How to check that a file has more than 1 line in a BASH conditional?

The command:

wc -l file.txt

will generate output like:

42 file.txt

with wc helpfully telling you the file name as well. It does this in case you're checking out a lot of files at once and want individual as well as total stats:

pax> wc -l *.txt 973 list_of_people_i_must_kill_if_i_find_out_i_have_cancer.txt   2 major_acheivements_of_my_life.txt 975 total

You can stop wc from doing this by providing its data on standard input, so it doesn't know the file name:

if [[ $(wc -l <file.txt) -ge 2 ]]

The following transcript shows this in action:

pax> wc -l qq.c26 qq.cpax> wc -l <qq.c26

As an aside, you'll notice I've also switched to using [[ ]] and $().

I prefer the former because it has less issues due to backward compatibility (mostly to do with with string splitting) and the latter because it's far easier to nest executables.

A pure bash (≥4) possibility using mapfile:

#!/bin/bashmapfile -n 2 < file.txtif ((${#MAPFILE[@]}>1)); then    echo "This file has more than 1 line."fi

The mapfile builtin stores what it reads from stdin in an array (MAPFILE by default), one line per field. Using -n 2 makes it read at most two lines (for efficiency). After that, you only need to check whether the array MAPFILE has more that one field. This method is very efficient.

As a byproduct, the first line of the file is stored in ${MAPFILE[0]}, in case you need it. You'll find out that the trailing newline character is not trimmed. If you need to remove the trailing newline character, use the -t option:

mapfile -t -n 2 < file.txt

if [ `wc -l file.txt | awk '{print $1}'` -ge "2" ]...

You should always check what each subcommand returns. Command wc -l file.txt returns output in the following format:

12 file.txt

You need first column - you can extract it with awk or cut or any other utility of your choice.