Your syntax is fine, but not all commands/programs accept input from the pipe, or more accurately from stdin. Looking at the man page for file (https://www.man7.org/linux/man-pages/man1/file.1.html) I can’t see a stdin option, so you have to pass each of the files from your head output as arguments to file.

When you use the pipe you’re passing stdout from the left command to stdin on the right command.

file takes a filename as an argument, it does not read stdin (by default).

The first command works because head does read from stdin and then echoes the first line, but as an argument to file using command substitution - the $( … ) bit. Command substitution is neat but this is not really the best use-case for it.

A better way is to use find with -print0 which makes it use a zero-byte instead of a newline character (otherwise the command will fail in case a filename contains a newline - newlines and spaces in filenames break a lot of scripts if you aren’t careful) and -quit, which makes find exit after the first match. And then pass it to xargs which is a utility that transfers stdin to command arguments - In this case the file binary - and handles zero-bytes as seperators when using the -0 arg.

$ find /usr/lib -maxdepth 1 -type l -print0 -quit | xargs -0 file

Removing -quit from the find command also works as expected (ie. on each found link).

@emotional_soup_88
You could use “-” as argument for file:

find /usr/lib -maxdepth 1 -type l | head -n 1 | file -

which means, ‘file’ will read its input from the pipe. But that is probably not what you want, because ‘file’ parses the input and does not accept it as an argument.

You use a pipe to pass the output of a program to another program, in order to do something with that input. Like:

cat unsorted_list.txt | sort -

deleted by creator