UNIX for Dummies Questions & Answers

I have a line like:

"Jun 19 12:56:22 routername 45454:"

I want to keep all information except the seconds of the time. I tried:

Code:

sed 's/..:..:../..:../g'

but apparently I'm on the wrong track, because although that matches on the time, it replaces it with the literal ..:..

How can I retain the hours and minutes?

Code:

sed 's/\(.*:\).. \(.*\)/\1 \2/'

Thanks, as always. Let me try to walk through it and please correct me where I'm wrong.

1.) substitute
2.) group of any # of characters before a :
3.) 2 more characters (im a bit confused on the space)
4.) any grouping of characters?
5.) keep fields 1 and 2?

eh, little more confused on it than I thought... any help would be appriciated. Also, it's keeping the second : making it look like 12:57: .. any way to avoid that?

Earnstaf,
You are right about the "s" command sequence.
The space is to catch the space after the second.
If it is not there, it may get confused with the other ":".
I thought you just want to remove the number of seconds.
If you want to also remove the ":" before it:

Code:

sed 's/\(.*\):.. \(.*\)/\1 \2/'

Thanks ... and I'll read up more on the backreferences (\1, \2, \3 ...) ...

Earnstaf,
The "\n" (ie \1, \2, \3) is pattern substitution.
First you must save it as: \(RE\)
Where RE is Regular Expression.
After you saved it, you just use it again as substitution:
\1 for the first saved.
\2 for the second saved.
\n for the nth saved.

Ahhh got it...

thanks.