On Mon, 2006-06-19 at 10:18 +0200, Lars Kneschke wrote: > Am Montag, 19. Juni 2006 09:52 schrieb Paul J Stevens: > > Aaron Stone wrote: > > > Just for fun, try "SELECT * FROM dbmail_headername". I don't think you > > > should see more than 50-100 entries. I have 12729. The majority of them > > > are like this: > > > > Do you also get this for newly inserted messages, or just for converted > > 'old' messages? I'm guessing the latter, which may make this a > > retrieval/decoding issue.
Yes, after running dbmail-util -by. > I get 928 rows, but only valid headers. > > It's funny to see how many different emailheaders gets used. Yes, very! Two interesting things about my results: mysql> select count(*) from dbmail_headername where id not in (select distinct headername_id from dbmail_headervalue); +----------+ | count(*) | +----------+ | 11428 | +----------+ mysql> select * from dbmail_headervalue where headername_id = (select id from dbmail_headername where headername like 'Content-Type'); Empty set (0.02 sec) So I figure that this is a GMime bug in parsing Content-Type headers. Aaron
