@shoulders
I found and read the parser code. He is a quick over view.
IP Block Lists
There is are no defined supported formats. All files are treated as text files and then an IPv4 and IPv6 regex is applied and all of the IPs are extracted allowing all formats and more (json, csv, xml, text)
DNSBL
CSV can only be used if they match one of the internally coded formats which are feed specific.
each line is processed as a single record
if a valid (non 0.0.0.0 / 127.0.0.1) address is found on the same as a domain, then the domain is ignored.
domains are extracted (via various clean ups) from the record and then processed as you expect.
Hope this helps