In a bash script, loop through a bash array for input to a binary

You can use printf to add newlines after each entry in the array:

myarray=("Alice" "Bob" "Charlie" "q") mybinary -f inputfile.txt -o outputfile.txt<<EOF $(printf '%s\n' "${myarray[@]}") EOF 

As doneal24 and steeldriver pointed out, you could also use a pipe rather than a here-document:

printf '%s\n' "${myarray[@]}" | mybinary -f inputfile.txt -o outputfile.txt 

If I understand well your needs:

#!/bin/bash myarray=("Alice" "Bob" "Charlie" "q") # q quits the binary mybinary -f inputfile.txt -o outputfile.txt < <(printf '%s\n' "${myarray[@]}") 

If your binary is well coded, it should be deal with STDIN transparently.

To print the elements of an array one per line, in zsh, you do:

print -rC1 -- "$array[@]" 

Or:

print -rNC1 -- "$array[@]" 

for NUL-delimited records instead of lines.

(print -rC1 -- $array would only print the non-empty elements).

bash has a print builtin (initially from ksh), but only as an optional loadable builtin and it doesn't support the -C option (nor -N).

printf '%s\n' "${array[@]}" doesn't work correctly for empty arrays as it prints an empty line (as if the array had one empty element) instead of no line at all.

The common way in bash (or POSIX shells¹ in general) is to define some helper functions:

println() { [ "$#" -eq 0 ] || printf '%s\n' "$@"; } print0() { [ "$#" -eq 0 ] || printf '%s\0' "$@"; } 

Then, you can do:

println "${array[@]}" q | your-cmd 

^{¹ beware though that some POSIX shells (mostly ksh88 and some pdksh-based ones) don't have printf builtin, so won't be able to output large lists as you'll soon reach the limit on the size of the arguments that can be passed to an external command.}