awk question

keenboy · 10-21-2009, 06:10 AM

Hello,

I have a tab separated text file which I'd like to separate by commas instead.

I use the following command:

Code:

awk -v OFS=':' '$1=$1' file

The issue I have is the spaces also get separated as well as the tabs which mean that column data with spaces in them are split into different columns.

How do I just replace the TABs but not the spaces?

Thanks,

acid_kewpie · 10-21-2009, 06:23 AM

Make tab the input delimiter:

-F\\t

or

-v FS='\t'

simpler and more readable to not use awk though;

sed -e 's/\t/:/g' file

catkin · 10-21-2009, 06:25 AM

sed would be a good tool for that

ghostdog74 · 10-21-2009, 06:49 AM

awk

Code:

awk '{gsub("\t",":")}1' file

keenboy · 10-21-2009, 07:15 AM

Thanks, ghostdog74

Unfortunately this doesn't seem to work for me.

Quote:

awk '{gsub("\t",":")}1' file

The result is the file still tab separated not comma separated.

Is there anything else I can try?

Thanks,

druuna · 10-21-2009, 07:25 AM

Hi,

These all work for me:

cat infile | tr "\t" ":"

sed 's/\t/:/g' infile --> acid_kewpie's solution from post #2

awk '{gsub("\t",":")}1' infile --> ghostdog74's solution from port #4

ghostdog74 · 10-21-2009, 07:40 AM

@OP if you want more "precise" answers, provide more information about your data files and the output you want. Help us to help yourself.

acid_kewpie · 10-21-2009, 10:03 AM

why do some people only ever read the last reply??? What was wrong with the two solutions i gave??