Input/Output in C++
- C++ iostream.h instead of stdio.h
- Why change?
- Input/output routines in iostream can be extended to new types declared by the user
- The routines are in some senses easier to use
- Some aspects of the routines can be set without having to repeat them (e.g., setting the desired precision for printing floating point values)
- Readings:
2.1-2.11, 15.1-15.4, 17.1-17.8, 16.1-16.7, 18.1-18.6
Outline
Simple input/output (iostream.h)
cout, cin, cerr
output
insertion operator (<<) and chaining
int, float, string
input
extraction operator (>>) and chaining
int string
Advanced input/output
object flags (setf, unsetf)
input status bits
manipulators (iomanip.h)
file input/output (fstream.h)
Using iostream.h
- Include iostream.h instead of stdio.h
- Standard iostream objects:
cout - object providing a connection to the monitor
cin - object providing a connection to the keyboard
cerr - object providing a connection to error streem
- To perform input and output we send messages to one of these objects (or one that is connected to a file)
The Insertion Operator (<<)
- To send output to the screen we use the insertion operator on the object cout
- Format: cout << Expression;
- The compiler figures out the type of the object and prints it out appropriately
cout << 5; // Outputs 5
cout << 4.1; // Outputs 4.1
cout << “String”; // Outputs String
cout << ‘\n’; // Outputs a newline
The Extraction Operator (>>)
- To get input from the keyboard we use the extraction operator and the object cin
- No need for & in front of variable
- The compiler figures out the type of the variable and reads in the appropriate type
int X;
float Y;
cin >> X; // Reads in an integer
cin >> Y; // Reads in a float
Chaining Calls
- Multiple uses of the insertion and extraction operator can be chained together:
cout << E1 << E2 << E3 << … ;
cin >> V1 >> V2 >> V3 >> …;
- Equivalent to performing the set of insertion or extraction operators one at a time
- Example
cout << “Total sales are $” << sales << ‘\n’;
cin >> Sales1 >> Sales2 >> Sales3;
Setting Format Flags
- The object cout has flags that determine how objects are printed, to change how things are printed we access and change these flags
- To set a flag(s) we use the setf function which is associated with objects such as cout and cin
- To call setf we say
cout.setf(flags)
- the setf function is a field of the object cout
- Q: But what flags? A: C++ predefines them
Setting Format Flags (cont)
- But in order to be able to set flags we often have to unset other flags first, to do so we use the unsetf function:
- C++ also provides a short-hand to combine both operations:
cout.setf(OnFlags,OffFlags)
- First turns off the flags OffFlags
- Then turns on the flags OnFlags
Integer Base and Format Flags
Choosing the base to print out an integer in:
Flags to use:
ios::dec - show ints as decimal (the default)
ios::oct - show ints as octal
ios::hex - show ints as hexadecimal
Should only have one on at a time
To change, turn the others off and set one on
cout.unsetf(ios::dec);
cout.unsetf(ios::oct);
cout.unsetf(ios::hex);
cout.setf(ios::oct);
Integer Base and Format Flags (cont)
One can combine flags using | operator
cout.unsetf(ios::dec | ios::oct | ios::hex);
cout.setf(ios::oct);
or
cout.setf(ios::oct,ios::dec | ios::oct | ios::hex);
C++ also includes a shorthand for the second (combination) flag: ios::basefield:
cout.setf(ios::oct,ios::basefield);
Turns all of the base flags off and the octal flag on
Integer Base Example
cout.setf(ios::oct,ios::basefield);
cout << x << '\n'; // Outputs 52\n
cout.setf(ios::hex,ios::basefield);
cout << x << '\n'; // Outputs 2a\n
cout.setf(ios::dec,ios::basefield);
cout << x << '\n'; // Outputs 42\n
Showing the Base
The flag ios::showbase can be set (its default is off), it results in integers being printed in a way that demonstrates their base
- decimal - no change
- octal - leading 0
- hexadecimal - leading 0x
int x = 42;
cout.setf(ios::showbase);
cout.setf(ios::oct,ios::basefield);
cout << x << '\n'; // Outputs 052\n
cout.setf(ios::hex,ios::basefield);
cout << x << '\n'; // Outputs 0x2a\n
cout.setf(ios::dec,ios::basefield);
cout << x << '\n'; // Outputs 42\n
Showing the Plus Sign
The flag ios::showpos can be set (its default is off) to print a + sign when a positive integer or floating point value is printed
int x = 42;
int y = 3.1415;
cout.setf(ios::showpos);
cout << x << '\n'; // Outputs +42\n
cout << y << ‘\n’; // Outputs +3.1415\n
Showing Upper Case Hex Ints
The flag ios::uppercase (default off) can be used to indicate that the letters making up hexadecimal numbers should be shown as upper case:
int x = 42;
cout.setf(ios::uppercase);
cout.setf(ios::hex,ios::basefield);
cout << x << '\n'; // Outputs 2A\n
Setting the Width
- You can use the width(int) function to set the width for printing a value, but it only works for the next insertion command (more on this later):
int x = 42;
cout.width(5);
cout << x << ‘\n’; // Outputs 42
cout << x << ‘\n’; // Outputs 42
Setting the Fill Character
Use the fill(char) function to set the fill character. The character remains as the fill character until set again.
int x = 42;
cout.width(5);
cout.fill(‘*’);
cout << x << ‘\n’; // Outputs ***42
Justification
Set justification using flags ios::left, ios::right, and ios::internal (after sign or base) - only one
Use ios::adjustfield to turn all three flags off
int x = 42;
cout.setf(ios::showpos);
cout.fill('*');
cout.setf(ios::right,ios::adjustfield);
cout.width(6);
cout << x << '\n'; // Outputs ***+42
cout.setf(ios::left,ios::adjustfield);
cout.width(6);
cout << x << '\n'; // Outputs +42***
cout.setf(ios::internal,ios::adjustfield);
cout.width(6);
cout << x << '\n'; // Outputs +***42
Decimal Points in Floats
Set flag ios::showpoint to make sure decimal point appears in output (C++ only shows significant digits in default)
float y = 3.0;
cout << y << ‘\n’; // Outputs 3
cout.setf(ios::showpoint);
cout << y << ‘\n’; // Outputs 3.00000
Format of Float
Floating point values are printed out in fixed or scientific notation based on how they are stored/initialized:
cout << 2.3; // Outputs 2.3
cout << 5.67e8; // Outputs 5.67e+08
cout << 0.0; // Outputs 0
Significant Digits in Float
Use function precision(int) to set the number of significant digits printed (may convert from fixed to scientific to print):
float y = 23.1415;
cout.precision(1);
cout << y << '\n'; // Outputs 2e+01
cout.precision(2);
cout << y << '\n'; // Outputs 23
cout.precision(3);
cout << y << '\n'; // Outputs 23.1
Floating Point Format
- Can use flags ios::scientific and ios::fixed to force floating point output in scientific or fixed format
- Only one flag at a time, ios::floatfield to turn off
cout.setf(ios::scientific,ios::floatfield);
cout << 123.45 << ‘\n’; // Outputs 1.2345e+02
cout.setf(ios::fixed,ios::floatfield);
cout << 5.67E1 << ‘\n’; // Outputs 56.7
- Effect of precision depends on format
- scientific (total significant digits)
- fixed (how many digits after decimal point)
Displaying bools
- Variables of type bool print out as 0 (false) or 1 (true)
- To print out words (false, true) use flag ios::boolalpha
bool b = true;
cout.setf(ios::boolalpha);
cout << b << ‘\n’; // Outputs true
cout << (!b) << ‘\n’; // Outputs false
Manipulators
- Isn’t that all kind of involved??
- Plus, what’s that with width only counting for one arg?
- A solution - manipulators
- A manipulator is a simple function that can be included in an insertion or extraction chain
- C++ manipulators
- must include iomanip.h to use
- several are provided to do useful things
- you can also create your own (see 17.3, 17.5, 17.6, 17.8)
Output Manipulators (no args)
Manipulators included like arguments in extraction
endl - outputs a new line character, flushes output
dec - sets int output to decimal
hex - sets int output to hexadecimal
oct - sets int output to octal
Example:
#include <iostream.h>
#include <iomanip.h>
int x = 42;
cout << oct << x << endl; // Outputs 52\n
cout << hex << x << endl; // Outputs 2a\n
cout << dec << x << endl; // Outputs 42\n
Output Manipulators (1 arg)
Manipulators taking 1 argument
setw(int) - sets the width to int value
setfill(char) - sets fill char to char value
setprecision(int) - sets precision to int value
setbase(int) - sets int output to hex if int is 16, oct if int is 8, dec if int is 0 or 10
setiosflags(flags) - set flags on
resetiosflags(flags) - sets flags off
cout << resetiosflags(ios::floatfield) << setiosflags(ios::fixed | ios::showpoint) << setw(7) << setprecision(2) << setfill(‘_’) << 34.267 << endl; // outputs __34.27
Input Status Flags
- When performing input, certain problems may occur, we can determine if an error has occurred by checking these flags:
eof() - end-of-file occurred during input
fail() - input operation failed
good() - no flags set (not eof or any of fail flags)
- Flags stay set and all input fails until clear() function called
Testing Status Flags
cout << “Total is “ << total << endl;
Testing Status Flags
Extraction is an operator, returns cin object (can check eof() or other flags after operation):
while (!(cin >> x).eof())
cout << “Total is “ << total << endl;
Integer Input
- If none of the flags hex, dec, oct set then we can indicate how an int is formatted with value typed:
42 - decimal 42
052 - octal 52
0x2a - hexadecimal 2a
- If any of these flags set, all input will be treated as being of only that type
- note, in explicit decimal format, 052 read as 52, 0x2a read as 0
Character Input
- The extraction operator when applied to a character ignores whitespace
- To read any character use the get(char) function, can also provide no argument (works like getchar)
char ch;
cin >> ch; // Reads next non-whitespace char
cin.get(ch); // Reads next character (any)
while (cin.get() != ‘\n’); // Reads to newline
String Input
- Can use arguments of string type like any other variable
- like scanf with %s reads as many chars as typed (may be too many)
- can control by using width(int) function or setw(int) manipulator
- ignores leading whitespace
- stops at first whitespace
- Example
char string[101];
cint >> setw(100) >> string;
String Input with Whitespace
Use function get(stringloc,size,delimitchar)
- reads string into array at stringloc taking in at most size chars and stopping at delimitchar (default is ‘\n’ -- you can leave delimitchar off)
- stops before delimitchar
Use function getline to also read newline character
Example:
char theline[101];
cin.get(theline,100,’\n’); // or
cin.get(theline,100);
File Input/Output
- Done with the same operations (insertion, extraction) as keyboard input and monitor output
- Simply open input or output object with connection to a file and use it where you would use cin or cout
- To use
- include <fstream.h>
- create input object of type ifstream
- or output object of type ofstream
Opening Files
- Use open function or include file name when declaring variable:
ifstream inobj1;
inobj1.open(“in1.dat”)
ifstream inobj2(“in2.dat”);
- To check if file successfully opened check object in condition:
if (!inobj1)
cout << “Unable to open file in1.dat” << endl;
Opening Output Files
Standard open routine opens as in “w” in fopen in C
- existing file deleted if it already exists
Other open routines -- add second argument and combination of certain flags:
ios::out - open as output connection (must include)
ios::append - append to existing file
ios::nocreate - file must exist, error otherwise
ios::noreplace - file must not exist, error otherwise
Example
ofstream out(“outf”,ios::out | ios::append);
// out is an append connection to outf
Closing a File
Use close() on object to close connection to file:
ifstream in(“in.dat”);
…
in.close();
File Example
cout << ”File to copy from: ";
cout << "Unable to open " << infname << endl;
File Example (cont)
cout << "File to copy to: ";
ofstream out(outfname,ios::out | ios::noreplace);
cout << "Unable to open " << outfname << " -- already exists!" << endl;