Endian Conversion Functions

Introduction

Header boost/endian/conversion.hpp provides byte order reversal and conversion functions that convert objects of the multi-byte built-in integer types, and also types float and double, between native, big, or little endian byte ordering. User defined types are also supported.

conversion.hpp FAQ

Is the implementation header only?

Yes.

Does the implementation use compiler intrinsic built-in byte swapping?

Yes, if available. See Intrinsic built-in support below.

Why are the template versions of reverse() and reverse_value() in a detail namespace?

They are unsafe for general use. Consider reversing the bytes of a std::pair as a whole - the bytes from first would end up in second and visa versa, and this is totally wrong!

Why are both value returning and modify-in-place functions provided?

Returning the result by value is the standard C and C++ idiom for functions that compute a value from an argument. Modify-in-place functions allow cleaner code in many real-world endian use cases and are more efficient for user defined types that have members such as string data that do not need to be reversed. Thus both forms are provided.

What are the limitations of floating point support?

The only supported types are four byte float and eight byte double. Even after endianness has been accounted for, floating point values will not be portable between systems that use different floating point formats. Systems where the integer endianness and floating point endianness are not supported.

What are the limitations of integer support?

Only 16-bit, 32-bit, and 64-bit integers are supported. Tests have been performed on machines that use two's complement arithmetic.

Reference

Functions are implemented inline if appropriate. noexcept is elided for compilers that do not support it. Boost scoped enum emulation is used so that the library still works for compilers that do not support scoped enums.

Synopsis

#define BOOST_ENDIAN_INTRINSIC_MSG "message describing presence or absence of intrinsics"

namespace boost
{
namespace endian
{
  enum class order {big, little, native};

  // reverse byte order (i.e. endianness)
  int16_t  reverse_value(int16_t x) noexcept;
  int32_t  reverse_value(int32_t x) noexcept;
  int64_t  reverse_value(int64_t x) noexcept;
  uint16_t reverse_value(uint16_t x) noexcept;
  uint32_t reverse_value(uint32_t x) noexcept;
  uint64_t reverse_value(uint64_t x) noexcept;
  float    reverse_value(float x) noexcept;
  double   reverse_value(double x) noexcept;

  void     reverse(int16_t& x) noexcept;
  void     reverse(int32_t& x) noexcept;
  void     reverse(int64_t& x) noexcept;
  void     reverse(uint16_t& x) noexcept;
  void     reverse(uint32_t& x) noexcept;
  void     reverse(uint64_t& x) noexcept;
  void     reverse(float& x) noexcept;
  void     reverse(double& x) noexcept; 

  // reverse byte order unless native endianness is big
  template <class ReversibleValue >
    ReversibleValue big_endian_value(ReversibleValue x) noexcept; 
  template <class Reversible>
    void big_endian(Reversible& x) noexcept; 

  // reverse byte order unless native endianness is little
  template <class ReversibleValue >
    ReversibleValue little_endian_value(ReversibleValue x) noexcept; 
  template <class Reversible>
    void little_endian(Reversible& x) noexcept; 

  // synonyms, based on names popularized by BSD (e.g. OS X, Linux) endian.h
  //  "h" for "host" (i.e. native), "be" for "big endian",
  //  "le" for "little endian", "m" for "modify" in place
  template <class T> T bswap(T x) noexcept      {return reverse_value(x);}
  template <class T> T htobe(T host) noexcept   {return big_endian_value(host);}
  template <class T> T htole(T host) noexcept   {return little_endian_value(host);}
  template <class T> T betoh(T big) noexcept    {return big_endian_value(big);}
  template <class T> T letoh(T little) noexcept {return little_endian_value(little);}

  template <class T> void bswapm(T& x) noexcept      {reverse(x);}
  template <class T> void htobem(T& host) noexcept   {big_endian(host);}
  template <class T> void htole(mT& host noexcept)   {little_endian(host);}
  template <class T> void betohm(T& big) noexcept    {big_endian(big);}
  template <class T> void letohm(T& little) noexcept {little_endian(little);}

  // generic byte order conversion
  template <order From, order To, class ReversibleValue>
    ReversibleValue convert_value(ReversibleValue from) noexcept;
  template <order From, order To, class Reversible>
    void convert(Reversible& x) noexcept; 

  // runtime effective byte order determination
  order effective_order(order x) noexcept;

  // runtime byte-order conversion
  template <class ReversibleValue>
    ReversibleValue convert_value(ReversibleValue from,
      order from_order, order to_order) noexcept;
  template <class Reversible>
    void convert(Reversible& x,
      order from_order, order to_order) noexcept;

} // namespace endian
} // namespace boost

Requirements

The template definitions in this header refer to named requirements whose details are set out in this section. User defined types may be used in the function templates in this header only if they meet the function's template parameter requirements.

ReversibleValue requirements

ReversibleValue is an object type to be supplied by a C++ program instantiating a template; x is a value of type (possibly const) ReversibleValue.

Reversible requirements

Reversible is an object type to be supplied by a C++ program instantiating a template; x is a modifiable lvalue of type Reversible.

Functions

int16_t  reverse_value(int16_t x) noexcept;
int32_t  reverse_value(int32_t x) noexcept;
int64_t  reverse_value(int64_t x) noexcept;
uint16_t reverse_value(uint16_t x) noexcept;
uint32_t reverse_value(uint32_t x) noexcept;
uint64_t reverse_value(uint64_t x) noexcept;
float    reverse_value(float x) noexcept;
double   reverse_value(double x) noexcept;

Returns: x, with the order of its constituent bytes reversed.

void reverse(int16_t& x) noexcept;
void reverse(int32_t& x) noexcept;
void reverse(int64_t& x) noexcept;
void reverse(uint16_t& x) noexcept;
void reverse(uint32_t& x) noexcept;
void reverse(uint64_t& x) noexcept;
void reverse(float& x) noexcept;
void reverse(double& x) noexcept;

Postconditions: The order of the constituent bytes of x are reversed.

template <class ReversibleValue >
  ReversibleValue big_endian_value(ReversibleValue x) noexcept; 
template <class Reversible>
  void big_endian(Reversible& x) noexcept;

Returns (first form): x if the native byte order is big endian, otherwise reverse_value(x).

Effects (second form): None if the native byte order is big endian, otherwise reverse(x).

Example:

int32_t x = some-value;
big_endian(x);  // reverses the byte order of x, unless
                // the native byte order is big-endian

template <class ReversibleValue >
  ReversibleValue little_endian_value(ReversibleValue x) noexcept; 
template <class Reversible>
  void little_endian(Reversible& x) noexcept;

Returns (first form): x if the native byte order is little endian, otherwise reverse_value(x).

Effects (second form): None if the native byte order is little endian, otherwise reverse(x).

Example:

int32_t x = some-value;
int32_t y(little_endian(x));
// y has been set to x; the byte order is reversed unless
// the native byte order is little-endian.

template <order From, order To, class ReversibleValue>
  ReversibleValue convert_value(ReversibleValue from) noexcept;
template <order From, order To, class Reversible>
  void convert(Reversible& x) noexcept;

The effective order of an order template parameter is the same as the order template parameter if the parameter is not order::native, otherwise it is the constant order::big or order::little that represents the actual native byte order.

Returns (first form): from if From and To have the same effective order, otherwise reverse_value(from).

Effects (second form): None if From and To have the same effective order, otherwise reverse(x).

Example:

int32_t x;
... read an external big-endian value into x
convert<order::big, order::native>(x);  // more generic equivalent of big_endian(x);

order effective_order(order x) noexcept;
Returns: x if x != order::native, otherwise the order constant for the actual native byte order.
Example:
effective_order(order::big);     // returns order::big
effective_order(order::little);  // returns order::little
effective_order(order::native);  // returns order::big if the native order
                                 // is big-endian, otherwise order::little
template <class ReversibleValue>
  ReversibleValue convert_value(ReversibleValue from,
    order from_order, order to_order) noexcept;
template <class Reversible>
  void convert(Reversible& x,
    order from_order, order to_order) noexcept;
Returns (first form): from if effect_order(from_order) == effective_order(to_order), otherwise reverse_value(from).
  
Effects (second form): None if effect_order(from_order) == effective_order(to_order), otherwise reverse(x).
  
Example:
  

    
int32_t x;
... read an external value of an endianness know only at runtime into x
convert(x, some_order, order::native);  // convert to native byte order if needed
  
Intrinsic built-in support

Recent compilers, including GCC, Clang, and Microsoft, supply intrinsic built-in support for byte swapping. Such support is automatically detected and used since it results in smaller and much faster generated code for release builds.
Defining BOOST_ENDIAN_NO_INTRINSICS will suppress use of the intrinsics. Please try defining it if you get compiler errors, such as header byteswap.h not being found.
The macro BOOST_ENDIAN_INTRINSIC_MSG is defined as either "no byte swap intrinsics" or a string describing the particular set of intrinsics being used.
Acknowledgements
Tomas Puverle was instrumental in identifying and articulating the need to 
support endian conversion as separate from endian integer types. Phil Endecott suggested the form of the value returning signatures. Vicente Botet and other reviewers suggested supporting floating point types and user defined types. General reverse template implementation approach using std::reverse suggested by Mathias Gaunard. Portable implementation approach for 16, 32, and 64-bit integers suggested by tymofey, with avoidance of undefined behavior as suggested by Giovanni Piero Deretta, and a further refinement suggested by Pyry Jahkola. Intrinsic builtins implementation approach for 16, 32, and 64-bit integers suggested by several reviewers, and by David Stone, who provided his Boost licensed macro implementation that became the starting point for boost/endian/detail/intrinsic.hpp.

Last revised: 20 May, 2013
© Copyright Beman Dawes, 2011
Distributed under the Boost Software License, Version 1.0. See www.boost.org/ LICENSE_1_0.txt