API ID in header or similar

[tbtbuk]

Hi All,

We've built a custom inhouse spam filter which is working as a custom script well. However, we'd like our third party tool to be able to release emails from quarantine after a while (other checks have to happen first), however our script isnt aware of the ID that gets set (looks like C1R1465***), seems to be different to the MSGID etc, so we can't pass that over to our check to use the API to release it. I guess we could filter all emails at the set time it came in to release, but there are a lot of downsides to that. So I was wondering, how does that C1..... ID get set for the API, and can I add it to a header in the email (to parse and find) or access it another way in our custom script?

Thanks for any assistance.

Tom

 

[Stoiko Ivanov]

however our script isnt aware of the ID that gets set (looks like C1R1465***), seems to be different to the MSGID etc,

the id is in the form 'CXRYYYYYTZZZZZZZ' (X represents the pmg clusternode - YYYYYY is a number from a postgres sequence and ZZZZZZZZ is a timestamp) - it is not available at the time the custom check script runs (the script provides information for the rulesystem in form of spam-points or virus information - and the rule system later creates this id, if the mail is put into quarantine.

You could get the potential mails via REST-API - usually the returned fields from:
https://pmg.proxmox.com/pmg-docs/api-viewer/index.html#/quarantine/spam
should be enough to uniquely identify a mail and then release it

Else you could get the complete message content via the REST-API and match the message-id headers of the original mail to the one:
https://pmg.proxmox.com/pmg-docs/api-viewer/index.html#/quarantine/content

In both cases this should happen outside of the custom check script.

I hope this helps!