CVS prob

[Francis J. Lacoste]

On Thu, Jan 24, 2002 at 04:38:54PM +0100, Arnaud Gaillard wrote:
> Hello all,
> 
> First we don't know if this is the right place to post this kind of things
> but we try!!

This is the perfect place for that kind of discussions!

> We managed to compile the unstable version from the CVS, but there is
> currently the following error in the code.
> 
> On the sendmail2dlf file the following line should be removed: (I think they
> were used prior the caculation of the timestamp in the Syslog.pm)
> line 283-289 starting with:
> unless ( defined $log->{'month'} && defined $log->{'day'} &&

You are totally right. When I updated the various 2dlf scripts for the
new timestamp attribute in Lire::Syslog, I may have forgotten some place
where the month, day, and time attributes might be tested. (I just
basically looked for place where the script called syslog2cal.)

I fixed it and commited to CVS.

> 
> I think the same prob might occur in the other 2dlf file

I'll do a search for those.

> 
> In the same routine we are surprised when we debug the perl code to see that
> the $rec variable contains a hash on attributes like 
>    to, pri, etc.
> and that the values are duplicated in the 
>    flags hash. 
> So we are reading from the perl debugger:
> 
>        DB<36> x $rec
>     0  HASH(....)
>        'content' => ...
>        'delay' => ...
>        [...]
>        'flags' => HASH(....)
>            'delay' =>
>            [...]
>            'to' =>
>        [...]
>        'to' =>
> 
> So why do we need to duplicate these data?
> 

Hmm, the code responsible for this is in Lire::Email lines 63-77. It's
not really needed and it could be reworked for better efficiency.

Thanks for spotting those!

-- 
Francis J. Lacoste
francis at Contre.COM

-- 
To UNSUBSCRIBE, email to development-request at logreport.org with a subject of
"unsubscribe". Trouble? Send an email with subject "help" to
development-request at logreport.org