Related articles 

help on activation tree steve.bagnall@gmail.com (steve_bagnall) (20060911) 
From:  "steve_bagnall" <steve.bagnall@gmail.com> 
Newsgroups:  comp.compilers 
Date:  11 Sep 2006 15:53:36 0400 
Organization:  http://groups.google.com 
Keywords:  code, question 
PostedDate:  11 Sep 2006 15:53:36 EDT 
Hi,
The following is a question I'm struggling with, it's an abbreviation
of Q2.5b from the Dragon book, I was wondering if anyone can help me
get my answer right:
In a language that has dynamic binding of names to memory addresses,
employing deep binding and a scoping rule that allows a reference to
point to the closest ancestor in the activation tree that holds a
declaration. Given the following functions; what is the tree of
activation records when A is evaluated. I have used F * G to denote
the composition of two functions F and G.
function F(X);
function X, Y;
Y := X * H;
return Y;
end F;
function H();
return SIN;
end H;
function G(Z)
function Z, W;
W := ARCTAN * Z;
return W;
end G;
function MAIN();
real A;
function U, V;
V := F(G);
U := V();
A := U(PI / 2);
print A;
end MAIN;
At the moment my thinking is as follows:

 MAIN 

 A = 0.7854 [ ARCTAN(SIN(PI / 2)) ] 
 U = Y 
 V = Y 



 F 

 X = G 
 Y = X * H 



 Y [ ARCTAN * SIN ] 

But I'm not sure about this, because I'm not sure that the activation
record for the dynamic function that I've called Y should be called Y,
and also what data it should hold.
Also, the statement: U := V() seems to suggest that you should call the
function V() and assign the return value to U, but at that point V is
the composition of ARCTAN and SIN and I don't see how that can take no
parameters, so I've just assigned the function that V points to to U?
Plus in general, this doesn't look much like a tree, which the question
seemed to suggest I would create, making me think I've got it totally
wrong.
I'd appreciate any pointers at all on this one. By the way I'm not a
student or anything so this is not cheating.
Cheers,
Steve.
Return to the
comp.compilers page.
Search the
comp.compilers archives again.