I just noticed something when I stepped through one of my examples - in this case: $to:=New object $from:=New object $i:=100 $from.ptr:=->$i $to.ptr:=->$k $to.ptr->:=$from.ptr-> // $k = 100 at this line The value of $k changed as soon as this line runs.
I also wondered what would happen if I initialized $to.ptr to a nil pointer. The answer is C_POINTER($nil) $to.ptr:=$nil // $to.ptr = null Nil is regarded as null in v17. On Thu, Feb 28, 2019 at 8:22 AM Kirk Brooks <[email protected]> wrote: > Jeremy, > > > On Wed, Feb 27, 2019 at 11:53 PM Jeremy Roussak via 4D_Tech < > [email protected]> wrote: > >> Objects pre-date ORDA, as you know. > > True. In this case what we are really talking about is 4D's object > notation. I tend to use 'ORDA' for all the new capabilities and should be > more specific. > > >> I was simply using an object as a way of permitting a method to return >> several values at the same time. Those values happened to be pointers to >> variables (created by OBJECT DUPLICATE and so necessarily referenced using >> pointers). I wanted then to copy the value in the variable pointed to by >> one pointer into the variable pointed to by another pointer. The two >> variables were of identical type, one having been created by duplicating >> the other. >> > I find this an excellent application of using objects as containers for > passing parameters to and retrieving values from a method. > > Since you mention OBJECT DUPLICATE you must be working with variables on a > form. In that context using Form removes the need to work with pointers > altogether. > >> There was therefore no reason why >> >> $to.ptr-> := $from.ptr-> >> >> should have created any error at all; yet it did, and it was an error for >> which I can find no logical explanation, particularly given that the >> assignment did actually take place. Further, if I complicate the code by >> writing >> > Perhaps you didn't create $to and $from with New object before using them? > Try running this: > > C_OBJECT($to;$from;$obj) > C_LONGINT($i;$j;$k) > > $to:=New object > $from:=New object > > $i:=100 > > $from.ptr:=->$i > $to.ptr:=$from.ptr > > $k:=$to.ptr-> // $k = 100 > > > To use the code as you wrote it you must also initialize the property you > are writing to: > > $from.ptr:=->$i > $to.ptr:=->$k > > $to.ptr->:=$from.ptr-> // $k = 100 > > > But if this happens > > $i:=100 > $from.ptr:=->$i > $to.ptr:=->$k > $to.ptr->:=$from.ptr-> > > $i:=200 > > $k:=$from.ptr-> // $k = 200 > > > This was a useful exercise for me. I didn't realize you could use pointers > this way in objects. I don't think it's a particularly good way to work > with them but it looks like there is some fairly rigorous type checking > going on under the hood. For example, I expected this to work: > > $i:=100 > $from.ptr:=->$i > $k:=$from.ptr > > but it doesn't because $k isn't a pointer. However: > > $i:=100 > $from.ptr:=->$i > $to:=OB Copy($from;True) > > $i:=200 > > $k:=$to.ptr // $k = 100 > > does work because I use the True flag to resolve the pointers to their > values. This also points out (discovered pun) the fact that within an > object properties are dynamically typed ($to.ptr is currently a number). OB > Copy was a great tool in pre-ORDA days as a way to build a template of an > object based on fields and then use it to create data objects. > > C_POINTER(p1; p2) >> p1:=$from.ptr >> p2:=$to.ptr >> p2-> := p1-> >> >> I get no errors, interpreted or complied. >> > Mixing 'classic' 4D code with 'ORDA' code is tricky. You can wind up > spending so much time converting between the two there's just not much > benefit. Where it excels is when you start with a new method and use ORDA > to refactor some task in old code. Generally I will have 2/3 the code to > accomplish the same results. Less code is better. > > And some things are still just faster in 'classic' 4D. Large queries for > instance. "Large" meaning tables with millions of records or extensive > relations. Or long loops where you are doing queries. OTOH I hear from guys > with big applications ORDA is generally faster than SQL. In situations > where I'm working with smaller sets (thousands or less of records) I love > working with ORDA. And forms are just amazing using ORDA, Form, dynamic > form variables and object notation. > > -- > Kirk Brooks > San Francisco, CA > ======================= > > What can be said, can be said clearly, > and what you can’t say, you should shut up about > > *Wittgenstein and the Computer * > > -- Kirk Brooks San Francisco, CA ======================= What can be said, can be said clearly, and what you can’t say, you should shut up about *Wittgenstein and the Computer * ********************************************************************** 4D Internet Users Group (4D iNUG) Archive: http://lists.4d.com/archives.html Options: https://lists.4d.com/mailman/options/4d_tech Unsub: mailto:[email protected] **********************************************************************
