#include <cuda_runtime.h>
#include <device_functions.h>
#include <cuda.h>
#include <device_launch_parameters.h>
#include <stdio.h>
#include <opencv2/core/cuda.hpp>

#define IPAD 2
#define SHARED_MEMORY_SIZE 32
#define SHARED_MEMORY_SIZE_Y 32
#define BLOCK_SIZE SHARED_MEMORY_SIZE

namespace YUVTailorAndBlender {

	enum  COLORSPACE
	{
		RGB = 0x0000,
		I420 = 0x0001,
		NV12 = 0x0002,
		NV21 = 0x0003,
		YV12 = I420,
		YCbCr = NV12,
	};

	extern __host__ void TurnNV12ToI420Classics(unsigned char* pNV12, int nWidth, int nHeight,
		unsigned char* pI420, int nPitch);

	extern __host__ void TurnNV12ToI420Dissociative(unsigned char* pNV12, int nWidth, int nHeight,
				unsigned char* pI420, int nPitch);

	extern __host__ void TurnCUDAFormatToI420(unsigned char* dataY, unsigned char* dataUV, size_t pitchY, size_t pitchUV, unsigned char* dstImage, int width, int height,CUstream* pStream = nullptr);

	extern __host__ void CopyPlane(
		unsigned char* pSrc, int nSrcWidth, int nSrcHeight, int nSrcPitch,
		unsigned char* pDst, int nDstWidth, int nDstHeight, int nDstPitch,
		int nCopyWidth, int nCopyHeight,
		int nCopyStartX, int nCopyStartY,
		unsigned int Type,
		CUstream* pStream
	);

	extern __host__ void CropI420(
		unsigned char* pSrcY, int nSrcYPitch,
		unsigned char* pSrcU, int nSrcUPitch,
		unsigned char* pSrcV, int nSrcVPitch,
		unsigned char* pDstY, int nDstYPitch,
		unsigned char* pDstU, int nDstUPitch,
		unsigned char* pDstV, int nDstVPitch,
		int nCropWidth, int nCropHeight,
		int nCropStartX, int nCropStartY
	);


	extern __host__ void GradientBlenderYUV(
		unsigned char* pRelateY, int YRelateStride,
		unsigned char* pRelateU, int URelateStride,
		unsigned char* pRelateV, int VRelateStride,
		unsigned char* pTargetY, int YTargetStride,
		unsigned char* pTargetU, int UTargetStride,
		unsigned char* pTargetV, int VTargetStride,
		int width, int height,
		float* pRelateMask, float* pTargetMask, int MaskStride,
		unsigned char* pDstY, int DstYStride,
		unsigned char* pDstU, int DstUStride,
		unsigned char* pDstV, int DstVStride,
		int DstWidth, int DstHeight
	);

	void __global__ TurnNV12ToI420ClassicsKernal(unsigned char* pNV12, int nWidth, int nHeight,
		unsigned char* pI420, int nPitch);

	void __global__ CopyPlaneKernal(unsigned char* pNV12Y, int nWidth, int nHeight,
		unsigned char* pI420Y, int nPitch);

	void __global__ _CopyPlaneKernal(
		unsigned char* pSrc, int nSrcWidth, int nSrcHeight, int nSrcPitch,
		unsigned char* pDst, int nDstWidth, int nDstHeight, int nDstPitch,
		int nCopyWidth, int nCopyHeight,
		int nCopyStartX, int nCopyStartY,
		int nChannel);

	void __global__ TurnNV12ToI420DissociativeKernalUV(unsigned char* pNV12UV, int nWidth, int nHeight,
				unsigned char* pI420UV, int nPitch);

	void __global__ TurnCUDAFormatToI420Kernal(unsigned char* dataY, unsigned char* dataUV, size_t pitchY, size_t pitchUV, unsigned char* I420, int width, int height);

	void __global__ GradientBlenderYUVKernal(
		unsigned char* pRelateY, int YRelateStride,
		unsigned char* pRelateU, int URelateStride,
		unsigned char* pRelateV, int VRelateStride,
		unsigned char* pTargetY, int YTargetStride,
		unsigned char* pTargetU, int UTargetStride,
		unsigned char* pTargetV, int VTargetStride,
		int width, int height,
		float* pRelateMask, float* pTargetMask, int MaskStride,
		unsigned char* pDstY, int DstYStride,
		unsigned char* pDstU, int DstUStride,
		unsigned char* pDstV, int DstVStride
		);

	void __global__ CropI420Kernel(
		unsigned char* pRelateY, int YRelateStride,
		unsigned char* pRelateU, int URelateStride,
		unsigned char* pRelateV, int VRelateStride,
		unsigned char* pTargetY, int YTargetStride,
		unsigned char* pTargetU, int UTargetStride,
		unsigned char* pTargetV, int VTargetStride,
		int nCropWidth, int nCropHeight,
		int nCropX,int nCropY,
		int DstWidth, int DstHeight
	);

	cudaError_t setColorSpace2(float hue);
}