Difference between using - and -- as options to set while setting positional params

In man bash it is mentioned that set has two options - and --

I was wondering if there is any difference while using - and -- as options to set while setting positional parameters.

I could not find any big difference mentioned in man bash when it comes to their usage in setting positi开发者_运维技巧onal params.

---

You use -- (double minus) to indicate the end of shell options and the start of arguments. Hence, for example:

set -- -a -b -f somefile

After this, $1 is -a, $2 is -b, $3 is -f and $4 is somefile.

Without the --, the shell would interpret -a, -b and -f as shell options, and would set $1 to somefile.

The Bash (4.1) manual says (of set):

-- If no arguments follow this option, then the positional parameters are unset. Otherwise, the positional parameters are set to the arguments, even if some of them begin with a ‘-’.

- Signal the end of options, cause all remaining arguments to be assigned to the positional parameters. The ‘-x’ and ‘-v’ options are turned off. If there are no arguments, the positional parameters remain unchanged.

---

The bash(1) man page for 4.1.5(1) says:

--      If no arguments follow this option, then the  positional
        parameters are unset.  Otherwise, the positional parame‐
        ters are set to the args, even if  some  of  them  begin
        with a -.
-       Signal  the  end of options, cause all remaining args to
        be assigned to the positional parameters.  The -x and -v
        options are turned off.  If there are no args, the posi‐
        tional parameters remain unchanged.

The first difference is when there are no arguments after the - or --. For the former, the existing positional parameters will be unchanged. For the latter, the positional parameters will be cleared.

So set -- clears the positional parameters and set - is a no-op.

The -v and -x settings may be modified by set - .... So, if you had set -v turned on (which causes the shell to print input lines as they are read), it will be turned off by the set - ... command. set -- ... will leave it unchanged.

set -x is more common that set -v in that set -x is often used to debug scripts to see exactly what commands are being run. Quite often when debugging a shell script, you would run it with bash -x <script>. Knowing that set - ... turns -x off, you'd probably want to use set -- ..., since it would be quite unexpected to have -x turned off as a side effect of another command.

---

The convention is to use a single - for single-letter arguments, e.g. -i and double -- for their full-worded counterpart, e.g. --interactive.

See here for more information.

This has been answered very well elsewhere on the StackExchange network.

UPDATE

The double-dash -- prefixes long (verbatim) options to commands, but when used with a Bash builtin (like set), it means the end of options to that particular command.

E.g. if you want to create a file that starts with a dash:

touch -dashed

...you'll get an error:

touch: illegal option -- h
usage: touch [-acfm] [-r file] [-t [[CC]YY]MMDDhhmm[.SS]] file ...

However, try it with --, et voila:

touch -- -dashed

...and then ls to see -dashed in your current directory.

See here for more information.

---

In my bash manpage there is a sentence: An argument of - is equivalent to --.