Problems with the "Kill off-site Images" filter

LWC · Trooper Joined: Feb 13, 2004 Posts: 27 Location: Israel

This filter kills any form's <input fields, which have value=http:// entries
in them, but only if they lead to external URLs.

However, it also kills them if there's no URL after "http://" at all.
Now, which sites use "http://" with no URL, you ask?
Well, you must know those forms which ask you to provide an URL and
hints you to use "http://"...like in Google's translation page, for example
(i.e. ).

So the bottom line is this: said filter thinks it's ok if the URL is not
external. How can I convince it to think it's ok if there's
no URL at all too?

z12 · Sergeant Joined: Jul 17, 2002 Posts: 135 Location: USA

Hi LWC,

Modify the matching expression like so:

LWC · Trooper Joined: Feb 13, 2004 Posts: 27 Location: Israel

Hmm, the original filter has only two lines in "match".

I see you haven't touched the third one, changed
about half of the first one and added a new second line.

This filter has so many code in it that it looks likes chinese...can you
explain a little what you did? For example, why did you change
the "alt" part?

But most importantly, you've added a new parameter (number 3), but
haven't provided a new "replace" line so it's not even used, is it?

I think the only thing that needs to be changed is the first line:

z12 · Sergeant Joined: Jul 17, 2002 Posts: 135 Location: USA

Hi LWC,

hmm, It seems that maybe we have different filters. I always backup the default config before I switch to mine, but it's possible that I've modifed it.

Here is the filter I was referring to:

LWC · Trooper Joined: Feb 13, 2004 Posts: 27 Location: Israel

Well, I see now. Your version is sort of a compromize.
When it's not the same host, it still accepts it if there's no slash in it
(i.e. domain alone).

It's a lot better than before, but can't there be a solution that
doesn't even accept a domain?

Also, the Google page does work now, but Altavista's translation
page still doesn't work because the code there is:

Lepus · Trooper Joined: Mar 02, 2004 Posts: 15 Location: USA

one quick'n'dirty fix might be to change...

z12 · Sergeant Joined: Jul 17, 2002 Posts: 135 Location: USA

Lepus, thanks for jumpin in!

Yeah, I was going the quick & dirty route with [^/]+{1,*}/

In most all of my replacement code, I use something like <proxo killed blah blah /> so I don't know where my head was at. I guess I assumed there would always be a path to an image.

I tried (^\h)[a-z0-9] on my filter but no joy. It sounds interesting, perhaps you can post something. The old \h never seems to work the way I think it would.

Anyway, heres a new filter to test drive:

Lepus · Trooper Joined: Mar 02, 2004 Posts: 15 Location: USA

LWC · Trooper Joined: Feb 13, 2004 Posts: 27 Location: Israel

Lepus, your reverse logic worked ("must meet a positive value" instead of
"must ignore a negative value")!

Your 8 letters fixed both Altavista's and Google's translation pages.
I wonder if it fixed any other "http://" only input tags pages, but I can't
think of any others right now.

z12 · Sergeant Joined: Jul 17, 2002 Posts: 135 Location: USA

Hi,

Simplicity, gotta love it. That works great Lepus.


	Home ˇ Topics ˇ Submit News ˇ Top 10 This entire site, Cops Themes, and Computer Cops are Š 2002 - 2004 Computer Cops, LLC. All rights reserved. You can syndicate our news using the file RSS 0.91, ultramode.txt, or RSS 1.0. Acceptable Use Policy. Use signifies your agreement. Engine Copyright Š 2002 by PHP-Nuke, GNU/GPL Licensed. ICRA Member. Paul Laudanski, Member of Computer Cops, LLC Server Load: 1334 pages served in previous 5 minutes. Page Rendered: 0.400 seconds.