Specific regular exp. creating

[HanSatan]

Dear Spamihilator Coder.

I would like to create reg exp. to eliminate sucha line(s):

New York, NY, 60262

or

New York, NY, 49482

So I created such reg exp.:

(?i)New\sYork,\sNY,\s[\0123456789]{1,5}

But it also deletes lines without zip code at the end. Could you give me a tip how to create working reg exp to eliminate specific line including zip code? What am I doing wrong?

Regards.

[HanSatan]

Ok I fixed it, now it works fine:

(?i)New\sYork,\sNY,\s[0123456789]{5,20}

But maybe there's some other (better way) to filter it out?

Thx in advance for any tips...

[Chactory]

Hi,

this is cool, you asked and answered yourself. I'm delighted about somebody else using RegExes with the Spam-Word Filter.

For me, the line you've created seems to be perfect. My only objection would be, that it's a relatively unspecific entry. But I understand that this may be a help, if it belongs to a recurrently reappearing spam mail.

Perhaps you want to help me test my new spam-words list? If you want then download my updated list, unpack it, copy it into the folder "C:\Programs\Spamihilator", end and restart Spamihilator, and send me a feedback about how it works with you.

It could be set too critical for english users, but I would be glad someone in an anglophone country to try it. But it could be better to save your old "spampoints.ini", e.g. as "spampoint.ini_old".

Regards, Chactory

[HanSatan]

Hi,

Yes indeed I did

And yeah it doesn't get use like "universal" spam word but approx. 25-40% of my spam mails contain this line so it's helpfull

[Chactory]

Hi! Please read my posting again, because in the same moment as you wrote your new posting, I added a further question. Regards, Chactory

[Chactory]

Hi,

approx. 25-40% of my spam mails contain this line so it's helpfull

this is strange ... where do those mails come from?

Regards, Chactory

[HanSatan]

I cant post links here I dunno why, board script doesn't lemme throught.

chactory.de/ownloads/spampoints.zip right? if yes I will check it throught few next days.

Concerning my spam emails, they come from many hosts pc-114-102-162-190.cm.vtr.net,
c9522df7.virtua.com . br etc. Anyway most of those cases could be solved if my mail server would be patched properly, because most of them contain:

From: "Kaanxmjgrm" <hansatan@hansatan.com>
Envelope-to: hansatan@hansatan.com

(my own e-mail address as sender)...

I need to talk with server admin

BTW

What do you think about following priority list:

2009-02-21_144126.jpg (13.9 KiB) 1719-mal betrachtet

Is it fine? And 1x thing, url filter = link filter right ?

[Chactory]

Hi,

excuse me, the hyperlink is http://www.chactory.de/downloads/spampoints.zip, of course. The board likes it more, when you use BBCode to post hyperlinks: spampoints.ini.

Thank you for testing my spampoints.ini in the next days.

The cm.vtr.net server seems to be in Chile, the virtua.com is registered in Connecticut, US ...

(my own e-mail address as sender)... I need to talk with server admin

Regards, Chactory

[Chactory]

Hi,
my filter sequence is rather akin to your's. I just don't use the attachment filter. The AIR Filter caused false positive sieving with me, because in my company, there are listed similar mail addresses in non-spam mails often.
Yes, the Link Filter works as an url filter. I'm really enthusiastic about it!
You may want to try the DCC Filter. When you enter a higher threshold than default (I'm using 1200), then you won't achieve false positives often.
Regards, Chactory

[Quellcore]

(?i)New\sYork,\sNY,\s[0123456789]{5,20}

But maybe there's some other (better way) to filter it out?

The number part iat the end could be cleaned up:

Code: Alles auswählen

(?i)New\sYork,\sNY,\s[0-9]{5}\D

This will require the ZIP Code part of your example to have exactly 5 digits.

Regards,
Quellcore

[HanSatan]

Indeed,

but lemme ask, why did you add \D in the end of the line ?

I checked it with regex coach but it doesn't seem to work with \D

[Quellcore]

Hallo HanSatan!

I also tried it with Regex Coach and here it works (EDIT: if it's not at the very end of the searched string).
"/D" stands for any Non-Digit, without that the Regular expression would also kick in for "New York, NY, 123456"
Sorry, it looks like it doesn't work at the very end of a text.
Try

Code: Alles auswählen

(?i)New\sYork,\sNY,\s[0-9]{5}\b

instead.


Regards,
Quellcore

[HanSatan]

Thanks for support Quellcore.

Tell me 1x thing, if I would like to disable lines like "knock here" or "click here" can I use such a reg exp:

Code: Alles auswählen

(?i).\shere

Yes I know that it will also remove "same here",

or maybe

Code: Alles auswählen

(?i)(knock)?(click)?\shere

Or maybe some other ideas/suggestions?

[Quellcore]

Tell me 1x thing, if I would like to disable lines like "knock here" or "click here"

How about:

Code: Alles auswählen

(?i)(knock|click)\shere\b

Regards,
Quellcore

[HanSatan]

Thanks a lot,

it seems to match perfectly !

I got a 1x question ( I know I know I'm buggy)

I would like to filter out "beding" word but I don't want to filter "bed" word, so how should I set reg exp?

smthing like this:

Code: Alles auswählen

\bbed(ing)\b

or maybe there is other, better way with "." sign ??