std::num_put - cppreference.com (original) (raw)

Class std::num_put encapsulates the rules for formatting numeric values as strings. Specifically, the types bool, long, unsigned long, long long, unsigned long long(since C++11), double, long double, void*, and of all types implicitly convertible to these (such as int or float) are supported. The standard formatting output operators (such as cout << n;) use the std::num_put facet of the I/O stream's locale to generate text representation of numbers.

std-num put-inheritance.svg

Inheritance diagram

If a std::num_put specialization is not guaranteed to be provided by the standard library (see below), the behaviors of its put() and do_put() are not guaranteed as specified.

Contents

[edit] Specializations

The standard library is guaranteed to provide the following specializations (they are required to be implemented by any locale object):

std::num_put<char> creates narrow string representations of numbers
std::num_put<wchar_t> creates wide string representations of numbers

In addition, the standard library is also guaranteed to provide every specialization that satisfies the following type requirements:

[edit] Nested types

Type Definition
char_type CharT
iter_type OutputIt

[edit] Data members

[edit] Member functions

| | constructs a new num_put facet (public member function) | | ----------------------------------------------------------- | | | invokes do_put (public member function) |

[edit] Protected member functions

| | destructs a num_put facet (protected member function) | | ----------------------------------------------------------------------------------- | | | formats a number and writes to output stream (virtual protected member function) |

[edit] Example

Possible output:

Direct conversion to string: 1234567.890000 Output using a german locale: 1.234.567,890000 Output using an american locale: 1,234,567.890000

[edit] Defect reports

The following behavior-changing defect reports were applied retroactively to previously published C++ standards.

DR Applied to Behavior as published Correct behavior
LWG 427 C++98 num_put was guaranteed to accept any CharT thatmeets the requirements for a character on whichany of the iostream components can be instantiated only guarantees to accept char,wchar_t and other implementation-defined character types
LWG 2392 C++98 only character type CharT could beguaranteed to be accepted by num_put can guarantee to accept implementation-defined character container types

[edit] See also

| | defines numeric punctuation rules (class template) [edit] | | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | | parses numeric values from an input character sequence (class template) [edit] | | | converts an integral or floating-point value to string (function) [edit] | | | converts an integral or floating-point value to wstring (function) [edit] |