Shell Programming and Scripting

When you run ssh, you get a brand new, independent shell.

You set variables in it the exact same way you would locally: VARNAME="..."

You create functions in it the exact same way you would locally:

Code:

function_name
{
...
}

There's no "exporting" of anthing, it does exactly what you tell it to, no more, no less.

When you want to send big blocks of code over ssh, and make sure your code does exactly what it looks like, it can be useful to do something like this:

Code:

ssh username@host exec /bin/bash -s "$ARG1" "$ARG2" <<"EOF"
function_name
{
        echo "inside function"
}

echo "ARG1 was $1"
echo "ARG2 was $2"

function_name
EOF

The -s argument, followed by "$ARG1", "$ARG2", etc. lets you set the $1 $2 ... variables for code running on the remote server. (It works on any shell, not just bash.) Nothing else at all from the local side carries over, giving you precise control.

<<"EOF" instead of <<EOF prevents any expansion whatsoever from happening inside the here-document, guaranteeing that everything in it gets sent over raw, dollar-signs and quotes and whitespace and anything else purely unmangled.

An unfortunate side-effect of this is that your remote shell code won't be able to use read. Standard input has been redirected from the contents of the here-document, you can't read from the terminal.