in decoding an author name or log message, then fall back to using
encoding in lossy 'replace' mode. Use of this option may cause
information to be lost, but at least it allows the conversion to run to
completion. This option only affects the encoding of log messages and
author names; there is no fallback encoding for filenames. (By using an
--options file, it is possible to specify a fallback encoding for
filenames.) Default: disabled.
- If a file appears both inside an outside of the CVS attic,
retain the attic version in an SVN subdirectory called 'Attic'. (Normally
this situation is treated as a fatal error.)
- Transform RCS/CVS symbol names before entering them into
Subversion. pattern is a Python regexp pattern that is matches
against the entire symbol name; replacement is a replacement using
Python's regexp reference syntax. You may specify any number of these
options; they will be applied in the order given on the command line.
- Read symbol conversion hints from path. The format
of path is the same as the format output by
--write-symbol-info, namely a text file with four
whitespace-separated columns: project-id, symbol,
conversion, and parent-lod-name. project-id is the
numerical ID of the project to which the symbol belongs, counting from 0.
project-id can be set to '.' if project-specificity is not needed.
symbol-name is the name of the symbol being specified.
conversion specifies how the symbol should be converted, and can be
one of the values 'branch', 'tag', or ´exclude'. If
conversion is '.', then this rule does not affect how the symbol is
converted. parent-lod-name is the name of the symbol from which
this symbol should sprout, or '.trunk.' if the symbol should sprout from
trunk. If parent-lod-name is omitted or '.', then this rule does
not affect the preferred parent of this symbol. The file may contain blank
lines or comment lines (lines whose first non-whitespace character is
- Specify how to convert ambiguous symbols (those that appear
in the CVS archive as both branches and tags). opt must be
'heuristic' (decide how to treat each ambiguous symbol based on whether it
was used more often as a branch/tag in CVS), 'strict' (no default; every
ambiguous symbol has to be resolved manually using --force-branch,
--force-tag, or --exclude), 'branch' (treat every ambiguous
symbol as a branch), 'tag' (treat every ambiguous symbol as a tag), or
'exclude' (do not convert ambiguous symbols). The default is
- Force symbols whose names match regexp to be
branches. regexp must match the whole symbol name.
- Force symbols whose names match regexp to be tags.
regexp must match the whole symbol name.
- Exclude branches and tags whose names match regexp
from the conversion. regexp must match the whole symbol name.
- Do not exclude branches that were only used for a single
import. (By default such branches are excluded because they are usually
created by the inappropriate use of cvs import.)
- Set the default username to name when cvs2svn needs
to generate a commit for which CVS does not record the original username.
This happens when a branch or tag is created. The default is to use no
author at all for such commits.
- Specify a file in the format of Subversion's config file,
whose [auto-props] section can be used to set arbitrary properties on
files in the Subversion repository based on their filenames. (The
[auto-props] section header must be present; other sections of the config
file, including the enable-auto-props setting, are ignored.) Filenames are
matched to the filename patterns case-insensitively.
- Specify an apache-style mime.types file for setting
- For files that don't have the kb expansion mode but have a
known mime type, set the eol-style based on the mime type. For such files,
set svn:eol-style to "native" if the mime type begins with
"text/", and leave it unset (i.e., no EOL translation)
otherwise. Files with unknown mime types are not affected by this option.
This option has no effect unless the --mime-types option is also
- Set svn:eol-style to style for files that don't have
the CVS ´kb' expansion mode and whose end-of-line translation mode
hasn't been determined by one of the other options. style must be
'binary' (default), 'native', 'CRLF', 'LF', or 'CR'.
- By default, cvs2svn sets svn:keywords on CVS files to
"author id date" if the mode of the RCS file in question is
either kv, kvl or unset. If you use the --keywords-off switch, cvs2svn
will not set svn:keywords for any file. While this will not touch the
keywords in the contents of your files, Subversion will not expand
- Include .cvsignore files in the output. (Normally
they are unneeded because cvs2svn sets the corresponding svn:ignore
- Record CVS revision numbers as file properties in the
Subversion repository. (Note that unless it is removed explicitly, the
last CVS revision number will remain associated with the file even after
the file is changed within Subversion.)
- Keyword mode to force, or no to use the default.
opt can be 'untouched' (keep literally as they are recorded in the
RCS file), ´collapsed' (retain the key but not the value),
'expanded' (fake the way CVS expands them, even for binaries), 'kept'
(just let CVS deal with keywords).
- Use CVS to extract revision contents. This option is slower
than --use-internal-co or --use-rcs.
- Use RCS 'co' to extract revision contents. This option is
faster than --use-cvs but fails in some cases.
- Use an external Python program to extract the file revision
contents from the RCS files and output them to the blobfile. This option
is much faster than --use-rcs or --use-cvs but leaves
keywords unexpanded and requires a separate, seekable blob file to write
to in parallel to the main cvs2git script.
- Set the path to use for temporary data. Default is a
directory called cvs2svn-tmp under the current directory.
- Path to the co program. (co is needed if the
--use-rcs option is used.)
- Path to the cvs program. (cvs is needed if
the --use-cvs option is used.)
- Execute only pass pass of the conversion.
pass can be specified by name or by number (see
- -p [start]:[end],
- Execute passes start through end of the
conversion (inclusive). start and end can be specified by
name or by number (see --help-passes). If start or
end is missing, it defaults to the first or last pass,
respectively. For this to work the earlier passes must have been completed
before on the same CVS repository, and the generated data files must be in
the temporary directory (see --tmpdir).
- Print the version number.
- -h, --help
- Print the usage message and exit with success.
- Print the numbers and names of the conversion passes and
exit with success.
- Output the unix-style manpage for this program to standard
- -v, --verbose
- Print more information while running. This option may be
specified twice to output voluminous debugging information.
- -q, --quiet
- Print less information while running. This option may be
specified twice to suppress all non-error output.
- Write to path symbol statistics and information
about how symbols were converted during CollateSymbolsPass.
- Prevent the deletion of temporary files.
- Profile with 'cProfile' (into file
A directory called cvs2svn-tmp
(or the directory specified by
) is used as scratch space for temporary data files.
Main authors are:
C. Michael Pilato <email@example.com>
Greg Stein <firstname.lastname@example.org>
Branko Čibej <email@example.com>
Blair Zajac <firstname.lastname@example.org>
Max Bowsher <email@example.com>
Brian Fitzpatrick <firstname.lastname@example.org>
Tobias Ringström <email@example.com>
Karl Fogel <firstname.lastname@example.org>
Erik Hülsmann <email@example.com>
David Summers <firstname.lastname@example.org>
Michael Haggerty <email@example.com>
Manpage was written for the Debian GNU/Linux system by Laszlo 'GCS' Boszormenyi
<firstname.lastname@example.org> (but may be used by others).
cvs(1), git(1), git-fast-import(1)