Join arbitrary lines to output as one line

Using sed

As I understand it, you have a file text which is something like:

$ cat text 1 2 3 4 5 

And, you want to extract the 3rd and 5th lines and put them in the middle of some echo statements. In that case, try:

$ { echo 'Some text'; echo 'Some text'; echo -n 'Some text'; echo -n "$(sed -n 3p text)$(sed -n 5p text)"; echo 'Some text'; echo 'Some text'; } >file $ cat file Some text Some text Some text35Some text Some text 

Simplification

The above can be accomplished with a single echo command:

$ echo $'Some text\nSome text\nSome text'"$(sed -n 3p text)$(sed -n 5p text)"$'Some text\nSome text' >file $ cat file Some text Some text Some text35Some text Some text 

Or, as a single printf command:

$ printf 'Some text\nSome text\nSome text%s%sSome text\nSome text\n' "$(sed -n 3p text)" "$(sed -n 5p text)" >file $ cat file Some text Some text Some text35Some text Some text 

How the sed command works

The above uses sed commands such as:

sed -n 3p text 

Here, -n is an option that tells sed not print anything unless we explicitly ask it to. 3p is a command that tells sed to print the third line. Similarly, we use 5p if we want to print the fifth line.

Using awk

An awk solution (hat tip: Runium):

awk ' BEGIN{ printf "%s","Some text\nSome text\nSome text" } NR>5 {exit} NR==3 || NR==5 {printf "%s", $0} END{ printf "%s","Some text\nSome text\n" }' text