SA Bugzilla – Bug 7734
Migrating DNSBL functions to using AskDNS
Last modified: 2020-07-17 04:39:25 UTC
I propose trying to deprecate as much legacy stuff as possible for 4.0.0, like DNSEval.pm, some URIDNSBL, HashBL functions etc and start preferring AskDNS and _TAGS_, which I've found quite intuitive to use. I've already implemented locally some stuff like HEADER() support for AskDNS, so we can for example askdns FOO_REPLYTO _HEADER(Reply-To:addr:domain)_.list.bar.com A /^127/ I added :host and :domain support to get(), but probably additionally something like this would be better: askdns_flags FOO_REPLYTO tag_domain (tag_ip, tag_revip, tag_host...) HashBL/URIDNSBL could be deprecated with with _URIS_, _SUBJECTURIS_, _BODYURIS_, _BODYEMAILS_, flags like tag_lc tag_sha1 tag_fulluri tag_schemelessuri etc. Feel free to submit any ideas.
*** Bug 7733 has been marked as a duplicate of this bug. ***
Committed: - AskDNS _HEADER()_ support - header :host :domain :ip :revip modifiers Sending trunk/UPGRADE Sending trunk/lib/Mail/SpamAssassin/Conf.pm Sending trunk/lib/Mail/SpamAssassin/PerMsgStatus.pm Sending trunk/lib/Mail/SpamAssassin/Plugin/AskDNS.pm Transmitting file data ....done Committing transaction... Committed revision 1862893.
Oh yes and I pretty much rewrote the AskDNS code/logic, it was nightmare to understand. Now it's structures are per-rule based.
From users On Thu, Jul 11, 2019 at 01:26:59AM +0100, RW wrote: > > There's a problem here that __HELO_RECORD looks to be only checking > IPv4. The HELO has an AAAA record. Heh, this is getting real hard.. so askdns would need to have the knowledge if resolver is even AAAA capable. Or even A capable for that matter, on a IPv6 only server. Askdns can only return hit or not hit, and meta can only know whether subrule is run or not run. Too much mind bending for mornings..