Mats Nilsson wrote:
> Max Bowsher wrote:
>> OK. Thinking it over, I think we might be able to achieve
>> this quite easily.
>>
>> Index: cvs2svn.py
>> ===================================================================
>> --- cvs2svn.py (revision 6941)
>> +++ cvs2svn.py (working copy)
>> @@ -173,6 +173,10 @@
>> else:
>> op = OP_CHANGE
>>
>> + if self.get_branch_name(revision) is None and not
>> trunk_rev.match(revision): + sys.stderr.write("Revision '%s' has no
>> branch name.\n" % revision) + sys.exit(1)
>
> It would be better to just log and disregard the superfluous revisions,
> than sys.exit(1). There might exist repositories with such stuff, but
> being otherwise well-formed. I'd be disappointed if it didn't at least
> convert the things it knew how to handle.
True, but it's a lot easier to bail out, than to make sure that skipping a
revision won't violate invariants that other parts of the code rely on. I
*think* it would actually be OK, so, I'll check, and if it looks reasonably
easy to do, just make it throw a VERY LOUD warning, and continue.
>> If you would like to file an ENHANCEMENT issue for this, I
>> will attach a
>> patch after I've checked it a little more carefully.
>
> http://subversion.tigris.org/issues/show_bug.cgi?id=1505
Thanks.
>> I'm unfamiliar with MKS, so I didn't know that mks2cvs was basing the
>> numbers on something internal to MKS.
>
> MKS Source Integrity is more or less RCS with some extra fields in the
> file header. It also uses some particular encoding for deltas in files
> marked as binary, I'm told.
>
>> Just a thought, but couldn't you just multiply each branch
>> number by 2 to
>> convert MKS->CVS?
>
> Clever. That might just work, and I'll look into it later on.
Good luck.
Max.
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Mon Sep 1 16:51:23 2003