SQLBlackWhitelist using wildcards

[William A. Knob]

       Ok, I understand... but so, how can I block "subdomains" on my 
Blacklist without wildcards? On the ruleset file "spam.blacklist.rules" 
I put something like that: "FromorTo:    *@*.domain.com     yes"... In 
the SQL stuff how this works?

    Regards;



Julian Field escreveu:
> Unfortunately you can't do that without slowing it down a lot. The 
> SQLBlackWhiteList stuff, instead of allowing wildcards and hence 
> having to check every entry in the list for every message, reduces the 
> whole problem to a couple of hash table lookups which are very fast, 
> as it knows that there aren't any wildcards.
>
> If you allow the use of wildcards, every entry has to be matched 
> against every address of every message. This is slow and is why 
> MailScanner rulesets shouldn't ideally have more than several hundred 
> (or maybe a thousand) entries. The SQL stuff does not allow wildcards 
> much, with the result that it can just do table lookups to find if the 
> address is listed or not. This is enormously faster than searching 
> every entry of a ruleset.
>
> The reason the SQL black+whitelist support is fast, not because of it 
> being SQL (which actually makes it run slower) but because it doesn't 
> support wildcards.
>
> I hope that explains my design philosophy a bit for this feature.
>
> Jules.
>
> William A. Knob wrote:
>>       Hi all!
>>
>>    People, I want to use "wildcards" on my black/whitelist SQL tables 
>> to use with Mailscanner... Anybody knows how can I do that? Or anyone 
>> has made a modification on the "SQLBlackWhiteList.pm" script to do 
>> that stuff?
>>
>>    Regards;
>>
>>
>
> Jules
>


-- 
	
*William A. Knob - Divisão Desenvolvimento*
Raidbr Soluções em Informática Ltda.
Rua José Albino Reuse, 1125. Cinquentenário. Caxias do Sul - RS
Fone/ Fax: (54) 3223.7074

Visite nosso site:
www.raidbr.com.br <http://www.raidbr.com.br>


-- 
Esta mensagem foi verificada pelo sistema de antivírus e acredita-se estar livre de perigo.