Quick Reference

Book Cover

Introduction

This quick reference is intended for readers of book, The C++ Standard Library from Scratch. It is not a full reference to the standard library. Descriptions, where provided are terse and are not intended to replace the more complete descriptions found in the book. Some elements are listed using actual class or function declarations; other elements are listed using a shorthand to make things simpler and/or more concise. Some elements of the standard library were omitted entirely because they are not included in the book. Conversely, many of the topics covered in the book are covered in even more detail here.

Numerous concepts were simplified in this quick reference. For example, in the Standard, string is actually a typedef for basic_string<char, char_traits<char>, allocator<char> >. If you look up a function in the standard and find that it has different arguments or a different return type than I've specified here, it's probably because I wanted to make it simpler. Also, there are defects in the standard. For situations I was aware of, I tried to present the likely resolution to the defect, rather than the original defective definition.

Important Note: Except where otherwise noted, all of the components in this quick reference are in namespace std.


Table of Contents


Strings

String Argument Conventions

The member functions in this section take arguments with the following names:

The single argument pos or pos1 signifies a position within the this string.

The two arguments, pos and len, signify a substring of the this string starting at pos with length len. If len is npos or is more than size(), then the substring goes from pos to the end of the string.

The arguments, pos1 and len1 have the same meaning as pos and len and are used in functions where there is a second string involved.

The arguments, s2, pos2 and len2 signify a substring of s2 starting at pos2 with length len2.

The argument, nts, signifies a Null-Terminated String (a C-style string).

The arguments, buf and buflen signify a character buffer, buf, of length buflen. (A null character, is treated like any other character).

The arguments, repetitions and c signify repetitions copies of the signal character, c.

The two arguments, start and finish signify an iterator range, [start, finish), comprising a substring of the this string.

The string Class

#include <string>
class string;

String Nested Types and Constants

The following types are nested within the string class:

Types

size_type
difference_type
iterator
const_iterator
reverse_iterator
const_reverse_iterator

Constants

const size_type npos;   // Maximum size_type

String Member Functions

Constructors

string();
string(const string& s2); // Copy constructor
string(const string& s2, size_type pos2, size_type len2);
string(const char* nts);        // convert from C-string
string(const char* buf, size_type bufsize);
string(size_type repetitions, char c);

Assignment Operators and Functions

string& operator=(const string& s2); // normal assignment
string& operator=(const char* nts); // convert from C-string
string& operator=(char c); // assign a single character
string& assign(const string& s2);
string& assign(const string& s2, size_type pos2, size_type len2);
string& assign(const char* nts);
string& assign(const char* buf, size_type buflen);
string& assign(size_type repetitions, char c);
void swap(string& s2);

Iterator Functions

iterator               begin();
iterator               end();
const_iterator         begin() const;
const_iterator         end() const;
reverse_iterator       rbegin();
reverse_iterator       rend();
const_reverse_iterator rbegin() const;
const_reverse_iterator rend()   const;

Size and Capacity

size_type size() const;
size_type length() const; // same as size()
size_type max_size() const;
void resize(size_type size, char c = '\0');
void clear();
bool empty() const;
size_type capacity() const;
void reserve(size_type capacity = 0);

Element Access

char&       operator[](size_type pos);
const char& operator[](size_type pos) const;
char&       at(size_type pos);
const char& at(size_type pos) const;

Append Functions and Operators

string& append(const string& s2);
string& append(const string& s2, size_type pos2, size_type len2);
string& append(const char* nts);
string& append(const char* buf, size_type buflen);
string& append(size_type repetitions, char c);
string& operator+=(const string& s2);
string& operator+=(const char* nts);
string& operator+=(char c);

Insert Functions

string& insert(size_type pos1, const string& s2);
string& insert(size_type pos1, const string& s2,
               size_type pos2, size_type len2);
string& insert(size_type pos1, const char* nts);
string& insert(size_type pos1, const char* buf, size_type buflen);
string& insert(size_type pos1, size_type repetitions, char c);
iterator insert(iterator pos1, char c);
void     insert(iterator pos1, size_type repetitions, char c);

Erase Functions

string& erase(size_type pos = 0, size_type len = npos);
iterator& erase(iterator pos);
iterator& erase(iterator start, iterator finish);

Replace Functions

string& replace(size_type pos1, size_type len1, const string& s2);
string& replace(size_type pos1, size_type len1, const string& s2,
                size_type pos2, size_type len2);
string& replace(size_type pos1, size_type len1, const char* nts);
string& replace(size_type pos1, size_type len1,
                const char* buf, size_type buflen);
string& replace(size_type pos1, size_type len1,
                size_type repetitions, char c);
string& replace(iterator start, iterator finish, const string& s2);
string& replace(iterator start, iterator finish, const char* nts);
string& replace(iterator start, iterator finish,
                const char* buf, size_type buflen);
string& replace(iterator start, iterator finish,
                size_type repetitions, char c);

Comparison Functions

// compare function returns -1 if *this < s2, 0 if *this == s2
// and +1 if *this > s2.
int compare(const string& s2) const;
int compare(size_type pos1, size_type len1, const string& s2);
int compare(size_type pos1, size_type len1, const string& s2,
            size_type pos2, size_type len2);
int compare(const char* nts);
int compare(size_type pos1, size_type len1, const char* nts);
int compare(size_type pos1, size_type len1,
            const char* buf, size_type buflen);

Substrings

string substr(pos = 0, len = npos) const;
// This constructor is repeated here because it creates substring
string(const string& s2, size_type pos2, size_type len2);

Conversions

// This constructor is repeated here because it converts
// a C string to a C++ string.
string(const char* nts);
const char* c_str() const; // Returns a null-terminated string
const char* data() const;  // Not null-terminated
// Copy characters to an array. Does not append a null-terminator
size_type copy(char* buf, size_type bufsize, size_type pos1 = 0) const;

Search Functions

size_type find(const string& s2, size_type pos1);
size_type find(const char* nts, size_type pos1);
size_type find(const char* buf, size_type pos1, size_type bufsize);
size_type find(char c, size_type pos1);
size_type rfind(const string& s2, size_type pos1);
size_type rfind(const char* nts, size_type pos1);
size_type rfind(const char* buf, size_type pos1, size_type bufsize);
size_type rfind(char c, size_type pos1);
size_type find_first_of(const string& s2, size_type pos1);
size_type find_first_of(const char* nts, size_type pos1);
size_type find_first_of(const char* buf, size_type pos1, size_type bufsize);
size_type find_first_of(char c, size_type pos1);
size_type find_last_of(const string& s2, size_type pos1);
size_type find_last_of(const char* nts, size_type pos1);
size_type find_last_of(const char* buf, size_type pos1, size_type bufsize);
size_type find_last_of(char c, size_type pos1);
size_type find_first_not_of(const string& s2, size_type pos1);
size_type find_first_not_of(const char* nts, size_type pos1);
size_type find_first_not_of(const char* buf, size_type pos1, 
                            size_type bufsize);
size_type find_first_not_of(char c, size_type pos1);
size_type find_last_not_of(const string& s2, size_type pos1);
size_type find_last_not_of(const char* nts, size_type pos1);
size_type find_last_not_of(const char* buf, size_type pos1, 
                           size_type bufsize);
size_type find_last_not_of(char c, size_type pos1);

String-Related Global Operators and Functions

The following functions and operators are related to the string class but are not members of string. Along with the string class, the following definitions are in the <string> header:
// String concatenation:
string operator+(const string& s1, const string& s2);
string operator+(const char* nts, const string& s2);
string operator+(char c, const string& s2);
string operator+(const string& s1, const char* nts);
string operator+(const string& s1, char c);
// Equal
bool operator==(const string& s1, const string& s2);
bool operator==(const char* nts, const string& s2);
bool operator==(const string& s1, const char* nts);
// Not-equal
bool operator!=(const string& s1, const string& s2);
bool operator!=(const char* nts, const string& s2);
bool operator!=(const string& s1, const char* nts);
// Ordering:
bool operator<(const string& s1, const string& s2);
bool operator<(const char* nts, const string& s2);
bool operator<(const string& s1, const char* nts);
// Greater than:
bool operator>(const string& s1, const string& s2);
bool operator>(const char* nts, const string& s2);
bool operator>(const string& s1, const char* nts);
// Less-than or equal-to:
bool operator<=(const string& s1, const string& s2);
bool operator<=(const char* nts, const string& s2);
bool operator<=(const string& s1, const char* nts);
// Greater-than or equal-to:
bool operator>=(const string& s1, const string& s2);
bool operator>=(const char* nts, const string& s2);
bool operator>=(const string& s1, const char* nts);
// Input and output
ostream& operator<<(ostream& os, const string& s);
istream& operator>>(istream& is, string& s);
istream& getline(istream& is, string& s, char delimiter = '\n');

Containers

Container Types

Note: All container classes have an additional allocator template parameter that defaults to class allocator<pair<const K, T> > for map and multimap and allocator<T> for all other containers. The allocator parameter is omitted here for clarity.

Sequence Containers

 
Class definition Header file
template <class T> class vector; <vector>
template <class T> class deque; <deque>
template <class T> class list; <list>
class string; (see also separate section on strings)
(A string can be used as a sequence container with element type char)
<string>

Sorted Associative Containers

 
Class definition Header file
template <class T> class set; <set>
template <class T> class multiset; <set>
template <class K, class T, class Comp> class map; <map>
template <class K, class T, class Comp> class multimap; <map>

Nested Types for All Containers

The following types are defined within all container classes.
value_type
iterator
const_iterator
reverse_iterator
const_reverse_iterator
size_type
difference_type

Member Functions for All Containers

The following operations are common for all container types unless otherwise noted. The word container means one of the eight container classes listed above. Performance characteristics are in terms of the number of element operations involved. For example, the performance of insert depends on the number of elements that must be moved to perform the insertion. 
Function Performance
container (); (Default Constructor) constant
container (const container & c2); (Copy constructor) linear
~container (); (Destructor) linear
container operator=(const container& c2); (Assignment operator) linear
void swap(const container & c2); constant
iterator begin();
const_iterator begin() const;
constant
iterator end();
const_iterator end() const;
constant
reverse_iterator rbegin();
const_reverse_iterator rbegin() const;
constant
reverse_iterator rend();
const_reverse_iterator rend() const;
constant
size_type size() const; constant
size_type max_size() const; constant
bool empty() const; constant
void clear(); linear

Container-Related Global Operators

The following operators are not members of any container class, but are implemented for all standard containers. The word container means one of the eight standard container classes. Both arguments must be of the same type.. They all take time that is proportional to the number of elements in the smaller argument (i.e. linear time).

bool operator==(const container & c1, const container & c2);
bool operator!=(const container & c1, const container & c2);
bool operator<(const container & c1, const container & c2);
bool operator>(const container & c1, const container & c2);
bool operator<=(const container & c1, const container & c2);
bool operator>=(const container & c1, const container & c2);

Sequence Containers Member Functions

Not all of the following operations are available for all sequence containers. The containers column contains a combination of the following codes to indicate which containers support the specified operation: V = vector, D = deque, L = list, S = string.

The InIter parameters to the template member functions are not necessarily of the same type as iterator. These functions copy elements from another container that may not be of the same type. Compilers that do not support template members will not support these functions. 
Function Performance containers
container (size_type num, const T& val = T()); (construct n copies of t) linear VDLS
template<class InIter>
container(InIter start, InIter finish); (construct from iterator range)
linear VDLS
void assign(size_type num, const T& val); linear VDLS
template<class InIter>
void assign(InIter start, InIter finish);
linear VDLS
void resize(size_type num, const T& val = T()); linear VDLS
iterator insert(iterator pos, const T& t); constant for list,
linear for others
VDLS
void insert(iterator pos, size_type num, const T& t); constant for list,
linear for others
VDLS
template<class InIter>
void insert(iterator pos, InIter start, InIter finish);
linear VDLS
iterator erase(iterator pos); constant for list,
linear for others
VDLS
iterator erase(iterator start, iterator finish); linear VDLS
T& front();
const T& front() const;
constant VDL-
T& back();
const T& back() const;
constant VDL-
void push_front(const T& t); constant -DL-
void push_back(const T& t); constant VDLS
void pop_front(); constant -DL-
void pop_back(); constant VDL-
T& operator[](size_type index);
const T& operator[](size_type index) const;
constant VD-S
T& at(size_type index);
const T& at(size_type index) const;
constant VD-S
size_type capacity() const; constant V--S
void reserve(size_type num); linear V--S

Sorted Associative Containers

Nested Types

 
set<T> and multiset<T> map<K,T> and multimap<K,T>
typedef T key_type; typedef K key_type;
typedef T value_type; typedef pair<const K, t> value_type;
typedef Comp key_compare; typedef Comp key_compare;

Member Functions

Not all of the following operations are available for all associative containers. The containers column contains a combination of the following codes to indicate which containers support the specified operation: s = set, S = multiset, m = map, M = multimap.

The InIter parameters to the template member functions are not necessarily of the same type as iterator. These functions copy elements from another container that may not be of the same type. Compilers that do not support template members will not support these functions. 
Function Performance containers
container (Comp c = Comp()); (Constructor) constant sSmM
template <class InIter>
container(InIter start, InIter finish, Comp c = Comp());
N log(N) sSmM
key_compare key_comp() const; constant sSmM
value_compare value_comp() const; constant sSmM
pair<iterator, bool> insert(const value_type& v);
(unique insert)
logarithmic s-m-
iterator insert(const value_type& v); (non-unique insert) logarithmic -S-M
template <class InIter>
insert(InIter start, InIter finish)
N Log(N)
(approx.)
sSmM
iterator find(const K& k);
const_iterator find(const K& k) const;
logarithmic sSmM
iterator lower_bound(const K& k);
const_iterator lower_bound(const K& k) const;
logarithmic sSmM
iterator upper_bound(const K& k);
const_iterator upper_bound(const K& k) const;
logarithmic sSmM
pair<iterator,iterator> equal_range(const K& k);
pair<const_iterator,const_iterator>
equal_range(const K& k) const;
logarithmic sSmM
T& operator[](const K& k); logarithmic --m-
size_type erase(const K& k); logarithmic sSmM
size_type count(const K& k) const; logarithmic sSmM

Special List Functions

The list container provides special functions to manipulate the contents of the list without actually copying the elements. This can provide a significant efficiency gain in the (relatively few) cases where this kind of manipulation is needed. Here is a concise description of these special functions:
void splice(iterator pos, list& list2); // Move all elements of list2
void splice(iterator pos, list& list2, iterator i2);
void splice(iterator pos, list& list2, iterator start2, iterator end2);
The splice functions removes elements from list2 and inserts them into *this. The data structures are modified in such a way that the elements are simply moved between the lists, not copied or deleted.
void remove(const T& value);
template <class UnaryPred> void remove(UnaryPred pred);
Remove all elements that match the specified value or predicate.
void unique();
template <class BinPred> unique(BinPred eq);
For each sequence of equal elements within the list, removes all but the first. If a predicate argument is supplied, it is used instead of == to compare elements.
void sort();
template <class Compare> sort(Compare cmp);
Sorts the list. If a Compare predicate is supplied, it is used to compare the elements instead of the < operator.
void merge(list& list2);
template <class Compare> merge(list& list2, Compare cmp);
Merges two sorted lists into one. The elements are removed from list2 and added to *this. If a Compare predicate is supplied, it is used to compare the elements instead of the < operator.
void reverse();
Reverses the order of elements in the list.

Container Adaptors

Stack

#include <stack>
template <class T, class Container = deque<T> >
class stack
{
public:
  // Constructor
  explicit stack(const Container& c = Container());
  bool empty() const;
  size_type size() const;
  value_type& top();
  const value_type& top() const;
  void push(const value_type& t);
  void pop();
};

Queue

#include <queue>
template <class T, class Container = deque<T> >
class queue
{
public:
  // Constructor
  explicit queue(const Container& c = Container());
  bool empty() const;
  size_type size() const;
  value_type& front();
  const value_type& front() const;
  value_type& back();
  const value_type& back() const;
  void push(const value_type& t);
  void pop();
};

Priority_queue

#include <queue>
template <class T, class Container = vector<T>,
          class Compare = less<Container::value_type> >
class priority_queue
{
public:
  // Constructors
  explicit priority_queue(const Compare& comp = Compare(),
                          const Container& c = Container());
  template <class InIter>
  priority_queue(InIter start, InIter finish,
                 const Compare& comp = Compare(),
                 const Container& c = Container());
  bool empty() const;
  size_type size() const;
  const value_type& top() const;  // const version only
  void push(const value_type& t);
  void pop();
};

Iterators

Insertion Iterators

In each function prototype below, output-iterator represents a standard library classes that is not usually used directly. The result of calling one of these functions is generally passed to an algorithm.
#include <iterator>
output-iterator back_inserter(Container& c);
output-iterator front_inserter(Container& c);
output-iterator inserter(Container& c, Iterator i);

Iterator Functions

#include <iterator>
template<class Iter, difference_type>
void advance(Iter& i, difference_type d);
template<class Iter>
difference_type distance(Iter start, Iter finish);

Function Objects

Function Object Base Classes

#include <functional>
template <class arg, class ret>
struct unary_function
{
  typedef arg argument_type;
  typedef ret result_type;
};
template <class arg1, class arg2, class ret>
struct binary_function
{
  typedef arg1 first_argument_type;
  typedef arg2 second_argument_type;
  typedef ret  result_type;
};

Standard Function Objects

#include <functional>
// Arithmetic function objects
template <class T> struct plus;       // binary +
template <class T> struct minus;      // binary -
template <class T> struct multiplies; // binary *
template <class T> struct divides;    // binary /
template <class T> struct modulus;    // binary %
template <class T> struct negate;     // unary -
// Comparisons
template <class T> struct equal_to;      // ==
template <class T> struct not_equal_to;  // !=
template <class T> struct less;          // <
template <class T> struct greater;       // >
template <class T> struct less_equal;    // <=
template <class T> struct greater_equal; // >=
// Logical operations
template <class T> struct logical_and; // binary &&
template <class T> struct logical_or;  // binary ||
template <class T> struct logical_not; // unary !

Negator and Binder Functions

The result of calling one of these adaptors is generally passed to an algorithm. In each prototype below, unary-function-object or binary-function-object represents a standard library template class that is not usually referenced directly but is just passed to another template function.
#include <functional>
template <class UnaryPred>
unary-function-object not1(const UnaryPred& pred);

template <class BinaryPred>
binary-function-object not2(const BinaryPred& pred);

template <class BinaryPred, class T>
unary-function-object bind1st(const BinaryPred& pred, const T& val);

template <class BinaryPred, class T>
unary-function-object bind2nd(const BinaryPred& pred, const T& val);

Function Pointer Adaptors

These adaptors convert function pointers into function objects. The result of calling one of these adaptors is generally passed to an algorithm. In the prototypes below, function-object represents a standard library template class that is not usually referenced directly but is just passed to another template function

Pointer-to-Function Adaptors

The result of calling ptr_fun will be a unary function object if func takes one argument or a binary function object if func takes two arguments.

template <class ret, class Arg>
unary-function-object ptr_fun(ret (*func)(Arg));

template <class ret, class Arg1, class Arg2>
binary-function-object ptr_fun(ret (*func)(Arg1, arg2));

Pointer-to-Member-Function Adaptors

The result of calling mem_fun is function object that takes a pointer to T as its first argument. The result of calling mem_fun_ref is a function object that takes a reference to T as its first argument. The first argument is used as the this value in the resulting member function call. If memfunc takes an argument, it is passed as the second argument to the function object. The memfunc pointer can point to a const or a non-const member function (although some non-conforming implementations do not yet support the const version).

// Returns a unary function object taking a T* argument
template <class ret, class T>
unary-function-object mem_fun(ret (T::*f)());

// Returns a binary function object taking T* and Arg arguments
template <class ret, class T, class Arg>
binary-function-object mem_fun(ret (T::*f)(Arg));

// Returns a unary function object taking a T& argument
template <class ret, class T>
unary-function-object mem_fun_ref(ret (T::*f)());

// Returns a binary function object taking T& and Arg arguments
template <class ret, class T, class Arg>
binary-function-object mem_fun_ref(ret (T::*f)(Arg));

// Returns a unary function object taking a const T* argument
template <class ret, class T>
unary-function-object mem_fun(ret (T::*f)() const);

// Returns a binary function object taking const T* and Arg arguments
template <class ret, class T, class Arg>
binary-function-object mem_fun(ret (T::*f)(Arg) const);

// Returns a unary function object taking a const T& argument
template <class ret, class T>
unary-function-object mem_fun_ref(ret (T::*f)() const);

// Returns a binary function object taking const T& and Arg arguments
template <class ret, class T, class Arg>
binary-function-object mem_fun_ref(ret (T::*f)(Arg) const);

Algorithms

Others have already done a good job of creating references for the standard algorithms. You can see a few at these locations:

http://www.cs.rpi.edu/projects/STL/htdocs/stl.html

http://www.sgi.com/Technology/STL/

The STL algorithms changed a bit when they were incorporated into the C++ Standard Library, one notable change from the above sources is in the count and count_if algorithms. Instead of passing the result in by reference, the new versions return the result directly. The return type, difference_type, is potentially different for each iterator type. In practice, however, long int will suffice for almost all cases. The new definitions are as follows:

template <class InIter, class T>
difference_type count(InIter start, InIter finish, const T& value);
template <class InIter, class UnaryPred>
difference_type count_if(InIter start, InIter finish, UnaryPred pred);

Input/Output

Stream Classes

#include <iostream>
class ios;
class istream : virtual public ios;
class ostream : virtual public ios;
class iostream : public istream, public ostream;
#include <sstream>
class stringstream : public iostream;
class istringstream : public istream;
class ostringstream : public ostream;
#include <fstream>
class fstream : public iostream;
class ifstream : public istream;
class ofstream : public ostream;
#include <strstream> // (deprecated)
class strstream : public iostream
class istrstream : public istream
class ostrstream : public ostream

ios Nested Types

The following types are defined within class ios and are inherited by all of the stream classes. They are all integral types that work as either bit masks or enumerations.
// Bitmask types: value can be the bit-wise OR of one or more of 
// the constants on the right
ios::fmtflags = (see below)
ios::iostate  = ios::goodbit = 0, ios::badbit, ios::eofbit, ios::failbit
ios::openmode = ios::in, ios::out, ios::trunc,
                ios::app, ios::ate, ios::binary
// Enumerated type: value can be one of the constants on the right
ios::seekdir  = ios::beg, ios::cur, ios::end

Format Flags

// The value of a fmtflags object is the bitwise OR of one or more
// of the following flags:
ios::boolalpha  // show bools as "true" and "false" rather than 1 and 0
ios::showbase   // show "0x" or "0" in front of hex or octal numbers
ios::showpos    // put a "+" in front of positive numbers
ios::showpoint  // show decimal point on floats that have no fraction
ios::skipws     // skip whitespace on input (default)
ios::unitbuf    // flush buffer after each operation
ios::uppercase  // Use uppercase hex digits, floating-point exponents, etc.
// Only one of the following justification adjustment flags should be set
ios::left       // left justify output in field width
ios::right      // right justify output in field width
ios::internal   // right justify but put prefix (sign, base, currency) at left
// Only one of the following base flags should be set
ios::dec        // show integral values in decimal
ios::hex        // show integral values in hexadecimal
ios::oct        // show integral values in octal
// Only one of the following floating point format flags should be set
ios::fixed      // fixed-point notation
ios::scientific // scientific notation
// The following masks are used to group mutually-exclusive flags
ios::adjustfield = ios::left | ios::right | ios::internal
ios::basefield   = ios::dec | ios::hex | ios::oct
ios::floatfield  = ios::fixed | ios::scientific

ios Member Functions

The following member functions of the ios base class are inherited by istream and ostream.
// Get/set format flags
fmtflags flags() const;        // get all format flags
fmtflags flags(fmtflags f);    // set all format flags
fmtflags setf(fmtflags f);     // set one or more format flags
fmtflags setf(fmtflags f, fmtflags g); // Set one format flag of a
                                       // mutually-exclusive group
void unsetf(fmtflags mask);    // clear one or more format flags
// Get/set format variables (set functions return old value)
int precision() const;         // get floating-point precision
int precision(int p);          // set floating-point precision
int width() const;             // get field width
int width(int w);              // set field width
char fill() const;             // get fill character
char fill(char f);             // set fill character
// Get/set error state
bool good();       // stream is in a good state
bool eof();        // end of file (end of stream) detected
bool fail();       // Last operation failed (e.g. format error)
bool bad();        // I/O error
void clear(iostate = ios::goodbit);  // Clear/set all error bits
void setstate(iostate state);        // set one or more error bits

Unformatted Output Functions

The following functions are available in ostream, iostream, and derived classes:
ostream& put(char c);
ostream& write(const char*, streamsize);
ostream& flush();

Unformatted Input Functions

The following functions are available in istream, iostream, and derived classes:
int get();
istream& get(char& c);
istream& get(char*, streamsize, term = '\n');     // doesn't read term
istream& getline(char*, streamsize, term = '\n'); // discards term
istream& ignore(streamsize, term = eof);
istream& read(char*, streamsize);

fstream Member Functions

In addition to the members it inherits from iostream and ios, fstream supplies the following member functions. In the constructors and open functions, default-mode is ios::in for ifstream, ios::out for ofstream, and ios::in|ios::out for fstream:
// Constructors (same constructors available for ifstream and ofstream)
fstream();
explicit fstream(const char* filename, ios::openmode mode = default-mode);
// Functions to open and close files
void open(const char* filename, ios::openmode mode = default-mode);
void close();
bool is_open();

stringstream Member Functions

In addition to the members it inherits from iostream and ios, stringstream supplies the following member functions. In the constructors, default-mode is ios::in for istringstream, ios::out for ostringstream, and ios::in|ios::out for stringstream:
// Constructors (same constructors available for istringstream & ostringstream)
explicit stringstream(ios::openmode mode = default-mode);
explicit stringstream(const string& s, ios::openmode mode = default-mode);
// Get/set string
string str() const;
void str(const string& s);

strstream Member Functions

The strstream classes are deprecated, meaning that they may not be part of a future C++ standard. The stringstring classes are generally preferred. However, some legacy code still uses strstream and some library implementations do not yet support stringstream.
// istrstream constructors:
explicit istrstream(const char* s);   // null-terminated
istrstream(const char* s, streamsize size);
// ostrstream constructors:
ostrstream(); // dynamically-allocated string
// string allocated by caller:
ostrstream(char* s, streamsize size, ios::openmode = ios::out);
// strstream constructors:
strstream(); // dynamically-allocated string
// string allocated by caller:
ostrstream(char* s, streamsize size, ios::openmode = ios::in|ios::out);
// istrstream, ostrstream and strstream members:
char* str();  // Calls freeze(true) & returns buffer. Call freeze(false) to thaw
// ostream and strstream (but not istrsteam) members:
int pcount() const;         // Number of chars in output buffer
void freeze(bool f = true); // prevent stream from freeing or changing buffer

I/O Manipulators

// fmtflags manipulators that take no arguments
boolalpha, noboolalpha  // set/clear ios::boolalpha
showbase, noshowbase    // set/clear ios::showbase
showpoint, noshowpoint  // set/clear ios::showpoint
showpos, noshowpos,     // set/clear ios::showpos
skipws, noskipws,       // set/clear ios::skipws
uppercase, nouppercase  // set/clear ios::uppercase
unitbuf, nounitbuf      // set/clear ios::unitbuf
left, right, internal   // set ios::left, ios::right or ios::internal
dec, hex, oct           // set ios::dec, ios::hex or ios::oct
fixed, scientific       // set ios::fixed or ios::scientific
// Manipulators that take an argument require <iomanip> header
#include <iomanip>
setiosflags(fmtflags f)    // Set specified flags
resetiosflags(fmtflags f)  // Clear specified flags
setbase(int b)             // b may be 10 (dec), 16 (hex), or 8 (oct)
setfill(char f)            // Set fill character
setprecision(int p)        // Set floating point output precision
setw(int w)                // set field width for next output operation

C Library Functions Used in the Book

Null-Terminated String Functions

#include <cstring>

// null-terminated string copy
char* strcpy(char* to, const char* from);
char* strncpy(char* to, const char* from, size_t maxlen);

// null-terminated string comparison.
// Returns -1 (s1<s2), 0 (s1==s2), or 1(s1>s2)
int strcmp(const char* s1, const char* s2);

// null-terminated string length
size_t strlen(char* s);

Character Operations

#include <cctype>

// Case conversion
char toupper(char c);
char tolower(char c);

// Character tests
bool isalnum(char c);
bool isalpha(char c);
bool isctrl(char c);
bool isdigit(char c);
bool isgraph(char c);
bool islower(char c);
bool isprint(char c);
bool ispunct(char c);
bool isspace(char c);
bool isupper(char c);
bool isxdigit(char c);

Random Numbers

#include <cstdlib>

// Random-number generation
int rand();
void srand(int seed);

Time and Date

#include <ctime>
struct tm
{
  int tm_sec;    /* seconds after the minute -- [0,60] */
  int tm_min;    /* minutes after the hour -- [0,59] */
  int tm_hour;   /* hours since midnight -- [0,23] */
  int tm_mday;   /* day of the month -- [1,31] */
  int tm_mon;    /* months since January -- [0,11] */
  int tm_year;   /* years since 1900 */
  int tm_wday;   /* days since Sunday -- [0,6] */
  int tm_yday;   /* days since January 1 -- [0,365] */
  int tm_isdst;  /* Daylight Savings Time flag */
};
// Get system time
time_t time();

// Compose time
time_t mktime(struct tm* timeptr);

// Decompose time
struct tm* gmtime(const time_t* timeval);
struct tm* localtime(const time_t* timeval);

// format time into a null-terminated string
size_t strftime(char* nts, size_t ntsmax, const char* format,
                struct tm* timeptr);
// Format string contains the following conversion codes:
//   %a  abbreviated weekday name
//   %A  full weekday name
//   %b  abbreviated month name
//   %B  full month name
//   %c  locale's date and time representation
//   %d  month as a number (01-31)
//   %H  hour in 24-hour format (00-23)
//   %I  hour in 12-hour format (01-12)
//   %j  day of the year (001-366)
//   %m  month as a number (01-12)
//   %M  minute (00-59)
//   %p  AM or PM (locale-specific)
//   %S  second (00-59)
//   %U  week of the year starting Sunday (0-53)
//   %w  weekday as a number (0=Sunday to 6=Saturday)
//   %W  week of the year starting Monday (0-53)
//   %x  locale's date representation
//   %X  locale's time representation
//   %y  two-digit year (00-99)
//   %Y  four-digit year (1900+)
//   %z  time zone name
//   %%  The percent symbol (%)

Debugging Assertions

#include <cassert>
// NOTE: since assert is a macro, it is not in namespace std
#define assert(cond) ...