Both of those messages I clearly prefixed with "all personal opinion".
However, this is NOT legal (throws an error about double-declaring local var
'n'):
<cffunction name="foo">
<cfargument name="n" type="numeric" />
<cfset var n = "forty two" />
<cfreturn n & arguments.n />
</cffunction>
Cheers,
barneyb
> -----Original Message-----
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED] On Behalf Of Sean A Corfield
> Sent: Friday, March 19, 2004 3:16 PM
> To: [EMAIL PROTECTED]
> Subject: Re: [CFCDev] Unscoped references in CFSET within CFFUNCTION
>
> On Mar 19, 2004, at 11:07 AM, Barney Boisvert wrote:
> > First, the separate arguments scope is an oddity that I've not seen
> > except
> > in CF. Granted, I don't have familiarity with a terribly
> large number
> > of
> > different languages
>
> CF is unusual in having named scopes in the first place. Most
> languages
> operate using simple block scoping instead - which allows you to hide
> declarations with other declarations and then usually not be able to
> get at the original declaration:
>
> /* C code */
> char* foo(int n) {
> char* n = "forty two";
> return n;
> }
> int main() {
> printf("foo(42) = %s\n", foo(42) );
> return 0;
> }
>
> You get a compiler warning - but it is legal and char* n hides the
> argument int n.
>
> > Second, you can't use argument names in the local variables scope
> > (because
> > CF makes them local variables behind the scenes).
> Consequently, you'll
> > never have a local variable and an argument with the same name, so
> > there is
> > no need to differentiate the arguments with a scope; they
> are already
> > differentiated.
>
> Er, that's not actually true:
>
> <cffunction name="foo">
> <cfargument name="n" type="numeric" />
> <cfset var n = "forty two" />
> <cfreturn n & arguments.n />
> </cffunction>
>
> That shows that var n and arguments.n both exist independently...
>
> > Third, why complicate things with additional scopes? The named
> > 'arguments'
> > scope is nothing more than a crutch so you don't have to look at the
> > CFARGUMENT tags at the top of the function to know what was
> passed in.
>
> And yet lots of people advocate using the scope qualifier everywhere,
> even if it isn't really needed, to avoid relying on CF's somewhat
> complex and non-intuitive scope stack searching. Can you - WITHOUT
> LOOKING IT UP - list all of the scopes that CF searches in
> the correct
> order for an unscoped variable name?
>
> > Fourth, having two copies of anything is bad because sooner
> or later,
> > you'll
> > screw up and use the wrong one.
>
> Well, that's a mildly plausible argument but of course the same issue
> exists between "variables" scope and 'var' scope - if you forget a
> "variables" prefix you may get a local scope variable or an unnamed
> scope variable. Hence, best practice is to use "variables" everywhere
> to avoid confusion - or would you rather mix 'var' scope, arguments
> scope and the unnamed scope altogether?
>
> > Fifth, I'm a lazy bastard, and those extra 10 characters add up. ;)
>
> "first of all come great dreams, then a feeling of laziness, and
> finally a witty or clever excuse for remaining in bed." (Kierkegaard
> 1813-1855; "The Present Age")
>
> Regards,
> Sean
>
> ----------------------------------------------------------
> You are subscribed to cfcdev. To unsubscribe, send an email
> to [EMAIL PROTECTED] with the words 'unsubscribe cfcdev'
> in the message of the email.
>
> CFCDev is run by CFCZone (www.cfczone.org) and supported
> by Mindtool, Corporation (www.mindtool.com).
>
> An archive of the CFCDev list is available at
> www.mail-archive.com/[EMAIL PROTECTED]
>
----------------------------------------------------------
You are subscribed to cfcdev. To unsubscribe, send an email
to [EMAIL PROTECTED] with the words 'unsubscribe cfcdev'
in the message of the email.
CFCDev is run by CFCZone (www.cfczone.org) and supported
by Mindtool, Corporation (www.mindtool.com).
An archive of the CFCDev list is available at www.mail-archive.com/[EMAIL PROTECTED]
