The purpose of cvs2git is to convert a CVS repository to a Git repository. All mechanism of conversion is based on reposurgeon tool. The conversion uses the version 3.1.2 of mercurial provided on jessie.
docker run -it lroche/cvs2git modulecvs cvs_remote_url
docker run -it lroche/cvs2git a2ps cvs.savannah.gnu.org
At the end of process, the script launchs a bash session to allow to check the result, here from the example you can get the a2ps-git directory by using docker cp command :
docker cp <containerId>:/work/a2ps/a2ps-git .
Better way can be to use a bind mount volume during your run:
docker run [-it] -v ~/gitconversion/a2ps:/work/a2ps lroche/cvs2git a2ps cvs.savannah.gnu.org
So here the result git repository will be in ~/gitconversion/a2ps/a2ps-git directory on host machine.
-
Reposurgeon uses rsync to connect on remote server so sometimes you could get some issues with it: there is a workaround, only if you have access on cvs server files (,v files), you can copy them on host machine in a directory named ${project-name}-mirror. If your project is a2ps and your workdir is ~/gitconversion/a2ps:
- create a directory named : ~/gitconversion/a2ps/a2ps-mirror
- place all your files ,v in a directory named ~/gitconversion/a2ps/a2ps-mirror/a2ps
- create a empty directory named CVSROOT here: ~/gitconversion/a2ps/a2ps-mirror/CVSROOT
You should have in ~/gitconversion/a2ps/ this following structure now:
├── a2ps-mirror
│ ├── CVSROOT
│ └── a2ps
└── a2ps.map
- The author map allows to specify a full name and email address for each local user ID in the repo you are converting. The expected name file is ${moduleCVSName}.map
So to set a author map file, create a map file in your host machine and use a bind mount volume to get it in container:
docker run [-it] -v ~/gitconversion/a2ps:/work/a2ps lroche/cvs2git a2ps cvs.savannah.gnu.org
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent