doc/cstdint.qbk

[section:cstdint Standard Integer Types]

[section Overview]

The header [^[@../../../../boost/cstdint.hpp <boost/cstdint.hpp>]] provides the typedef's useful
for writing portable code that requires certain integer widths. All typedef's are in namespace boost.

The specifications for these types are based on the ISO/IEC 9899:1999 C Language standard header <stdint.h>.
The 64-bit types required by the C standard are ['not required] in the boost header,
and may not be supplied for all platforms/compilers, because [^long long] is not [yet] included in the C++ standard.

See [@../../test/cstdint_test.cpp cstdint_test.cpp] for a test program.

[endsect]

[section:rationale Rationale]

The organization of the Boost.Integer headers and classes is designed to take advantage of <stdint.h> types from the
1999 C standard without causing undefined behavior in terms of the 1998 C++ standard.
The header <boost/cstdint.hpp> makes the standard integer types safely available in namespace [^boost]
without placing any names in namespace [^std]. The intension is to complement rather than compete
with the C++ Standard Library. Should some future C++ standard include <stdint.h> and <cstdint>,
then <boost/cstdint.hpp> will continue to function, but will become redundant and may be safely deprecated.

Because these are boost headers, their names conform to boost header naming conventions rather than
C++ Standard Library header naming conventions.

[endsect]

[section:ce ['Caveat emptor]]

As an implementation artifact, certain C <limits.h> macro names may possibly be
visible to users of <boost/cstdint.hpp>. Don't use these macros; they are not part of
any Boost-specified interface. Use [^boost::integer_traits<>] or [^std::numeric_limits<>] instead.

As another implementation artifact, certain C <stdint.h> typedef names may possibly be visible
in the global namespace to users of <boost/cstdint.hpp>. Don't use these names, they are not part of
any Boost-specified interface. Use the respective names in namespace [^boost] instead.

[endsect]

[section Exact-width integer types]

The typedef [^int#_t], with # replaced by the width, designates a signed integer type of exactly # bits;
for example [^int8_t] denotes an 8-bit signed integer type. Similarly, the typedef [^uint#_t] designates an unsigned
integer type of exactly # bits.

These types are optional. However, if a platform supports integer types with widths of
8, 16, 32, 64, or any combination thereof, then <boost/cstdint.hpp> does provide the
corresponding typedefs.

The absence of int64_t and uint64_t is indicated by the macro `BOOST_NO_INT64_T`.

[endsect] 

[section Minimum-width integer types]

The typedef [^int_least#_t], with # replaced by the width, designates a signed integer type with a width
of at least # bits, such that no signed integer type with lesser size has at least the specified width.
Thus, [^int_least32_t] denotes the smallest signed integer type with a width of at least 32 bits.
Similarly, the typedef name [^uint_least#_t] designates an unsigned integer type with a width of at least # bits,
such that no unsigned integer type with lesser size has at least the specified width.

The following minimum-width integer types are provided for all platforms:

* [^int_least8_t]
* [^int_least16_t]
* [^int_least32_t]
* [^uint_least8_t]
* [^uint_least16_t]
* [^uint_least32_t]

The following types are available only if, after including <boost/cstdint.hpp>, the macro BOOST_NO_INT64_T is not defined:

* [^int_least64_t]
* [^uint_least64_t]


All other minimum-width integer types are optional.

[endsect] 

[section Fastest minimum-width integer types]

The typedef [^int_fast#_t], with # replaced by the width, designates the fastest signed integer type
with a width of at least # bits. Similarly, the typedef name [^uint_fast#_t] designates the fastest
unsigned integer type with a width of at least # bits.

There is no guarantee that these types are fastest for all purposes. In any case, however, they satisfy
the signedness and width requirements.

The following fastest minimum-width integer types are provided for all platforms:

* [^int_fast8_t]
* [^int_fast16_t]
* [^int_fast32_t]
* [^uint_fast8_t]
* [^uint_fast16_t]
* [^uint_fast32_t]

The following types are available only if, after including <boost/cstdint.hpp>, the macro BOOST_NO_INT64_T is not defined:

* [^int_fast64_t]
* [^uint_fast64_t]

All other fastest minimum-width integer types are optional.

[endsect] 

[section Greatest-width integer types]

The typedef [^intmax_t ]designates a signed integer type capable of representing any value of any signed integer type.

The typedef [^uintmax_t] designates an unsigned integer type capable of representing any value of any unsigned integer type.

These types are provided for all platforms.

[endsect]

[section Integer Constant Macros]

The following macros are always defined after inclusion of this header, these allow
integer constants of at least the specified width to be declared:
INT8_C, UINT8_C, INT16_C, UINT16_C, INT32_C, UINT32_C, INTMAX_C, UINTMAX_C.

The macros INT64_C and UINT64_C are also defined if the the macro BOOST_NO_INT64_T is not defined.

The C99 macro __STDC_CONSTANT_MACROS is also defined as an artifact of the implementation.

For example:

   #include <boost/cstdint.hpp>
   
   // Here the constant 0x1FFFFFFFF has the correct suffix applied:
   static const boost::uint64_t c = INT64_C(0x1FFFFFFFF);

[endsect]

[endsect]