[1644] in Kerberos_V5_Development
Re: preliminary appdefaults patch
daemon@ATHENA.MIT.EDU (Ken Raeburn)
Tue Aug 27 15:52:00 1996
To: Sam Hartman <hartmans@MIT.EDU>
Cc: Mark Eichin <eichin@cygnus.com>, "E. Jay Berkenbilt" <ejb@ql.org>,
kenh@cmf.nrl.navy.mil, krbdev@MIT.EDU
From: Ken Raeburn <raeburn@cygnus.com>
Date: 27 Aug 1996 15:51:15 -0400
In-Reply-To: Sam Hartman's message of 27 Aug 1996 13:29:26 -0400
> Mark> It looks like rlogin uses:
> Mark> { "noforwardable", 0, &off_option, 'F' },
> Mark> { "noforward", 0, &off_option, 'f' },
> Mark> { "noencrypt", 0, &off_option, 'x' },
> Mark> { "forwardable", 0, NULL, 'F' },
> Mark> { "forward", 0, NULL, 'f' },
> Mark> { "encrypt", 0, NULL, 'x' },
> Mark> { "noflow", 0, NULL, 0 },
Sigh. Those should've been "no-", IMO. (Plus "noflow" for backwards
compatibility.)
> I'm not quite sure what the meaning of that is; it sounds like
> you use the same option to tuern something off that you do to turn it
> on.
The getopt_long interface is kind of weird. If the third (pointer)
field is non-null, the pointed-to int gets the specified value ('F' or
whatever) and zero is returned. Also, an extra pointer argument to
getopt_long can be used to pass back the index into the option table.
So you get something like this:
while ((i = getopt_long(argc, argv, "LD:l:de:cat:78k:xfFn",
long_options, NULL)) != EOF) {
switch(i) {
case 0:
switch(off_option) {
default:
goto usage;
break;
case 0:
flow = 0;
break;
#ifdef KERBEROS
case 'x':
encrypt_flag = 0;
encrypt_done = 1;
break;
case 'f':
forward_flag = 0;
forward_done = 1;
case 'F':
forwardable_flag = 0;
forwardable_done = 1;
break;
#endif
}
break;
....
Ken
