[pmg-devel] [PATCH pmg-api v2 1/3] pmgqm: add plain subject/from fields for reports

[Hannes Laimer]

On 19.09.25 01:21, Stoiko Ivanov wrote:
> On Thu, 18 Sep 2025 16:54:01 +0200
> Thomas Lamprecht <t.lamprecht at proxmox.com> wrote:
> 
>> Am 18.09.25 um 16:32 schrieb Hannes Laimer:
>>> We don't want plain-text reports to contain HTML-escaped chars,
>>> this skips html encoding for the plain-text part.
>>>
>>> Signed-off-by: Hannes Laimer <h.laimer at proxmox.com>
>>> ---
>>> v2:
>>> - explicitly utf8 encode after `decode_rfc1522`
>>>
>>>   src/PMG/CLI/pmgqm.pm | 16 ++++++++++++----
>>>   1 file changed, 12 insertions(+), 4 deletions(-)
>>>
>>> diff --git a/src/PMG/CLI/pmgqm.pm b/src/PMG/CLI/pmgqm.pm
>>> index 41f9f1a..c26b545 100755
>>> --- a/src/PMG/CLI/pmgqm.pm
>>> +++ b/src/PMG/CLI/pmgqm.pm
>>> @@ -45,17 +45,25 @@ sub get_item_data {
>>>   
>>>       $item->{id} = sprintf("C%dR%dT%d", $ref->{cid}, $ref->{rid}, $ref->{ticketid});
>>>   
>>> -    $item->{subject} =
>>> -        PMG::Utils::rfc1522_to_html(PVE::Tools::trim($head->get('subject')) || 'No Subject');
>>> +    my $raw_subject = PVE::Tools::trim($head->get('subject')) || 'No Subject';
>>> +    $item->{subject} = PMG::Utils::rfc1522_to_html($raw_subject);
>>> +    $item->{subject_plain} = Encode::encode('UTF-8', PMG::Utils::decode_rfc1522($raw_subject));
>>
>> it's a bit borderline, but might be worth adding a rfc1522_to_utf8 (or
>> rfc1522_to_plain_utf8?) helper for this. For one there would be a tiny
>> bit better code re-use, but it would make it slightly clearer for what
>> happens here.
> FWIW: quickly looked through the code and currently think the encoding
> here should do the sensible thing - and would agree that the helper sub
> could help future reviewers see faster what's happening here.
> 
> one thing that is still passed through encode_entities is pmail (tried by
> sending testmails to `discardįšų@test.domain`).
> 
> additionally it seems we do a double-encoding somewhere (probably in SMTP.pm)
> but this should be unrelated to your patch.
> 

We could also encode right before giving it to the template, so probably
in finalize. Since the template is actually the only thing that requires
this to be utf8 encoded.

>>
>>>   
>>> -    my $from = PMG::Utils::rfc1522_to_html(PVE::Tools::trim($head->get('from') // $ref->{sender}));
>>> -    my $sender = PMG::Utils::rfc1522_to_html(PVE::Tools::trim($head->get('sender')));
>>> +    my $raw_from = PVE::Tools::trim($head->get('from') // $ref->{sender});
>>> +    my $from = PMG::Utils::rfc1522_to_html($raw_from);
>>> +    my $from_plain = Encode::encode('UTF-8', PMG::Utils::decode_rfc1522($raw_from));
>>> +
>>> +    my $raw_sender = PVE::Tools::trim($head->get('sender'));
>>> +    my $sender = PMG::Utils::rfc1522_to_html($raw_sender);
>>> +    my $sender_plain = Encode::encode('UTF-8', PMG::Utils::decode_rfc1522($raw_sender));
>>>   
>>>       if ($sender) {
>>>           $item->{sender} = $sender;
>>>           $item->{from} = sprintf("%s on behalf of %s", $sender, $from);
>>> +        $item->{from_plain} = sprintf("%s on behalf of %s", $sender_plain, $from_plain);
>>>       } else {
>>>           $item->{from} = $from;
>>> +        $item->{from_plain} = $from_plain;
>>>       }
>>>   
>>>       $item->{envelope_sender} = $ref->{sender};
>>
>>
>>
>> _______________________________________________
>> pmg-devel mailing list
>> pmg-devel at lists.proxmox.com
>> https://lists.proxmox.com/cgi-bin/mailman/listinfo/pmg-devel
>>
>>
>