SYNTAX DIRECTED TRANSLATION COMPILERS 29TH OCTOBER 2013

SDD Semantic Rule A  X1X2X3 A.x = f ( X1.a, X2.b, X3.c )

SDT A  X1X2{A.x = f( X1.a, X2.b) }X3 Embedded

When we perform reduction, we evaluate the attribute to get the result S – attributed SDD SDT A  X1X2X3 {a} When we perform reduction, we evaluate the attribute to get the result

We have the stack X1X2X3  symbols and states X1. a, X2. b, X3 We have the stack X1X2X3  symbols and states X1.a, X2.b, X3.c  values

After Reduction A  replaces X1X2X3 A After Reduction A  replaces X1X2X3 A.x  evaluated by the function f( X1.a, X2.b, X3.c )

L  E { Print (E. val) } E  E + T { E. val = E. val + T L  E { Print (E.val) } E  E + T { E.val = E.val + T.val } E  T {E.val = T.val} T  T * F { T.val = T.val * F.val } T  F { T.val = F.val } F  id { F.val = id }

Stack Symbol Stack Value Input String Syntax Action $ 3*4 $ 3 $3 *4 Semantic Action $ 3*4 $ 3 $3 *4 F  id F.val = id $ F T  F T.Val = F.val $ T Shift $ T * $ 3* 4 $ T*4 $ 3*4 $ T*F T  T * F T.Val = T.val * F.val $ 12

To evaluate T. val = T. val. F To evaluate T.val = T.val * F.val , we need the two values; which we get form the value stack.

L – attributed SDD is suitable for top-down parsing LL (1)  synthesized attributes  inherited attributes

A non-terminal can have multiple inherited attributes ; for each will have some defined action In case of synthesize attribute, we append it to the end of the rule For predictive parsers, for each non-terminal, we have a function

E. G : Recursive Descent Parser A  X1X2X3 A(). {. if ( a == X1 ) E.G : Recursive Descent Parser A  X1X2X3 A() { if ( a == X1 ) move to next symbol if ( X1 is a NT) X1() }

This function invokes the functions of the body of the production To implement predictive parsing, we need to implement those functions

Made By : Mohit Bhura 11CS30019