Compiler Design (CoSc3102)

Chapter Four
Syntax Directed Translation
Outline
4.1. Syntax directed definitions
4.2. Types of Attributes
4.3. Dependency graph and evaluation order
4.4. S-attributed definitions
4.5. L-attributed definitions

1
10/31/202 WCU-CS Compiled by TM.
 Syntax Directed Translations
 In syntax analysis phase, we have learnt how a parser constructs parse trees.
 But the plain parse-tree constructed in that phase is generally not used for a
compiler as it is,
 Because, it does not carry any information about of how to evaluate the tree.
 Means that, the productions of CFG makes the rules of the language, but do
not accommodate how to interpret them.
 For example: E → E + T

 This CFG production has no any semantic rule associated with it, and it
cannot help in making any sense of the production.
 Semantic information can be represented by associating CFG used in SA
with semantic rules
 Then the result is a syntax-directed translation,

 i.e., SDT=Context free Grammar(production rule) + Semantic Rules

 Grammar symbols are associated with attributes to associate
information with the programming language constructs.

10/31/202 WCU-CS Compiled by TM. 2

 Syntax directed Translations cont’d..
 What is Semantics?
 Semantics of a language provide meaning to its constructs, like tokens and
syntax structure.
 Semantics help to interpret symbols, their types and relations with each other

 Semantic analysis judges whether the syntax structure constructed in the

source program derives any meaning or not.
For example: int a = “value”;
 This is not an error issue in LA and SA phase, as it is lexically and
structurally correct,
 But it should generate a semantic error as the type of the assignment differs.
 Semantic Errors
 Some of the semantics errors that the semantic analyzer is expected to
recognize:
 Type mismatch
 Undeclared variable
 Reserved identifier misuse.
 Multiple declaration of variable in a scope.
 Accessing an out of scope variable.

10/31/202 WCU-CS Compiled by TM. 3

 Syntax directed Translations cont’d..
 Attributes for expressions can be a:
 type of value: int, float, double, char, string,...
 type of construct: variable, constant, operations, ...
 Values of these attributes are evaluated by the semantic rules associated with
the production rules, which is set by grammar of the language
 Evaluation of these semantic rules:
 may generate intermediate codes
 may put information into the symbol table
 may perform type checking
 may issue error messages
 may perform some other activities
 The following tasks should be performed in semantic analysis:
 The source program will be checked for semantic errors
 Scope resolution
 Type checking: type information will be collected for the code generation
 Array-bound checking

10/31/202 WCU-CS Compiled by TM. 4

 SDD and SDT
 When we associate semantic rules with productions, we use two
notations/approaches
1. syntax-directed definition (SDD)
 associates a semantic actions with each grammar production rule,
 the rule states how attributes are calculated, but not say when they
will be evaluated
 give high-level specifications for translations.
 hide many implementation details such as order of evaluation of
semantic actions.
2. Translation scheme/syntax-directed translation(SDT)
 uses semantic actions embedded anywhere in the bodies of
productions.
 actions may perform arbitrary computations, such as appending
output; they are carried out in left-to-right order during parsing
 indicate the order of evaluation of semantic actions associated with a
productions.
10/31/202 WCU-CS Compiled by TM. 5
 Syntax directed definitions cont’d….
 In SDT, every non-terminals can get zero or more attributes depend on the
type of attribute.
 Example: newval := oldval + 12 here, the type of the identifier
newval must match with type of the expression (oldval+12)
 In Semantic rule, attributes (such as data type, string, number, memory
location, etc.) can be represented by VAL.
 A SDD specifies the values of attributes by associating semantic rules with
the grammar productions.
 Example:
 Production Semantic Rule
EE+T E.val=E.val + T.val
ET E.val=T.val
TE*T T.val=E.val * T.val
TF T.val=F.val
FNUM F.val=num.lexval (this attribute return by LA)
 We may also alternatively insert the semantic actions inside the grammar.
10/31/202  Example: E  E+T {print ‘+’} 6
 4.2. Types of Attribute
 Attribute Grammar
 It is a special form of CFG, where some additional information (attributes)
are appended to one or more of its non-terminals
 This provide context-sensitive information.
 Attribute grammar is a medium to provide semantics to the context-free
grammar.
 Attribute grammar (when viewed as a parse-tree) can pass values or
information among the nodes of a tree.
 Example: E → E + T { E.value = E.value + T.value }
 The right part of the CFG contains the semantic rules that specify how the
grammar should be interpreted.
 Here, the values of E and T are added together and the result is copied to
E.
 Based on the way the attributes get their values, they can be broadly divided
into two categories:
i. synthesized attributes and
ii. inherited attributes.
10/31/202 7
 Types of Attribute cont’d
i. synthesized, meaning that its value is obtained
 from attributes of child nodes in the parse tree,

 or from other attributes of the same node in the parse tree

 Example: S → ABC

 If S is taking values from its child nodes (A,B,C),

 then it is said to be a synthesized attribute, as the values of ABC are

synthesized to S.
ii. inherited, meaning that its value is obtained
 from the parent node in the parse tree,

 or from sibling nodes in the parse tree

 useful when there is mismatch between grammar for parsing, and

“abstract syntax” for translation

 Example S → ABC

 A can get values from S, B and C. B can take values from S, A, and C.

Likewise, C can take values from S, A, and B

10/31/202 8
 Types of Attribute cont’d
 Expansion: When a non-terminal is expanded to terminals as per a
grammatical rule

 Reduction: When a terminal is reduced to its corresponding non-terminal

according to grammar rules.
 Syntax trees are parsed top-down and left to right.

 Whenever reduction occurs, we apply its corresponding semantic rules

(actions).
 Semantic analysis uses Syntax Directed Translations to perform the above
tasks.
 Semantic analyzer receives AST (Abstract Syntax Tree) from its previous
stage (syntax analysis).
 Semantic analyzer attaches attribute information with AST, which are called
Attributed AST.
 Attributes are two tuple value, <attribute name, attribute value>
 For example: int value = 5;<type, “integer”><presentvalue, “5”>

 For every production, we attach a semantic rule. 9

 Types of Attribute cont’d…..
 Example: synthesized attributed SDD for simple calculator’s Arithmetic
Expression; “9+3*5\n”;
Production Semantic Rule
L  E \n L.val = E.val
E  E1 + T E.val = E1.val + T.val L.val=24
E T E.val = T.val
T  T1 * F T.val = T1.val * F.val
T F T.val = F.val E.val=24 \n
F (E) F.val = E.val
F  digit F.val = digit.lexval
+
E.val=9 T.val=15

T.val=3 * F.val=5
T.val=9

 Symbols E, T, and F are This is an annotated parse tree

associated with a synthesized F.val=9 Note that no complete parse tree need
attribute val. actually be constructed.
 The token digit has a synthesized
attribute lexval (it is assumed that digit.lexval=9
it is evaluated by the lexical
analyzer)
10
 4.3. Dependency graph and evaluation order
 Dependencies of Attributes
 In semantic rule: b := f(c1, c2, ..., ck) we say b depends on c1, c2, ..., ck
 The semantic rule for b must be evaluated after the semantic
rules for c1, c2, ..., ck
 This dependencies of attributes can be represented by a directed graph
called dependency graph
 Dependency Graphs and Evaluation order of SDD’s
 It indicate the flow of information among the instances of attributes in a
given parse tree.
 An edge from one attribute instance to another
 means that the value of the first is needed to compute the second.
 Example: if dependency graph has an edge from M to N then M must be
evaluated before the attribute of N
 Thus the only allowable orders of evaluation are those sequence of nodes
N1,N2,…,Nk such that if there is an edge from Ni to Nj then i<j
 Such an ordering is called a topological sort of a graph, which produces a
possible sequential order of evaluation
10/31/202 11
Dependency graph & evaluation order cont’d…
 Evaluation order for SDD includes how the SDD(Syntax Directed
Definition) is evaluated with the help of attributes, dependency graphs,
semantic rules, and S and L attributed definitions.
 A dependency graph is used to determine the order of evaluation of
attributes with the help of edges.
 An edge from, the first node to the second node attribute gives the
information that evaluation of first node attribute is required for the
evaluation of the second node attribute.
 Edges represent the semantic rules of the corresponding production.
 Values of synthesized attributes may be calculated in any order
 For a S-attributed SDD, any bottom-up order be sufficient

 Whereas for inherited attributes, order of evaluation is important

 it is possible to write rules for which no order of evaluation is
possible
 In a L-attributed SDD, dependencies always go left-to-right, so no
circularity is possible.

10/31/202 12
 4.4. S-attributed definitions
 S-attributed SDD
 If an SDT uses only synthesized attributes, it is called as S-attributed
SDT.
 These attributes are evaluated using S-attributed SDTs that have their
semantic actions written after the production (right hand side).

 As depicted above, attributes in S-attributed SDTs are evaluated in

bottom-up parsing,
 as the values of the parent nodes depend upon the values of the child
nodes.
10/31/202 13
 4.5. L-attributed definitions
 L-attributed SDD
 This form of SDT uses both synthesized and inherited attributes with
restriction of not taking values from right siblings.
 In L-attributed SDTs, a non-terminal can get values from its parent, child,
and sibling nodes.
 See the following production
S → ABC S can take values from A, B, and C (synthesized).
 A can take values from S only.
 B can take values from S and A.
 C can get values from S, A, and B.
 Attributes in L-attributed SDTs are evaluated by depth-first and left-to-right
parsing manner.

10/31/202 14
 Next
Chapter Five
Type Checking

Outline
5.1. Type systems
5.2. Specifications of a type checker
5.3. Equivalence of types
5.4. Type conversion

15
10/31/202 WCU-CS Compiled by TM.