On Monday 20 Apr 2009 09:28:49 Arun Khan wrote:
> On Monday 20 Apr 2009, Mukund Deshmukh wrote:
> > > /sbin/ifconfig eth0 | egrep "RX bytes" | sed -e 's/[ ]\{2,
> > > \}//g' -e 's/([0-9]\{1,\}\.[0-9]\{1,\} [KM]b)//g' >> <filename>
> >
> > I feel people who can write regular expression are the most brilliant
> > people on this earth :-))

Actually, It could take more effort to read and decipher a complex regex 
written by someone else :P

> The above is quite simple,

Well, I figure that its all the extra \ that BRE makes you put in makes it 
look `impressive' to newbies ;) For instance, with ERE, one could use just + 
instead of \{1,\}. GNU's sed allows ERE with -r btw. Not portable, of course.

> some of the stuff that I seen from PERL geeks
> boggles my mind

Hehehe, like a `proper' regex for matching email addresses :P
http://www.regular-expressions.info/email.html

> but testing our regular expressions before putting it
> into action/production is a must :D

Agreed.

Mrugesh
-- 
http://mm.glug-bom.org/mailman/listinfo/linuxers

Reply via email to