/*=========================================================================

  Program:   Visualization Toolkit
  Module:    vtkImageDifference.cxx

  Copyright (c) Ken Martin, Will Schroeder, Bill Lorensen
  All rights reserved.
  See Copyright.txt or http://www.kitware.com/Copyright.htm for details.

     This software is distributed WITHOUT ANY WARRANTY; without even
     the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR
     PURPOSE.  See the above copyright notice for more information.

=========================================================================*/
#include "vtkImageDifference.h"

#include "vtkImageData.h"
#include "vtkInformation.h"
#include "vtkInformationVector.h"
#include "vtkObjectFactory.h"
#include "vtkStreamingDemandDrivenPipeline.h"
#include "vtkSMPTools.h"

vtkStandardNewMacro(vtkImageDifference);

// Thread-local data needed for each thread.
class vtkImageDifferenceThreadData
{
public:
  vtkImageDifferenceThreadData()
    : ErrorMessage(nullptr), Error(0.0), ThresholdedError(0.0) {}

  const char *ErrorMessage;
  double Error;
  double ThresholdedError;
};

// Holds thread-local data for all threads.
class vtkImageDifferenceSMPThreadLocal
  : public vtkSMPThreadLocal<vtkImageDifferenceThreadData>
{
public:
  typedef vtkSMPThreadLocal<vtkImageDifferenceThreadData>::iterator iterator;
};

// Construct object to extract all of the input data.
vtkImageDifference::vtkImageDifference()
{
  this->Threshold = 16;
  this->AllowShift = 1;
  this->Averaging = 1;

  this->ErrorMessage = nullptr;
  this->Error = 0.0;
  this->ThresholdedError = 0.0;

  this->ThreadData = nullptr;
  this->SMPThreadData = nullptr;

  this->SetNumberOfInputPorts(2);
}

// not so simple macro for calculating error
#define vtkImageDifferenceComputeError(c1,c2) \
/* compute the pixel to pixel difference first */ \
  r1 = abs((static_cast<int>((c1)[0]) - static_cast<int>((c2)[0])));    \
  g1 = abs((static_cast<int>((c1)[1]) - static_cast<int>((c2)[1])));    \
  b1 = abs((static_cast<int>((c1)[2]) - static_cast<int>((c2)[2])));    \
if ((r1+g1+b1) < (tr+tg+tb)) { tr = r1; tg = g1; tb = b1; } \
/* if averaging is on and we have neighbor info then compute */ \
/* input1 to avg(input2) */ \
if (this->Averaging && \
    (idx0 > inMinX + 1) && (idx0 < inMaxX - 1) && \
    (idx1 > inMinY + 1) && (idx1 < inMaxY - 1)) \
{\
    ar1 = static_cast<int>((c1)[0]);            \
    ag1 = static_cast<int>((c1)[1]);            \
    ab1 = static_cast<int>((c1)[2]);                                    \
    ar2 = static_cast<int>((c2)[0]) + static_cast<int>((c2 - in2Inc0)[0]) + static_cast<int>((c2 + in2Inc0)[0]) + \
      static_cast<int>((c2-in2Inc1)[0]) + static_cast<int>((c2-in2Inc1-in2Inc0)[0]) + static_cast<int>((c2-in2Inc1+in2Inc0)[0]) + \
      static_cast<int>((c2+in2Inc1)[0]) + static_cast<int>((c2+in2Inc1-in2Inc0)[0]) + static_cast<int>((c2+in2Inc1+in2Inc0)[0]); \
    ag2 = static_cast<int>((c2)[1]) + static_cast<int>((c2 - in2Inc0)[1]) + static_cast<int>((c2 + in2Inc0)[1]) + \
      static_cast<int>((c2-in2Inc1)[1]) + static_cast<int>((c2-in2Inc1-in2Inc0)[1]) + static_cast<int>((c2-in2Inc1+in2Inc0)[1]) + \
      static_cast<int>((c2+in2Inc1)[1]) + static_cast<int>((c2+in2Inc1-in2Inc0)[1]) + static_cast<int>((c2+in2Inc1+in2Inc0)[1]); \
    ab2 = static_cast<int>((c2)[2]) + static_cast<int>((c2 - in2Inc0)[2]) + static_cast<int>((c2 + in2Inc0)[2]) + \
      static_cast<int>((c2-in2Inc1)[2]) + static_cast<int>((c2-in2Inc1-in2Inc0)[2]) + static_cast<int>((c2-in2Inc1+in2Inc0)[2]) + \
      static_cast<int>((c2+in2Inc1)[2]) + static_cast<int>((c2+in2Inc1-in2Inc0)[2]) + static_cast<int>((c2+in2Inc1+in2Inc0)[2]); \
  r1 = abs(ar1 - ar2/9); \
  g1 = abs(ag1 - ag2/9); \
  b1 = abs(ab1 - ab2/9); \
  if ((r1+g1+b1) < (tr+tg+tb)) { tr = r1; tg = g1; tb = b1; } \
  /* Now compute the avg(input1) to avg(input2) comparison */ \
  ar1 = static_cast<int>((c1)[0]) + static_cast<int>((c1 - in1Inc0)[0]) + static_cast<int>((c1 + in1Inc0)[0]) + \
    static_cast<int>((c1-in1Inc1)[0]) + static_cast<int>((c1-in1Inc1-in1Inc0)[0]) + static_cast<int>((c1-in1Inc1+in1Inc0)[0]) + \
    static_cast<int>((c1+in1Inc1)[0]) + static_cast<int>((c1+in1Inc1-in1Inc0)[0]) + static_cast<int>((c1+in1Inc1+in1Inc0)[0]); \
  ag1 = static_cast<int>((c1)[1]) + static_cast<int>((c1 - in1Inc0)[1]) + static_cast<int>((c1 + in1Inc0)[1]) + \
    static_cast<int>((c1-in1Inc1)[1]) + static_cast<int>((c1-in1Inc1-in1Inc0)[1]) + static_cast<int>((c1-in1Inc1+in1Inc0)[1]) + \
    static_cast<int>((c1+in1Inc1)[1]) + static_cast<int>((c1+in1Inc1-in1Inc0)[1]) + static_cast<int>((c1+in1Inc1+in1Inc0)[1]); \
  ab1 = static_cast<int>((c1)[2]) + static_cast<int>((c1 - in1Inc0)[2]) + static_cast<int>((c1 + in1Inc0)[2]) + \
    static_cast<int>((c1-in1Inc1)[2]) + static_cast<int>((c1-in1Inc1-in1Inc0)[2]) + static_cast<int>((c1-in1Inc1+in1Inc0)[2]) + \
    static_cast<int>((c1+in1Inc1)[2]) + static_cast<int>((c1+in1Inc1-in1Inc0)[2]) + static_cast<int>((c1+in1Inc1+in1Inc0)[2]); \
  r1 = abs(ar1/9 - ar2/9); \
  g1 = abs(ag1/9 - ag2/9); \
  b1 = abs(ab1/9 - ab2/9); \
  if ((r1+g1+b1) < (tr+tg+tb)) { tr = r1; tg = g1; tb = b1; } \
  /* finally compute avg(input1) to input2) */ \
  ar2 = static_cast<int>((c2)[0]);             \
  ag2 = static_cast<int>((c2)[1]);             \
  ab2 = static_cast<int>((c2)[2]);             \
  r1 = abs(ar1/9 - ar2); \
  g1 = abs(ag1/9 - ag2); \
  b1 = abs(ab1/9 - ab2); \
  if ((r1+g1+b1) < (tr+tg+tb)) { tr = r1; tg = g1; tb = b1; } \
}

//----------------------------------------------------------------------------
// This functor is used with vtkSMPTools to execute the algorithm in pieces
// split over the extent of the data.
class vtkImageDifferenceSMPFunctor
{
public:
  // Create the functor by providing all of the information that will be
  // needed by the ThreadedRequestData method that the functor will call.
  vtkImageDifferenceSMPFunctor(
    vtkImageDifference *algo, vtkImageData ***inputs, vtkImageData **outputs,
    int *extent, vtkIdType pieces)
    : Algorithm(algo), Inputs(inputs), Outputs(outputs),
      Extent(extent), NumberOfPieces(pieces) {}

  void Initialize() {}
  void operator()(vtkIdType begin, vtkIdType end);
  void Reduce();

private:
  vtkImageDifferenceSMPFunctor();

  vtkImageDifference *Algorithm;
  vtkImageData ***Inputs;
  vtkImageData **Outputs;
  int *Extent;
  vtkIdType NumberOfPieces;
};

//----------------------------------------------------------------------------
void vtkImageDifferenceSMPFunctor::operator()(vtkIdType begin, vtkIdType end)
{
  this->Algorithm->SMPRequestData(
    nullptr, nullptr, nullptr, this->Inputs, this->Outputs,
    begin, end, this->NumberOfPieces, this->Extent);
}

//----------------------------------------------------------------------------
// Used with vtkSMPTools to compute the error
void vtkImageDifferenceSMPFunctor::Reduce()
{
  const char *errorMessage = nullptr;
  double error = 0.0;
  double thresholdedError = 0.0;

  for (vtkImageDifferenceSMPThreadLocal::iterator
       iter = this->Algorithm->SMPThreadData->begin();
       iter != this->Algorithm->SMPThreadData->end();
       ++iter)
  {
    errorMessage = iter->ErrorMessage;
    if (errorMessage)
    {
      break;
    }
    error += iter->Error;
    thresholdedError += iter->ThresholdedError;
  }

  this->Algorithm->ErrorMessage = errorMessage;
  this->Algorithm->Error = error;
  this->Algorithm->ThresholdedError = thresholdedError;
}

//----------------------------------------------------------------------------
// This method computes the input extent necessary to generate the output.
int vtkImageDifference::RequestUpdateExtent(
  vtkInformation * vtkNotUsed(request),
  vtkInformationVector **inputVector,
  vtkInformationVector *outputVector)
{
  int idx;

  // get the info objects
  vtkInformation* outInfo = outputVector->GetInformationObject(0);
  vtkInformation *inInfo = inputVector[0]->GetInformationObject(0);

  int *wholeExtent =
    inInfo->Get(vtkStreamingDemandDrivenPipeline::WHOLE_EXTENT());

  int uExt[6];
  outInfo->Get(vtkStreamingDemandDrivenPipeline::UPDATE_EXTENT(),uExt);

  // grow input whole extent.
  for (idx = 0; idx < 2; ++idx)
  {
    uExt[idx*2] -= 2;
    uExt[idx*2+1] += 2;

    // we must clip extent with whole extent is we hanlde boundaries.
    if (uExt[idx*2] < wholeExtent[idx*2])
    {
      uExt[idx*2] = wholeExtent[idx*2];
    }
    if (uExt[idx*2 + 1] > wholeExtent[idx*2 + 1])
    {
      uExt[idx*2 + 1] = wholeExtent[idx*2 + 1];
    }
  }
  inInfo->Set(vtkStreamingDemandDrivenPipeline::UPDATE_EXTENT(),uExt,6);

  // now do the second input
  inInfo = inputVector[1]->GetInformationObject(0);
  wholeExtent =
    inInfo->Get(vtkStreamingDemandDrivenPipeline::WHOLE_EXTENT());
  outInfo->Get(vtkStreamingDemandDrivenPipeline::UPDATE_EXTENT(),uExt);

  // grow input whole extent.
  for (idx = 0; idx < 2; ++idx)
  {
    uExt[idx*2] -= 2;
    uExt[idx*2+1] += 2;

    // we must clip extent with whole extent is we hanlde boundaries.
    if (uExt[idx*2] < wholeExtent[idx*2])
    {
      uExt[idx*2] = wholeExtent[idx*2];
    }
    if (uExt[idx*2 + 1] > wholeExtent[idx*2 + 1])
    {
      uExt[idx*2 + 1] = wholeExtent[idx*2 + 1];
    }
  }
  inInfo->Set(vtkStreamingDemandDrivenPipeline::UPDATE_EXTENT(),uExt,6);

  return 1;
}

//----------------------------------------------------------------------------
void vtkImageDifference::ThreadedRequestData(
  vtkInformation * vtkNotUsed( request ),
  vtkInformationVector ** vtkNotUsed( inputVector ),
  vtkInformationVector * vtkNotUsed( outputVector ),
  vtkImageData ***inData,
  vtkImageData **outData,
  int outExt[6], int id)
{
  unsigned char *in1Ptr0, *in1Ptr1, *in1Ptr2;
  unsigned char *in2Ptr0, *in2Ptr1, *in2Ptr2;
  unsigned char *outPtr0, *outPtr1, *outPtr2;
  int min0, max0, min1, max1, min2, max2;
  int idx0, idx1, idx2;
  vtkIdType in1Inc0, in1Inc1, in1Inc2;
  vtkIdType in2Inc0, in2Inc1, in2Inc2;
  vtkIdType outInc0, outInc1, outInc2;
  int tr, tg, tb, r1, g1, b1;
  int ar1, ag1, ab1, ar2, ag2, ab2;
  int inMinX, inMaxX, inMinY, inMaxY;
  int *inExt;
  int matched;
  unsigned long count = 0;
  unsigned long target;
  double error = 0.0;
  double thresholdedError = 0.0;
  vtkImageDifferenceThreadData *threadData = nullptr;

  if (this->EnableSMP)
  {
    threadData = &this->SMPThreadData->Local();
  }
  else
  {
    threadData = &this->ThreadData[id];
  }

  // If an error has occurred, then do not continue.
  if (threadData->ErrorMessage)
  {
    return;
  }

  if (inData[0] == nullptr || inData[1] == nullptr || outData == nullptr)
  {
    threadData->ErrorMessage = "Missing data";
    return;
  }

  if (inData[0][0]->GetNumberOfScalarComponents() != 3 ||
      inData[1][0]->GetNumberOfScalarComponents() != 3 ||
      outData[0]->GetNumberOfScalarComponents() != 3)
  {
    threadData->ErrorMessage = "Expecting 3 components (RGB)";
    return;
  }

  // this filter expects that input is the same type as output.
  if (inData[0][0]->GetScalarType() != VTK_UNSIGNED_CHAR ||
      inData[1][0]->GetScalarType() != VTK_UNSIGNED_CHAR ||
      outData[0]->GetScalarType() != VTK_UNSIGNED_CHAR)
  {
    threadData->ErrorMessage = "All ScalarTypes must be unsigned char";
    return;
  }

  in1Ptr2 = static_cast<unsigned char *>(
    inData[0][0]->GetScalarPointerForExtent(outExt));
  in2Ptr2 = static_cast<unsigned char *>(
    inData[1][0]->GetScalarPointerForExtent(outExt));
  outPtr2 = static_cast<unsigned char *>(
    outData[0]->GetScalarPointerForExtent(outExt));

  inData[0][0]->GetIncrements(in1Inc0, in1Inc1, in1Inc2);
  inData[1][0]->GetIncrements(in2Inc0, in2Inc1, in2Inc2);
  outData[0]->GetIncrements(outInc0, outInc1, outInc2);

  min0 = outExt[0];  max0 = outExt[1];
  min1 = outExt[2];  max1 = outExt[3];
  min2 = outExt[4];  max2 = outExt[5];

  inExt = inData[0][0]->GetExtent();
  // we set min and Max to be one pixel in from actual values to support
  // the 3x3 averaging we do
  inMinX = inExt[0]; inMaxX = inExt[1];
  inMinY = inExt[2]; inMaxY = inExt[3];

  target = static_cast<unsigned long>((max2 - min2 +1)*(max1 - min1 + 1)/50.0);
  target++;

  for (idx2 = min2; idx2 <= max2; ++idx2)
  {
    in1Ptr1 = in1Ptr2;
    in2Ptr1 = in2Ptr2;
    outPtr1 = outPtr2;
    for (idx1 = min1; !this->AbortExecute && idx1 <= max1; ++idx1)
    {
      if (!id)
      {
        if (!(count%target))
        {
          this->UpdateProgress(count/(50.0*target));
        }
        count++;
      }
      in1Ptr0 = in1Ptr1;
      in2Ptr0 = in2Ptr1;
      outPtr0 = outPtr1;
      for (idx0 = min0; idx0 <= max0; ++idx0)
      {
        tr = 1000;
        tg = 1000;
        tb = 1000;

        /* check the exact match pixel */
        vtkImageDifferenceComputeError(in1Ptr0,in2Ptr0);

        // do a quick check to see if this match is exact, if so
        // we can save some seious time by skipping the eight
        // connected neighbors
        matched = 0;
        if ((tr <= 0)&&(tg <= 0)&&(tb <= 0))
        {
          matched = 1;
        }

        /* If AllowShift, then we examine neighboring pixels to
           find the least difference.  This feature is used to
           allow images to shift slightly between different graphics
           systems, like between opengl and starbase. */
        if (!matched && this->AllowShift)
        {
          /* lower row */
          if (idx1 > inMinY)
          {
            vtkImageDifferenceComputeError(in1Ptr0-in1Inc1,in2Ptr0);
            if (idx0 > inMinX)
            {
              vtkImageDifferenceComputeError(in1Ptr0-in1Inc0-in1Inc1,in2Ptr0);
            }
            if (idx0 < inMaxX)
            {
              vtkImageDifferenceComputeError(in1Ptr0+in1Inc0-in1Inc1,in2Ptr0);
            }
          }
          /* middle row (center already considered) */
          if (idx0 > inMinX)
          {
            vtkImageDifferenceComputeError(in1Ptr0-in1Inc0,in2Ptr0);
          }
          if (idx0 < inMaxX)
          {
            vtkImageDifferenceComputeError(in1Ptr0+in1Inc0,in2Ptr0);
          }
          /* upper row */
          if (idx1 < inMaxY)
          {
            vtkImageDifferenceComputeError(in1Ptr0+in1Inc1,in2Ptr0);
            if (idx0 > inMinX)
            {
              vtkImageDifferenceComputeError(in1Ptr0-in1Inc0+in1Inc1,in2Ptr0);
            }
            if (idx0 < inMaxX)
            {
              vtkImageDifferenceComputeError(in1Ptr0+in1Inc0+in1Inc1,in2Ptr0);
            }
          }
        }

        error += (tr + tg + tb)/(3.0*255);
        tr -= this->Threshold;
        if (tr < 0)
        {
          tr = 0;
        }
        tg -= this->Threshold;
        if (tg < 0)
        {
          tg = 0;
        }
        tb -= this->Threshold;
        if (tb < 0)
        {
          tb = 0;
        }
        *outPtr0++ = static_cast<unsigned char>(tr);
        *outPtr0++ = static_cast<unsigned char>(tg);
        *outPtr0++ = static_cast<unsigned char>(tb);
        thresholdedError += (tr + tg + tb)/(3.0*255.0);

        in1Ptr0 += in1Inc0;
        in2Ptr0 += in2Inc0;
      }
      outPtr1 += outInc1;
      in1Ptr1 += in1Inc1;
      in2Ptr1 += in2Inc1;
    }
    outPtr2 += outInc2;
    in1Ptr2 += in1Inc2;
    in2Ptr2 += in2Inc2;
  }

  // Add the results to the thread-local total.
  threadData->Error += error;
  threadData->ThresholdedError += thresholdedError;
}

//----------------------------------------------------------------------------
// Create thread-local objects before initiating the multithreading
int vtkImageDifference::RequestData(
  vtkInformation *request,
  vtkInformationVector **inputVector,
  vtkInformationVector *outputVector)
{
  int r = 1;

  if (this->EnableSMP) // For vtkSMPTools implementation.
  {
    // Get the input and output data objects
    vtkImageData *inDataPointer[2];
    vtkImageData **inData[2] = { &inDataPointer[0], &inDataPointer[1] };
    vtkImageData *outData[1];
    this->PrepareImageData(inputVector, outputVector, inData, outData);

    // Get the extent
    int extent[6];
    outData[0]->GetExtent(extent);

    // Do a dummy execution of SplitExtent to compute the number of pieces
    vtkIdType pieces = this->SplitExtent(nullptr, extent, 0, this->NumberOfThreads);

    // Use vtkSMPTools to multithread the functor
    vtkImageDifferenceSMPThreadLocal threadData;
    vtkImageDifferenceSMPFunctor functor(
      this, inData, outData, extent, pieces);
    this->SMPThreadData = &threadData;
    bool debug = this->Debug;
    this->Debug = false;
    vtkSMPTools::For(0, pieces, functor);
    this->Debug = debug;
    this->SMPThreadData = nullptr;
  }
  else
  {
    // For vtkMultiThreader implementation.
    int n = this->NumberOfThreads;
    this->ThreadData = new vtkImageDifferenceThreadData[n];

    // The superclass will call ThreadedRequestData
    r = this->Superclass::RequestData(request, inputVector, outputVector);

    // Compute error sums here.
    this->Error = 0.0;
    this->ThresholdedError = 0.0;
    for (int i = 0; i < n; i++)
    {
      this->Error += this->ThreadData[i].Error;
      this->ThresholdedError += this->ThreadData[i].ThresholdedError;
      this->ErrorMessage = this->ThreadData[i].ErrorMessage;
      if (this->ErrorMessage)
      {
        break;
      }
    }

    delete [] this->ThreadData;
    this->ThreadData = nullptr;
  }

  if (this->ErrorMessage)
  {
    // Report errors here, do not report errors while multithreading!
    vtkErrorMacro("RequestData: " << this->ErrorMessage);
    this->ErrorMessage = nullptr;
    this->Error = 1000.0;
    this->ThresholdedError = 1000.0;
    r = 0;
  }

  return r;
}

//----------------------------------------------------------------------------
//Make the output the intersection of the inputs, of course the inputs better
//be the same size
int vtkImageDifference::RequestInformation (
  vtkInformation * vtkNotUsed(request),
  vtkInformationVector **inputVector,
  vtkInformationVector *outputVector)
{
  // get the info objects
  vtkInformation* outInfo = outputVector->GetInformationObject(0);
  vtkInformation *inInfo1 = inputVector[0]->GetInformationObject(0);
  vtkInformation *inInfo2 = inputVector[1]->GetInformationObject(0);

  int *in1Ext = inInfo1->Get(vtkStreamingDemandDrivenPipeline::WHOLE_EXTENT());
  int *in2Ext = inInfo2->Get(vtkStreamingDemandDrivenPipeline::WHOLE_EXTENT());

  int i;
  if (in1Ext[0] != in2Ext[0] || in1Ext[1] != in2Ext[1] ||
      in1Ext[2] != in2Ext[2] || in1Ext[3] != in2Ext[3] ||
      in1Ext[4] != in2Ext[4] || in1Ext[5] != in2Ext[5])
  {
    this->Error = 1000.0;
    this->ThresholdedError = 1000.0;

    vtkErrorMacro("ExecuteInformation: Input are not the same size.\n"
      << " Input1 is: " << in1Ext[0] << "," << in1Ext[1] << ","
                        << in1Ext[2] << "," << in1Ext[3] << ","
                        << in1Ext[4] << "," << in1Ext[5] << "\n"
      << " Input2 is: " << in2Ext[0] << "," << in2Ext[1] << ","
                        << in2Ext[2] << "," << in2Ext[3] << ","
                        << in2Ext[4] << "," << in2Ext[5] );
  }

  // We still need to set the whole extent to be the intersection.
  // Otherwise the execute may crash.
  int ext[6];
  for (i = 0; i < 3; ++i)
  {
    ext[i*2] = in1Ext[i*2];
    if (ext[i*2] < in2Ext[i*2])
    {
      ext[i*2] = in2Ext[i*2];
    }
    ext[i*2+1] = in1Ext[i*2+1];
    if (ext[i*2+1] > in2Ext[i*2+1])
    {
      ext[i*2+1] = in2Ext[i*2+1];
    }
  }
  outInfo->Set(vtkStreamingDemandDrivenPipeline::WHOLE_EXTENT(),ext,6);

  return 1;
}

//----------------------------------------------------------------------------
vtkImageData *vtkImageDifference::GetImage()
{
  if (this->GetNumberOfInputConnections(1) < 1)
  {
    return nullptr;
  }
  return vtkImageData::SafeDownCast(
    this->GetExecutive()->GetInputData(1, 0));
}

//----------------------------------------------------------------------------
void vtkImageDifference::PrintSelf(ostream& os, vtkIndent indent)
{
  this->Superclass::PrintSelf(os,indent);

  os << indent << "Error: " << this->Error << "\n";
  os << indent << "ThresholdedError: " << this->ThresholdedError << "\n";
  os << indent << "Threshold: " << this->Threshold << "\n";
  os << indent << "AllowShift: " << this->AllowShift << "\n";
  os << indent << "Averaging: " << this->Averaging << "\n";
}