Thanks! That works exactly the way I want it. I had misread the docs on
regex, I thought that \. would match any character, guess I had better have
another read.
Cheers,
> On Sun, 21 May 2000, Simon Bryan wrote:
>
> > Hi,
> > I have the following in my squid .conf, can anyone tell me whay it isn't
> > working? That is the sites are still served up.
> >
> > acl chatrooms2 url_regex "/etc/squid/acls/chatrooms2"
> > http_access deny chatrooms2
> >
> > The file chatrooms2 is:
> > \.LeagueChat\.
> > \.Chat\.
> > \.chat\.
> > \.chat_email\.
> > \.Chat_Rooms\.
> > \.chatroom\.
> > \.General_Chat\.
> > \.hnchat\.
> > \.voicechat\.
> > \.Teenchat\.
> > \.memberchat\.
> >
> > I know it is being parsed by Squid as if I make a deliberate mistake in
> > the file Squid complains. There is nothing in the log except the entry
> > where the pages are being served up.
> >
> > http://www.blueandgoldarmy.com/Chat/chat.htm simon
> > DIRECT/www.blueandgoldarmy.com text/html
> > 958877654.304 147 192.0.0.52 TCP_HIT/000 0 GET
> >
> > Shouldn't this site be barred?
>
> No, because neither "/Chat/" or "/chat." match ".Chat." or ".chat.".
> If you replace the '\.' with '/' (or '\/' if escaping is required in
> this case) it should match the first part. Also, if you use:
>
> acl chatrooms url_regex -i "/foo/filename"
>
> then you don't need to put multiple regexps in there just for
> differences in capitalization.
>
> I like to use "\b" in my regexps which denotes a "word boundary".
> For example, if you want to match either a '.' or '/' or any other
> non-word character, use the following as a regexp:
>
> \bchat\b
>
> This would match both "/Chat/" and "/chat." in the URL above if
> case-insensitive matching is enabled with the -i option. The main
> advantage of doing this rather than using the catch-all '.' for
> any-character matching is that you won't accidentally match any other word
> that just happens to have "chat" in the middle of it (like "achate"...
> yes, its a real word).
>
> By the way, depending on what regex libraries Squid was compiled with,
> some of the regex functions may not work. For some reason, the regexp
> libraries included with FreeBSD don't seem to understand "\b", so I had to
> configure Squid with --enable-gnuregex so it would use the GNU regex
> libraries.
>
>
> -- Chris Dillon - cdillon@wolves.k12.mo.us - cdillon@inter-linc.net
> FreeBSD: The fastest and most stable server OS on the planet.
> For Intel x86 and Alpha architectures. ( http://www.freebsd.org )
>
>
-- Simon Bryan sbryan@olmc.nsw.edu.au Information Technology Manager sbryan@mpx.com.au OLMC ParramattaReceived on Sun May 21 2000 - 16:16:52 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:53:32 MST