On Fri, Sep 03, 2004 at 01:49:51AM +0100,
Roy Badami <roy(_at_)gnomon(_dot_)org(_dot_)uk> wrote
a message of 18 lines which said:
I'm tending to the opinion that the words 'non-empty' should simply
be deleted from section 2 of -pra. They seem to serve no purpose
other that to add complexity,
The added complexity is very small (using the length of the code as a
metric). The real issue is the definition of "non-empty".
The sendmail milter tests "strlen(hdr->hdr_val) > 0". Mark Kramer uses
the Perl \s expression. I used Python's definition of "empty". They do
not agree and I do not find an authoritative definition.
empty field bodies are illegal, and I don't believe they are
prevelent in the wild.
Never assume anything about the Real World when it is the Internet :-)
I thought that the provisions of -pra like:
If that header is malformed (e.g. it appears to contain multiple
mailboxes, or the single mailbox is hopelessly malformed, or the
single mailbox does not contain a domain name),
were unecessary until I ran the PRA algorithm against my spam mailbox
and find very strange addresses.
"Locate the first Resend-Sender field in the message that has a
non-empty field body"
Not complete: is a set of whitespaces empty? Is "(Sender not
displayed)" empty or malformed?