Programming Languages:Introduction and Overview

Alan Borning

borning@cs.washington.edu

University of Washington, Seattle

CSE 341, Spring 2000

Course topics

• Scheme

• Perl

• Java

• Miranda (a pure functional language)

• CLP(R) (constraint logic programming)

• General programming language concepts

Required work

• Moderate sized programs in Scheme, Perl, Miranda, CLP(R)

• Java project

• Course project (implementation and paper) on a language of your choosing (other than Java)

• Midterm, final

• Some written homework

Organizational

• Instructor: Alan Borning

• TA: Keunwoo Lee

• when in doubt, check the class web page

• please sign up for the cse341 mailing list:mail to majordomo@cs.washington.eduin body of message: subscribe cse341

History of Programming Languages

1955

1960

1965

Fortran

Algol 60

Lisp

CPL

Snobol

History ofProgramming Languages

1965

1970

1975

BCPL

Algol 68

B

Prolog

C

Unix

rewritten

in C

Simula

Scheme

Smalltalk-72

History ofProgramming Languages

1975

1980

1985

K&R Published

The C Programming

Language

ANSI X3J11

convened to

standardize C

C with Classes

CPre

C++

Objective

C

Smalltalk-80

Icon

History ofProgramming Languages

1985

1990

1995

Stroustrup’s

The C++ Programming

Language

ANSI X3J16

Formed for

C++ Standard

Oak

Project

C++ PL

2nd Ed.

Java

Haskell

Perl

Perl5

What is a programming language for?

• Instructing machines?

• Communicating among programmers?

• Expressing high level designs?

• Notation for algorithms?

• Tool for experimentation?Languages are for both humansand computers!

Effective Use of Programming Languages

“Learning the fundamentals of a programming language is one thing: learning how to design and write effective programs in that language is something else entirely.”

—Scott Meyers

Why do we care?

• Whorf-Sapir hypothesis for natural languages

• Tradeoffs among languages
  • reusability, maintainability
  • performance, robustness
  • flexibility, dynamicism
  • libraries
  • aesthetics (i.e., “fun-ness”)

Language classification

• Imperative (Fortran, Algol, C)

• Functional (“Pure” Scheme/Lisp, Miranda)

• Logic/Constraint (Prolog, CLP(R))

• Object-oriented (Smalltalk, Java, C++)

• Languages may encourage a certain style even if they do not force it on you!

Complexity vs. Expressiveness

Complexity

Expressiveness

C

C++

Java

Scheme

Haskell

Miranda

What’s wrong with imperative?

int i = 7;

printf("%d\n",i*2);

• What gets printed?

Assignments makereasoning difficult!

int i = 7;

i = 3;

printf("%d\n",i*2);

Imperative programming

• Nice for execution, translation… BUT:

• Harder for humans tounderstand and reason about

• Harder for sophisticated software tools
  • Proving correctness is harder
  • Restricts code motion, limits optimizer(especially important for parallel machines)

The Functional Approach

• Eliminates assignments (side effects),focus on expressions

• Tell what to compute, not how(leave order of computation unspecified)

• Higher level programming model—leave more details to machine

Scheme

• Very simple syntactically

• Still an imperative language, though

• But encourages a functional style

• Can write in a purely functional subset
  • we will do this in the beginning
  • still has assignments

• Dynamically typed

Miranda (and Haskell)

• Pure functional languages

• Statically-typed

• “Lazy” evaluationSample Miranda function definition:factorial n = product [1..n]

Constraint Logic Programming

• Metaphor: theorem proving and equation solving

• Again, no side effects

• Variables are like those in mathematicsSample CLP(R) rule:centigrade_fahrenheit(C,F) :- 1.8*C=F-32.Use:?- centigrade_fahrenheit(X,212).