Get Specific Parameters 1

[dvknn]

Hi..I have a file which looks like this:

2003.11.13,(h p computer) - FILTER: 1
2003.11.13,(marketing tool) - FILTER: 1
2003.11.13,(market tool) - FILTER: 1
2003.11.13,(market item) - FILTER: 1
2003.11.13,(avionics) - FILTER: 4
2003.11.13,(Proofs (printing)) - FILTER: 1
2003.11.13,(heart medication) - FILTER: 2
2003.11.13,(FOLIOS) - FILTER: 1
2003.11.13,(molded plastic) - FILTER: 1
2003.11.13,(PORTFOLIOS) - FILTER: 1
2003.11.13,(paint equipment) - FILTER: 1
2003.11.13,(portfolios) - FILTER: 1

In my output, I should see the lines without the starting "(" and ending &quot". So, my file should look like :

2003.11.13,h p computer - FILTER: 1
2003.11.13,marketing tool - FILTER: 1
2003.11.13,market tool - FILTER: 1
2003.11.13,market item - FILTER: 1
2003.11.13,avionics - FILTER: 4
2003.11.13,Proofs (printing) - FILTER: 1
2003.11.13,heart medication - FILTER: 2
2003.11.13,FOLIOS - FILTER: 1
2003.11.13,molded plastic - FILTER: 1
2003.11.13,PORTFOLIOS - FILTER: 1
2003.11.13,paint equipment - FILTER: 1
2003.11.13,portfolios - FILTER: 1

I have tried an awk..but, my awk results show me without "("..so, for
2003.11.13,Proofs (printing) - FILTER: 1, I see my output as 2003.11.13,Proofs printing)) - FILTER: 1

Can anybody help me please?
Thanks

 

[Salem]

Try this

#!/bin/awk -f

{
  match($0,"\\(.*\\)")
  left = substr($0,1,RSTART-1)
  mid  = substr($0,RSTART,RLENGTH)
  right= substr($0,RSTART+RLENGTH)
  sub("^\\(","",mid)
  sub("\\)$","",mid)
  print left mid right
}

--

http://www.catb.org/~esr/faqs/smart-questions.html

 

[PHV]

Try something like this:

sed -e 's!(!!;s!\(.*\))!\1!' <input >output

Hope This Help
PH.

 

[dvknn]

I have this 'awk' script.

BEGIN {
FS="[()]"
OFS=","
}
{
split($1, a, " "
print a[1],$2
}

the line
2003.11.13,(Proofs (printing)) - FILTER: 1

comes out
2003.11.13,Proofs printing

but, I want to get
2003.11.13,Proofs (printing) - FILTER: 1

Is this possible?

 

[PHV]

dvknn, have you tried the sed script I posted ?

 

[dvknn]

Sorry PHV, Could you please explain the script before I try...it will make more sense if I understand it..

 

[PHV]

>sed -e 's!(!!;s!\(.*\))!\1!' <input >output 
s!(!!          Suppress the 1st &quot;(&quot;
s!\(.*\))!\1!  Suppress the last &quot;)&quot;

 

[dvknn]

Do I need to have the &quot;<&quot; and &quot;>&quot; in the command?

 

[dvknn]

I got it..

But still not what I want...Let me try some other thing and get back with you guys.

 

[PHV]

Don't understand why still not what you want.
Tested on my machine with your input example, got your output example. What' wrong ?

 

[dvknn]

Sorry PHV...your script works..Let me tell you the story..

The file that I have looks like this:

2003.11.13 15:59:08 [Thread-65761] toa1.sac.fedex.com [5] gemstone DocumentDeterminationServer.keywordSearch NO KEYWORD MATCH: avionics - FILTER: 4
2003.11.13 15:59:34 [Thread-65801] toa1.sac.fedex.com [5] gemstone DocumentDeterminationServer.keywordSearch NO KEYWORD MATCH: Proofs (printing) - FILTER: 1
2003.11.13 15:59:34 [Thread-65801] toa1.sac.fedex.com [5] gemstone DocumentDeterminationServer.keywordSearch NO KEYWORD MATCH: Proofs (printing) - FILTER: 1
2003.11.13 16:00:56 [Thread-65863] toa1.sac.fedex.com [5] gemstone DocumentDeterminationServer.keywordSearch NO KEYWORD MATCH: heart medication - FILTER: 2

now, out of this how can I get

2003.11.13,avionics - FILTER: 4
2003.11.13,Proofs (printing) - FILTER: 1
2003.11.13,Proofs (printing) - FILTER: 1
2003.11.13,heart medication - FILTER: 2

thanks

 

[PHV]

Try this:

sed -e 's! .*MATCH:!,!'

Hope This Help
PH.

 

[dvknn]

Where is the filename ..?

 

[PHV]

sed -e 's! .*MATCH:!,!' </path/to/input >/path/to/output

 

[dvknn]

I got it....everything LOOKS GOOD NOW....
Thanks for answering my questions so patiently..

 

[RFC1795]

Some good tips. But I'm struggling to get something working for my problem.

I'm trying to remove everything between the ( ) from the following line:
cisco 7206 (NPE200) processor (revision B) with 114688K/16384K bytes of memory.

The only part I need to see out the line is the &quot;114688K/16384K&quot; part. Easy to do if all lines are of the same length. Problem is some lines look like this :
cisco RSP4 (R5000) processor with 131072K/2072K bytes of memory.

The options I see, are to remove the () and anything inside them somehow and then awk '{print $5}' for example.
The other option is to somhow revers the word order and print the column needed.

I know somebody knows this ;-)

Thanks all.

 

[Ygor]

That other option would be...

awk '{print $(NF-3)}' file1

 

[RFC1795]

Thanks

Many ways to kill the cat .. managed to find a command similar to the one you mantioned Ygor.

I take it NF means something like Near Field? -3 from 3rd last.

I tried awk -Fwith '{print $NF}' | awk '{print $1}' and was left with the entry I needed.

Thank you all .. another good tip for the memory banks.

 

[PHV]

Try something like this:

sed -e 's!.with !!;s! bytes.*!!'

Hope This Help
PH.

 

[vgersh99]

echo 'cisco RSP4 (R5000) processor with 131072K/2072K bytes of memory.' | nawk -F'(with)|(bytes)' '{print $2}'

vlad
+----------------------------+
| #include<disclaimer.h> |
+----------------------------+