duplicate subject lines in headers (again)

[Warwick Brown]

> So it is definitely a MailScanner issue, but it was supposed to be fixed
> a long time ago.
> 
> Look at
> <https://github.com/MailScanner/v4/blob/master/mailscanner/bin/MailSca
> nner/Exim.pm>.
> At line 845 is the DeleteHeader subroutine.
> 
> The previous issue was that lines 863 and 864 used to be
> 
>     next if !$usingregexp &&
>             lc($metadata->{headers}[$hdrnum]{name}) ne lc $key;
> 
> instead of
> 
>     next if !$usingregexp &&
>             lc(quotemeta($metadata->{headers}[$hdrnum]{name})) ne lc $key;
> 
> 
> This is the fix that Jules referred to at
> <http://lists.mailscanner.info/pipermail/mailscanner/2009-
> September/093266.html>.
> 
> Compare that routine to the one in your MailScanner/Exim.pm and see if
> you are missing the 'quotemeta' or if there are other differences.
> 
> ...
> > So I'm afraid it does look like MailScanner rather than exim....
> 
> Yes, I think so.

Hi Mark,

I have the latter line, as I built the whole platform from fresh source code only 2 months ago.

I notice that in Exim.pm, from line 645 to line 854 there is some acknowledgement to the issue of trailing whitespace in relation to ISO encoded subject lines...I wonder if that 'fixup' may be more appropriate in a broader sense. I'm kicking myself for not being a perlmonger and while I can almost 'read' the code, I'm by no means competent to change it with any assurance. But from my 'spider-sense' I am wondering if all subjects should be treated in the same way?

Kind regards,

Warwick