PRINCIPLES OF PROGRAMMING LANGUAGES 1

CSC 132
Syllabus

 Unit – 1: Syntax and Semantics

 Unit – 2: Data, Data Types and Basic Statements

 Unit – 3: Subprograms and its Implementation

 Unit – 4: Object-Orientation, Concurrency, Event Handling

 Unit – 5: Functional Programming Languages (Scheme, ML)

 Unit – 6: Logic Programming Languages (Prolog)

Why Study Programming Languages?

 Increased capacity to express ideas

 Improved background for choosing appropriate language
 Increased ability to learn new languages
 Better understanding of the significance of implementation
 Better use of languages that are already known
 Overall advancement of computing
Programming Domains

 Scientific Applications (Fortran)

 Business Applications (COBOL)
 Artificial Intelligence (LISP, Prolog)
 Systems Programming (C)
 Web Software (HTML, PHP, Java)
 Language Evaluation Criteria
 Readability
 Writability
 Reliability
 Cost

Adopted from Concepts of Programming Languages - Sebesta

Readability

 Overall simplicity
 Orthogonality
 Data Types
 Syntax Design
 Special words (while, if etc...)
 Form and meaning (Semantics should follow directly from the syntax. Ex: static has different
meaning based on context in C)
Readability – Overall Simplicity

 A language with more number of basic constructs is more difficult to learn than one with a
smaller number.

 Feature multiplicity (having more than one way to accomplish a particular operation).
 Ex: incrementing by 1

 Operator overloading (ex: using + for adding arrays or difference b/w first elements in the
arrays).

 Too much simplicity makes program less readable (ex: assembly language).
Readability – Orthogonality

 Orthogonality is the ability of combining a small set of primitives to build control and data
structures.

 The more orthogonal the design of a language, more is the simplicity.

 Most orthogonal programming language is ALGOL 68.

 Functional languages offer potentially the greatest overall simplicity (as the same
construct, a function can be used for all operations).
Readability – Orthogonality (cont...)

 IBM mainframe instructions for adding two integers that reside in main memory or
registers:
A reg1, memory_cell
AR reg1, reg2

 VAX minicomputer instruction:

ADDL operand_1, operand_2

 In the above example, VAX instruction is more orthogonal as it supports all four
combinations with single instruction.
Readability – Data Types

 The presence of meaningful data types aids readability.

 If a numeric type is used to indicate Boolean conditions, the statement will be unclear.
Ex: timeOut = 1
Writability

 Simplicity and Orthogonality

 A language with small set of primitives is better than one with a large set of primitives
 Too much orthogonality decreases writability
 Support for Abstraction
 Expressivity
Writability – Support for Abstraction

 Abstraction is the ability to define and use complicated structures or operations in ways
that allow many of the details to be ignored.

 Programming languages can support two types of abstraction: process and data.

 Example for process abstraction is to use a subprogram for sorting that is required several
times in a program.

 Example for data abstraction is a binary tree that stores integer data in its nodes. It is better
to implement a binary tree in C++ or Java using classes rather than in Fortran77 using
pointers and dynamic memory management.
Writability – Expressivity

 Ability of a language that allows a great deal of computation to be accomplished with a

very small program.

 In C, count++ is more convenient than writing count = count + 1.

 Inclusion of for statement in Java makes writing counting loops easier than with the use of
while.
Reliability

 A program is said to be reliable if it performs to its specifications under all conditions.

 Type Checking
 Exception Handling
 Aliasing
 Readability and Writability
Reliability – Type Checking

 Type checking is simply testing for type errors either at compile time or run-time.

 As run-time type checking is expensive, compile time checking is desirable. (Ex: Java)
Reliability – Exception Handling

 Ability of a program to intercept run-time errors and take corrective measures is known as
exception handling.

 Ada, C++, Java and C# provides extensive capabilities for exception handling.
Reliability - Aliasing

 Aliasing is having two or more distinct names that can be used to access the same memory
cell (Ex: Pointers in C and C++).

 Many languages greatly restrict aliasing to increase their reliability.

Cost

 Total cost of a programming language is a function of:

 Cost of training the programmers
 Cost of writing the programs
 Cost of compiling the programs in the language
 Cost of executing programs
 Cost of language implementation system (Ex: JVM)
 Cost of poor reliability
 Cost of maintenance

 Other criteria that can be used to evaluate a programming language are: portability, generality
and well-definedness.
Influences on Language Design

 Computer Architecture
 Programming Design Methodologies
Computer Architecture

 Imperative languages have been designed around the computer architecture called the von
Neumann architecture.

 In this architecture, data and programs are stored in memory and they are executed by the
CPU.

 Central features of imperative languages are variables, which model memory cells;
assignment statements, which are based on the piping operation; and iterative form of
repetition, which is the most efficient way to implement repetition on this architecture.
Computer Architecture (cont...)

Adopted from Concepts of Programming Languages - Sebesta

Programming Design Methodologies

 Intense analysis begun in large part by the structured programming movement in the late
1960s.

 New software development methodologies that emerged as a result of the research in

1970s were called top-down design or stepwise refinement.

 In the late 1970s, a shift from procedure-oriented to data-oriented program design

methodologies began.
Programming Design Methodologies (cont...)

 First language to provide a limited support for data abstraction is SIMULA 67.

 Latest step in the evolution of data-oriented software development, which began in the
early 1980s, is object-oriented design.

 First language to include object-oriented concepts was Smalltalk.

Programming Design Methodologies (cont...)

 Imperative language like Ada 95, C++, Java and C# support object-orientation.

 Functional languages like CLOS and F# also support object-orientation.

 Logic programming language like Prolog++ supports object-orientation.

Language Categories

 Imperative languages
 Object-oriented languages
 Functional languages
 Logical languages
 Visual languages
 Scripting languages
 Markup languages
Language Design Trade-offs

 Reliability vs. Cost of Execution

 Ex: Arrays in Java are more reliable than in C (no bounds checking); whereas cost of execution of
C arrays is less than that of Java.

 Readability vs. Writability

 Ex: APL provides rich set of operators for array operations. This allows to write a complex
computation in less number of lines. But the readability of program decreases. (Daniel
McCracken took four hours to read and understand a four-line APL program)
Implementation Methods

 A language implementation system cannot be the only software on a computer. Requires an

operating system also.

 The operating system and language implementations are layered over the machine
language interface of a computer.
Implementation Methods (cont...)

Adopted from Concepts of Programming Languages - Sebesta

Implementation Methods - Compilation

 Programs can be translated to machine language and be directly executed on the computer.
This is called compiler implementation.

 Very fast program execution.

 Ex: C, C++, COBOL, Ada

Implementation Methods – Compilation
(cont...)

Adopted from Concepts of Programming Languages - Sebesta

Implementation Methods – Compilation (cont...)

 The user and system code together is called a load module or executable image.

 Process of collecting system programs and linking them to user programs is called linking
and loading. This is done by a system program called as linker.

 The speed of connection between a computer’s memory and its processor is known as von
Neumann bottleneck. This is the primary motivation for research and development of
parallel computers.
Implementation Methods – Pure
Interpretation
 Programs are interpreted by another program called an interpreter, with no translation.

 Advantage is easy implementation of many source-level debugging operations.

 Disadvantage is this method is 10 to 100 times slower than compiled systems.

 Statement decoding is the bottleneck of a pure interpreter.

Implementation Methods – Pure Interpretation
(cont...)

Adopted from Concepts of Programming Languages - Sebesta

Implementation Methods – Pure Interpretation
(cont...)
 Another disadvantage is, this method requires more space (for including symbol table
along with source code).

 Ex: Earlier versions of APL, SNOBOL and LISP

Implementation Methods – Hybrid Approach

Adopted from Concepts of Programming Languages - Sebesta

Implementation Methods – Hybrid Approach
(cont...)
 High-level programs are translated to an intermediate language designed to allow easy
interpretation.

 This method is better than pure interpretation as the source language statements are
decoded only once.

 Ex: Perl, Java and .NET

Programming Environments

 A programming environment is a collection of tools used in the development of software.

 UNIX is an older programming environment. GUI versions of UNIX or Solaris CDE,

GNOME and KDE.

 Borland’s JBuilder is an IDE.

 Microsoft’s Visual Studio .NET

 NetBeans