Difference between revisions of "CVS2SVN"

Here I'll describe how we transformed our ScummVM CVS to SVN. It may be helpful for other big projects facing same challenge.

Of course, SF.net has this nice "import CVS repository", but that one runs automatic [http://cvs2svn.tigris.org/ cvs2svn] script but we wanted more:

* That includes merge of scummvm-old and scummvm modules. scummvm module was created when we performed major project directory tree restructuring

Generally connection between moved files in SVN could be restored as shown in following chunk of real ScummVM repository dump:

Note lines marked by '+'. Those were added. This successfully restores SVN history link. 53 is the last revision number where dirent.h file was altered.

To successfully fullfill the task I wrote several simple scripts. I had to do it in scripts, not manually because restoration of connections is pretty lengthy process, and I wanted to minimise repository freeze time. So I first modified dump file locally, and later reapplied my changes to fresh dump. Those scripts are intentionally kept simple and no error checking is performed.

Whole dump takes over 1.4G, so there is no way to edit the file directly. Hence I came with simple idea of extracting just those Revision numbers and Node-paths, so they could be later replaced in original dump.

This stage took about 1.5 hours. Bottleneck is the disk subsytem. Overall size of produced data is over 1.4G.

Now I opened dump.nodes.in files in XEmacs and started to add those links. First, I searched all words 'delete' and studied each case. I had to consult files layout and CVS log messages to see either those files simply killed or really moved or renamed. Due to fact that some files were really renamed, and there are name clashes between files in directories this was not possible to fully automate the task, however big chunk of it could be scripted.

So what I did is to specify that Node-copyfrom-path: manually and left Node-copyfrom-rev blank. Then I recorded some simple macro in XEmacs, as that was quicker to do than writing yet another script. The macro was something like this:

With this I saw revision number in another window. Doublecheck that this is correct place and put that revision number to Node-copyfrom-rev field. However I didn't see any inconsistencies here, so I guess it could insert those numbers fully automatic.

This one is simple. What it does is to merge back those inserted lines and modify all internal path, so it will put all modules into separate directories on SVN repository:

Second regexp here is tricky, since Node-copyfrom-path could contain either /trunk/scummvm/blah or trunk/scummvm/blah and we have to keep that leading slash of it is present.

So after this stage amount of data on disk doubles since we have both merged and non-merged dumps. I kept non-merget dumps, so pass2 could be redone without performing lengthy pass1 over again.

Then repository was dumped, bzipped and uploaded to sf.net. Now it is possible to import existing dump, but at time when we do it, we had to submit a PR.