Hi,

I am facing a problem.

I have a Sample output of a File myfile

Sample:

Apr 27 02:02:36 MEDIA 14 LINK 3 : AVAILABLE
Apr 27 02:02:36 MEDIA 14 LINK 7 : AVAILABLE
Apr 27 02:02:37 MEDIA 14 LINK 0 : AVAILABLE
Apr 27 02:02:37 MEDIA 14 LINK 1 : AVAILABLE
Apr 27 02:02:37 MEDIA 14 LINK 2 : AVAILABLE
Apr 27 02:02:37 MEDIA 12 LINK 4 : AVAILABLE
Apr 27 02:02:37 MEDIA 14 LINK 5 : AVAILABLE
Apr 27 02:02:37 MEDIA 14 LINK 6 : AVAILABLE
Apr 27 04:43:20 MEDIA 13 LINK 0 : UNAVAILABLE
Apr 27 04:43:20 MEDIA 03 LINK 1 : UNAVAILABLE
Apr 27 04:43:20 MEDIA 13 LINK 2 : UNAVAILABLE
Apr 27 04:43:20 MEDIA 13 LINK 3 : UNAVAILABLE
Apr 27 04:43:20 MEDIA 11 LINK 4 : UNAVAILABLE
Apr 27 04:43:20 MEDIA 13 LINK 5 : UNAVAILABLE
Apr 27 04:43:20 MEDIA 13 LINK 6 : UNAVAILABLE
Apr 27 04:43:20 MEDIA 13 LINK 7 : UNAVAILABLE
Apr 27 04:43:20 MEDIA 13 : UNAVAILABLE
Apr 27 04:43:20 MEDIA 14 LINK 0 : UNAVAILABLE
Apr 27 04:43:20 MEDIA 14 LINK 2 : UNAVAILABLE
Apr 27 04:43:20 MEDIA 07 LINK 5 : UNAVAILABLE
Apr 27 04:43:20 MEDIA 14 LINK 6 : UNAVAILABLE
Apr 27 04:43:20 MEDIA 14 LINK 7 : UNAVAILABLE
Apr 27 04:43:23 MEDIA 13 : AVAILABLE


The output I want is:

MEDIA 14
MEDIA 12
MEDIA 13
MEDIA 03
MEDIA 11
MEDIA 07

i.e Unique Media No.

Please Help.

Try something with associative arrays - awk maybe.
We're here to help, not write it for you. Show us what you try and what difficulties stump you.

awk ' { print $4 " " $5 } ' filename | sort -u
yes, it will also sort, but I do not think it is a problem for you

Hi,

Using arrays might be too complicated. Use awk's pattern search, print appropriate fields and pipe it through sort.

Hope this helps.

EDIT: @pan64: That will print all the fields records, not just those that are UNAVAILABLE

Druuna,

PAN64 is correct. awk has a default field separator of <space>. With this, by printing field 4 and 5, he will get the required fields.

PAN64: I had forgotten about -u on the sort, thank you for the reminder. The -u tells sort to print only unique records. Saves you piping the output to uniq.

Now the Unavalible:

Here is a link to some more information - http://www.math.utah.edu/docs/info/gawk_5.html

Thanx all for Quick Replies.

I have used "cat myfile |grep 'Apr 27.*MEDIA.*' " already to get to this output(Sample).

to get Unique Records i have used

"cat myfile |grep 'Apr 27.*MEDIA.*'|uniq " but i knw this will not work.

Any help is always useful..

fist, just use
grep 'Apr 27.*MEDIA.*' myfile
without cat and pipe.

second, uniq does not work because those lines contain times. So awk will drop unnecessary parts (awk can also be used instead of grep), and uniq or sort will do the rest.

@Nermal: pan64 isn't correct, he forgot to add the /STRING/ part. I was not talking about awk's separator
Your solution also works (if you provide input).

Hope this helps.

I can't see if AVAILABLE or UNAVAILABLE counts, and also what about the line MEDIA 07 and MEDIA 12 in the result (see first post example).

@pan64: You might have a good point!

I might have assumed incorrectly, maybe tarunshrivastav can answer that one.

Why not just make it easy:
Sort if you need to.

Is it midnight in Perth already ??? _

I can come up with some things without it being too late

and we do not need the grep also

Since there is this useless cat hanging around, I'll chime in with another way to skin it.

grail's example is obviously better, but i sometimes feel cut is under appreciated in favour of sed/awk etc so I thought id chime in with an example using cut.

The awk bit was half inched from http://www.pement.org/awk/awk1line.txt