|
|
Subscribe / Log in / New account

SIEVE Language for Mail Filtering Quick Guide

From:  "Alina POPESCU" <alina.popescu-AT-axigen.com>
To:  <lwn-AT-lwn.net>
Subject:  Article - SIEVE Language for Mail Filtering Quick Guide
Date:  Fri, 10 Nov 2006 11:34:52 +0200

What is SIEVE?

SIEVE is a language created and used for mail filtering that broadens
the filtering options generally provided by mail servers or
Antispam/Antivirus applications. They work basically by comparing
different keys using different comparators and comparison methods,
against headers of a mail message. Based on the result of the
comparison, you can apply different actions to the corresponding mail
message, i.e. reject, discard, redirect, etc.

This language provides an extremely flexible filtering methodology, as
users can define any number of script filters according to their needs.
Designed to be easily implemented on either a mail client or mail server
(such as Sendmail, Qmail, Axigen and so on), using SIEVE scripts does
not depend on access protocol, mail architecture, and operating system. 

SIEVE is designed as a proposed Internet Standard, as a result of a
multi-vendor effort that has been discussed in various technical and
standards-oriented public and private meetings since at least 1994.

Why Use SIEVE Filters

Mail traffic for most users has been increasing due to increased usage
of e-mail, the emergence of unsolicited email as a form of advertising,
and increased usage of mailing lists. 

There are a number of reasons to use the SIEVE filtering system: 
* You can create efficient and flexible rules. Scripts written in SIEVE
are executed during final delivery, when the message is moved to the
user-accessible mailbox. Therefore, it is reasonable to sort when the
MTA deposits mail into the user's mailbox.
* SIEVE scripts are a safe filtering method since they do not operate on
the mail content but only extract information from the mail header and
take actions according to the pre-defined rules. 
* As and addition to Antispam and Antivirus applications, you can use
SIEVE scripts to also filter all legitimate emails, thus gaining speed
and efficiency when using email communication.
The SIEVE Language

1. General aspects
SIEVE has a fixed form described as a standard but in can be improved by
using extensions. The extension mechanism works if the system implements
those extensions. In order to use an extension, it must be advertised at
the beginning of the file (script) with a require clause.
require "extension_name"  
or  
require ["extension_name1", "extension_name2"]

The structure of SIEVE as described in the standard defines 5 actions:
keep, fileinto, reject, discard, redirect which are self-explanatory. It
also defines 3 control commands:
* <stop> - which stops the processing to that point
* <if elsif else> structure
* require command - which defines an extension of the language.

The if structure has the form: 
if <test> <block>
elsif <test> <block>
else <block>
where a block is a block of commands (actions and control commands -
including other ifs)

In the standard form, without any extensions, the test can be one of the
following:
* address - tests a set of the address headers against a set of keys
using different comparison methods;
* envelope - optional test; 
* header - tests a set of the headers against a set of keys using
different comparison methods;
* true, false - constants;
* allof <other tests> - logic and between several tests;
* anyof <other tests> - logic or between several tests;
* not <test> - negation of a test;
* exists - test if a set of headers exist;
* size - test against the size of a message;
A test can take 2 values: true or false.

2. Examples 
A simple example of a SIEVE script that will move all mails that have
"Spam" in the subject or are received from "spammer@example.com", in the
Spam folder can be written as follows:
require "fileinto"
	
if anyof(address :is ["From", "Sender"] "spammer@example.com", header
:matches "Subject" "Spam")
{
	fileinto "Spam";
}
Another, more complex example is a filter that will reject all mails
that have a virus and are outgoing mails and not incoming mails. This
latter example uses several extensions that need to be implemented.
require
["envelope","virustest","relational","comparator-i;ascii-numeric","rejec
t"];

if allof(virustest :value "eq" :comparator "i;ascii-numeric" "5",
	envelope :contains "From" ["domain1.org", "domain2.org"],
	not envelope :contains "To" ["domain1.org", "domain2.org"])  {
		reject "This mail is from domain.org to the world and
contains a virus";
	}
For a more detailed presentation of the SIEVE language we also advise
you to read the dedicated RFC 3028.
SIEVE filters in the AXIGEN Mail Server

Currently, AXIGEN uses SIEVE language for script filter definition.
Different user-defined SIEVE scripts can be included in any AXIGEN
Filtering System. When activated in AXIGEN, each filter is assigned a
priority value. The notion of priority is used to define the order of
filters in the filtering chain. This means that filters with higher
priority will be applied first. All SIEVE filters can be activated at
multiple levels: server, domain or account/mail list. 

AXIGEN also implements the vacation SIEVE extension. This means that
SIEVE script files can be created and applied for generating
out-of-office type automatic replies. Thus, auto-generated messages can
be send when the user of the account for which the vacation applies, is
on vacation, out of office or in general away for an extended period of
time. Although it is not a security function, vacation extension is an
extra functionality available via script files. For an easy out-of-the
office implementation in our mail server, please see this example
available in our Knowledgebase:

For detailed instructions on SIEVE language and scripts implementation
in AXIGEN, please see our online documentation.

http://www.axigen.com/articles/sieve-language-for-mail-fi...



to post comments


Copyright © 2006, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds