You could do it that way. But that doesn't change the fact that flag_value()
ignores "junk" arguments.
And you could fix it your way, but it doesn't change the fact that testing a
string for flag names is a silly way to decide whether or not it's a command.
;)
In my humble shoddy-hack artist opinion ;), I don't think it is a good
coding practice to allow "junk" arguments to be ignored without at the
Short circuiting flag_value on bad flags isn't shoddy or hackish. It might be
a very good idea. I just don't think it fixes this particular problem, it
only hides it. My code would work fine with your code.
In fact, I should probably make a new function that does short circuit. But
not until I have the time to clean up error reporting when it returns NO_FLAG.
I'd have to make a new function because I use flag_value for a lot of things,
like reading area files. I'd like that to fail gracefully if I remove a flag
or something.
--Palrich.
----- Original Message -----
From: "Snafu Life. (Jonathan L. Potter)" <[EMAIL PROTECTED]>
To: <[email protected]>
Sent: Sunday, December 05, 2004 6:58 PM
Subject: Re: REDIT: north name door bug, searching archives...
I am sorry I should have explained my bug and code a little better. It's been
a long time since i wrote here. My apologies.
Mike Barton wrote:
We put in a "<direction> flag <flags>" command.
You could do it that way. But that doesn't change the fact that flag_value()
ignores "junk" arguments.
flag_value() accepts multiple flags, ignores anything that is not a flag,
and doesn't return NO_FLAG as long as at least one argument is a flag.
What I did was force it to bail if the first argument is not a flag which
more than likely means it's a command argument _not_ a flag, i.e. name, key
and dig. Depending on how other code is written it could affect other areas
of the code as well, I just didn't bother testing to find out. It is broke,
fix it. ;)
There is some thought that snips should contain at least one error, so that
people understand what the code is doing. I really don't feel that is the
way to do it. I would rather give a fix that works (with a better
explanation next time) and leave some ideas, which follow...
Mike Barton wrote:
> I'm not sure short circuiting flag_value is the best solution.
very least, logging it so you can find out it is happening. I also would not
advise allowing the code to make any changes at all as long as there are
junk arguments. In other words, don't short circuit flag_value, change it so
that it checks for junk arguments first, bailing if there is one or more,
and if there are no junk arguments, go back and start changing things.
