Cisco Cisco Email Security Appliance C170 Guia Do Utilizador

Bodies with MIME headers indicating encoding but no MIME parts

Multi-part MIME messages with encoding indicated

All of the above without the encoding specified in a MIME header

You can use regular expressions (regexes) to match on any part of the message or body, including 
matching attachments. The various attachment types include text, HTML, MS Word, Excel, and others. 
Examples of character sets of interest include gb2312, HZ, EUC, JIS, Shift-JIS, Big5, and Unicode. 
Message filter rules with regular expressions can be created through the content filter GUI, or using a 
text editor to generate a file that is then imported into the system. For more information, see 

 and 

It is important to begin a regular expression with a caret (

^

) and end it with a dollar sign (

$

) whenever 

you want to exactly match a string and not a prefix. 

When matching an empty string, do not use 

“”

 as that actually matches all strings. Instead use 

“^$”

. For 

an example, see the second example in 

.

It is also important to remember that if you want to match a literal period, you must use an escaped period 
in the regular expression. For example, the regular expression 

sun.com

 matches the string 

thegodsunocommando

, but the regular expression 

^sun\.com$

 only matched the string 

sun.com.

Technically, the style of regular expressions used are Python re Module style regular expressions. For 
a more detailed discussion of Python style regular expressions, consult the Python Regular Expression 
HOWTO, accessible from

:

http://www.python.org/doc/howto/

In some languages, the concepts of a word or word boundary, or case do not exist. 

Complex regular expressions that depend on concepts like what is or is not a character that would 
compose a word (represented as “

\w

” in regex syntax) cause problems when the locale is unknown or if 

the encoding is not known for certain. 

Regular expressions can be tested for matching using the sequence 

==

 and for non-matching using the 

sequence 

!=

. For example: 

Unless otherwise noted, regular expressions are case-sensitive. Thus, if your regular expression is 
searching for 

foo

, it does not match the pattern 

FOO

 or even 

Foo

.

rcpt-to == "^goober@dev\\.null\\....$" (matching)

rcpt-to != "^goober@dev\\.null\\....$" (non-matching)