Closed Uelb closed 7 years ago
I think it depends on the state of the art of the different email providers, mandrill does have the spam report attribute but I don't know for the others. If it's not present, the specific hash seems to be the way to go, if it's an information given by all providers, then I think it should be a top level attribute.
Mandrill does it : https://mandrill.zendesk.com/hc/en-us/articles/205583207-What-is-the-format-of-inbound-email-webhooks- Apparently sengrid sends it too : https://sendgrid.com/docs/Classroom/Basics/Inbound_Parse_Webhook/setting_up_the_inbound_parse_webhook.html#-Default-Parameters Cloudmailin also : http://docs.cloudmailin.com/features/spam_filtering/
Mailgun and Postmark also appear to do it. Ok, then I think we can keep this top level.
Do you want me to rebase my commits ?
Haha, you are making changes faster than I can comment. One sec!
Ok think I'm good. Sorry about all the comments, just having my coffee now. Just one more suggestion and I think it'll be good.
hey, mind rebasing against master and then I'll get this merged. Thanks!
@Uelb Github is still reporting conflicts so I can't perform a clean merge
It should be fine now
Wow, just saw hound, I'll fix that...
So, is the idea here to have the adapters extract their spam reports and send them in as a top-level attribute? Is that preferential vs having the adapters load that via a vendor-specific hash like that implemented in #285?