Shell Programming and Scripting

Hi, all: I have a question about "cleaning up" a huge file with regular expression(s) and sed:

The init file goes like this:

block1,blah-blah-blah-blah,numseries1,numseries2,numseries3,numseries4
block2,blah-blah-blah-blah-blah,numseries,numseries2,numseries3,numseries4
...
block99999,blah-blah-blah,numseries,numseries2,numseries3,numseries4


And the final output should result as:

blah-blah-blah-blah
blah-blah-blah-blah-blah
blah-blah-blah

In short, I would need to remove in every line:
1) everything from the '^' to the first ',' before blah-blahs
2) everything from the first ',' after blah-blahs to the '$'

I've tried it by using the following sed command without success:

sed -e "s/^.*,//g" -e "s/,.*//g" file.in > file.out

Thanks in advance!

if your input is symmetric.. i.e blah blah is always the second column then CUT should be a better option.

Please confirm if this is the case. If so,

Code:

 
cut -d"," -f2 <input_file> > <output_file

will work.

Code:

while(<DATA>){
	my @tmp=split(",",$_,3);
	print $tmp[1],"\n";
}
__DATA__
block1,blah-blah-blah-blah,numseries1,numseries2,numseries3,numseries4
block2,blah-blah-blah-blah-blah,numseries,numseries2,numseries3,numseries4
block99999,blah-blah-blah,numseries,numseries2,numseries3,numseries4

Both ideas work perfectly. As I'm only testing a couple of fast ideas by now, I'll keep the Perl code in the safe...

Thanks a lot!