SIMPL Semantic Analysis

This section discussed the form of the symbol table, the form of the syntax tree, and syntax tree attributes.

Symbol Table

The SIMPL compiler uses a syntax tree produced by the parser to perform semantic analysis. A central structure in the analysis is the symbol table. The form of a symbol table entry is as follows:

Field	Values	Description
name	String	The name of the variable.
value	Integer	The numeric value of a boolean constant
ioType	OUT IN LOCAL	One of three possible functions of the variable.
dtype	NUM BOOL	The data type of an expression node or a variable declaration
varNum	Integer	Each variable is assigned a local variable number (used by the JVM).

Syntax Tree

A syntax tree is built out of several node types. These are listed below:

Node Type	Children	Description
PROG	p1 p2 p3 p4	The top level program node, with name p1, input variables p2, local variables p3, and body p4.
ID		A node representing a variable reference or declaration.
SPACE	p1 p2	A list of variable declarations, with head p1, and tail p2.
NULL		The empty list node.
!	p1 p2	A list of statements, with head p1, and tail p2.
IF	p1 p2 p3	An if statement node, with test expression p1, then part p2, and else part p3.
WHILE	p1 p2	A while statement node, with test expression p1, and body p2.
<-	p1 p2	An assignment statement node, with variable l-value p1, and r-value expression p2.
>	p1 p2	A greater than operator, with left operand p1, and right operand p2.
=	p1 p2	An equals operator, with left operand p1, and right operand p2.
+	p1 p2	An addition or OR operator, with left operand p1, and right operand p2.
-	p1 p2	A subtraction operator, with left operand p1, and right operand p2.
*	p1 p2	A multiplication or AND operator, with left operand p1, and right operand p2.
/	p1 p2	A division operator, with left operand p1, and right operand p2.
NEG	p1	A unary negation or NOT operator with operand p1
NUM		A node representing a counting number constant.
BOOL		A node representing a boolean constant

Node Attributes

During semantic analysis several attributes are calculated, and added to the nodes of the syntax tree. The table below lists the attributes attached to each node.

Attribute	Values	Description	Calculation
name	String	The character string representation of a token.	Supplied by the lexical scanner.
value	Integer	The numeric value of a constant. (0 and 1 used for the boolean constants .F and .T)	Supplied by the lexical scanner.
dtype	NUM BOOL	The data type of an expression	Synthesized from the data types of the children
lineNum	Integer	The source line on which a token was scanned.	Supplied by the lexical scanner.
symTab	Symbol Table	A pointer to the global symbol table.	Synthesized by the PROG node p1, p2, and p3 children; inherited by the p4 child.
use	DECLARE REF	The use of a variable name; either declared or referenced.	Inherited through all children of the PROG node.
size	Integer	The number of leaf nodes in a nodes subtree.	Synthesized from the nodes children.
lab1	String	A generated label, used by control nodes.	Synthesized in the p4 child of the PROG node, using a label generator.
lab2	String	A generated label, used by control nodes.	Synthesized in the p4 child of the PROG node, using a label generator.