Bash how to append word to end of a line?

Question

I have executed a command in bash to retrieve some addresses from a file like this:

grep address file.txt | cut -d'=' -f2 | tr ':' ' '

yields:

xxx.xx.xx.xxx port1
xxx.xx.xx.xxx port2

and I would like to append ' eth0' to each of those output lines and then ideally for loop over the result to call a command with each line. Problem I'm having is getting that extra string in the end to each line. I tried:

| sed -e 's/\(.+)\n/\1 eth0/g'

which didn't work..and then supposing I got it there, if I wrap it in a for loop it won't pass in the full lines since they contain spaces. So how do I go about this?

Solution 1

You can match $ to append to a line, like:

sed -e 's/$/ eth0/'

EDIT:

To loop over the lines, I'd suggest using a while loop, like:

while read line
do
  # Do your thing with $line
done < <(grep address file.txt | cut -d'=' -f2 | tr ':' ' ' | sed -e 's/$/ eth0')

Solution 2

How about just using awk:

awk -F= '/address/{gsub(/:/," ");print $2,"eth0"}' file

Demo:

$ cat file
junk line
address=192.168.0.12:80
address=127.0.0.1:25
don not match this line

$ awk -F= '/address/{gsub(/:/," ");print $2,"eth0"}' file
192.168.0.12 80 eth0
127.0.0.1 25 eth0

Or just with sed:

$ sed -n '/address/{s/:/ /g;s/.*=//;s/$/ eth0/p}' file
192.168.0.12 80 eth0
127.0.0.1 80 eth0

Solution 3

All you need is:

awk -F'[=:]' '{print $2, $3, "eth0"}' file.txt |
while IFS= read -r ip port eth
do
   printf "ip=%s, port=%s, eth=%s\n" "$ip" "$port" "$eth"
done

Always use IFS= and -r when using read unless you have a very specific reason not to. google for why.

Solution 4

I came here looking for the same answer, but none of the above do it as clean as

sed -i 's/address=.*/& eth0/g' file

Search and replace inline with sed for lines begining with address, replace with the same line plus 'eth0'

eg.

sed -i 's/address=.*/& eth0/g' file; cat file
junk line
address=192.168.0.12:80 eth0
address=127.0.0.1:25 eth0
don not match this line

View more solutions

Comments

[Palace Chan]

Cool, and how do I loop over the results in the end? The spacing causes me to loop over each word and not each line!

[Chris Seymour]

Bash, grep, cut, tr and sed no no no... What you want can be achieved with either sed or awk alone.

[FatalError]

@sudo_O While I share your sentiment to an extent, this really isn't that bad :P. It's in the unix bloodline to chain together a ton of small utilities like that, that's why they exist.

[Palace Chan]

That's awesome, I always forget the <( stuff and was doing $(

[Ed Morton]

@FatalError UNIX is not about chaining together a ton of small utilities when one small utility will do the job.

[Calvin Taylor]

I see that this doesn't call anything per line as indicated, but this answers the title question.

[Vishnu Kumar]

what does & do here?

[Calvin Taylor]

it's basically shorthand for everything from the LHS into the replacement RHS. take another look at the example or try it for yourself. =) It's great for pre/post pending.