symbol table is a data structure kept by a translator that …kzhu/ei326/l2.pdfsymbol table a symbol...

SYMBOL TABLE

A symbol table is a data structure kept by a

translator that allows it to keep track of each

declared name and its binding.

Assume for now that each name is unique within

its local scope.

The data structure can be any implementation of

a dictionary, where the name is the key.

1

HANDLING NONLOCAL REFERENCES

1. Each time a scope is entered, push a new dictionary onto

the stack.

2. Each time a scope is exited, pop a dictionary off the top of

the stack.

3. For each name declared, generate an appropriate binding

and enter the name-binding pair into the dictionary on the

top of the stack.

4. Given a name reference, search the dictionary on top of

the stack:

a) If found, return the binding.

b) Otherwise, repeat the process on the next dictionary down in

the stack.

c) If the name is not found in any dictionary, report an error.

2

EXAMPLE DICTIONARY STACK

At line 4 and 11:

<j, 2> <i, 2> <size,1> <a, 1>

<sort, 1>

At line 7:

<t, 6>

<j, 2> <i, 2> <size,1> <a, 1>

<sort, 1>

3

1 void sort (float a[ ], int size) {

2 int i, j;

3 for (i = 0; i < size; i++)

4 for (j = i + 1; j < size; j++)

5 if (a[j] < a[i]) {

6 float t;

7 t = a[i];

8 a[i] = a[j];

9 a[j] = t;

10 }

11 }

STATIC SCOPING

For static scoping, the referencing environment for

a name is its defining scope and all nested

subscopes.

The referencing environment defines the set of

statements which can validly reference a name.

4

DYNAMIC SCOPING

In dynamic scoping, a name is bound to its most

recent declaration based on the program’s call

history.

Used be early Lisp, APL, Snobol, Perl.

Symbol table for each scope built at compile time,

but managed at run time.

Scope pushed/popped on stack when

entered/exited.

5

6

1 int h, i;

2 void B(int w) {

3 int j, k;

4 i = 2*w;

5 w = w+1;

6 ...

7 }

8 void A (int x, int y) {

9 float i, j;

10 B(h);

11 i = 3;

12 ...

13 }

14 void main() {

15 int a, b;

16 h = 5; a = 3; b = 2;

17 A(a, b);

18 B(h);

19 ...

20 }

RUNTIME SYMBOL TABLE

Call history

main (17) A (10) B

Function Dictionary

B <w, 2> <j, 3> <k, 3>

A <x, 8> <y, 8> <i, 9> <j, 9>

main <a, 15> <b, 15>

<h, 1> <i, 1> <B, 2> <A, 8> <main, 14>

Reference to i (4) resolves to <i, 9> in A.

7

8

1 int h, i;

2 void B(int w) {

3 int j, k;

4 i = 2*w;

5 w = w+1;

6 ...

7 }

8 void A (int x, int y) {

9 float i, j;

10 B(h);

11 i = 3;

12 ...

13 }

14 void main() {

15 int a, b;

16 h = 5; a = 3; b = 2;

17 A(a, b);

18 B(h);

19 ...

20 }

RUNTIME SYMBOL TABLE (AGAIN)

Same example: call history

main (17) B

Function Dictionary

B <w, 2> <j, 3> <k, 3>

main <a, 15> <b, 15>

<h, 1> <i, 1> <B, 2> <A, 8> <main, 14>

Reference to i (4) resolves to <i, 1> in global scope.

9

DISADVANTAGES OF DYNAMIC SCOPING

Compromises the ability to statically type check

references to non-local variables

Variables in a function visible to other functions

which call this function program less reliable

Access to non-local variable follows chains of

dynamic links more time consuming

Therefore most modern languages prefer static

scoping!

10

VISIBILITY

A name is visible if its referencing environment

includes the reference and the name is not

redeclared in an inner scope.

A name redeclared in an inner scope effectively

hides the outer declaration.

Some languages provide a mechanism for

referencing a hidden name; e.g.: this.x in

C++/Java.

11

EXAMPLE JAVA PROGRAM

1 public class Student {

2 private String name;

3 public Student (String name, ...) {

4 this.name = name;

5 ...

6 }

7 }

12

OVERLOADING

Overloading uses the number or type of

parameters to distinguish among identical

function names or operators.

Examples:

+, -, *, / can be float or int

+ can be float or int addition or string

concatenation in Java

System.out.print(x) in Java

13

LIFETIME

The lifetime of a variable is the time interval

during which the variable has been allocated a

block of memory.

Earliest languages used static allocation.

Memory assigned at compile time

Function only allocated space for arguments and

return value recursive function not supported!

Algol introduced the notion that memory should

be allocated/deallocated at scope entry/exit.

Modern languages are based on this idea but may

break scope equals lifetime rule. 14

TYPES

15

A type is a collection of values and operations on

those values.

Example: Integer type has values ..., -2, -1, 0, 1, 2, ...

and operations +, -, *, /, <, ...

The Boolean type has values true and false and

operations , , .

16

BASICS

BASICS

Computer types have a finite number of values due to

fixed size allocation; problematic for numeric types.

Exceptions:

Smalltalk uses unbounded fractions.

Haskell type Integer represents unbounded integers.

More problematic is the fixed sized floating point numbers

0.2 is not exact in binary.

So 0.2 * 5 is not exactly 1.0

Floating point is inconsistent with real numbers in mathematics.

17

BASICS

In the early languages, Fortran, Algol, Cobol, all of

the types were built in.

If needed a type color, could use integers; but what

does it mean to multiply two colors?

Purpose of types in programming languages is to

provide ways of effectively modeling a problem

solution.

18

TYPE ERRORS

Machine data carries no type information.

Basically, just a sequence of bits.

Example: 0100 0000 0101 1000 0000 0000 0000 0000

• The floating point number 3.375

• The 32-bit integer 1,079,508,992

• Two 16-bit integers 16472 and 0

• Four ASCII characters: @ X NUL NUL

19

TYPE ERRORS

A type error is any error that arises because an

operation is attempted on a data type for which it is

undefined.

Type errors are common in assembly language

programming.

High level languages reduce the number of type

errors.

A type system provides a basis for detecting type

errors.

20

STATIC AND DYNAMIC TYPING

A type system imposes constraints such as the

values used in an addition must be numeric.

Cannot be expressed syntactically in EBNF.

Some languages perform type checking at compile

time (e.g., C).

Other languages (e.g., Perl) perform type checking at

run time.

Still others (e.g., Java) do both.

21

STATIC AND DYNAMIC TYPING

A language is statically typed if the types of all

variables are fixed when they are declared at

compile time.

A language is dynamically typed if the type of a

variable can vary at run time depending on the

value assigned.

Can you give examples of each?

Static: C/C++, Java, ML, Haskell

Dynamic: Perl, Python, JavaScript, Prolog, Scheme

22

BASIC TYPES

Terminology in use with current 32-bit

computers:

Nibble: 4 bits

Byte: 8 bits

Half-word: 16 bits

Word: 32 bits

Double word: 64 bits

Quad word: 128 bits

23

FINITE SIZE IN TYPES

Unlike mathematics:

a + (b + c) (a + b) + c (why??)

In most languages, the numeric types are finite in

size.

So a + b may overflow the finite range.

Also in C-like languages, the equality and

relational operators produce an int, not a Boolean

24

OVERLOADING

An operator or function is overloaded when its

meaning varies depending on the types of its

operands or arguments or result.

Java: a + b (ignoring size)

integer add

floating point add

string concatenation

Mixed mode: one operand an int, the other floating

point

25

TYPE CONVERSION

Type conversion is implicit or explicit change of the type of a value to a different one.

Implicit conversion type coercion

Explicit conversion type casting

A type conversion is a narrowing conversion if the result type permits fewer bits, thus potentially losing information.

Otherwise it is termed a widening conversion.

Should languages ban implicit narrowing conversions?

Why?

Lose information

Unexpected results:

26

In PL/I:

declare (a) char (3);

…

a = ‘123’;

a = a + 1;

NONBASIC TYPES

Enumeration:

enum day {Monday, Tuesday, Wednesday, Thursday,

Friday, Saturday, Sunday};

enum day myDay = Wednesday;

In C/C++ the above values of this type are

0, ..., 6.

More powerful in Java:

for (day d : day.values())

Sytem.out.println(d); 27

POINTERS

C, C++, Ada, Pascal

Java???

Value is a memory address

Indirect referencing

Operator in C: *

28

EQUIVALENCE OF ARRAYS AND POINTERS IN C

float sum(float a[ ], int n) {

int i;

float s = 0.0;

for (i = 0; i<n; i++)

s += a[i];

return s;

float sum(float *a, int n) {

int i;

float s = 0.0;

for (i = 0; i<n; i++)

s += *a++;

return s;

29

STRCPY EXAMPLE

void strcpy(char *p, char *q) {

while (*p++ = *q++) ;

}

30

POINTER OPERATIONS

If T is a type and ref T is a pointer:

& : T → ref T

* : ref T → T

For an arbitrary variable x:

*(&x) = x

31

PITFALLS OF POINTERS

Bane of reliable software development

Error-prone

Buffer overflow, memory leaks

Particularly troublesome in C

Other languages such as Java, Haskell, ML and

Prolog completely remove pointers from the

vocabulary of the language

However internally these languages still makes

heavy use of pointers

Still can do dynamic allocation/deallocation of

memory 32

ARRAYS AND LISTS

Array: Indexed sequences of values of the same type.

Example:

int a[10];

float x[3][5]; /* odd syntax vs. math */

char s[40];

/* indices: 0 ... n-1 */

Default one-dimensional, but possible to have array

of arrays: x[3][5] is an array of 5 3-element arrays

(odd syntax!) 33

INDEXING

Only operation for many languages

Type signature

[ ] : T[ ] x int → T

Example

float x[3] [5];

type of x: float[ ][ ]

type of x[1]: float[ ]

type of x[1][2]: float

34

STRUCTURES

Analogous to a tuple in mathematics

Collection of elements of different types

Used first in Cobol, PL/I

Absent from Fortran, Algol 60

Common to Pascal-like, C-like languages

Omitted from Java as redundant

35

STRUCTURE (CONT’D)

struct employeeType {

int id;

char name[26];

int age;

float salary;

char dept;

};

struct employeeType employee;

...

employee.age = 45;

Number of bytes employeeType requires? (assuming 32-bit int)

Answer: 39 bytes

Memory allocation may have special requirement

Reverse order of allocation

32-bit int or float must be allocated at address multiple of 4

How many bytes are needed?

Answer: 44 bytes

36

UNIONS

C: union

Pascal: case-variant record

Logically: multiple views of same storage

Useful in some systems applications

37

FUNCTIONS AS TYPES

“First-class citizens”:

Can be assigned a value

Can be passed as an argument to a function

Pascal example:

function newton(a, b: real; function f: real): real;

Know that f returns a real value, but the arguments to f are unspecified.

Java interface class:

public interface RootSolvable {

double valueAt(double x);

}

public double Newton(double a, double b, RootSolvable f);

38

SUBTYPES

A subtype is a type that has certain constraints placed on its values or operations.

In Ada subtypes can be directly specified:

subtype one_to_ten is Integer range 1 .. 10;

type Day is (Monday, Tuesday, Wednesday, Thursday, Friday, Saturday, Sunday);

subtype Weekend is Day range Saturday .. Sunday;

type Salary is delta 0.01 digits 9

range 0.00 .. 9_999_999.99;

subtype Author_Salary is Salary digits 5

range 0.0 .. 999.99; 39

SUBTYPES (CONT’D)

Java implements subtypes using class hierarchy:

Integer i = new Integer(3);

...

Number v = i;

...

Integer x = (Integer) v;

//Integer is a subclass of Number,

// and therefore a subtype

In general, t: T, s: S, t = s is possible only if:

T = S, or

S is a subtype of T 40

Type cast required

POLYMORPHISM AND GENERICS

Greek: “poly” means many, “morph” means

form/shape

A function or operation is polymorphic if it can be

applied to any one of several related types and

achieve the same result.

An advantage of polymorphism is that it enables

code reuse.

41

POLYMORPHISM

Example: overloaded built-in operators and

functions

+ - * / == != ...

Java: + also used for string concatenation

Ada 83

Ada, C++: define + - ... for new types

Java overloaded methods: number or type of

parameters

Example: class PrintStream

print, println defined for:

boolean, char, int, long, float, double, char[ ]

String, Object

42

SEMANTICS

43

MOTIVATION

To provide an authoritative definition of the

meaning of all language constructs for:

Programmers

Compiler writers

Standards developers

A programming language is complete only

when its syntax, type system, and semantics

are well-defined.

44

MOTIVATION (CONT’D)

Semantics is a precise definition of the meaning of a

syntactically and type-wise correct program.

Ideas of “meaning”:

Operational Semantics:

The meaning attached by compiling using compiler C and

executing using machine M. Ex: Fortran on IBM 709.

Axiomatic Semantics:

Program Verification

Denotational Semantics:

Statements as state transforming functions

High level mathematic precision 45

EXPRESSION SEMANTICS

Expression Semantics:

Operators + associativity + precedence

Evaluation orders

Pure vs. Impure

46

AN EXAMPLE

Infix (a + b) – (c * d)

Infix uses associativity and precedence to disambiguate.

Polish Prefix: - + a b * c d

Polish Postfix: a b + c d * -

Same symbol can’t be used for operations of different

arities

Cambridge Polish: (- (+ a b) (* c d))

47

SHORT CIRCUIT EVALUATION

Traditionally, a op b is evaluated as

eval (a) and eval (b) and then op

Optimization:

a and b evaluated as:

if a then b else false

a or b evaluated as:

if a then true else b

Also known as lazy evaluation: b can be

undefined 48

SIDE EFFECT

A change to any non-local variable or I/O.

What is the value of:

i = 2; b = 2; c = 5;

a = b * i++ + c * i;

Depends on which one is evaluated first:

b * i++ or c * I

Side Effects make a language impure and should

be avoided if possible!

49

PROGRAM STATE (IMPERATIVE & OO)

The state of a program is the bindings of all

active variables and their current values.

Maps:

the pairing of active variable names with specific

memory locations environment

the pairing of active memory locations with their

current values memory

state = memory × environment

50

PROGRAM STATE (CONT’D)

The current statement (portion of an abstract syntax

tree) to be executed in a program is interpreted

relative to the current state.

The individual steps that occur during a program

run can be viewed as a series of state

transformations.

For the purposes of this lecture, use only a map from

a variable to its value; like a debugger watch

window, tied to a particular statement.

51

THE FACTORIAL PROGRAM

// compute the factorial of n

1 void main ( ) {

2 int n, i, f;

3 n = 3;

4 i = 1;

5 f = 1;

6 while (i < n) {

7 i = i + 1;

8 f = f * i;

9 }

10 }

52


1 void main ( ) {

2 int n, i, f;

3 n = 3;

4 i = 1;

5 f = 1;

6 while (i < n) {

7 i = i + 1;

8 f = f * i;

9 }

10 }

n i f

undef undef undef

53


1 void main ( ) {

2 int n, i, f;

3 n = 3;

4 i = 1;

5 f = 1;

6 while (i < n) {

7 i = i + 1;

8 f = f * i;

9 }

10 }

n i f

3 undef undef

54


1 void main ( ) {

2 int n, i, f;

3 n = 3;

4 i = 1;

5 f = 1;

6 while (i < n) {

7 i = i + 1;

8 f = f * i;

9 }

10 }

n i f

3 1 undef

55


1 void main ( ) {

2 int n, i, f;

3 n = 3;

4 i = 1;

5 f = 1;

6 while (i < n) {

7 i = i + 1;

8 f = f * i;

9 }

10 }

n i f

3 1 1

56


1 void main ( ) {

2 int n, i, f;

3 n = 3;

4 i = 1;

5 f = 1;

6 while (i < n) {

7 i = i + 1;

8 f = f * i;

9 }

10 }

n i f

3 1 1

57


1 void main ( ) {

2 int n, i, f;

3 n = 3;

4 i = 1;

5 f = 1;

6 while (i < n) {

7 i = i + 1;

8 f = f * i;

9 }

10 }

n i f

3 2 1

58


1 void main ( ) {

2 int n, i, f;

3 n = 3;

4 i = 1;

5 f = 1;

6 while (i < n) {

7 i = i + 1;

8 f = f * i;

9 }

10 }

n i f

3 2 2

59


1 void main ( ) {

2 int n, i, f;

3 n = 3;

4 i = 1;

5 f = 1;

6 while (i < n) {

7 i = i + 1;

8 f = f * i;

9 }

10 }

n i f

3 2 2

60

ASSIGNMENT SEMANTICS

Fundamental to imperative and object-oriented

programming

Issues

Multiple assignment

Assignment statement vs. expression

Copy vs. reference semantics

Semantics of assignment:

Evaluate the source expression a value

Replace the value of the target variable a state

61

COPY VS. REFERENCE SEMANTICS

Copy: a = b;

a, b have same value.

Changes to either have no effect on other.

Used in imperative languages.

Reference

a, b point to the same object.

A change in object state affects both

Used by many object-oriented languages.

62

COPY VS. REFERENCE SEMANTICS

public void add (Object word, Object number) {

Vector set = (Vector) dict.get(word);

if (set == null) { // not in Concordance

set = new Vector( );

dict.put(word, set);

}

if (allowDupl || !set.contains(number))

set.addElement(number);

}

//Under copy semantics, number will not be added to the dictionary, because set is an object referenced from dict.

63

CONTROL FLOW SEMANTICS

To be complete, an imperative language needs:

Statement sequencing

Conditional statement

Looping statement

64

SEQUENCE

s1 s2

Semantics: in the absence of a branch (return,

break, continue, etc.):

First execute s1

Then execute s2

Output state of s1 is the input state of s2

65

CONDITIONAL

IfStatement if ( Expresion ) Statement [ else Statement ]

Example: if (a > b)

z = a;

else

z = b;

If the test expression is true

then the output state of the conditional is the output

state of the then branch

Else:

the output state of the conditional is the output state of

the else branch

66

LOOPS

WhileStatement while ( Expression ) Statement

The expression is evaluated.

If it is true, first the statement is executed, and

then the loop is executed again.

Otherwise the loop terminates.

Foreach loop:

An iterator is any finite set of values over which a

loop can be repeated.

67

symbol table is a data structure kept by a translator that …kzhu/ei326/l2.pdfsymbol table a symbol...

Documents