Re: cvs2svn: raw or cooked?

Karl Fogel wrote:

> > 2. potential ambiguity if a log message's contents look like a
> > log message delimiter.
>
> Once a million years, but yeah, that's true.

I have seen log messages that said this:

        Copied something from somewhere else.
        Here are the the log entries from somewhere else.
        revision 1.7
        date: 2000/08/27 22:27:08; author: jrandom; state: Exp; lines: +3 -2
        A third log message.
        ----------------------------
        revision 1.6
        date: 2000/02/01 00:44:52; author: jrandom; state: Exp; lines: +2 -7
        Another log message.
        ----------------------------
        revision 1.5
        date: 1999/07/16 18:19:04; author: jrandom; state: Exp; lines: +7 -3
        A third log message.
        ----------------------------

Yeah, it's evil.

> You don't have worry about reading the XML, I think -- that's what
> we're working with right now, so you should get it "for free" by the
> time you're ready.

Next problem. I haven't started dealing with file contents yet, just
with the metadata. I have questions about vdelta. I'll start a new
thread with those questions.

> > BTW, here's my status. cvs2svn.pl now can read one RCS file and build
> > the correct data structure. I've written the directory tree walker as
> > a separate program, and I need to integrate it into cvs2svn.pl, then
> > it's time to check it all in.
>
> Congrats!

Yeah, but that's old status. Now I have a program that will suck in
the whole repository and build the right data structures. Next task
is to put some good test cases in.

> cd working_copy/foo/bar
> cvs add cvs2svn.pl
> cvs ci -m "Initial checkin." cvs2svn.pl

Yeah, that part's okay. I'm more concerned with making "make check"
do something. I'll start a new thread with those questions, too.

-- 
                                        K<bob>
kbob_at_jogger-egg.com, http://www.jogger-egg.com/