typedef endian_arithmetic<order::little, uint32_t, 32, align::yes> little_uint32_at;
typedef endian_arithmetic<order::little, uint64_t, 64, align::yes> little_uint64_at;
+ // aligned floating point types
+ typedef endian_arithmetic<order::big, float, 32, align::yes> big_float32_at;
+ typedef endian_arithmetic<order::big, double, 64, align::yes> big_float64_at;
+ typedef endian_arithmetic<order::little, float, 32, align::yes> little_float32_at;
+ typedef endian_arithmetic<order::little, double, 64, align::yes> little_float64_at;
+
// aligned native endian typedefs are not provided because
// <cstdint> types are superior for this use case
typedef little_uint64_t native_uint64_t;
# endif
+ // unaligned floating point types
+ typedef endian_arithmetic<order::big, float, 32, align::no> big_float32_t;
+ typedef endian_arithmetic<order::big, double, 64, align::no> big_float64_t;
+ typedef endian_arithmetic<order::little, float, 32, align::no> little_float32_t;
+ typedef endian_arithmetic<order::little, double, 64, align::no> little_float64_t;
+ typedef endian_arithmetic<order::native, float, 32, align::no> native_float32_t;
+ typedef endian_arithmetic<order::native, double, 64, align::no> native_float64_t;
+
//---------------------------------- end synopsis ------------------------------------//
template <BOOST_SCOPED_ENUM(order) Order, class T, std::size_t n_bits,