CS-445 Programming Languages Slides 2 –

Classification of Programming Languages

By Talha Waheed, UET Lahore

 Typical classifications of PL

 Low-level vs. High-level vs. Very High Level

 Language generations
 1st , 2nd , 3rd , 4th, 5th ...
 e.g. machine, assembly, procedural, application,
AI techniques & inference languages, neural nets
 General-purpose vs. special purpose
 C/C++, Scripting Languages,
Robot Programming Language
 Different Language Implementation
 Language Paradigms
2
 Communicating with Computers
 Non-Tech Definition of a PL
 way of communicating with computer to tell it what it should do.
 Challenge for half a century
 ENIAC to present

 Approaches to communicating with computers

 Driven by applications and their properties
 Also by computers and their properties
 But a "modern" approach should minimize this
 Still arguments based on "performance"

 But with 500 M operations per second available and doubling every
couple of years
 Performance can no longer be the dominant issue!

3
ENIAC To Present

Ruby on Mobile
Rails Development Machine Learning
4
First Generation of Computers

5
 Evolution of Communicating with Computers
• Communicating at all by
 Cables, Binary, ASSEMBLY LANGUAGE
• Communicating math (formulas)
 FORTRAN
• Communicating business math
 To non-mathematicians
 COBOL
• Communicating algorithms
 ALGOL

• Simulation
 SIMULA First Ever Computer Program??

6
 1st Generations of Languages -
Machine Languages
 An example code (von Neumann machine, 1946) for the program
Add numbers in locations 10 and 11 and store result in location 12:
00000010101111001010
00000010111111001000
00000011001110101000
 Operations are specified in terms of actual numeric code.
 Operand are specified in terms of their absolute address in memory
or as a register number.
 Everything is encoded as a sequence of 0s and 1s.
 As a result, programming requires deep knowledge of underlying
machine architecture and programs become unintelligible!

7 Programming Languages - Talha Waheed

 1st Generations of Languages -
Assembly Languages
 To make programs more intelligible,
 each operation and operand specified using symbolic name
e.g., load, add, store
 Assembler turns assembly code into machine code.
 Can have macros as shorthand for sequences of code
(macro assembler).
 Still Machine dependent, i.e., features depend on
hardware.
 Hard to read and to port to other machines.
 Programming is similar to programming in machine
languages, though somewhat less difficult.
8
 Generations of Languages -
High Level Languages
 Today's conventional languages (C, C++, Java, Lisp (List Processor)).
 Higher level constructs such as
 if and while statements, procedures, etc.
 much easier to write and read
 supporting consistency checks which help detect bugs
 preprogrammed libraries (reusability)
 machine independent, though sometimes implementation dependent
 Portable code
 programmer specifies instructions that a language processor
understands and translates or executes.

9
10
11
History of Language Development
 1950s - Exploiting Machine Power
 Machine Language, Assembly Languages, FORTRAN
 1960s - Increasing Expressive Power
 COBOL, Lisp, Algol 60, BASIC, SNOBOL, PL/I, Simula
 1970s - Portability and Reliability
 Pascal, Algol 68, C
 1980s - Managing Complexity
 Modula-2, Ada, Smalltalk, Eiffel, ML
 1990s - Parallelism, Distribution, Objects
 C++, Modula-3, Ada 95, . . . JAVA!!!, XML
 2000s - Web based, AI
 C#, Ruby, Python, R , What Next ???
12 Programming Languages - Talha Waheed
History: mid-1950s to early 1970s (1/2)
Main areas of application
 numerical computation
 1957-FORTRAN
 1958-Algol
 business
 1955-FLOWMATIC (Grace Hopper)
 1959-COBOL
 artificial intelligence
 about 1955-IPL, IPL-2, IPL-V (Newell, Shaw, Simon)
 Fall 1958-Lisp (McCarthy)
 for string processing: COMIT, SNOBOL4
 systems programming
 CPL, BCPL
 1972-C

13 Programming Languages - Talha Waheed

History: mid-1950s to early 1970s (2/2)

 Hardware: main frames running batch processes

 mostly file processing, punched cards
 error recovery and exception handling to reduce need
for reruns
 no timing constraints on programs

14 Programming Languages - Talha Waheed

History: 1970s to 1980s

 Hardware: time sharing main frames

 time sharing --- "an idea whose time has come and
gone"
 interactive computing; keyboard input
 I/O properties of keyboards different from files
 error recovery by interaction with user
 timing constraints (computer response time, user
pauses)

15 Programming Languages - Talha Waheed

History: 1980s to 1990s
 Hardware: PCs, micro-controllers, and embedded
systems
 Time sharing not needed (but reemerges to provide
multiprocessing)
 interactive graphics is important
 micro-controllers and embedded systems operate
without OS or use special real-time OS
 embedded systems must be reliable and correct; need
to operate in real time, recover from errors, ordinarily
never terminate, are often distributed systems
16 Programming Languages - Talha Waheed
History: 1990s to present
 Hardware: the world is one big computer (Internet)
 distributed computing is everywhere on local area networks,
the internet
 reemergence of central computers (servers)
 security becomes even more important than before
 making programs run on multiple platforms becomes an issue

 New applications
 (desktop) publishing: TEX, Postscript
 software agents and other processes, including scripts

17 Programming Languages - Talha Waheed

 A Simplified Family Tree of Programming
Languages

18
19
The Evolution of Human

20
21
 Factors Influencing Language Design
The language we "speak" influences what we can say
[Whorf]

 Application programmers' perspective

 Make programming as
easy and natural as possible
 Define a user-friendly
"virtual machine“

 Language implementers' perspective

 Make efficient implementation possible
 Define a reasonable mapping to a real machine
22
 Factors Influencing Language Design

 Hardware Architecture
 Application Domains
 New uses and users, Changing requirements
 Programming methodologies
 Managing large, complex systems
 Implementation methods
 Improved techniques
 Theoretical studies Hardware

 Assess strengths and weaknesses

 Standardization (formal or de facto)
 Portability & Re-use Applications Programming
Systems
 Strong promoter (usually an organization) Languages
Software
 Easy access to high quality compilers and
tools
23 Theory
 Translation

Converting a program written in a HLL into machine language.

Compilation: whole program is translated before execution. (C/C++)
Interpretation: Translate and execute, one statement at a time. (BASIC)

Source Source
Code Code

Compiler Input Interpreter Output

Input Target Output

Code
24
 Comparison of the two methods
Compilation Interpretation
 Brings the program down to  Brings the machine up to the
machine level program level.
 Can execute translated program  Must retranslate for every
many times because the entire execution.
translation is produced.
 Program execution is much faster  Program execution is much slower.
because the translator can do
optimization.
 Easier to provide feedback when
 Harder to provide feedback when
debugging because executing the
debugging because executing the
source code.
target code.

25
 Pseudocompilation

 A hybrid of compilation and interpretation

Source
Code

Compiler

Intermediate
Code

Input Interpreter Output

26
 Pseudo-compilation
 A compiler translates whole program before execution, but only
into intermediate code.

 An interpreter translates and executes intermediate code one

statement at a time.

 The intermediate code can be executed on any machine that has

interpreter for intermediate code.
 more portability and machine independence

 Java initially used this hybrid strategy. (intermediate code: bytecode)

27