On Mon, Feb 9, 2009 at 4:53 PM, Felipe Contreras
<address@hidden> wrote:
> Then I tried to update to your latest revision and I'm getting a
> failure from fast import. I'm attaching the report.
Ah, I read your comments in the code, so it fails when the author is
not defined... I think that should be optional.
How about this patch?
I think it will fail for any author not listed in the --authors-file that *does* have an email address with < and > around it because it will get wrapped in an extra set of < and >.
I know that creating an author map to handle all of the various authors is a pretty major PITA. I'm not sure what the best thing to do here is though. I didn't think the previous code that was trying to detect various types of author names and fix them was particularly robust. I'm open to suggestions but at the moment I think getting a list of authors and commiters from these queries and scripting up a proper author map is probably the best approach to the problem.
// 'select distinct keypair from revision_certs'
// 'select distinct value from revision_certs where name = "author"'
Cheers,
Derek