I would like to expand a little more on "Bash - How to pass arguments to a script that is read via standard input" post.
I would like to create a script that takes standard input and runs it remotely while passing arguments to it.
Simplified contents of the script that I'm building:
ssh server_name bash <&0
How do I take the following method of accepting arguments and apply it to my script?
cat script.sh | bash /dev/stdin arguments
Maybe I am doing this incorrectly, please provide alternate solutions as well.
Try this:
cat script.sh | ssh some_server bash -s - <arguments>
ssh
shouldn't make a difference:
$ cat do_x
#!/bin/sh
arg1=$1
arg2=$2
all_cmdline=$*
read arg2_from_stdin
echo "arg1: ${arg1}"
echo "arg2: ${arg2}"
echo "all_cmdline: ${all_cmdline}"
echo "arg2_from_stdin: ${arg2_from_stdin}"
$ echo 'a b c' > some_file
$ ./do_x 1 2 3 4 5 < some_file
arg1: 1
arg2: 2
all_cmdline: 1 2 3 4 5
arg2_from_stdin: a b c
$ ssh some-server do_x 1 2 3 4 5 < some_file
arg1: 1
arg2: 2
all_cmdline: 1 2 3 4 5
arg2_from_stdin: a b c
This variant on ccarton's answer also seems to work well:
ssh some_server bash -s - < script.sh <arguments>