[pmg-devel] [PATCH pmg-api] RuleDB/Notify: properly en-/decode the mail subject

[Stoiko Ivanov]

huge thanks for addressing this so quickly!

after quickly diving into the docs of the used perl-modules - seems like
the best approach here and works in my tests!

tiny nit/question:
On Wed,  5 Oct 2022 09:49:41 +0200
Dominik Csapak <d.csapak at proxmox.com> wrote:

> we need to mime decode the subject after reading it, so that we get
> the 'real' subject instead of the (possibly) mime encoded one (which
> might be base64 or quoted-printable encoded). To get a proper subject in
> the notification mail again, we have to encode it again before passing
> it MIME::Entity->build
> 
> Signed-off-by: Dominik Csapak <d.csapak at proxmox.com>
> ---
>  src/PMG/RuleDB/Notify.pm | 3 ++-
>  src/bin/pmg-smtp-filter  | 3 ++-
>  2 files changed, 4 insertions(+), 2 deletions(-)
> 
> diff --git a/src/PMG/RuleDB/Notify.pm b/src/PMG/RuleDB/Notify.pm
> index 6b964a6..af853a3 100644
> --- a/src/PMG/RuleDB/Notify.pm
> +++ b/src/PMG/RuleDB/Notify.pm
> @@ -6,6 +6,7 @@ use DBI;
>  use MIME::Body;
>  use MIME::Head;
>  use MIME::Entity;
> +use MIME::Words qw(encode_mimewords);
>  use Encode qw(decode encode);
>  
>  use PVE::SafeSyslog;
> @@ -228,7 +229,7 @@ sub execute {
>  	Charset => 'UTF-8',
>  	From    => $from,
>  	To      => $to,
> -	Subject => encode('UTF-8', $subject),
> +	Subject => encode_mimewords(encode('UTF-8', $subject), "Charset" => "UTF-8"),
>  	Data => encode('UTF-8', $body));
>  
>      if ($self->{attach} eq 'O') {
> diff --git a/src/bin/pmg-smtp-filter b/src/bin/pmg-smtp-filter
> index 45eb125..eaecd21 100755
> --- a/src/bin/pmg-smtp-filter
> +++ b/src/bin/pmg-smtp-filter
> @@ -9,6 +9,7 @@ use Time::HiRes qw (usleep gettimeofday tv_interval);
>  use POSIX qw(:sys_wait_h errno_h signal_h);
>  
>  use MIME::Parser;
> +use MIME::WordDecoder qw(mime_to_perl_string);
>  use File::Path;
>  use Net::Server::PreFork;
>  use Net::Server::SIG qw(register_sig check_sigs);
> @@ -152,7 +153,7 @@ sub get_prox_vars {
>      } if !$spaminfo;
>  
>      my $vars = {
> -	'SUBJECT' => $entity->head->get ('subject', 0) || 'No Subject',
> +	'SUBJECT' => mime_to_perl_string($entity->head->get ('subject', 0) || 'No Subject'),
any reason why 'No Subject' is inside the mime_to_perl_string call?
Works either way - and I'm fine with both options - but stumbled upon
reading it

>  	'RULE' => $rule->{name},
>  	'RULE_INFO' => $msginfo->{rule_info},
>  	'SENDER' => $msginfo->{sender},

with or without the potential change:
Tested-By: Stoiko Ivanov <s.ivanov at proxmox.com>
Reviewed-By: Stoiko Ivanov <s.ivanov at proxmox.com>