std::num_get (3) - Linux Man Pages

std::num_get: std::num_get


std::num_get - std::num_get


Defined in header <locale>
class CharT,
class InputIt = std::istreambuf_iterator<CharT>
> class num_get;

Class std::num_get encapsulates the rules for parsing string representations of numeric values. Specifically, types bool, unsigned short, unsigned int, long, unsigned long, long long, unsigned long long, float, double, long double, and void* are supported. The standard formatting input operators (such as cin >> n;) use the std::num_get facet of the I/O stream's locale to parse the text representations of the numbers.
 std-num get-inheritance.svg
Inheritance diagram

Type requirements

InputIt must meet the requirements of LegacyInputIterator.


Two standalone (locale-independent) full specializations and two partial specializations are provided by the standard library:

Defined in header <locale>
std::num_get<char> creates narrow string parsing of numbers
std::num_get<wchar_t> creates wide string parsing of numbers
std::num_get<char, InputIt> creates narrow string parsing of numbers using custom input iterator
std::num_get<wchar_t, InputIt> creates wide string parsing of numbers using custom input iterator

In addition, every locale object constructed in a C++ program implements its own (locale-specific) versions of these specializations.

Member types

Member type Definition
char_type CharT
iter_type InputIt

Member functions

              constructs a new num_get facet
constructor (public member function)
              destructs a num_get facet
destructor (protected member function)
              invokes do_get
get (public member function)

Member objects

                          id of the locale
static std::locale::id id (public member object)

Protected member functions

do_get parses a number from an input stream
          (virtual protected member function)


// Run this code

  #include <iostream>
  #include <locale>
  #include <string>
  #include <sstream>
  #include <iterator>

  int main()
      std::string de_double = "1.234.567,89";
      std::string us_double = "1,234,567.89";

      // parse using streams
      std::istringstream de_in(de_double);
      double f1;
      de_in >> f1;

      std::istringstream us_in(de_double);
      double f2;
      us_in >> f2;

      std::cout << "Parsing " << de_double << " as double gives " << std::fixed
                << f1 << " in de_DE locale and " << f2 << " in en_US\n";

      // use the facet directly
      std::istringstream s3(us_double);
      auto& f = std::use_facet<std::num_get<char>>(s3.getloc());
      std::istreambuf_iterator<char> beg(s3), end;
      double f3;
      std::ios::iostate err;
      f.get(beg, end, s3, err, f3);
      std::cout << "parsing " << us_double
                << " as double using raw en_US facet gives " << f3 << '\n';


  Parsing 1.234.567,89 as double gives 1234567.890000 in de_DE locale and 1.234000 in en_US
  parsing 1,234,567.89 as double using raw en_US facet gives 1234567.890000

See also

           defines numeric punctuation rules
numpunct (class template)
           formats numeric values for output as character sequence
num_put (class template)
           extracts formatted data
operator>> (public member function of std::basic_istream<CharT,Traits>)