I wonder - would it be a good idea to store the Globals in Ultralite
tables and do inserts and selects on them? It may not actually work out to
be as whacked out a way of handling information as it sounds... What do
you say? :)
In that case the accessors and mutators would be just a single method with
table name, column name and value as parameters...
Samir
Jeff Ishaq writes:
>
> Howdy. I'm writing a Globals class that stashes our app's globals into the
> storage heap, and simply extends accessors and mutators to get at them.
> Pretty straightforward stuff.
>
> I'm not a C++ wizard, so I'd appreciate some tips for minimizing code
> duplication for the accessors and mutators. Here's what I'm looking at, as
> an example:
>
> void SGlobals::Set_bGoingBack( Boolean bIn )
> {
> // Lock down the record handle:
> SGlobalsStructPtr pGlob = (SGlobalsStructPtr) MemHandleLock(
> m_recH );
> ErrFatalDisplayIf ( !pGlob, "Unable to lock down globals record.");
>
> Err err = DmWrite( pGlob,
> offsetof(SGlobalsStruct,
> bGoingBack),
> &bIn,
> sizeof( Boolean ) );
>
> ErrFatalDisplayIf( err, "Unable to mutate global.");
>
> // Unlock the record handle, return the value
> MemHandleUnlock( m_recH );
> }
>
> Boolean SGlobals::Get_bGoingBack( void )
> {
> // Lock down the record handle:
> SGlobalsStructPtr pGlob = (SGlobalsStructPtr) MemHandleLock(
> m_recH );
> ErrFatalDisplayIf ( !pGlob, "Unable to lock down globals record.");
>
> Boolean bRet = pGlob->bGoingBack;
>
> // Unlock the record handle, return the value
> MemHandleUnlock( m_recH );
> return bRet;
> }
>
> For each variable, I've cut-and-pasted the same function. Then I change the
> method name, the type of the intermediate variable (bRet in this example),
> the member (SGlobalsStruct.bGoingBack, in this case), and the return type of
> the method. Whenever I end up cutting-and-pasting, a little voice in my
> head says "you're not doing this the right way."
>
> I understand the concept of templates (and have the references to get smart
> on them real fast), but I'm not quite sure how to switch off of an
> argument. For example, this would be ideal:
>
> SGlobals.Set( offsetof(SGlobalsStruct, bGoingBack), (Boolean)true);
>
> Boolean bTmp = SGlobals.Get( offsetof( SGlobalsStruct, bGoingBack));
>
> All I can think of is a templatized second param and a big-ass switch
> statement inside. Is there a Better Way? Is this a pipe dream, and I
> should just cut-and-paste away and be done with it?
>
> Thanks!
> -Jeff Ishaq
> The Windward Group
>
>
