it this possible in bash script

packets · 08-02-2010, 11:50 PM

I'm a doing a script. There are some part of the script that I need to do but don't know how to do it.

If I have a text which contains:

Quote:

foo,fee,fii,fuu

Could I rearrange it to:

Quote:

foo
fee
fii
fuu

I don't know if sed can do this. There are some function of sed that I just recently discovered but I don't know if it can do the scenario above.

I'm currently checking a tool/command that could help me do this thing just wandering if there's anyone out there that could suggest on how I could accomplish this.

vxc69 · 08-02-2010, 11:57 PM

Try:

Code:

sed -i 's/,/\n/g' YOUR-FILE.txt

packets · 08-02-2010, 11:57 PM

I tried this one and it works

Quote:

echo "test,too,foo" | sed 's/\,/\n/g'

Quote:

test
too
foo

packets · 08-02-2010, 11:58 PM

Quote:

sed -i 's/,/\n/g' YOUR-FILE.txt

Thanks for this one.

catkin · 08-03-2010, 03:40 AM

Or it can be done entirely within bash which requires a few more lines of code but runs faster as long as the strings are small, say less than 2 kB

Code:

#!/bin/bash

list='foo,fee,fii,fuu'
IFS=,
array=($list)
echo ${array[*]}
unset IFS

Notes:

In the array assignment array=($list), $list is substituted and then split into array elements wherever a character from the IFS (=internal field separator?) variable is found.
unset IFS results in bash treating IFS as if it had the default value. It is a quicker and more concise way of restoring a default IFS than saving the default value and re-assigning it.

konsolebox · 08-03-2010, 09:04 AM

Code:

while read -d ,; do echo "$REPLY"; done <<< "foo,fee,fii,fuu"

Code:

while read -d ,; do echo "$REPLY"; done < file.txt

packets · 08-03-2010, 06:16 PM

Thanks to all replies

ghostdog74 · 08-03-2010, 09:07 PM

simply use set

Code:

IFS=","
set -- $list
echo $1, $2 ,$3 etc

packets · 08-04-2010, 02:19 AM

What if I want to do it reverse?

I have

Quote:

foo,
fee,
fii

And I want it to be

Quote:

foo,fee,fii

ghostdog74 · 08-04-2010, 02:26 AM

iterate the file, and use printf

Code:

while read -r line
do
  printf "%s" $line
done <"file"

or

Code:

$ var=$(<file)
$ echo $var # note: without double quotes
foo, fee, fii

the other method involve removing the newlines, which i will leave it to somebody else to show you

packets · 08-04-2010, 03:07 AM

I search also here in LWQ and it seems to work

Quote:

sed -e :a -e '$!N; s/\n//; ta' lists.txt

ghostdog74 · 08-04-2010, 03:18 AM

Quote:

Originally Posted by packets

I search also here in LWQ and it seems to work

using tr is even simpler

Code:

tr -d "\n" <file>newfile

and of course, awk

Code:

awk -vRS="\n" '1' ORS="" file

packets · 08-04-2010, 03:31 AM

@ghostdog74

Thanks. I'm actually talking to my web developer coz it seems he can parse the data before as a single line separated by comma before it pass to my script.