Hello, Paul!
You wrote to "DBMAIL Developers Mailinglist" <[email protected]> on Thu, 19 Apr 2007 20:17:21 +0200:

PJ> Did you try to compile with all modules and unit-testing enabled?
PJ> Sqlite is broken by the patch.

Sorry about sqlite and pg modules - I use mysql and remove it from configure not long ago.
Now I compile with all modules and unit tests.

PJ> Also, this is invalid:

PJ> @@ -1647,12 +1645,17 @@ int db_icheck_isheader(GList  **lost)
PJ>         char query[DEF_QUERYSIZE];
PJ>         memset(query,0,DEF_QUERYSIZE);

PJ> -       snprintf(query, DEF_QUERYSIZE,
PJ> +/*     snprintf(query, DEF_QUERYSIZE,
PJ>                         "SELECT MIN(messageblk_idnr),MAX(is_header) "
PJ>                         "FROM %smessageblks "
PJ>                      "GROUP BY physmessage_id HAVING MAX(is_header)=0",
PJ> +                       DBPFX);*/
PJ> +       snprintf(query, DEF_QUERYSIZE,
PJ> +                       "SELECT messageblk_idnr "
PJ> +                       "FROM %smessageblks "
PJ> +                       "WHERE is_header=0 AND "
PJ> +                       "physmessage_id NOT IN (SELECT physmessage_id
PJ> FROM dbmail_messageblks WHERE is_header=1)",
PJ>                         DBPFX);

PJ> That query is designed to find header blocks that are not marked as
PJ> such. This change will break that.

This my mistake. I thought that this select used to remove blocks without headers.

PJ> What is that #pragma directive in dbmail.h? It's causing a lot of
PJ> warnings.

This pragma used to suppress some warnings on intel compiler.



With best regards, Anton Zakatov. E-mail: [EMAIL PROTECTED]

Attachment: dbmail-2.2.2519-4.base.patch.gz
Description: Binary data

_______________________________________________
Dbmail-dev mailing list
[email protected]
http://twister.fastxs.net/mailman/listinfo/dbmail-dev

Reply via email to