lecture 8 semantic analysis

1

Lecture 8

Semantic Analysis

2

The Compiler So Far Lexical analysis

» Detects inputs with illegal tokens

Parsing» Detects inputs with ill-formed parse trees

Semantic analysis» Last “front end” phase» Catches all remaining errors

3

What’s Wrong? Example 1

int y = x + 3;

Example 2

String y = “abc” ; y ++ ;

4

Why a Separate Semantic Analysis?

Parsing cannot catch some errors

Some language constructs are not context-free» Example: All used variables must have been

declared (i.e. scoping)» ex: { int x { .. { .. x ..} ..} ..}» Example: A method must be invoked with

arguments of proper type (i.e. typing)» ex: int f(int, int) {…} called by f(‘a’, 2.3, 1)

5

More problems require semantic analysis

1. Is x a scalar, an array, or a function?2. Is x declared before it is used? 3. Is x defined before it is used?4. Are any names declared but not used?5. Which declaration of x does this reference?6. Is an expression type-consistent?7. Does the dimension of a reference match the

declaration?8. Where can x be stored? (heap, stack, . . . )9. Does *p reference the result of a malloc()?10. Is an array reference in bounds?11. Does function foo produce a constant value?

6

Why is semantic analysis hard?

need non-local information answers depend on values, not on syntax answers may involve computation

7

How can we answer these questions?

1. use context-sensitive grammars (CSG)» general problem is P-space complete

2. use attribute grammars(AG)» augment context-free grammar with rules» calculate attributes for grammar symbols

3. use ad hoc techniques» augment grammar with arbitrary code» execute code at corresponding reduction» store information in attributes, symbol

tables

8

Attribute Grammars A generalization of CFG An attribute grammar is a context free grammar

with associated attributes and semantic rules Each grammar symbol is associated with a set of

attributes Each production is associated with a set of

semantic rules for computing attributes

Also called Syntax-directed Definition in Dragon book.

9

Dependences between attributes

Attribute values are computed from constants & other attributes

synthesized attribute » value computed from children [&

constants] inherited attribute

» value computed from siblings & parent [& constants ]

induce dependency graph among attributes of parse tree nodes.

10

Synthesized Attributes

11

Synthesized Attributes

12

Inherited Attributes

13

Inherited Attributes

14

Example attribute grammarA grammar to evaluate signed

binary numbers Production Evaluation Rules1 NUM SIGN LIST LIST.pos = 0 NUM.val = SIGN.neg ? -LIST.val : LIST.val2 SIGN + SIGN.neg = false3 SIGN - SIGN.neg = true4 LIST BIT BIT.pos = LIST.pos LIST.val = BIT.val5 LIST LIST BIT LIST1.pos = LIST0.pos + 1 BIT.pos = LIST0.pos LIST0.val = LIST1.val + BIT.val6 BIT 0 BIT.val = 07 BIT 1 BIT.val = 2BIT.pos

15

Annotated Parse Tree

16

Annotated Parse Treeand Dependency graph

• val and neg are synthesized attributes• pos is an inherited attribute

17

Two Notations for AGs Syntax-Directed Definition Syntax-directed Translation [Scheme]

18

Syntax-directed definition Each grammar production Ais associated with a

set of semantic rules of the form b := f (c1, c2, …, ck)where f is a function and1. b is a synthesized attribute of A and c1, c2, …, ck are

attributes of grammar symbols in , or2. b is an inherited attribute of one of the grammar

symbols in and c1, c2, …, ck are attributes of A or grammar symbols in

19

Dependencies of Attributes• In the semantic ruleb := f(c1, c2, …, ck)we say b depends on c1, c2, …, ck• The semantic rule b must be evaluated after the

semantic rules for c1, c2, …, ck• The dependencies of attributes can be represented by a

directed graph called dependency graph

20

Dependency Graph

21

Evaluatino Order Apply topological sort on dependency graph

a1 := floata2 := a1addtype(a3, a2) /* a4 */a5 := a2addtype(a6, a5) /* a7 */a8 := a5addtype(a9, a8) /* a10 */

22

S-attributed Grammar S-attribute Grammar:

» all attributes are synthesized attributes.» can be evaluated in one-pass» useful in many context. (e.g. Calculator)» A good math to LR parsing.

23

Example

24

L-attributed Grammar A syntax-directed definition is L-attributed if each

attribute in each semantic rule for each production A X1 X2 … Xn is a synthesized attribute, or an inherited attribute of Xj, 1 j n, depending only on 1. the attributes of X1, X2, …, Xj-1 and/or

2. the inherited attributes of A

25

Example

26

Counter Example

27

pros and cons of AGs advantages:

» clean formalism» automatic generation of evaluator» high-level specification.

Disadvantages» efficiency determined by evaluation

strategy» increase space requirement» circular testing» handling non-local information

28

Syntax-directed Translation[scheme]

allow arbitrary actions can have global data structure can place actions among production

29

The definition A translation scheme is an attribute grammar in which

semantic rules are enclosed between braces { and }, and are inserted within the right sides of productions

The value of an attribute must be available when a semantic rule refers to it

30

Examples» YACC » A ::= B C { $$ = concat($1,$2); }» CUP» A ::= B:m C:p { RESULT = concat(m,p); }

Typical uses» build abstract syntax tree & symbol table» perform error/type checking

31

Example D T {L.in := T.type} L T int {T.type := integer} T float {T.type := float} L {L1.in := L.in} L1 ‘,’ id {addtype(id.entry,

L.in)} L id {addtype(id.entry, L.in)}

32

Example

33

Restriction on translation scheme

An inherited attribute for a symbol on the right side must be computed in a semantic rule before that symbol

A semantic rule must not refer to a synthesized attribute for a symbol to its right

A synthesized attribute for the symbol on the left can be computed after all attributes it depends on have been computed

34

From L-Attributed Definitionsto Translation Schemes

35

From L-Attributed Definitionsto Translation Schemes

S {B.ps := 10} B {S.ht := B.ht} B {B1.ps := B.ps} B1 {B2.ps := B.ps} B2 {B.ht :=

max(B1.ht, B2.ht)} B {B1.ps := B.ps} B1 sub {B2.ps := shrink(B.ps)}

B2 {B.ht := disp(B1.ht, B2.ht)} B text {B.ht := text.ht B.ps}

36

Abstract syntax tree An abstract syntax tree is a condensed form of parse tree

useful for representing constructs.» usually a parse tree with the nodes for most

non-terminal symbols removed.

This represents “x - 2 * y”. can use a linearized (operator) form of the

tree. » x 2 y * - in postfix form.

A popular intermediate representation.

37

Scope Matching identifier declarations with uses

» Important static analysis step in most languages

38

Scope (Cont.) The scope of an identifier is the portion of a

program in which that identifier is accessible

The same identifier may refer to different things in different parts of the program» Different scopes for same name don’t overlap

An identifier may have restricted scope

39

Static vs. Dynamic Scope Most languages have static scope

» Scope depends only on the program text, not run-time behavior

» C and Java have static scope

A few languages are dynamically scoped» Lisp, SNOBOL» Lisp has changed to mostly static scoping» Scope depends on execution of the program

40

Static Scoping Example { int x = 0;

x++;{ int x = 1; x++; }x++;}

41

Static Scoping Example (Cont.)

{ int x = 0;x++;{ int x = 1; x++; }x++;}

Uses of x refer to closest enclosing definition

42

Dynamic Scope A dynamically-scoped variable refers to the

closest enclosing binding in the execution of the program

Example(defun g (y) (let (a 4) (f 3)));(defun f (x) a);» (g 5) => 4

43

Symbol Tables Consider the block: B :{ int x = 0; Es} Idea:

» Before processing Es, add definition of x to current definitions, overriding any other definition of x

» After processing Es, remove definition of x and restore old definition of x

A symbol table is a data structure that tracks the current bindings of identifiers

44

Symbol tables A symbol table associates values or

attributes (e.g., types and values) with names.

What should be in a symbol table?» variable and procedure names» literal constants and strings

45

Symbol TablesWhat information might compiler need? textual name data type declaring procedure lexical level of declaration if array, number and size of dimensions if procedure, number and type of

parameters

46

Symbol TablesImplementation

» usually implemented as hash tables

How to handle nested lexical scoping?» when we ask about a name, we want the closest

lexical declaration

One solution» use one symbol table per scope» tables chained to enclosing scopes» insert names in table for current scope» name lookup starts in current table if needed,

checks enclosing scopes in order

47

B1

B2

TA

TB1

A

TB2

C

TC

scopes tables

48

Imperative-style Symbol Table

beginScope() start a new nested scope endScope() exit current scope addSymbol(x, value) add a symbol x to the table Object lookup(x) finds current x (or null) checkScope(x) true if x defined in current scope (needed for checking duplicate declarations in the

same scope)

49

class SymbolTableclass SymbolTabe { static class Entry { Symbol id; Object value; int scope; Entry(…){…} }private Map map = new HashMap(); private int curScope; …private Stack stack = new Stack();lookup(Symbol s) { Entry e = (Entry) map.get(s) ; if( e == null) return null; else return e.value; }checkScope(x) {

Entry e = (Entry) map.get(s) ; if( e == null) return null; else return e.scope = curScope; }beginScope() { curScope++; push(null); }endScope() { for( Symbol s = (Symbol) pop(); s != null ; s = (Symbol) pop() ) map.delete(s); curScope--; }addSymbol(Symbol x, Object v) { assert lookup(x) == null; push(x) ;

map.add(x,v); }……}

50

Types What is a type?

» The notion varies from language to language

Consensus» A set of values» A set of operations on those values

Classes are one instantiation of the modern notion of type

51

Why Do We Need Type Systems?

Consider the assembly language fragment

addi r1, r2, r3

What are the types of r1, r2, r3?

52

Types and Operations Certain operations are legal for values of each

type

» It doesn’t make sense to add a function pointer and an integer in C

» It does make sense to add two integers

» But both have the same assembly language implementation!

53

Type Systems A language’s type system specifies which

operations are valid for which types

The goal of type checking is to ensure that operations are used with the correct types» Enforces intended interpretation of values,

because nothing else will!

Type systems provide a concise formalization of the semantic checking rules

54

What Can Types do For Us? Can detect certain kinds of errors :

» “abc” ++ ; x = ar[ “abc”] ; int x = “abc” ; Memory errors:

» Reading from an invalid pointer, etc.» int x[50] ; x[50] = 3;

expressiveness (overloading, polymorphism)» help determine which methods/constructors would be

invoked.» Ex: add(Complex, Complex), add(int,int),

add(String,String),..» add(23,14) => add(int, int) invoked

provide information for code generation » ex: memory size

55

Type Checking Overview Three kinds of languages:

» Statically typed: All or almost all checking of types is done as part of compilation (C, Java, Cool)

» Dynamically typed: Almost all checking of types is done as part of program execution (Scheme)

» Untyped: No type checking (machine code)

56

Pros and cons Static typing:

» catches many programming errors at compile time

» Avoids overhead of runtime type checks Dynamic typing:

» Static type systems are restrictive» Rapid prototyping easier in a dynamic type

system

57

The practice Most code is written in statically typed languages

with an “escape” mechanism» Unsafe casts in C, Java» (Person) malloc(1024) // c» Object x = …;» Person p = (Person) x; // Java

It’s debatable whether this compromise represents the best or worst of both worlds

58

Type systems Types

» A type is a set of values that share a set of common properties.

» e.g. : int, InputStream, String, …» Defined by language (built-in types) and/or programmer

(user-defined types) Types are represented by type expressions Type system

» 1. A set of types in a programming language, and» 2. a collection of rules for assigning types to the» various parts of a program

A type checker enforces a type system on users’ programs

59

Type Systems (continued) Example type rules

» E1 and E2 are of the type int ==> E1+E2 is of the type int.

» x = E ==> x and E must have the same type.

60

Type checking Type checker

» enforces rules of type system» may be strong/weak, static/dynamic

Static type checking» performed at compile time» early detection, no run-time overhead» not always possible (e.g., A[i])

Dynamic type checking» performed at run time» more flexible, rapid prototyping» overhead to check run-time type tags

61

Type expressions Type expressions (v.s. value expressions)

» used to represent the type of a language construct

» describes both language and programmer types

Examples» basic types: int, float, char, ...» constructed types: arrays, records,

pointers, » functions,...

62

Type Expressions A basic type is a type expression

» boolean, char, integer, real, void, typeError A type constructor applied to type expressions is

a type expression» array: array(I, T)» record: T1 x T2 x …x Tn

» pointer: pointer(T)» function: D R

63

Type Expressions Constructing new types, if T, T1,…, Tn are type

expressions, I an non-negative integer then» array(I, T) // array type» T1 x T2 x … x Tn // record type» pointer(T) // pointer» T1 x T2 x …x Tn T // funciton

are type expressions

64

A simple type checker Using a synthesized attribute grammar, we will

describe a type checker for arrays, pointers, statements, and functions.

Grammar for source language:» P D ; S» D D ; D | id: T» T char | integer | array [num] of T | * T» E literal | num | id | E mod E | E[E] | E *» S id = E; | if E then S | while E do S | S ; S

65

Basic types char, integer, typeError assume all arrays start at 1, e.g., array [256] of

char results in the type expression array(256,char)

* builds a pointer type, so * integer results in the type expression pointer(integer)

66

Type Declarations D id: T { addType(id.entry, T.type) } T char { T.type = char } T integer { T.type = integer } T * T1 { T.type = pointer(T1.type) } T array [num] of T1

{ T.type = array(num.val, T1.type) }

67

Type checking of expressions E literal { E.type = char } E num { E.type = int } E id { E.type = lookup(id.entry) } E E1 mod E2 { E.type = (E1.type == int and

E2.type == int ) ? int : typeError } E E1 [ E2 ] {E.type = ( E1.type == array(s, t)

and E2.type == int)? t : typeError} E * E1 {E.type := (E1.type == pointer(t))? t :typeError}

68

Type checking of Statements P D “;” S S id “=” E {S.type = (lookup(id.entry) ==

E.type) ? void : typeError} S if E then S1 {S.type = (E.type == boolean)?

S1.type : typeError} S while E do S1 {S.type := (E.type ==

boolean) ? S1.type else typeError} S S1 “;” S2 {S.type := (S1.type == void and

S2.type == void)? void : typeError}

69

Type Checking of Functions T T1 T2 {T.type := T1.type T2 .type}

E E1 ( E2 ) {E.type := (E1.type == (s t and E2.type == s )? t : typeError}