Introduction to C #include <stdio.h> int main () { printf(“Welcome to CS 1621!n”); }
Outline II. Program Basics A. Program skeleton preprocessor directives global declarations functions local declarations statements B. Comments and Documentation C. Names (identifiers) reserved words
Outline (cont) II. Program Basics (cont) D. Variable declarations 1. Memory allocation 2. Atomic types void, int, float, char E. Constants 1. literal 2. defined 3. memory
Outline (cont) II. Program Basics (cont) F. Formatted input/output 1. Files 2. Printf (monitor output) a. format strings field specifications b. data list 3. Scanf (keyboard input) a. format strings b. address list 4. Prompting for Input
History of C 1960: ALGOL (ALGOrithmic Language) 1967: BCPL (Basic Combined Programming Language) 1970: B programming language (typeless) 1972: C: BCPL plus B with types 1978: Kernighan + Ritchie standard for C 1989: ANSI standard for C
C Program Structure • Program defined by: – global declarations – function definitions • May contain preprocessor directives • Always has one function named main, may contain others Preprocessor Directives Global Declarations Function Definitions int main () { } Local Declarations Statements
Parts of a Program #include <stdio.h> int x; int main () { int y; printf("Enter x and y: "); scanf(&x,&y); printf("Sum is %dn",x+y); } Preprocessor Directive Global Declaration Function Local Declaration Statements
Preprocessor Directives • Begin with # • Instruct compiler to perform some transformation to file before compiling • Example: #include <stdio.h> – add the header file stdio.h to this file – .h for header file – stdio.h defines useful input/output functions
Declarations • Global – visible throughout program – describes data used throughout program • Local – visible within function – describes data used only in function
Functions • Consists of header and body – header: int main () – body: contained between { and } • starts with location declarations • followed by series of statements • More than one function may be defined • Functions are called (invoked) - more later
Main Function • Every program has one function main • Header for main: int main () • Program is the sequence of statements between the { } following main • Statements are executed one at a time from the one immediately following to main to the one before the }
Comments • Text between /* and */ • Used to “document” the code for the human reader • Ignored by compiler (not part of program) • Have to be careful – comments may cover multiple lines – ends as soon as */ encountered (so no internal comments - /* An /* internal */ comment */)
Comment Example #include <stdio.h> /* This comment covers * multiple lines * in the program. */ int main () /* The main header */ { /* No local declarations */ printf(“Too many commentsn”); } /* end of main */
Documentation • Global - start of program, outlines overall solution, may include structure chart • Module - when using separate files, indication of what each file solves • Function - inputs, return values, and logic used in defining function • Add documentation for key (tough to understand) comments • Names of variables - should be chosen to be meaningful, make program readable
Syntax of C • Rules that define C language – Specify which tokens are valid – Also indicate the expected order of tokens • Some types of tokens: – reserved words: include printf int ... – identifiers: x y ... – literal constants: 5 ‘a’ 5.0 ... – punctuation: { } ; < > # /* */
Identifier • Names used for objects in C • Rules for identifiers in C: – first char alphabetic [a-z,A-Z] or underscore (_) – has only alphabetic, digit, underscore chars – first 31 characters are significant – cannot duplicate a reserved word – case (upper/lower) matters
Reserved Words • Identifiers that already have meaning in C • Examples: – include, main, printf, scanf, if, else, … – more as we cover C language
Valid/Invalid Identifiers Valid sum c4_5 A_NUMBER longnamewithmanychars TRUE _split_name Invalid 7of9 x-name name with spaces 1234a int AXYZ&
Program Execution • Global declarations set up • Function main executed – local declarations set up – each statement in statement section executed • executed in order (first to last) • changes made by one statement affect later statements
Variables • Named memory location • Variables declared in global or local declaration sections • Syntax: Type Name; • Examples: int sum; float avg; char dummy;
Variable Type • Indicates how much memory to set aside for the variable • Also determines how that space will be interpreted • Basic types: char, int, float – specify amount of space (bytes) to set aside – what can be stored in that space – what operations can be performed on those vars
Variable Name • Legal identifier • Not a reserved word • Must be unique: – not used before – variable names in functions (local declarations) considered to be qualified by function name – variable x in function main is different from x in function f1
Multiple Variable Declarations • Can create multiple variables of the same type in one statement: int x, y, z; is a shorthand for int x; int y; int z; - stylistically, the latter is often preferable
Variable Initialization • Giving a variable an initial value • Variables not necessarily initialized when declared (value is unpredictable - garbage) • Can initialize in declaration: • Syntax: Type Name = Value; • Example: int x = 0;
Initialization Values • Literal constant (token representing a value, like 5 representing the integer 5) • An expression (operation that calculates a value) • Function call • The value, however specified, must be of the correct type
Multiple Declaration Initialization • Can provide one value for variables initialized in one statement: int x, y, z = 0; • Each variable declared and then initialized with the value
Type • Set of possible values – defines size, how values stored, interpreted • Operations that can be performed on those possible values • Data types are associated with objects in C (variables, functions, etc.)
Standard Types • Atomic types (cannot be broken down) – void – char – int – float, double • Derived types – composed of other types
Literal Constants • Sequences of characters (tokens) that correspond to values from that type -35 is the integer -35 3.14159 is the floating pointer number 3.14159 ‘A’ is the character A • Can be used to initialize variables
Void Type • Type name: void • Possible values: none • Operations: none • Useful as a placeholder
Integer Type • Type name: – int – short int – long int • Possible values: whole numbers (within given ranges) as in 5, -35, 401 • Operations: arithmetic (addition, subtraction, multiplication, …), and others
Integer Types/Values Type Bytes Bits Min Val Max Val short int 2 16 -32768 32767 int 4 32 -2147483648 2147483647 long int 4 32 -2147483648 2147483647
Why Limited? • With a fixed number of bits, only a certain number of possible patterns • 16 bits, 65,536 possible patterns – 32768 negative numbers – 1 zero – 32767 positive numbers • Overflow: attempt to store a value to large in a variable (40000 in short int)
Two’s Complement Integers: positive number: 0, number in binary 97 in binary 1*64 + 1*32 + 1*1 (1100001) pad with leading zeroes (0 00000001100001) - 16 bits zero: 0, all zeroes negative number: 1, (inverse of number + 1) -97 (1, 111111110011110 + 1) 1 111111110011111
Unsigned Integers • Type: unsigned int • No negative values • unsigned int: – possible values: 0 to 65536 • Representation: binary number
Integer Literal Constants Syntax: 1 or more digits Optional leading sign (+ or -) Optional l or L at the end for long Optional u or U for unsigned Examples: 5, -35, 401, 4010L, -350L, 2000UL
Floating-Point Type • Type names: – float – double – long double • Possible values: floating point numbers, 5.0 - 3.5, 4.01 • Operations: arithmetic (addition, subtraction, multiplication, …), and others
Floating-Point Representation • float: 4 bytes, 32 bits • double: 8 bytes, 64 bits • long double: 10 bytes, 80 bits • Representation: – magnitude (some number of bits) plus exponent (remainder of bits) – 3.26 * 10^4 for 32600.0
Floating-Point Limitations • Maximum, minimum exponents – maximum possible value (largest positive magnitude, largest positive exponent) – minimum value (largest negative magnitude, largest positive exponent) – can have overflow, and underflow • Magnitude limited – cannot differentiate between values such as 1.00000000 and 1.00000001
Floating-Point Literals • Syntax: – Zero or more digits, decimal point, then zero or more digits (at least one digit) – Whole numbers also treated as float – Optional sign at start – Can be followed by e and whole number (to represent exponent) – f or F at end for float – l or L at end for long double • Examples: 5, .5, 0.5, -1.0, 2.1e+3, 5.1f
Character Type • Type name: char • Possible values: keys that can be typed at the keyboard • Representation: each character assigned a value (ASCII values), 8 bits – A - binary number 65 – a - binary number 97 – b - binary number 98 – 2 - binary number 50
Character Literals • Single key stroke between quote char ‘ • Examples: ‘A’, ‘a’, ‘b’, ‘1’, ‘@’ • Some special chars: – ‘0’ - null char – ‘t’ - tab char – ‘n’ - newline char – ‘’’ - single quote char – ‘’ - backslash char
String Literals • No string type (more later) • Contained between double quote chars (“) • Examples: “” - null string “A string” “String with newline n char in it” “String with a double quote ” in it”
Constants • Literal constants - tokens representing values from type • Defined constants – syntax: #define Name Value – preprocessor command, Name replaced by Value in program – example: #define MAX_NUMBER 100
Constants (cont) • Memory constants – declared similar to variables, type and name – const added before declaration – Example: const float PI = 3.14159; – Can be used as a variable, but one that cannot be changed – Since the value cannot be changed, it must be initialized
Formatted Input/Output • Input comes from files • Output sent to files • Other objects treated like files: – keyboard - standard input file (stdin) – monitor - standard output file (stdout) • Generally send/retrieve characters to/from files
Formatted Output • Command: printf - print formatted • Syntax: printf(Format String, Data List); – Format string any legal string – Characters sent (in order) to screen • Ex.: printf(“Welcome tonCS 1621!n”); causes Welcome to CS 1621! to appear on monitor
Formatted Output (cont) • Successive printf commands cause output to be added to previous output • Ex. printf(“Hi, how “); printf(“is it goingnin 1621?”); prints Hi, how is it going in 1621? To the monitor
Field Specifications • Format string may contain one or more field specifications – Syntax: %[Flag][Width][Prec][Size]Code – Codes: • c - data printed as character • d - data printed as integer • f - data printed as floating-point value – For each field specification, have one data value after format string, separated by commas
Field Specification Example printf(“%c %d %fn”,’A’,35,4.5); produces A 35 4.50000 (varies on different computers) Can have variables in place of literal constants (value of variable printed)
Width and Precision • When printing numbers, generally use width/precision to determine format – Width: how many character spaces to use in printing the field (minimum, if more needed, more used) – Precision: for floating point numbers, how many characters appear after the decimal point, width counts decimal point, number of digits after decimal, remainder before decimal
Width/Precision Example printf(“%5d%8.3fn”,753,4.1678); produces 753 4.168 values are right justified If not enough characters in width, minimum number used use 1 width to indicate minimum number of chars should be used
Left Justification (Flags) Put - after % to indicate value is left justified printf(“%-5d%-8.3fXn”,753,4.1678); produces 753 4.168 X For integers, put 0 after % to indicate should pad with 0’s printf(“%05d”,753); produces 00753
Size Indicator • Use hd for small integers • Use ld for long integers • Use Lf for long double • Determines how value is treated
Printf Notes • Important to have one value for each field specification – some C versions allow you to give too few values (garbage values are formatted and printed) • Values converted to proper type – printf(“%c”,97); produces the character a on the screen
Formatted Input • Command: scanf - scan formatted • Syntax: scanf(Format String, Address List); – Format string a string with one or more field specifications – Characters read from keyboard, stored in variables • scanf(“%c %d %f”,&cVar,&dVar,&fVar); attempts to read first a single character, then a whole number, then a floating point number from the keyboard
Formatted Input (cont) • Generally only have field specifications and spaces in string – any other character must be matched exactly (user must type that char or chars) – space characters indicate white-space is ignored – “white-space” - spaces, tabs, newlines – %d and %f generally ignore leading white space anyway (looking for numbers) – %d and %f read until next non-number char reached
Formatted Input (cont) • More notes – can use width in field specifications to indicate max number of characters to read for number – computer will not read input until return typed – if not enough input on this line, next line read, (and line after, etc.) – inappropriate chars result in run-time errors (x when number expected) – if end-of-file occurs while variable being read, an error occurs
Address Operator • & - address operator • Put before a variable (as in &x) • Tells the computer to store the value read at the location of the variable • More on address operators later
Scanf Rules • Conversion process continues until – end of file reached – maximum number of characters processed – non-number char found number processed – an error is detected (inappropriate char) • Field specification for each variable • Variable address for each field spec. • Any character other than whitespace must be matched exactly
Scanf Example scanf(“%d%c %f”,&x,&c,&y); and following typed: -543A 4.056 56 -543 stored in x, A stored in c, 4.056 stored in y, space and 56 still waiting (for next scanf)
Prompting for Input • Using output statements to inform the user what information is needed: printf(“Enter an integer: “); scanf(“%d”,&intToRead); • Output statement provides a cue to the user: Enter an integer: user types here

THE BASIC CONCEPTS OF C PROGRAMMINGPPT.PPT

  • 1.
    Introduction to C #include<stdio.h> int main () { printf(“Welcome to CS 1621!n”); }
  • 2.
    Outline II. Program Basics A.Program skeleton preprocessor directives global declarations functions local declarations statements B. Comments and Documentation C. Names (identifiers) reserved words
  • 3.
    Outline (cont) II. ProgramBasics (cont) D. Variable declarations 1. Memory allocation 2. Atomic types void, int, float, char E. Constants 1. literal 2. defined 3. memory
  • 4.
    Outline (cont) II. ProgramBasics (cont) F. Formatted input/output 1. Files 2. Printf (monitor output) a. format strings field specifications b. data list 3. Scanf (keyboard input) a. format strings b. address list 4. Prompting for Input
  • 5.
    History of C 1960:ALGOL (ALGOrithmic Language) 1967: BCPL (Basic Combined Programming Language) 1970: B programming language (typeless) 1972: C: BCPL plus B with types 1978: Kernighan + Ritchie standard for C 1989: ANSI standard for C
  • 6.
    C Program Structure •Program defined by: – global declarations – function definitions • May contain preprocessor directives • Always has one function named main, may contain others Preprocessor Directives Global Declarations Function Definitions int main () { } Local Declarations Statements
  • 7.
    Parts of aProgram #include <stdio.h> int x; int main () { int y; printf("Enter x and y: "); scanf(&x,&y); printf("Sum is %dn",x+y); } Preprocessor Directive Global Declaration Function Local Declaration Statements
  • 8.
    Preprocessor Directives • Beginwith # • Instruct compiler to perform some transformation to file before compiling • Example: #include <stdio.h> – add the header file stdio.h to this file – .h for header file – stdio.h defines useful input/output functions
  • 9.
    Declarations • Global – visiblethroughout program – describes data used throughout program • Local – visible within function – describes data used only in function
  • 10.
    Functions • Consists ofheader and body – header: int main () – body: contained between { and } • starts with location declarations • followed by series of statements • More than one function may be defined • Functions are called (invoked) - more later
  • 11.
    Main Function • Everyprogram has one function main • Header for main: int main () • Program is the sequence of statements between the { } following main • Statements are executed one at a time from the one immediately following to main to the one before the }
  • 12.
    Comments • Text between/* and */ • Used to “document” the code for the human reader • Ignored by compiler (not part of program) • Have to be careful – comments may cover multiple lines – ends as soon as */ encountered (so no internal comments - /* An /* internal */ comment */)
  • 13.
    Comment Example #include <stdio.h> /*This comment covers * multiple lines * in the program. */ int main () /* The main header */ { /* No local declarations */ printf(“Too many commentsn”); } /* end of main */
  • 14.
    Documentation • Global -start of program, outlines overall solution, may include structure chart • Module - when using separate files, indication of what each file solves • Function - inputs, return values, and logic used in defining function • Add documentation for key (tough to understand) comments • Names of variables - should be chosen to be meaningful, make program readable
  • 15.
    Syntax of C •Rules that define C language – Specify which tokens are valid – Also indicate the expected order of tokens • Some types of tokens: – reserved words: include printf int ... – identifiers: x y ... – literal constants: 5 ‘a’ 5.0 ... – punctuation: { } ; < > # /* */
  • 16.
    Identifier • Names usedfor objects in C • Rules for identifiers in C: – first char alphabetic [a-z,A-Z] or underscore (_) – has only alphabetic, digit, underscore chars – first 31 characters are significant – cannot duplicate a reserved word – case (upper/lower) matters
  • 17.
    Reserved Words • Identifiersthat already have meaning in C • Examples: – include, main, printf, scanf, if, else, … – more as we cover C language
  • 18.
  • 19.
    Program Execution • Globaldeclarations set up • Function main executed – local declarations set up – each statement in statement section executed • executed in order (first to last) • changes made by one statement affect later statements
  • 20.
    Variables • Named memorylocation • Variables declared in global or local declaration sections • Syntax: Type Name; • Examples: int sum; float avg; char dummy;
  • 21.
    Variable Type • Indicateshow much memory to set aside for the variable • Also determines how that space will be interpreted • Basic types: char, int, float – specify amount of space (bytes) to set aside – what can be stored in that space – what operations can be performed on those vars
  • 22.
    Variable Name • Legalidentifier • Not a reserved word • Must be unique: – not used before – variable names in functions (local declarations) considered to be qualified by function name – variable x in function main is different from x in function f1
  • 23.
    Multiple Variable Declarations •Can create multiple variables of the same type in one statement: int x, y, z; is a shorthand for int x; int y; int z; - stylistically, the latter is often preferable
  • 24.
    Variable Initialization • Givinga variable an initial value • Variables not necessarily initialized when declared (value is unpredictable - garbage) • Can initialize in declaration: • Syntax: Type Name = Value; • Example: int x = 0;
  • 25.
    Initialization Values • Literalconstant (token representing a value, like 5 representing the integer 5) • An expression (operation that calculates a value) • Function call • The value, however specified, must be of the correct type
  • 26.
    Multiple Declaration Initialization •Can provide one value for variables initialized in one statement: int x, y, z = 0; • Each variable declared and then initialized with the value
  • 27.
    Type • Set ofpossible values – defines size, how values stored, interpreted • Operations that can be performed on those possible values • Data types are associated with objects in C (variables, functions, etc.)
  • 28.
    Standard Types • Atomictypes (cannot be broken down) – void – char – int – float, double • Derived types – composed of other types
  • 29.
    Literal Constants • Sequencesof characters (tokens) that correspond to values from that type -35 is the integer -35 3.14159 is the floating pointer number 3.14159 ‘A’ is the character A • Can be used to initialize variables
  • 30.
    Void Type • Typename: void • Possible values: none • Operations: none • Useful as a placeholder
  • 31.
    Integer Type • Typename: – int – short int – long int • Possible values: whole numbers (within given ranges) as in 5, -35, 401 • Operations: arithmetic (addition, subtraction, multiplication, …), and others
  • 32.
    Integer Types/Values Type BytesBits Min Val Max Val short int 2 16 -32768 32767 int 4 32 -2147483648 2147483647 long int 4 32 -2147483648 2147483647
  • 33.
    Why Limited? • Witha fixed number of bits, only a certain number of possible patterns • 16 bits, 65,536 possible patterns – 32768 negative numbers – 1 zero – 32767 positive numbers • Overflow: attempt to store a value to large in a variable (40000 in short int)
  • 34.
    Two’s Complement Integers: positive number:0, number in binary 97 in binary 1*64 + 1*32 + 1*1 (1100001) pad with leading zeroes (0 00000001100001) - 16 bits zero: 0, all zeroes negative number: 1, (inverse of number + 1) -97 (1, 111111110011110 + 1) 1 111111110011111
  • 35.
    Unsigned Integers • Type:unsigned int • No negative values • unsigned int: – possible values: 0 to 65536 • Representation: binary number
  • 36.
    Integer Literal Constants Syntax: 1or more digits Optional leading sign (+ or -) Optional l or L at the end for long Optional u or U for unsigned Examples: 5, -35, 401, 4010L, -350L, 2000UL
  • 37.
    Floating-Point Type • Typenames: – float – double – long double • Possible values: floating point numbers, 5.0 - 3.5, 4.01 • Operations: arithmetic (addition, subtraction, multiplication, …), and others
  • 38.
    Floating-Point Representation • float:4 bytes, 32 bits • double: 8 bytes, 64 bits • long double: 10 bytes, 80 bits • Representation: – magnitude (some number of bits) plus exponent (remainder of bits) – 3.26 * 10^4 for 32600.0
  • 39.
    Floating-Point Limitations • Maximum,minimum exponents – maximum possible value (largest positive magnitude, largest positive exponent) – minimum value (largest negative magnitude, largest positive exponent) – can have overflow, and underflow • Magnitude limited – cannot differentiate between values such as 1.00000000 and 1.00000001
  • 40.
    Floating-Point Literals • Syntax: –Zero or more digits, decimal point, then zero or more digits (at least one digit) – Whole numbers also treated as float – Optional sign at start – Can be followed by e and whole number (to represent exponent) – f or F at end for float – l or L at end for long double • Examples: 5, .5, 0.5, -1.0, 2.1e+3, 5.1f
  • 41.
    Character Type • Typename: char • Possible values: keys that can be typed at the keyboard • Representation: each character assigned a value (ASCII values), 8 bits – A - binary number 65 – a - binary number 97 – b - binary number 98 – 2 - binary number 50
  • 42.
    Character Literals • Singlekey stroke between quote char ‘ • Examples: ‘A’, ‘a’, ‘b’, ‘1’, ‘@’ • Some special chars: – ‘0’ - null char – ‘t’ - tab char – ‘n’ - newline char – ‘’’ - single quote char – ‘’ - backslash char
  • 43.
    String Literals • Nostring type (more later) • Contained between double quote chars (“) • Examples: “” - null string “A string” “String with newline n char in it” “String with a double quote ” in it”
  • 44.
    Constants • Literal constants- tokens representing values from type • Defined constants – syntax: #define Name Value – preprocessor command, Name replaced by Value in program – example: #define MAX_NUMBER 100
  • 45.
    Constants (cont) • Memoryconstants – declared similar to variables, type and name – const added before declaration – Example: const float PI = 3.14159; – Can be used as a variable, but one that cannot be changed – Since the value cannot be changed, it must be initialized
  • 46.
    Formatted Input/Output • Inputcomes from files • Output sent to files • Other objects treated like files: – keyboard - standard input file (stdin) – monitor - standard output file (stdout) • Generally send/retrieve characters to/from files
  • 47.
    Formatted Output • Command:printf - print formatted • Syntax: printf(Format String, Data List); – Format string any legal string – Characters sent (in order) to screen • Ex.: printf(“Welcome tonCS 1621!n”); causes Welcome to CS 1621! to appear on monitor
  • 48.
    Formatted Output (cont) •Successive printf commands cause output to be added to previous output • Ex. printf(“Hi, how “); printf(“is it goingnin 1621?”); prints Hi, how is it going in 1621? To the monitor
  • 49.
    Field Specifications • Formatstring may contain one or more field specifications – Syntax: %[Flag][Width][Prec][Size]Code – Codes: • c - data printed as character • d - data printed as integer • f - data printed as floating-point value – For each field specification, have one data value after format string, separated by commas
  • 50.
    Field Specification Example printf(“%c%d %fn”,’A’,35,4.5); produces A 35 4.50000 (varies on different computers) Can have variables in place of literal constants (value of variable printed)
  • 51.
    Width and Precision •When printing numbers, generally use width/precision to determine format – Width: how many character spaces to use in printing the field (minimum, if more needed, more used) – Precision: for floating point numbers, how many characters appear after the decimal point, width counts decimal point, number of digits after decimal, remainder before decimal
  • 52.
    Width/Precision Example printf(“%5d%8.3fn”,753,4.1678); produces 753 4.168 valuesare right justified If not enough characters in width, minimum number used use 1 width to indicate minimum number of chars should be used
  • 53.
    Left Justification (Flags) Put- after % to indicate value is left justified printf(“%-5d%-8.3fXn”,753,4.1678); produces 753 4.168 X For integers, put 0 after % to indicate should pad with 0’s printf(“%05d”,753); produces 00753
  • 54.
    Size Indicator • Usehd for small integers • Use ld for long integers • Use Lf for long double • Determines how value is treated
  • 55.
    Printf Notes • Importantto have one value for each field specification – some C versions allow you to give too few values (garbage values are formatted and printed) • Values converted to proper type – printf(“%c”,97); produces the character a on the screen
  • 56.
    Formatted Input • Command:scanf - scan formatted • Syntax: scanf(Format String, Address List); – Format string a string with one or more field specifications – Characters read from keyboard, stored in variables • scanf(“%c %d %f”,&cVar,&dVar,&fVar); attempts to read first a single character, then a whole number, then a floating point number from the keyboard
  • 57.
    Formatted Input (cont) •Generally only have field specifications and spaces in string – any other character must be matched exactly (user must type that char or chars) – space characters indicate white-space is ignored – “white-space” - spaces, tabs, newlines – %d and %f generally ignore leading white space anyway (looking for numbers) – %d and %f read until next non-number char reached
  • 58.
    Formatted Input (cont) •More notes – can use width in field specifications to indicate max number of characters to read for number – computer will not read input until return typed – if not enough input on this line, next line read, (and line after, etc.) – inappropriate chars result in run-time errors (x when number expected) – if end-of-file occurs while variable being read, an error occurs
  • 59.
    Address Operator • &- address operator • Put before a variable (as in &x) • Tells the computer to store the value read at the location of the variable • More on address operators later
  • 60.
    Scanf Rules • Conversionprocess continues until – end of file reached – maximum number of characters processed – non-number char found number processed – an error is detected (inappropriate char) • Field specification for each variable • Variable address for each field spec. • Any character other than whitespace must be matched exactly
  • 61.
    Scanf Example scanf(“%d%c %f”,&x,&c,&y); andfollowing typed: -543A 4.056 56 -543 stored in x, A stored in c, 4.056 stored in y, space and 56 still waiting (for next scanf)
  • 62.
    Prompting for Input •Using output statements to inform the user what information is needed: printf(“Enter an integer: “); scanf(“%d”,&intToRead); • Output statement provides a cue to the user: Enter an integer: user types here