Transcript 강의 내용 및 방법 - 부산대학교 인공지능

강의 내용 및 방법
 접근방법
– Proof of Correctness는 제외
– Lambda Calculus: Python과 관련하여 조금 설명
– 리포트 2주일에 프로그램 1개 정도, term project는 없
음, 각 장의 문제풀기 및 다른 문제 제공
 시험
– 중간, 기말 및 1~2회 정도의 쪽지시험
 수업은 75분(14:30-15:45), 혹시 부족하면 보강
참고자료
 Http://borame.cs.pusan.ac.kr/lecture
–
–
–
–
강의참고자료 (한글)
2002년 강의 내용
2002년 리포트, 기타 참고사항
2003, 2004, 2005, 2006, 2007, 2008, 2009
년, 2010, 2011, 2012년 강의노트
– 다른 대학교, 자료, 우균 교수 자료(한글) 참고
 강의 중에 이해가 안 되는 부분은 언제나 홈페
이지의 묻고 답하기에 질문할 수 있음
프로그램 개발
 어떻게 하면 프로그램 개발과 유지보수 비용을 줄이
느냐?
– 프로그램에서 재사용할 수 있는 부분과 아닌 부분을 나
누어서 개발하고, 재사용을 극대화 (개인별 라이브러리
보관)
– 많은 라이브러리를 보유하고 활용
– high-level data types built in, such as flexible
arrays and dictionaries
 그러면서 어떻게 하면 메모리 사용량을 줄이고, 속도
를 높이느냐?
– Runtime checking을 최소화
프로그래밍 언어, 어떤 것을 쓰나?

Paul Graham은 "Revenge of the Nerds"와 "Beating for average"에서
언어의 관성을 언급하면서, 익숙해지면 그 언어의 틀 속에서 자신의 생
각을 가두게 되어 편견에 사로잡혀 언어의 활용을 효과적이지 못하게
된다면서, 익숙한 언어보다는 적합한 프로그래밍 언어를 선택할 줄 알
아야 한다고 했죠.
즉, 최소한 몇 가지 언어를 알고 자신의 업무에 적합하게 사용할 줄 알
아야 한다는 말입니다.

GitHub: JavaScript, Ruby, Python, PHP, Java, C#, C, Perl, C++,
Objective C
StackOverflow: C#, Java, PHP, JavaScript, C++, Python,
Objective-C, C, Ruby, Perl, Delphi

프로그래밍 언어 사용 빈도 (Script 언어 제외)
Tiobe.com (Skilled Engineer)
1. Java – 19.1%
2. C – 15.2%
3. C++ – 10.1%
4. PHP – 8.7%
5. Visual Basic – 8.4%
6. Perl – 6.2%
7. Python – 3.8%
8. C# – 3.7%
9. JavaScript – 3.1%
10. Ruby – 2.6%
11. Delphi – 2.1%
Dice.com (2006년)
1. Java – 35.7%
2. C, C++ – 15.3%
3. C# – 12.7%
4. Perl – 11.9%
5. JavaScript – 10.9%
6. Visual Basic .NET –
5.2%
7. PHP – 2.9%
8. Ajax – 2.7%
9. Python – 2.0%
10. Ruby – 0.7%
왜 프로그래밍 언어론을 배우나?
 수많은 프로그래밍 언어 중 필요에 따라 적합
한 것을 골라 쓸 수 있는 능력
 사용하는 프로그래밍 언어의 특성을 파악하여
가장 효과적으로 사용할 수 있게 함
 사용하는 프로그래밍 언어에 없지만 유용한
개념을 도입하여 활용
 프로그래밍 언어 그 자체로도 의미가 있음
언어란?

생각을 표현하는 도구!! 정보를 교환하는 도구!!
– 개념화 : 물체(Object), 물체와 관계, 추상화
• 중력의 법칙, 예측

– 인간은 전체를 듣지 않고도 부분을 이해함, 스스로 보완하여 인식
함
– 불완전하며 모호함
프로그래밍언어 !!
– 우리의 생각을 컴퓨터가 처리할 수 있게 표현하는 도구
– 조금만 틀려도 이해 못함(수학적 명확성 필요)
– 오류가 없는 프로그램은 만들 수 없으며, 인간이 1,000만 줄 이상
프로그램은 만들 수 없다. (조던 폴락)
– 기계와 인간 간의 간격을 줄일수록 바람직함
• 추상화 단계가 높아져야 짧은 프로그램을 명확한 프로그래밍이 가
능
– 기계뿐 아니라 인간도 이해해야 함 (쉽게 프로그램할 수 있어야)
Organization of Programming
Languages


Understand how languages are designed and implemented
구조
–
–
–
–
Lexicon – What sorts of words are legal?
Syntax -- What a program looks like?
Semantics -- What a program means?
Pragmatics – 실행 과정에서 해석? C언어의 ‘int’형
 가장 기본 단위
– 문법요소, 단어, 고유명사 등 명칭
– 문법요소, keywords, 명칭
 기본요소가 결합하는 방법
 결합한 요소의 뜻 파악

Implementation -- How a program executes
프로그래밍 언어의 모형

Understand most appropriate language for solving
specific problems, For example:
– Pascal, C, Perl -- procedural, statement oriented
– C++, Java, Smalltalk -- Object oriented
– ML, Lisp -- Functional
– Prolog -- Rule-based
– JavaScript, PHP, ASP.net, C# – Procedural, Eventdriven accessing
Language Goals
 During 1950s--1960s - Compile programs to
execute efficiently.
– There is a direct connection between
language features and hardware - integers,
reals, goto statements
– Programmers cheap; Machines expensive; Keep
the machine busy
 But today
– Compile programs that are built efficiently
– CPU power and memory very cheap
– Direct connection between language features
and design concepts - encapsulation, records,
inheritance, functionality, assertions
– Event-driven programming, Service-oriented
approach, Web-service, Very High-level
Language
Python



Python is a general-purpose, high-level programming language. Its
design philosophy emphasizes programmer productivity and code
readability. Python's core syntax and semantics are minimalist,
while the standard library is large and comprehensive.
Python supports multiple programming paradigms (primarily
functional, object oriented and imperative), and features a fully
dynamic type system and automatic memory management; it is
thus similar to Perl, Ruby, Scheme, and Tcl.
Python was first released by Guido van Rossum in 1991.[3] The
language has an open, community-based development model
managed by the non-profit Python Software Foundation. While
various parts of the language have formal specifications and
standards, the language as a whole is not formally specified. The
de facto standard for the language is the CPython implementation.
Era
Application
Major
languages
Other
languages
1960s
Business
Scientific
COBOL
FORTRAN
System
Artificial
intelligence
Business
Assembler
LISP
Assembler
Algol,
BASIC,
APL
JOVIAL, Forth
SNOBOL
Today
Scientific
System
Artificial
intelligence
Publishing
Process
New paradigms
COBOL,
C++, C, PL/I, 4GLs
Java, spreadsheet
FORTRAN,
C, BASIC
C++, Java
C, C++, Java
Ada,
BASIC,
Modula
LISP, Prolog
TeX,
Postscript,
word processing
UNIX shell, TCL, AWK,
Perl, Javascript
SED
ML, Smalltalk
Eiffel
Marvel,
Why study programming
languages? (1)
 To improve your ability to develop effective
algorithms
– Improper use of recursion
– Object-oriented programming, logic programming,
concurrent programming
 To improve your use of your existing
programming language
– Data structures for arrays, strings, lists,
records, set, bag, table (associative array)
– Malloc()  garbage collection
– Implementation details of recursion, object
classes, subroutine calls, exception(event)
handling …
– Service-oriented developing
• Web-service, Semantic web, Script language
Why study programming
languages? (2)
 To increase your vocabulary of useful
programming constructs
– Increase programming vocabulary and its implementation tech.
– Coroutine, Semaphore, event-driven programming, associative
array
– Named parameter (keyword argument) (Ada, Python, Smalltalk,
Common Lisp)
– Name mangling (name decoration): _Z on C, C++
• Filename mangling : Unix file names can contain colons or backslashes
 To allow a better choice of programming
language
– Numeric computation : C, FORTRAN, Ada
– AI : LISP, Prolog
– Internet applications : Perl, Java, HTML, XML, Web
Service(SOAP), PHP, ASP.net, Python
Why study programming
languages? (3)
 To make it easier to learn a new
language
– Web programming (JavaScript, PHP), Semantic web
– Python
 To make it easier to design a new
language
– User interface design
 *** C, COBOL, SMALLTALK의 덧셈의 속도차
이를 초래하는 언어개념 및 구현의 차이
Evolution of software architecture

1950s - Large expensive mainframe computers ran
single programs (Batch processing)

1960s - Interactive programming (time-sharing) on
mainframes

1970s - Development of Minicomputers and first
microcomputers. Apple II. Early work on windows,
icons, and PCs at XEROX PARC

1980s - Personal computer - Microprocessor, IBM PC
and Apple Macintosh. Use of windows, icons and mouse

1990s - Client-server computing - Networking, The
Internet, the World Wide Web

2000s - P2P, Grid Computing, Web Service, Eventdriven approach, Service-oriented programming, Script
language, Cloud Computing …
Attributes of a good language (1)
 Clarity, simplicity, and unity provides both a framework for thinking
about algorithms and a means of
expressing those algorithms
– Conceptual integrity
– 나쁜 예: APL, SNOBOL4
 Orthogonality - every combination of
features is meaningful
– Fewer exceptions  C언어에서 !!!
– Logical errors and inefficiency
Attributes of a good language(2)
 Naturalness for the application program structure reflects the logical
structure of algorithm
– Sequential algorithm, concurrent algorithm, logic
algorithm, non-deterministic algorithm
– Appropriate data structures, operations, control
structures, natural syntax
 Support for abstraction - program data
reflects problem being solved
– Data abstraction <D,O,C>
– Encapsulation
Attributes of a good language (3)
 Ease of program verification - verifying
that program correctly performs its
required function
– Verification/validation
• standard, or specification requirements
• the needs of the intended end-user or customer
– Comments, assert()
– Design specification
 Programming environment - external
support for the language
–
–
–
–
Debugger, syntax-directed editor
Supporting function, platforms
Smalltalk
Supporting all the software lifecycle phases
Attributes of a good language
(continued)
 Portability of programs transportability of the resulting
programs from the computer on which
they are developed to other computer
systems
– Transportability
– C, C++, Pascal  Java (Byte-code)
– ML : single source implementation
 Cost of use - program execution,
program translation, program creation,
and program maintenance
–
Code optimization, (Smalltalk, Perl), lifecycle costs
Language paradigms
 Imperative languages
– Goal is to understand a machine state (set of
memory locations, each containing a value)
– Statement oriented languages that change machine
state (C, Pascal, FORTRAN, COBOL, C++, Java)
– Syntax: S1, S2, S3, ...
 Applicative (functional) languages
– Goal is to understand the function that produces
the answer
– Function composition is major operation (ML, LISP)
– Syntax: P1(P2(P3(X)))
– Programming consists of building the function that
computes the answer
example1.f
*234567******************* example1.f **************************
*
A tiny program to plot my running times using GrADS
*
****************************************************************
*
PARAMETER(NDAYS=365, UNDEF=-9.99, EMPTY=0.00)
C
REAL MIN(NDAYS)
Changing to REAL
DIMENSION SEC(NDAYS)
C
CHARACTER*27 HEADER
CHARACTER*3 DAY
CHARACTER*28 PATH
C
PATH='/data/temp4/alfredo/fortran/'
C
OPEN(1,FILE=PATH//'runtimes.data',FORM='FORMATTED',
STATUS=‘OLD')
Nonexecutable

















Sequential (by default)
example1.f




















Undefined value
READ (1,2) HEADER
WRITE(*,2) HEADER
WRITE(*,2) '----------------------------'
runtimes.data
SEC
-9.99
C
-9.99
43.00
ND = 0
33.00
1 CONTINUE
33.00
ND = ND+1
27.00
READ (1,3,END=4) DAY,DATE,MIN(ND),SEC(ND)
-9.99
WRITE(*,3) DAY,DATE,MIN(ND),SEC(ND)
46.00
IF((MIN(ND).EQ.EMPTY).AND.(SEC(ND).EQ.EMPTY)) GO TO 4
...
GO TO 1
TUE 15102002 29.00 7.00
WED 16102002 -9.99 -9.99
4 CONTINUE
THU 17102002
ND = ND - 1 !Taking away the last line because is blank
FRI 18102002
CLOSE(1)
SAT 19102002
3 FORMAT(A3,2X,I8,2X,F5.2,2X,F5.2)
SUN 20102002
WRITE(*,*)'-------------------------------------------'
MON 21102002
WRITE(*,*)'I AM RUNNING SINCE JULY 1, 2002!!!!!'
TUE 22102002
WRITE(*,*)'THAT IS',ND,' DAYS AGO'
WED 23102002
THU 24102002
C
FRI 25102002
SAT 26102002
SUN 27102002
No value
DAY DATE
MON 01072002
TUE 02072002
WED 03072002
THU 04072002
FRI 05072002
SAT 06072002
SUN 07072002
MON 08072002
MIN
-9.99
-9.99
19.00
19.00
19.00
19.00
-9.99
19.00
example1.f
















OPEN(2,FILE=PATH//'runtimes_gr.data',ACCESS='DIRECT',
STATUS='UNKNOWN',FORM='UNFORMATTED',RECL=1)
C
NDNOR = 0
NDSIR = 0
DO N = 1, ND
IF((MIN(N).EQ.UNDEF).OR.(SEC(N).EQ.UNDEF)) THEN
TIME = UNDEF
NDNOR = NDNOR + 1
ELSE
TIME = MIN(N)+SEC(N)/60.
NDSIR = NDSIR + 1
ENDIF
WRITE(2,REC=N) TIME
ENDDO
example1.f














WRITE(2,REC=ND+1) UNDEF !Just adding 1 and 2 extra
WRITE(2,REC=ND+2) UNDEF !blank lines for plotting purposes
CLOSE(2)
WRITE(*,*)'WELL,',NDSIR,' DAYS LEAVING FOR A RUN'
WRITE(*,*)'AND',NDNOR,' DAYS JUST BEING LAZY'
WRITE(*,*)'------------------------------------------'
WRITE(*,*)'make ',ND+2,' days in the script file:'
WRITE(*,*)'grads -blc "run runtimes.gs"'
C
2 FORMAT(A27)
C
C f77 example1.f
C ./a.out
END
COBOL II

















000010 IDENTIFICATION DIVISION.
000020 PROGRAM-ID. SAMPLE.
000030 AUTHOR. J.P.E. HODGSON.
000040 DATE-WRITTEN. 4 February 2000
000041
000042* A sample program just to show the form.
000043* The program copies its input to the output,
000044* and counts the number of records.
000045* At the end this number is printed.
000046
000050 ENVIRONMENT DIVISION.
000060 INPUT-OUTPUT SECTION.
000070 FILE-CONTROL.
000080
SELECT STUDENT-FILE ASSIGN TO SYSIN
000090
ORGANIZATION IS LINE SEQUENTIAL.
000100
SELECT PRINT-FILE ASSIGN TO SYSOUT
000110
ORGANIZATION IS LINE SEQUENTIAL.
COBOL II

















000120
000130 DATA DIVISION.
000140 FILE SECTION.
000150 FD STUDENT-FILE
000160
RECORD CONTAINS 43 CHARACTERS
000170
DATA RECORD IS STUDENT-IN.
000180 01
STUDENT-IN PIC X(43).
000190
000200 FD PRINT-FILE
000210
RECORD CONTAINS 80 CHARACTERS
000220
DATA RECORD IS PRINT-LINE.
000230 01
PRINT-LINE PIC X(80).
000240
000250 WORKING-STORAGE SECTION.
000260 01
DATA-REMAINS-SWITCH PIC X(2) VALUE SPACES.
000261 01
RECORDS-WRITTEN PIC 99.
000270
COBOL III













000270
000280 01
000290
000300
000310
000311
000312 01
000313
000314
000315
000316
000318
000319
DETAIL-LINE.
05 FILLER PIC X(7) VALUE SPACES.
05 RECORD-IMAGE PIC X(43).
05 FILLER PIC X(30) VALUE SPACES.
SUMMARY-LINE.
05 FILLER PIC X(7) VALUE SPACES.
05 TOTAL-READ PIC 99.
05 FILLER PIC X VALUE SPACE.
05 FILLER PIC X(17) VALUE 'Records were read'.
05 FILLER PIC X(53) VALUE SPACES.
COBOL IV








000320 PROCEDURE DIVISION.
000330 PREPARE-SENIOR-REPORT.
000340
OPEN INPUT STUDENT-FILE.
000350
OUTPUT PRINT-FILE.
000351
MOVE ZERO TO RECORDS-WRITTEN.
000360
READ STUDENT-FILE
000370
AT END MOVE 'NO' TO DATA-REMAINS-SWITCH.
000380
END-READ.
Language paradigms (continued)

Rule-based languages
– Specify rule that specifies problem solution (Prolog,
BNF Parsing)
– Other examples: Decision procedures, Grammar rules
(BNF)
– Syntax: Answer  specification rule
– Programming consists of specifying the attributes of
the answer

Object-oriented languages
– Imperative languages that merge applicative design
with imperative statements (Java, C++, Smalltalk)
– Syntax: Set of objects (classes) containing data
(imperative concepts) and methods (applicative
concepts)
New Paradigm
 Internet, Home Computing, Ubiquitous Computing,
Embedded Computing, etc
 Languages and Compilers that are smart, small and safe
– Moveable code  agents
– Independent from machines and programming
environments
– Supporting pragmatics or intelligence
 Java, XML, Mobile computing. Mobile code
 Event driven approach, Service-oriented Architecture, Web
Service
 Supports multiple paradigms : Python
 다양한 플래트폼, 응용시스템과 환경을 결합하여 사용자에 최적
화한 느낌과 경험을 제공하는 통합된 응용시스템을 개발하는
환경
다음을 조사하라!!
(1) Orthogonality가 C언어에서 발생시킬 수 있는 오류의 예를 들
어라!
(2)우리가 사용하는 언어는 왜 프로그래밍 언어로 부적합한가?
(3) 스마트폰이 일반화하면서 ‘Platform’이란 말이 일반화하고 있
다. ‘Platform’이란 무엇이며, 일반 응용프로그램과 어떤 차이가
나는가? 현재 널리 쓰이는 smartphone platform을 조사하고,
이 platform을 만드는 데 사용한 프로그래밍언어와 응용소프트
웨어(앱)과 연결 가능한 언어 및 연결 방법을 조사하라.