Macros min/max conflicts on Windows with Visual Studio 2015

[nauhygon]

I wanted to test out SBE so I tried the cppbuild on Windows with Visual Studio 2015. The benchlet-sbe-*-runner failed to compile due to issues like the one below.

Error	C2589	'(': illegal token on right side of '::'	benchlet-sbe-car-runner	
D:\sbe\cppbuild\Release\generated\uk_co_real_logic_sbe_benchmarks\Car.h	870	

It turned out that the SBE_NULLVALUE_* macros were using std::numeric_limits<T>::min() and std::numeric_limits<T>::max, which conflicted with the min and max macros in Windows header winnt.h. The following change to suppress the macro expansions made the problem go away.

Old:

#define SBE_NULLVALUE_INT8 std::numeric_limits<std::int8_t>::min()
#define SBE_NULLVALUE_INT16 std::numeric_limits<std::int16_t>::min()
#define SBE_NULLVALUE_INT32 std::numeric_limits<std::int32_t>::min()
#define SBE_NULLVALUE_INT64 std::numeric_limits<std::int64_t>::min()
#define SBE_NULLVALUE_UINT8 std::numeric_limits<std::uint8_t>::max()
#define SBE_NULLVALUE_UINT16 std::numeric_limits<std::uint16_t>::max()
#define SBE_NULLVALUE_UINT32 std::numeric_limits<std::uint32_t>::max()
#define SBE_NULLVALUE_UINT64 std::numeric_limits<std::uint64_t>::max()

New:

#define SBE_NULLVALUE_INT8 (std::numeric_limits<std::int8_t>::min)()
#define SBE_NULLVALUE_INT16 (std::numeric_limits<std::int16_t>::min)()
#define SBE_NULLVALUE_INT32 (std::numeric_limits<std::int32_t>::min)()
#define SBE_NULLVALUE_INT64 (std::numeric_limits<std::int64_t>::min)()
#define SBE_NULLVALUE_UINT8 (std::numeric_limits<std::uint8_t>::max)()
#define SBE_NULLVALUE_UINT16 (std::numeric_limits<std::uint16_t>::max)()
#define SBE_NULLVALUE_UINT32 (std::numeric_limits<std::uint32_t>::max)()
#define SBE_NULLVALUE_UINT64 (std::numeric_limits<std::uint64_t>::max)()

BTW, awesome work on SBE and Aeron. Many thanks!

[tmontgomery]

Thanks! Feel free to send in a PR for this.

[nauhygon]

Thanks for merging the PR. Closing this issue.

[tmontgomery]

Thanks for submitting the PR!