Hello,
i am using sed (gnused for windows)with a batch file. I would like to extract from a text file the following informations: 1,2,3 (second column)
A 3
B 3
C 1
C 2
C 3
I have done a code which gives a list instead of a summary of the data inside the list:
3
3
1
2
3
How can i have only the "one number" (1,2,3)?
my command is:
sed -n "s/[A-Z]*//w Post_temp_list.txt" s_post.txt

use the right tool for the job. sed is not really suited for this type of "column based" problem.
You can use gawk instead. Download and install gawk from here: http://gnuwin32.sourceforge.net/pac...

the code:

{
 a[$2]
}
END{
 printf "("
 for ( i in a) printf "%s," , i
 print ")" 
}

save as script.awk and from command line

c:\test> gawk -f script.awk inputfile

sed 's/[A-Z]//g' testfile

Thank you all for your help,
ghostdog gave me an excellent script which work very well. I have to learn awk now!