Sorry for all of the last minute breakage, folks. It looks like deliveries should be working again, using fgetc in read_header and fread in store_message_tmp. Please check out CVS and test it out!
There's one case left that I'm a little worried about, which is if, while reading in the body of a message during an LMTP session, fread reads past the body and grabs LMTP commands. This *shouldn't* be a problem, because once the client (eg Postfix) has sent the whole message, it waits for the server to report back the delivery status. Worst case, some systems might block for a short period during the fread, waiting for more data from the client, then timing out (should be a sub-second timeout) and returning from fread. We'll just have to try it in the real world to see if it causes trouble... Aaron ""Aaron Stone"" <[EMAIL PROTECTED]> said: > > ""Jesse Norell"" <[EMAIL PROTECTED]> said: > > > Ok, I just thought this through again, and at the moment our only option > > > is > > > fgetc with one character at a time. Anything that's line-at-a-time is not > > > binary safe (correct this assumption and we're home free, though). Block at a > > > time, like fread, is not sensitive to the header/body boundary. It is > > > inevitable (and I must have missed it by testing with my usual dumb three line > > > headers) that read_header() will fread a bit into the body. So we'd have > > > to > > > either unread the extra stuff or pass an initial body along the delivery > > > chain. Neither is a pretty option (how much you can push back, if at all, > > > is > > > highly system dependent). > > > > Perhaps you could fgetc() through the headers, and once you hit the body > > use fread() for performance? > > Not a bad idea at all! I'll try just chaning the read_header function, and > folks please check CVS again in a few hours to see if it's working right. > > Aaron > > > > > > > > -- > > Jesse Norell > > > > [EMAIL PROTECTED] is not my email address; > > change "administrator" to my first name. > > -- > > > > _______________________________________________ > > Dbmail-dev mailing list > > Dbmail-dev@dbmail.org > > http://twister.fastxs.net/mailman/listinfo/dbmail-dev > > > > > > -- > > > > _______________________________________________ > Dbmail-dev mailing list > Dbmail-dev@dbmail.org > http://twister.fastxs.net/mailman/listinfo/dbmail-dev > --
