I was working with SVN quite a long time and was happy about it. But number of active projects increased and some of them were needed to be supported even while travels. Since that time I moved to decentralised version control system and have control over my code-sources even when I’m travelling and without internet connection.
For code hosting I use SourceForge.net for opencource projects and BitBucket.org for private projects. BitBucket.org supports both Git and Mercurial and allows 5 repo users for free. I prefer to work with Mercurial.
How to move a code repository from remote SVN server to Mercurial BitBucket?
First of all we need to SVN command line utilities called
svnadmin.They are installed together with TortoiseSVN, if an option was checked in the installer.
If we access SVN through SSH then we need following:
- PuTTY utilities
plink.exemust be in PATH
- It is probably needed a system variable SVN_SSH with full path to
plink.exe. I have for example SVN_SSH = “C://Program Files (x86)//PuTTY//plink.exe” Note double slashes here.
- In the file
[tunnels]we should uncomment line
ssh = $SVN_SSH ssh -q
pageant.exeshould be started with a private key
"C:\Program Files (x86)\PuTTY\pageant.exe" c:\key_private.ppk
This is probably the most difficult part to make SVN work through SSH.
And now we are ready to move the repository.
1. Make a dump of the remote SVN repo to local file repo.dmp (this works with SVN from version 1.7)
svnrdump dump http://example.com/svn-repo/ > repo.dump
2. Create a new local SVN repo and restore data from dump into it
svnadmin create c:\svn-repo svnadmin load c:\svn-repo < repo.dump
3. Convert local SVN repo to Mercurial repo
hg convert c:\svn-repo c:\hg-repo
hg convert command does not work it means
convert hg extension is not enabled. To enable it open Hg Workbench and go to
File->Settings->Extensions check convert option there.
4. Transfer the created Mercurial repo to BitBucket
First create a new repo on BitBucket service. Then just push into it.
cd /path/to/my/new-local-repo hg push ssh://email@example.com/username/new-repo
This is the longest operation and it is nice to have a fast Internet connection to complete it fast.
If we do this push in Hg Workbench there is a progress there on how many files are pushed.
5. Make the BitBucket repo default.
There is a file
.hg/hgrc in repo folder, it has a
[paths] section. Write the following line there:
default = ssh://firstname.lastname@example.org/username/new-repo
If plink complains about server’s host key is not cached in the registry:
The server's host key is not cached in the registry. You have no guarantee that the server is the computer you think it is. The server's rsa2 key fingerprint is: ssh-rsa 2048 хх:хх:хх:хх:хх:хх:хх:хх:хх Connection abandoned
Then we need to run plink directly in command line and confirm the key caching:
plink -agent bitbucket.org The server's host key is not cached in the registry. You have no guarantee that the server is the computer you think it is. The server's rsa2 key fingerprint is: ssh-rsa 2048 97:8c:1b:f2:6f:14:6b:5c:3b:ec:aa:46:46:74:7c:40 If you trust this host, enter "y" to add the key to PuTTY's cache and carry on connecting. If you want to carry on connecting just once, without adding the key to the cache, enter "n". If you do not trust this host, press Return to abandon the connection. Store key in cache? (y/n) y