I have tried writing a rule to detect a specific To Header:
header FULLNAMEINTOHEADER To =~ /\"Full\sName\"\s<[EMAIL PROTECTED]>/
When i tested this rule against an email with the string, the result i
get is:
* -2.0 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
* [score: 0.0000]
* 0.0 HTML_MESSAGE BODY: HTML included in message
* 0.5 MISSING_SUBJECT Missing Subject: header
* 2.3 TO_CC_NONE No To: or Cc: header
* 0.9 FM_NO_STYLE FM_NO_STYLE
* 2.5 FM_NO_FROM_OR_TO FM_NO_FROM_OR_TO
* -0.0 NO_RECEIVED Informational: message has no Received headers
* 0.5 FM_NO_TO FM_NO_TO
* 1.0 KAM_BLANK01 Blank emails
I noticed in the debug info it has the following errors:
[2768] warn: Possible unintended interpolation of @ntlworld in string at
F:\Perl\site/etc/mail/spamassassin/70_ben_misc.cf, rule
FULLNAMEINTOHEADER, line 1.
[2768] warn: rules: failed to run header tests, skipping some: Global
symbol "@ntlworld" requires explicit package name at
F:\Perl\site/etc/mail/spamassassin/70_ben_misc.cf, rule
FULLNAMEINTOHEADER, line 1.
Surely this must be some kind of bug?
I understand if i write a bad rule then it will cause problems, but why
does it claim not to be able to do *any* header tests? Surely it should
ignore the bad rule and just use everything else.
Secondly, it claims to not be able to run header tests, and yet it still
seems to run many header tests, which it runs on nothing (I presume it
just drops all the headers). So i get a whole load of hits for no To or
CC or Received headers, however they are clearly there.
Obviously it isn't going to be a problem with good rules, but i can't
see why this happens when i have a bad rule.
Is this something that could be fixed or do you deserve all you get if
you don't spamassassin --lint your new rules?
Thanks
Ben
