Syntax directed translation, unable to understand the concept

Question

Hya ,

I am currently studying compiler creation , in which i am studying a chapter called "syntax Directed Translation" and unable to grasp one concept from this chapter.

I have already attached the scanned pages of book and the topic in the book is "Top-Down translation"

Here are the scanned pages:

http://rapidshare.com/files/314805022/Scanned.rar.html

My questions are :

1. How the author written the semantic rules?
2. How the author written the rule "P1.in=P.in + T.val" , is this book forcefully trying to create this rule to make the reader understand top down? , since in L-attrib or inherited attrib. gets the value from siblings or parents?

also in diagram . i can see how how the value is flowing in tree but still want to know why passing the value from t.val to p.in?

3.Again in "Construction of syntax tree for the translation scheme" , How the rule "{P1.in = mknode('+',P.in,T.ptr)}" ?

I know everything previous topic but unable to digest this concept .

Any help is very very .. appreciated)))