[nmh-workers] Refactoring smatch

[Ken Hornstein]

I have decided to start working on my to-do list for nmh again, and at
the top of that list was "refactor smatch()".  Specifically, I would
like to have smatch do the following things:

- Automatically handle switch arguments.  There are a lot of these cases
  in the code:

    if (!(cp = *argp++) || *cp == '-')
        die("missing argument to %s", argp[-2]);

  And I have to ask myself ... why isn't smatch() just handling that??
  There's not really a reason it shouldn't do a little bit of that, since
  that code is duplicated over and over.

  It seems the existing convention is that if there is a space in the
  switch name it takes an argument.  E.g.:

    X("src +folder", 0, SRCSW) \
    X("file file", 0, FILESW) \
    X("rmmproc program", 0, RPROCSW) \
    X("normmproc", 0, NRPRCSW) \

  The switch matching code already knows that a space is the end of the
  argument name, so it seems reasonable to just go with that as an indicator
  that a switch takes an argument and if it does then process the argument
  and return back.

- Handle subcommands.  Right now there is a lot of work when you add a new
  switch to post(8) because switches to post are passed down from send(1)
  and whatnow(1) and whom(1) (there are probably other commands that do
  this a lot as well).  So being able to pass in an additional smatch array
  and have it build up an argument vector would be a huge win.  How the
  exact API would work there is a bit of an open question, because I could
  also imagine wanting to handle the case where several programs handled
  the same set of arguments.  So maybe that would be something like ...

  int smatch(const char **, const struct swit *, ...);

  Where the additional arguments might be something like:

  enum { SMATCH_ARGV, SMATCH_ADDITIONAL };

  and you'd do:

  smatch(&argp, switches, SMATCH_ARGV, post_switches, post_argv);

  Or, you know, something like that.

Thoughts?

--Ken