That's all you need to modify to have the GA pick it up for scoring.  But 
everyone who's submitted nonspam for the corpus needs to re-run their mass-check 
to pick this new rule up.  Once I have it in CVS that is.

C

Greg Ward wrote:

> Date: Fri, 1 Mar 2002 08:25:14 -0500
> From: Greg Ward <[EMAIL PROTECTED]>
> To: Craig Hughes <[EMAIL PROTECTED]>
> Cc: [EMAIL PROTECTED]
> Subject: Re: [SAtalk] Rule idea: "real name" == local part
> 
> On 28 February 2002, Craig Hughes said:
> > Greg, I think I'm going to be busy for a few days just with scoring
> > refinements.  If you could solidify the regex I'd appreciate it.
> 
> OK, here's a patch relative to CVS:
> 
> --- rules/20_head_tests.cf      26 Feb 2002 12:04:21 -0000      1.33
> +++ rules/20_head_tests.cf      1 Mar 2002 13:20:15 -0000
> @@ -20,6 +20,9 @@
>  header TO_MALFORMED            To !~ 
>/(?:(?:\"[^\"]+\"|\S+)\@\S+\.\S+|undisclosed-recipients:)/ [if-unset: 
>[EMAIL PROTECTED]]
>  describe TO_MALFORMED          To: has a malformed address
>  
> +header TO_LOCALPART_EQ_REAL     To =~ 
>/^\s*(\"?)([\w%\+\-=_\.]+)\1\s*<\2\@[\w%\+\-=_\.]+>/i
> +describe TO_LOCALPART_EQ_REAL   To: repeats local-part as real name
> +
>  header FROM_MALFORMED          From !~ /(?:\"[^\"]+\"|\S+)\@\S+\.\S+/ [if-unset: 
>[EMAIL PROTECTED]]
>  describe FROM_MALFORMED                From: has a malformed address
> 
> > I think as a general guide, the rules should be RFC-compliant, unless
> > there's a special reason not to be.  Also, 2822 should be your guide, not
> > 822.
> 
> OK.  Note that the character classes used to recognize "local part" and
> "domain" in the regexes are subsets of the character set allowed by RFC
> 2822; you see - + _ . in local parts all the time, but & ! { } etc. seem
> pretty unlikely, and adding them all really bulked up the regex.
> 
> Implementation question: is the \s* at the front necessary?  Or does SA
> strip whitespace after the colon before feeding headers to rules?
> 
> And do I need to modify any other file to make the GA pay attention to
> this rule and evolve its score?
> 
>         Greg
> 


_______________________________________________
Spamassassin-talk mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/spamassassin-talk

Reply via email to