On 2025-01-27 at 16:29:29 UTC-0500 (Mon, 27 Jan 2025 21:29:29 +0000)
Kent Oyer <k...@mxguardian.net>
is rumored to have said:
Here's a couple more that I've worked on but am unable to close:
6043
I don't see a resolution there. Does the cited issue no longer occur
in 4.x?
Yes I believe that is the case. The question is from 15 years ago and
the OP was asking how to determine if a URI came from a link or if it
was parsed from text (i.e. "non-linkified"). The solution is to check
the link type:
if ( $info->{'types'}->{'a'} ) { # this URI came from a clickable
link } if ( $info->{'types'}->{'parsed'} ) { # this URI was parsed
from text }
The specific ask of the bug has been fulfilled: get_uri_detail() returns
the fact that a URI is "unlinked" (not in a form that a MUA is likely to
linkify) and/or "parsed" (URL in the body text, not in a HTML tag)
and/or "schemeless" (such as an email address in body text) so a plugin
*could* use that.
It seemed to me that it was also a request to make it possible for
URIBL* rules to adjust scores based on the provenance of the "URI" which
I don't believe is possible without a substantial amount of new code &
documentation.
I've closed it with a bit of discussion. I don't think the high-level
problem has been entirely resolved (filenames being checked as URI
domains causing FPs) but I'm not sure that is a big enough problem to
demand a fix.
--
Bill Cole
b...@scconsult.com or billc...@apache.org
(AKA @grumpybozo@toad.social and many *@billmail.scconsult.com
addresses)
Not Currently Available For Hire
