[Mailman-Users] Filter for exact domain names

[Mark Sapiro]

Cyndi Norwitz wrote:

>I am getting a lot of spam post attempts from a domain tom.com.  I  
>would like to do an automatic discard in discard_these_nonmembers for  
>this domain.
>
>I can add this line of code: ^[^@]+@(.*\.)?tom\.com$
>
>But I am concerned that this will discard all .com domain names that  
>end with tom, which may be some legit ones.


It won't


>I do not understand the code as it's different from anything I've  
>worked with and the above recipe (which I mucked up several times on  
>my own despite some trips to the FAQ) came from Mark.  My guess is I  
>need to remove the question mark, but I'd like confirmation first.


It you remove the question mark, the regexp will match only sub-domains
of tom.com. I.e., it will match x at y.tom.com, but not x at tom.com. As it
is, it will match either of those, and it won't match x at tomtom.com
because the piece preceding tom.com doesn't end with a period.

Here's what it says:

^[^@]+ matches the beginning of the line and 1 or more non-@
characters, so ^[^@]+@ matches everything up to the first @ as long as
there is at least one character before the @.

Then (.*\.)? matches 0 or 1 occurrences of anything that ends with a
literal period - the ? says 0 or 1 of the preceding.

Then tom\.com$ matches if the remainder ends with 'tom.com'.

So the whole regexp matches

at least one non-@ followed by @ optionally followed by anything that
ends with a period followed by tom.com at the end.

See <http://docs.python.org/library/re.html>.

-- 
Mark Sapiro <mark at msapiro.net>        The highway is for gamblers,
San Francisco Bay Area, California    better use your sense - B. Dylan