// jpgd.h - C++ class for JPEG decompression.
// Rich Geldreich <richgel99@gmail.com>
// Alex Evans: Linear memory allocator (taken from jpge.h).
// v1.04, May. 19, 2012: Code tweaks to fix VS2008 static code analysis warnings (all looked harmless)
// D translation by Ketmar // Invisible Vector
//
// This is free and unencumbered software released into the public domain.
//
// Anyone is free to copy, modify, publish, use, compile, sell, or
// distribute this software, either in source code form or as a compiled
// binary, for any purpose, commercial or non-commercial, and by any
// means.
//
// In jurisdictions that recognize copyright laws, the author or authors
// of this software dedicate any and all copyright interest in the
// software to the public domain. We make this dedication for the benefit
// of the public at large and to the detriment of our heirs and
// successors. We intend this dedication to be an overt act of
// relinquishment in perpetuity of all present and future rights to this
// software under copyright law.
//
// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
// EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
// MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
// IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR
// OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
// ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
// OTHER DEALINGS IN THE SOFTWARE.
//
// For more information, please refer to <http://unlicense.org/>
//
// Supports progressive and baseline sequential JPEG image files, and the most common chroma subsampling factors: Y, H1V1, H2V1, H1V2, and H2V2.
//
// Chroma upsampling quality: H2V2 is upsampled in the frequency domain, H2V1 and H1V2 are upsampled using point sampling.
// Chroma upsampling reference: "Fast Scheme for Image Size Change in the Compressed Domain"
// http://vision.ai.uiuc.edu/~dugad/research/dct/index.html
/**
 * Loads a JPEG image from a memory buffer or a file.
 * req_comps can be 1 (grayscale), 3 (RGB), or 4 (RGBA).
 * On return, width/height will be set to the image's dimensions, and actual_comps will be set to the either 1 (grayscale) or 3 (RGB).
 * Requesting a 8 or 32bpp image is currently a little faster than 24bpp because the jpeg_decoder class itself currently always unpacks to either 8 or 32bpp.
 */
/// JPEG image loading.
module gamut.codecs.jpegload;

import gamut.types;
import gamut.internals.binop;
import core.stdc.string : memcpy, memset;
import core.stdc.stdlib : malloc, free;
import inteli.emmintrin;

version(decodeJPEG):

nothrow:
@nogc:

// Set to 1 to enable freq. domain chroma upsampling on images using H2V2 subsampling (0=faster nearest neighbor sampling).
// This is slower, but results in higher quality on images with highly saturated colors.
version = JPGD_SUPPORT_FREQ_DOMAIN_UPSAMPLING;

/// Input stream interface.
/// This function is called when the internal input buffer is empty.
/// Parameters:
///   pBuf - input buffer
///   max_bytes_to_read - maximum bytes that can be written to pBuf
///   pEOF_flag - set this to true if at end of stream (no more bytes remaining)
///   userData - user context for being used as closure.
///   Returns -1 on error, otherwise return the number of bytes actually written to the buffer (which may be 0).
///   Notes: This delegate will be called in a loop until you set *pEOF_flag to true or the internal buffer is full.
alias JpegStreamReadFunc = int function(void* pBuf, int max_bytes_to_read, bool* pEOF_flag, void* userData);


// ////////////////////////////////////////////////////////////////////////// //
private:

void *jpgd_malloc (size_t nSize) 
{ 
    return malloc(nSize);
}

void jpgd_free (void *p) 
{ 
    free(p);
}

// Success/failure error codes.
alias jpgd_status = int;
enum /*jpgd_status*/ {
  JPGD_SUCCESS = 0, JPGD_FAILED = -1, JPGD_DONE = 1,
  JPGD_BAD_DHT_COUNTS = -256, JPGD_BAD_DHT_INDEX, JPGD_BAD_DHT_MARKER, JPGD_BAD_DQT_MARKER, JPGD_BAD_DQT_TABLE,
  JPGD_BAD_PRECISION, JPGD_BAD_HEIGHT, JPGD_BAD_WIDTH, JPGD_TOO_MANY_COMPONENTS,
  JPGD_BAD_SOF_LENGTH, JPGD_BAD_VARIABLE_MARKER, JPGD_BAD_DRI_LENGTH, JPGD_BAD_SOS_LENGTH,
  JPGD_BAD_SOS_COMP_ID, JPGD_W_EXTRA_BYTES_BEFORE_MARKER, JPGD_NO_ARITHMITIC_SUPPORT, JPGD_UNEXPECTED_MARKER,
  JPGD_NOT_JPEG, JPGD_UNSUPPORTED_MARKER, JPGD_BAD_DQT_LENGTH, JPGD_TOO_MANY_BLOCKS,
  JPGD_UNDEFINED_QUANT_TABLE, JPGD_UNDEFINED_HUFF_TABLE, JPGD_NOT_SINGLE_SCAN, JPGD_UNSUPPORTED_COLORSPACE,
  JPGD_UNSUPPORTED_SAMP_FACTORS, JPGD_DECODE_ERROR, JPGD_BAD_RESTART_MARKER, JPGD_ASSERTION_ERROR,
  JPGD_BAD_SOS_SPECTRAL, JPGD_BAD_SOS_SUCCESSIVE, JPGD_STREAM_READ, JPGD_NOTENOUGHMEM,
}

enum {
  JPGD_IN_BUF_SIZE = 8192, JPGD_MAX_BLOCKS_PER_MCU = 10, JPGD_MAX_HUFF_TABLES = 8, JPGD_MAX_QUANT_TABLES = 4,
  JPGD_MAX_COMPONENTS = 4, JPGD_MAX_COMPS_IN_SCAN = 4, JPGD_MAX_BLOCKS_PER_ROW = 8192, JPGD_MAX_HEIGHT = 16384, JPGD_MAX_WIDTH = 16384,
}

// DCT coefficients are stored in this sequence.
static immutable int[64] g_ZAG = [  0,1,8,16,9,2,3,10,17,24,32,25,18,11,4,5,12,19,26,33,40,48,41,34,27,20,13,6,7,14,21,28,35,42,49,56,57,50,43,36,29,22,15,23,30,37,44,51,58,59,52,45,38,31,39,46,53,60,61,54,47,55,62,63 ];

alias JPEG_MARKER = int;
enum /*JPEG_MARKER*/ {
  M_SOF0  = 0xC0, M_SOF1  = 0xC1, M_SOF2  = 0xC2, M_SOF3  = 0xC3, M_SOF5  = 0xC5, M_SOF6  = 0xC6, M_SOF7  = 0xC7, M_JPG   = 0xC8,
  M_SOF9  = 0xC9, M_SOF10 = 0xCA, M_SOF11 = 0xCB, M_SOF13 = 0xCD, M_SOF14 = 0xCE, M_SOF15 = 0xCF, M_DHT   = 0xC4, M_DAC   = 0xCC,
  M_RST0  = 0xD0, M_RST1  = 0xD1, M_RST2  = 0xD2, M_RST3  = 0xD3, M_RST4  = 0xD4, M_RST5  = 0xD5, M_RST6  = 0xD6, M_RST7  = 0xD7,
  M_SOI   = 0xD8, M_EOI   = 0xD9, M_SOS   = 0xDA, M_DQT   = 0xDB, M_DNL   = 0xDC, M_DRI   = 0xDD, M_DHP   = 0xDE, M_EXP   = 0xDF,
  M_APP0  = 0xE0, M_APP15 = 0xEF, M_JPG0  = 0xF0, M_JPG13 = 0xFD, M_COM   = 0xFE, M_TEM   = 0x01, M_ERROR = 0x100, RST0   = 0xD0,
}

alias JPEG_SUBSAMPLING = int;
enum /*JPEG_SUBSAMPLING*/ { JPGD_GRAYSCALE = 0, JPGD_YH1V1, JPGD_YH2V1, JPGD_YH1V2, JPGD_YH2V2 };

enum CONST_BITS = 13;
enum PASS1_BITS = 2;
enum SCALEDONE = cast(int)1;

enum FIX_0_298631336 = cast(int)2446;  /* FIX(0.298631336) */
enum FIX_0_390180644 = cast(int)3196;  /* FIX(0.390180644) */
enum FIX_0_541196100 = cast(int)4433;  /* FIX(0.541196100) */
enum FIX_0_765366865 = cast(int)6270;  /* FIX(0.765366865) */
enum FIX_0_899976223 = cast(int)7373;  /* FIX(0.899976223) */
enum FIX_1_175875602 = cast(int)9633;  /* FIX(1.175875602) */
enum FIX_1_501321110 = cast(int)12299; /* FIX(1.501321110) */
enum FIX_1_847759065 = cast(int)15137; /* FIX(1.847759065) */
enum FIX_1_961570560 = cast(int)16069; /* FIX(1.961570560) */
enum FIX_2_053119869 = cast(int)16819; /* FIX(2.053119869) */
enum FIX_2_562915447 = cast(int)20995; /* FIX(2.562915447) */
enum FIX_3_072711026 = cast(int)25172; /* FIX(3.072711026) */

int DESCALE() (int x, int n) 
{ 
    return ((x + (SCALEDONE << (n-1))) >> n); 
}

int DESCALE_ZEROSHIFT() (int x, int n) 
{ 
    pragma(inline, true); return (((x) + (128 << (n)) + (SCALEDONE << ((n)-1))) >> (n)); 
}

ubyte CLAMP() (int i) 
{ 
    if (i < 0) i = 0;
    if (i > 255) i = 255;
    return cast(ubyte)i;
}


// Compiler creates a fast path 1D IDCT for X non-zero columns
struct Row(int NONZERO_COLS) {
pure nothrow @trusted @nogc:
  static void idct(int* pTemp, const(jpeg_decoder.jpgd_block_t)* pSrc) {
    static if (NONZERO_COLS == 0) {
      // nothing
    } else static if (NONZERO_COLS == 1) {
      immutable int dcval = (pSrc[0] << PASS1_BITS);
      pTemp[0] = dcval;
      pTemp[1] = dcval;
      pTemp[2] = dcval;
      pTemp[3] = dcval;
      pTemp[4] = dcval;
      pTemp[5] = dcval;
      pTemp[6] = dcval;
      pTemp[7] = dcval;
    } else {
      // ACCESS_COL() will be optimized at compile time to either an array access, or 0.
      //#define ACCESS_COL(x) (((x) < NONZERO_COLS) ? (int)pSrc[x] : 0)
      template ACCESS_COL(int x) {
        static if (x < NONZERO_COLS) enum ACCESS_COL = "cast(int)pSrc["~x.stringof~"]"; else enum ACCESS_COL = "0";
      }

      immutable int z2 = mixin(ACCESS_COL!2), z3 = mixin(ACCESS_COL!6);

      immutable int z1 = (z2 + z3)*FIX_0_541196100;
      immutable int tmp2 = z1 + z3*(-FIX_1_847759065);
      immutable int tmp3 = z1 + z2*FIX_0_765366865;

      immutable int tmp0 = (mixin(ACCESS_COL!0) + mixin(ACCESS_COL!4)) << CONST_BITS;
      immutable int tmp1 = (mixin(ACCESS_COL!0) - mixin(ACCESS_COL!4)) << CONST_BITS;

      immutable int tmp10 = tmp0 + tmp3, tmp13 = tmp0 - tmp3, tmp11 = tmp1 + tmp2, tmp12 = tmp1 - tmp2;

      immutable int atmp0 = mixin(ACCESS_COL!7), atmp1 = mixin(ACCESS_COL!5), atmp2 = mixin(ACCESS_COL!3), atmp3 = mixin(ACCESS_COL!1);

      immutable int bz1 = atmp0 + atmp3, bz2 = atmp1 + atmp2, bz3 = atmp0 + atmp2, bz4 = atmp1 + atmp3;
      immutable int bz5 = (bz3 + bz4)*FIX_1_175875602;

      immutable int az1 = bz1*(-FIX_0_899976223);
      immutable int az2 = bz2*(-FIX_2_562915447);
      immutable int az3 = bz3*(-FIX_1_961570560) + bz5;
      immutable int az4 = bz4*(-FIX_0_390180644) + bz5;

      immutable int btmp0 = atmp0*FIX_0_298631336 + az1 + az3;
      immutable int btmp1 = atmp1*FIX_2_053119869 + az2 + az4;
      immutable int btmp2 = atmp2*FIX_3_072711026 + az2 + az3;
      immutable int btmp3 = atmp3*FIX_1_501321110 + az1 + az4;

      pTemp[0] = DESCALE(tmp10 + btmp3, CONST_BITS-PASS1_BITS);
      pTemp[7] = DESCALE(tmp10 - btmp3, CONST_BITS-PASS1_BITS);
      pTemp[1] = DESCALE(tmp11 + btmp2, CONST_BITS-PASS1_BITS);
      pTemp[6] = DESCALE(tmp11 - btmp2, CONST_BITS-PASS1_BITS);
      pTemp[2] = DESCALE(tmp12 + btmp1, CONST_BITS-PASS1_BITS);
      pTemp[5] = DESCALE(tmp12 - btmp1, CONST_BITS-PASS1_BITS);
      pTemp[3] = DESCALE(tmp13 + btmp0, CONST_BITS-PASS1_BITS);
      pTemp[4] = DESCALE(tmp13 - btmp0, CONST_BITS-PASS1_BITS);
    }
  }
}


// Compiler creates a fast path 1D IDCT for X non-zero rows
struct Col (int NONZERO_ROWS) {
pure nothrow @trusted @nogc:
  static void idct(ubyte* pDst_ptr, const(int)* pTemp) {
    static assert(NONZERO_ROWS > 0);
    static if (NONZERO_ROWS == 1) {
      int dcval = DESCALE_ZEROSHIFT(pTemp[0], PASS1_BITS+3);
      immutable ubyte dcval_clamped = cast(ubyte)CLAMP(dcval);
      pDst_ptr[0*8] = dcval_clamped;
      pDst_ptr[1*8] = dcval_clamped;
      pDst_ptr[2*8] = dcval_clamped;
      pDst_ptr[3*8] = dcval_clamped;
      pDst_ptr[4*8] = dcval_clamped;
      pDst_ptr[5*8] = dcval_clamped;
      pDst_ptr[6*8] = dcval_clamped;
      pDst_ptr[7*8] = dcval_clamped;
    } else {
      // ACCESS_ROW() will be optimized at compile time to either an array access, or 0.
      //#define ACCESS_ROW(x) (((x) < NONZERO_ROWS) ? pTemp[x * 8] : 0)
      template ACCESS_ROW(int x) {
        static if (x < NONZERO_ROWS) enum ACCESS_ROW = "pTemp["~(x*8).stringof~"]"; else enum ACCESS_ROW = "0";
      }

      immutable int z2 = mixin(ACCESS_ROW!2);
      immutable int z3 = mixin(ACCESS_ROW!6);

      immutable int z1 = (z2 + z3)*FIX_0_541196100;
      immutable int tmp2 = z1 + z3*(-FIX_1_847759065);
      immutable int tmp3 = z1 + z2*FIX_0_765366865;

      immutable int tmp0 = (mixin(ACCESS_ROW!0) + mixin(ACCESS_ROW!4)) << CONST_BITS;
      immutable int tmp1 = (mixin(ACCESS_ROW!0) - mixin(ACCESS_ROW!4)) << CONST_BITS;

      immutable int tmp10 = tmp0 + tmp3, tmp13 = tmp0 - tmp3, tmp11 = tmp1 + tmp2, tmp12 = tmp1 - tmp2;

      immutable int atmp0 = mixin(ACCESS_ROW!7), atmp1 = mixin(ACCESS_ROW!5), atmp2 = mixin(ACCESS_ROW!3), atmp3 = mixin(ACCESS_ROW!1);

      immutable int bz1 = atmp0 + atmp3, bz2 = atmp1 + atmp2, bz3 = atmp0 + atmp2, bz4 = atmp1 + atmp3;
      immutable int bz5 = (bz3 + bz4)*FIX_1_175875602;

      immutable int az1 = bz1*(-FIX_0_899976223);
      immutable int az2 = bz2*(-FIX_2_562915447);
      immutable int az3 = bz3*(-FIX_1_961570560) + bz5;
      immutable int az4 = bz4*(-FIX_0_390180644) + bz5;

      immutable int btmp0 = atmp0*FIX_0_298631336 + az1 + az3;
      immutable int btmp1 = atmp1*FIX_2_053119869 + az2 + az4;
      immutable int btmp2 = atmp2*FIX_3_072711026 + az2 + az3;
      immutable int btmp3 = atmp3*FIX_1_501321110 + az1 + az4;

      int i = DESCALE_ZEROSHIFT(tmp10 + btmp3, CONST_BITS+PASS1_BITS+3);
      pDst_ptr[8*0] = cast(ubyte)CLAMP(i);

      i = DESCALE_ZEROSHIFT(tmp10 - btmp3, CONST_BITS+PASS1_BITS+3);
      pDst_ptr[8*7] = cast(ubyte)CLAMP(i);

      i = DESCALE_ZEROSHIFT(tmp11 + btmp2, CONST_BITS+PASS1_BITS+3);
      pDst_ptr[8*1] = cast(ubyte)CLAMP(i);

      i = DESCALE_ZEROSHIFT(tmp11 - btmp2, CONST_BITS+PASS1_BITS+3);
      pDst_ptr[8*6] = cast(ubyte)CLAMP(i);

      i = DESCALE_ZEROSHIFT(tmp12 + btmp1, CONST_BITS+PASS1_BITS+3);
      pDst_ptr[8*2] = cast(ubyte)CLAMP(i);

      i = DESCALE_ZEROSHIFT(tmp12 - btmp1, CONST_BITS+PASS1_BITS+3);
      pDst_ptr[8*5] = cast(ubyte)CLAMP(i);

      i = DESCALE_ZEROSHIFT(tmp13 + btmp0, CONST_BITS+PASS1_BITS+3);
      pDst_ptr[8*3] = cast(ubyte)CLAMP(i);

      i = DESCALE_ZEROSHIFT(tmp13 - btmp0, CONST_BITS+PASS1_BITS+3);
      pDst_ptr[8*4] = cast(ubyte)CLAMP(i);
    }
  }
}


static immutable ubyte[512] s_idct_row_table = [
  1,0,0,0,0,0,0,0, 2,0,0,0,0,0,0,0, 2,1,0,0,0,0,0,0, 2,1,1,0,0,0,0,0, 2,2,1,0,0,0,0,0, 3,2,1,0,0,0,0,0, 4,2,1,0,0,0,0,0, 4,3,1,0,0,0,0,0,
  4,3,2,0,0,0,0,0, 4,3,2,1,0,0,0,0, 4,3,2,1,1,0,0,0, 4,3,2,2,1,0,0,0, 4,3,3,2,1,0,0,0, 4,4,3,2,1,0,0,0, 5,4,3,2,1,0,0,0, 6,4,3,2,1,0,0,0,
  6,5,3,2,1,0,0,0, 6,5,4,2,1,0,0,0, 6,5,4,3,1,0,0,0, 6,5,4,3,2,0,0,0, 6,5,4,3,2,1,0,0, 6,5,4,3,2,1,1,0, 6,5,4,3,2,2,1,0, 6,5,4,3,3,2,1,0,
  6,5,4,4,3,2,1,0, 6,5,5,4,3,2,1,0, 6,6,5,4,3,2,1,0, 7,6,5,4,3,2,1,0, 8,6,5,4,3,2,1,0, 8,7,5,4,3,2,1,0, 8,7,6,4,3,2,1,0, 8,7,6,5,3,2,1,0,
  8,7,6,5,4,2,1,0, 8,7,6,5,4,3,1,0, 8,7,6,5,4,3,2,0, 8,7,6,5,4,3,2,1, 8,7,6,5,4,3,2,2, 8,7,6,5,4,3,3,2, 8,7,6,5,4,4,3,2, 8,7,6,5,5,4,3,2,
  8,7,6,6,5,4,3,2, 8,7,7,6,5,4,3,2, 8,8,7,6,5,4,3,2, 8,8,8,6,5,4,3,2, 8,8,8,7,5,4,3,2, 8,8,8,7,6,4,3,2, 8,8,8,7,6,5,3,2, 8,8,8,7,6,5,4,2,
  8,8,8,7,6,5,4,3, 8,8,8,7,6,5,4,4, 8,8,8,7,6,5,5,4, 8,8,8,7,6,6,5,4, 8,8,8,7,7,6,5,4, 8,8,8,8,7,6,5,4, 8,8,8,8,8,6,5,4, 8,8,8,8,8,7,5,4,
  8,8,8,8,8,7,6,4, 8,8,8,8,8,7,6,5, 8,8,8,8,8,7,6,6, 8,8,8,8,8,7,7,6, 8,8,8,8,8,8,7,6, 8,8,8,8,8,8,8,6, 8,8,8,8,8,8,8,7, 8,8,8,8,8,8,8,8,
];

static immutable ubyte[64] s_idct_col_table = [ 1, 1, 2, 3, 3, 3, 3, 3, 3, 4, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 6, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8 ];

void idct() (const(jpeg_decoder.jpgd_block_t)* pSrc_ptr, ubyte* pDst_ptr, int block_max_zag) {
  assert(block_max_zag >= 1);
  assert(block_max_zag <= 64);

  if (block_max_zag <= 1)
  {
    int k = ((pSrc_ptr[0] + 4) >> 3) + 128;
    k = CLAMP(k);
    k = k | (k<<8);
    k = k | (k<<16);

    for (int i = 8; i > 0; i--)
    {
      *cast(int*)&pDst_ptr[0] = k;
      *cast(int*)&pDst_ptr[4] = k;
      pDst_ptr += 8;
    }
    return;
  }

  int[64] temp;

  const(jpeg_decoder.jpgd_block_t)* pSrc = pSrc_ptr;
  int* pTemp = temp.ptr;

  const(ubyte)* pRow_tab = &s_idct_row_table.ptr[(block_max_zag - 1) * 8];
  int i;
  for (i = 8; i > 0; i--, pRow_tab++)
  {
    switch (*pRow_tab)
    {
      case 0: Row!(0).idct(pTemp, pSrc); break;
      case 1: Row!(1).idct(pTemp, pSrc); break;
      case 2: Row!(2).idct(pTemp, pSrc); break;
      case 3: Row!(3).idct(pTemp, pSrc); break;
      case 4: Row!(4).idct(pTemp, pSrc); break;
      case 5: Row!(5).idct(pTemp, pSrc); break;
      case 6: Row!(6).idct(pTemp, pSrc); break;
      case 7: Row!(7).idct(pTemp, pSrc); break;
      case 8: Row!(8).idct(pTemp, pSrc); break;
      default: assert(0);
    }

    pSrc += 8;
    pTemp += 8;
  }

  pTemp = temp.ptr;

  immutable int nonzero_rows = s_idct_col_table.ptr[block_max_zag - 1];
  for (i = 8; i > 0; i--)
  {
    switch (nonzero_rows)
    {
      case 1: Col!(1).idct(pDst_ptr, pTemp); break;
      case 2: Col!(2).idct(pDst_ptr, pTemp); break;
      case 3: Col!(3).idct(pDst_ptr, pTemp); break;
      case 4: Col!(4).idct(pDst_ptr, pTemp); break;
      case 5: Col!(5).idct(pDst_ptr, pTemp); break;
      case 6: Col!(6).idct(pDst_ptr, pTemp); break;
      case 7: Col!(7).idct(pDst_ptr, pTemp); break;
      case 8: Col!(8).idct(pDst_ptr, pTemp); break;
      default: assert(0);
    }

    pTemp++;
    pDst_ptr++;
  }
}

void idct_4x4() (const(jpeg_decoder.jpgd_block_t)* pSrc_ptr, ubyte* pDst_ptr) {
  int[64] temp;
  int* pTemp = temp.ptr;
  const(jpeg_decoder.jpgd_block_t)* pSrc = pSrc_ptr;

  for (int i = 4; i > 0; i--)
  {
    Row!(4).idct(pTemp, pSrc);
    pSrc += 8;
    pTemp += 8;
  }

  pTemp = temp.ptr;
  for (int i = 8; i > 0; i--)
  {
    Col!(4).idct(pDst_ptr, pTemp);
    pTemp++;
    pDst_ptr++;
  }
}


// ////////////////////////////////////////////////////////////////////////// //
struct jpeg_decoder {
nothrow:
@nogc:

private:
  static auto JPGD_MIN(T) (T a, T b) pure nothrow @safe @nogc { pragma(inline, true); return (a < b ? a : b); }
  static auto JPGD_MAX(T) (T a, T b) pure nothrow @safe @nogc { pragma(inline, true); return (a > b ? a : b); }

  alias jpgd_quant_t = short;
  alias jpgd_block_t = short;
  alias pDecode_block_func = void function (ref jpeg_decoder, int, int, int);

  static struct huff_tables {
    bool ac_table;
    uint[256] look_up;
    uint[256] look_up2;
    ubyte[256] code_size;
    uint[512] tree;
  }

  static struct coeff_buf {
    ubyte* pData;
    int block_num_x, block_num_y;
    int block_len_x, block_len_y;
    int block_size;
  }

  static struct mem_block {
    mem_block* m_pNext;
    size_t m_used_count;
    size_t m_size;
    char[1] m_data;
  }

  mem_block* m_pMem_blocks;
  int m_image_x_size;
  int m_image_y_size;
  JpegStreamReadFunc readfn;
  void* userData;
  int m_progressive_flag;
  ubyte[JPGD_MAX_HUFF_TABLES] m_huff_ac;
  ubyte*[JPGD_MAX_HUFF_TABLES] m_huff_num;      // pointer to number of Huffman codes per bit size
  ubyte*[JPGD_MAX_HUFF_TABLES] m_huff_val;      // pointer to Huffman codes per bit size
  jpgd_quant_t*[JPGD_MAX_QUANT_TABLES] m_quant; // pointer to quantization tables
  int m_scan_type;                              // Gray, Yh1v1, Yh1v2, Yh2v1, Yh2v2 (CMYK111, CMYK4114 no longer supported)
  int m_comps_in_frame;                         // # of components in frame
  int[JPGD_MAX_COMPONENTS] m_comp_h_samp;       // component's horizontal sampling factor
  int[JPGD_MAX_COMPONENTS] m_comp_v_samp;       // component's vertical sampling factor
  int[JPGD_MAX_COMPONENTS] m_comp_quant;        // component's quantization table selector
  int[JPGD_MAX_COMPONENTS] m_comp_ident;        // component's ID
  int[JPGD_MAX_COMPONENTS] m_comp_h_blocks;
  int[JPGD_MAX_COMPONENTS] m_comp_v_blocks;
  int m_comps_in_scan;                          // # of components in scan
  int[JPGD_MAX_COMPS_IN_SCAN] m_comp_list;      // components in this scan
  int[JPGD_MAX_COMPONENTS] m_comp_dc_tab;       // component's DC Huffman coding table selector
  int[JPGD_MAX_COMPONENTS] m_comp_ac_tab;       // component's AC Huffman coding table selector
  int m_spectral_start;                         // spectral selection start
  int m_spectral_end;                           // spectral selection end
  int m_successive_low;                         // successive approximation low
  int m_successive_high;                        // successive approximation high
  int m_max_mcu_x_size;                         // MCU's max. X size in pixels
  int m_max_mcu_y_size;                         // MCU's max. Y size in pixels
  int m_blocks_per_mcu;
  int m_max_blocks_per_row;
  int m_mcus_per_row, m_mcus_per_col;
  int[JPGD_MAX_BLOCKS_PER_MCU] m_mcu_org;
  int m_total_lines_left;                       // total # lines left in image
  int m_mcu_lines_left;                         // total # lines left in this MCU
  int m_real_dest_bytes_per_scan_line;
  int m_dest_bytes_per_scan_line;               // rounded up
  int m_dest_bytes_per_pixel;                   // 4 (RGB) or 1 (Y)
  huff_tables*[JPGD_MAX_HUFF_TABLES] m_pHuff_tabs;
  coeff_buf*[JPGD_MAX_COMPONENTS] m_dc_coeffs;
  coeff_buf*[JPGD_MAX_COMPONENTS] m_ac_coeffs;
  int m_eob_run;
  int[JPGD_MAX_COMPONENTS] m_block_y_mcu;
  ubyte* m_pIn_buf_ofs;
  int m_in_buf_left;
  int m_tem_flag;
  bool m_eof_flag;
  ubyte[128] m_in_buf_pad_start;
  ubyte[JPGD_IN_BUF_SIZE+128] m_in_buf;
  ubyte[128] m_in_buf_pad_end;
  int m_bits_left;
  uint m_bit_buf;
  int m_restart_interval;
  int m_restarts_left;
  int m_next_restart_num;
  int m_max_mcus_per_row;
  int m_max_blocks_per_mcu;
  int m_expanded_blocks_per_mcu;
  int m_expanded_blocks_per_row;
  int m_expanded_blocks_per_component;
  bool m_freq_domain_chroma_upsample;
  int m_max_mcus_per_col;
  uint[JPGD_MAX_COMPONENTS] m_last_dc_val;
  jpgd_block_t* m_pMCU_coefficients;
  int[JPGD_MAX_BLOCKS_PER_MCU] m_mcu_block_max_zag;
  ubyte* m_pSample_buf;
  int[256] m_crr;
  int[256] m_cbb;
  int[256] m_crg;
  int[256] m_cbg;
  ubyte* m_pScan_line_0;
  ubyte* m_pScan_line_1;
  jpgd_status m_error_code;
  bool m_ready_flag;
  int m_total_bytes_read;

  float m_pixelsPerInchX;
  float m_pixelsPerInchY; // -1 if not available
  float m_pixelAspectRatio; // -1 if not available

public:
  // Inspect `error_code` after constructing to determine if the stream is valid or not. You may look at the `width`, `height`, etc.
  // methods after the constructor is called. You may then either destruct the object, or begin decoding the image by calling begin_decoding(), then decode() on each scanline.
  this (JpegStreamReadFunc rfn, void* userData) 
  { 
    decode_init(rfn, userData); 
  }

  ~this () { free_all_blocks(); }

  @disable this (this); // no copies

  // Call this method after constructing the object to begin decompression.
  // If JPGD_SUCCESS is returned you may then call decode() on each scanline.
  int begin_decoding () {
    if (m_ready_flag) return JPGD_SUCCESS;
    if (m_error_code) return JPGD_FAILED;

    decode_start();
    m_ready_flag = true;
    return JPGD_SUCCESS;
  }

  // Returns the next scan line.
  // For grayscale images, pScan_line will point to a buffer containing 8-bit pixels (`bytes_per_pixel` will return 1).
  // Otherwise, it will always point to a buffer containing 32-bit RGBA pixels (A will always be 255, and `bytes_per_pixel` will return 4).
  // Returns JPGD_SUCCESS if a scan line has been returned.
  // Returns JPGD_DONE if all scan lines have been returned.
  // Returns JPGD_FAILED if an error occurred. Inspect `error_code` for a more info.
  int decode (/*const void** */void** pScan_line, uint* pScan_line_len) {
    if (m_error_code || !m_ready_flag) return JPGD_FAILED;
    if (m_total_lines_left == 0) return JPGD_DONE;

      if (m_mcu_lines_left == 0) {
        if (m_progressive_flag) load_next_row(); else decode_next_row();
        // Find the EOI marker if that was the last row.
        if (m_total_lines_left <= m_max_mcu_y_size) find_eoi();
        m_mcu_lines_left = m_max_mcu_y_size;
      }
      if (m_freq_domain_chroma_upsample) {
        expanded_convert();
        *pScan_line = m_pScan_line_0;
      } else {
        switch (m_scan_type) {
          case JPGD_YH2V2:
            if ((m_mcu_lines_left & 1) == 0) {
              H2V2Convert();
              *pScan_line = m_pScan_line_0;
            } else {
              *pScan_line = m_pScan_line_1;
            }
            break;
          case JPGD_YH2V1:
            H2V1Convert();
            *pScan_line = m_pScan_line_0;
            break;
          case JPGD_YH1V2:
            if ((m_mcu_lines_left & 1) == 0) {
              H1V2Convert();
              *pScan_line = m_pScan_line_0;
            } else {
              *pScan_line = m_pScan_line_1;
            }
            break;
          case JPGD_YH1V1:
            H1V1Convert();
            *pScan_line = m_pScan_line_0;
            break;
          case JPGD_GRAYSCALE:
            gray_convert();
            *pScan_line = m_pScan_line_0;
            break;
          default:
        }
      }
      *pScan_line_len = m_real_dest_bytes_per_scan_line;
      --m_mcu_lines_left;
      --m_total_lines_left;
      return JPGD_SUCCESS;
  }

  @property const pure nothrow @safe @nogc {
    jpgd_status error_code () { return m_error_code; }

    int width () { return m_image_x_size; }
    int height () { return m_image_y_size; }

    int num_components () { return m_comps_in_frame; }

    int bytes_per_pixel () { return m_dest_bytes_per_pixel; }
    int bytes_per_scan_line () { return m_image_x_size * bytes_per_pixel(); }

    // Returns the total number of bytes actually consumed by the decoder (which should equal the actual size of the JPEG file).
    int total_bytes_read () { return m_total_bytes_read; }
  }

private:
  // Retrieve one character from the input stream.
  uint get_char () {
    // Any bytes remaining in buffer?
    if (!m_in_buf_left) {
      // Try to get more bytes.
      prep_in_buffer();
      // Still nothing to get?
      if (!m_in_buf_left) {
        // Pad the end of the stream with 0xFF 0xD9 (EOI marker)
        int t = m_tem_flag;
        m_tem_flag ^= 1;
        return (t ? 0xD9 : 0xFF);
      }
    }
    uint c = *m_pIn_buf_ofs++;
    --m_in_buf_left;
    return c;
  }

  // Same as previous method, except can indicate if the character is a pad character or not.
  uint get_char (bool* pPadding_flag) {
    if (!m_in_buf_left) {
      prep_in_buffer();
      if (!m_in_buf_left) {
        *pPadding_flag = true;
        int t = m_tem_flag;
        m_tem_flag ^= 1;
        return (t ? 0xD9 : 0xFF);
      }
    }
    *pPadding_flag = false;
    uint c = *m_pIn_buf_ofs++;
    --m_in_buf_left;
    return c;
  }

  // Inserts a previously retrieved character back into the input buffer.
  void stuff_char (ubyte q) {
    *(--m_pIn_buf_ofs) = q;
    m_in_buf_left++;
  }

  // Retrieves one character from the input stream, but does not read past markers. Will continue to return 0xFF when a marker is encountered.
  ubyte get_octet () {
    bool padding_flag;
    int c = get_char(&padding_flag);
    if (c == 0xFF) {
      if (padding_flag) return 0xFF;
      c = get_char(&padding_flag);
      if (padding_flag) { stuff_char(0xFF); return 0xFF; }
      if (c == 0x00) return 0xFF;
      stuff_char(cast(ubyte)(c));
      stuff_char(0xFF);
      return 0xFF;
    }
    return cast(ubyte)(c);
  }

  // Retrieves a variable number of bits from the input stream. Does not recognize markers.
  uint get_bits (int num_bits) {
    if (!num_bits) return 0;
    uint i = m_bit_buf >> (32 - num_bits);
    if ((m_bits_left -= num_bits) <= 0) {
      m_bit_buf <<= (num_bits += m_bits_left);
      uint c1 = get_char();
      uint c2 = get_char();
      m_bit_buf = (m_bit_buf & 0xFFFF0000) | (c1 << 8) | c2;
      m_bit_buf <<= -m_bits_left;
      m_bits_left += 16;
      assert(m_bits_left >= 0);
    } else {
      m_bit_buf <<= num_bits;
    }
    return i;
  }

  // Retrieves a variable number of bits from the input stream. Markers will not be read into the input bit buffer. Instead, an infinite number of all 1's will be returned when a marker is encountered.
  uint get_bits_no_markers (int num_bits) {
    if (!num_bits) return 0;
    uint i = m_bit_buf >> (32 - num_bits);
    if ((m_bits_left -= num_bits) <= 0) {
      m_bit_buf <<= (num_bits += m_bits_left);
      if (m_in_buf_left < 2 || m_pIn_buf_ofs[0] == 0xFF || m_pIn_buf_ofs[1] == 0xFF) {
        uint c1 = get_octet();
        uint c2 = get_octet();
        m_bit_buf |= (c1 << 8) | c2;
      } else {
        m_bit_buf |= (cast(uint)m_pIn_buf_ofs[0] << 8) | m_pIn_buf_ofs[1];
        m_in_buf_left -= 2;
        m_pIn_buf_ofs += 2;
      }
      m_bit_buf <<= -m_bits_left;
      m_bits_left += 16;
      assert(m_bits_left >= 0);
    } else {
      m_bit_buf <<= num_bits;
    }
    return i;
  }

  // Decodes a Huffman encoded symbol.
  int huff_decode (huff_tables *pH) {
    int symbol;
    // Check first 8-bits: do we have a complete symbol?
    if ((symbol = pH.look_up.ptr[m_bit_buf >> 24]) < 0) {
      // Decode more bits, use a tree traversal to find symbol.
      int ofs = 23;
      do {
        symbol = pH.tree.ptr[-cast(int)(symbol + ((m_bit_buf >> ofs) & 1))];
        --ofs;
      } while (symbol < 0);
      get_bits_no_markers(8 + (23 - ofs));
    } else {
      get_bits_no_markers(pH.code_size.ptr[symbol]);
    }
    return symbol;
  }

  // Decodes a Huffman encoded symbol.
  int huff_decode (huff_tables *pH, ref int extra_bits) {
    int symbol;
    // Check first 8-bits: do we have a complete symbol?
    if ((symbol = pH.look_up2.ptr[m_bit_buf >> 24]) < 0) {
      // Use a tree traversal to find symbol.
      int ofs = 23;
      do {
        symbol = pH.tree.ptr[-cast(int)(symbol + ((m_bit_buf >> ofs) & 1))];
        --ofs;
      } while (symbol < 0);
      get_bits_no_markers(8 + (23 - ofs));
      extra_bits = get_bits_no_markers(symbol & 0xF);
    } else {
      assert(((symbol >> 8) & 31) == pH.code_size.ptr[symbol & 255] + ((symbol & 0x8000) ? (symbol & 15) : 0));
      if (symbol & 0x8000) {
        get_bits_no_markers((symbol >> 8) & 31);
        extra_bits = symbol >> 16;
      } else {
        int code_size = (symbol >> 8) & 31;
        int num_extra_bits = symbol & 0xF;
        int bits = code_size + num_extra_bits;
        if (bits <= (m_bits_left + 16)) {
          extra_bits = get_bits_no_markers(bits) & ((1 << num_extra_bits) - 1);
        } else {
          get_bits_no_markers(code_size);
          extra_bits = get_bits_no_markers(num_extra_bits);
        }
      }
      symbol &= 0xFF;
    }
    return symbol;
  }

  // Tables and macro used to fully decode the DPCM differences.
  static immutable int[16] s_extend_test = [ 0, 0x0001, 0x0002, 0x0004, 0x0008, 0x0010, 0x0020, 0x0040, 0x0080, 0x0100, 0x0200, 0x0400, 0x0800, 0x1000, 0x2000, 0x4000 ];
  static immutable int[16] s_extend_offset = [ 0, ((-1)<<1) + 1, ((-1)<<2) + 1, ((-1)<<3) + 1, ((-1)<<4) + 1, ((-1)<<5) + 1, ((-1)<<6) + 1, ((-1)<<7) + 1, ((-1)<<8) + 1, ((-1)<<9) + 1, ((-1)<<10) + 1, ((-1)<<11) + 1, ((-1)<<12) + 1, ((-1)<<13) + 1, ((-1)<<14) + 1, ((-1)<<15) + 1 ];
  
  static int JPGD_HUFF_EXTEND (int x, int s) nothrow @trusted @nogc 
  { 
      return (((x) < s_extend_test.ptr[s]) ? ((x) + s_extend_offset.ptr[s]) : (x)); 
  }

  // Clamps a value between 0-255.
  alias clamp = CLAMP;

  static struct DCT_Upsample {
  static:
    static struct Matrix44 {
    pure nothrow @trusted @nogc:
      alias Element_Type = int;
      enum { NUM_ROWS = 4, NUM_COLS = 4 }

      Element_Type[NUM_COLS][NUM_ROWS] v;

      this() (in Matrix44 m) {
        foreach (immutable r; 0..NUM_ROWS) v[r][] = m.v[r][];
      }

      ref inout(Element_Type) at (int r, int c) inout { pragma(inline, true); return v.ptr[r].ptr[c]; }

      ref Matrix44 opOpAssign(string op:"+") (in Matrix44 a) {
        foreach (int r; 0..NUM_ROWS) {
          at(r, 0) += a.at(r, 0);
          at(r, 1) += a.at(r, 1);
          at(r, 2) += a.at(r, 2);
          at(r, 3) += a.at(r, 3);
        }
        return this;
      }

      ref Matrix44 opOpAssign(string op:"-") (in Matrix44 a) {
        foreach (int r; 0..NUM_ROWS) {
          at(r, 0) -= a.at(r, 0);
          at(r, 1) -= a.at(r, 1);
          at(r, 2) -= a.at(r, 2);
          at(r, 3) -= a.at(r, 3);
        }
        return this;
      }

      Matrix44 opBinary(string op:"+") (in Matrix44 b) const {
        alias a = this;
        Matrix44 ret;
        foreach (int r; 0..NUM_ROWS) {
          ret.at(r, 0) = a.at(r, 0) + b.at(r, 0);
          ret.at(r, 1) = a.at(r, 1) + b.at(r, 1);
          ret.at(r, 2) = a.at(r, 2) + b.at(r, 2);
          ret.at(r, 3) = a.at(r, 3) + b.at(r, 3);
        }
        return ret;
      }

      Matrix44 opBinary(string op:"-") (in Matrix44 b) const {
        alias a = this;
        Matrix44 ret;
        foreach (int r; 0..NUM_ROWS) {
          ret.at(r, 0) = a.at(r, 0) - b.at(r, 0);
          ret.at(r, 1) = a.at(r, 1) - b.at(r, 1);
          ret.at(r, 2) = a.at(r, 2) - b.at(r, 2);
          ret.at(r, 3) = a.at(r, 3) - b.at(r, 3);
        }
        return ret;
      }

      static void add_and_store() (jpgd_block_t* pDst, in Matrix44 a, in Matrix44 b) {
        foreach (int r; 0..4) {
          pDst[0*8 + r] = cast(jpgd_block_t)(a.at(r, 0) + b.at(r, 0));
          pDst[1*8 + r] = cast(jpgd_block_t)(a.at(r, 1) + b.at(r, 1));
          pDst[2*8 + r] = cast(jpgd_block_t)(a.at(r, 2) + b.at(r, 2));
          pDst[3*8 + r] = cast(jpgd_block_t)(a.at(r, 3) + b.at(r, 3));
        }
      }

      static void sub_and_store() (jpgd_block_t* pDst, in Matrix44 a, in Matrix44 b) {
        foreach (int r; 0..4) {
          pDst[0*8 + r] = cast(jpgd_block_t)(a.at(r, 0) - b.at(r, 0));
          pDst[1*8 + r] = cast(jpgd_block_t)(a.at(r, 1) - b.at(r, 1));
          pDst[2*8 + r] = cast(jpgd_block_t)(a.at(r, 2) - b.at(r, 2));
          pDst[3*8 + r] = cast(jpgd_block_t)(a.at(r, 3) - b.at(r, 3));
        }
      }
    }

    enum FRACT_BITS = 10;
    enum SCALE = 1 << FRACT_BITS;

    alias Temp_Type = int;

    static int D(T) (T i) { pragma(inline, true); return (((i) + (SCALE >> 1)) >> FRACT_BITS); }
    enum F(float i) = (cast(int)((i) * SCALE + 0.5f));

    // NUM_ROWS/NUM_COLS = # of non-zero rows/cols in input matrix
    static struct P_Q(int NUM_ROWS, int NUM_COLS) {
      static void calc (ref Matrix44 P, ref Matrix44 Q, const(jpgd_block_t)* pSrc) {
        //auto AT (int c, int r) nothrow @trusted @nogc { return (c >= NUM_COLS || r >= NUM_ROWS ? 0 : pSrc[c+r*8]); }
        template AT(int c, int r) {
          static if (c >= NUM_COLS || r >= NUM_ROWS) enum AT = "0"; else enum AT = "pSrc["~c.stringof~"+"~r.stringof~"*8]";
        }
        // 4x8 = 4x8 times 8x8, matrix 0 is constant
        immutable Temp_Type X000 = mixin(AT!(0, 0));
        immutable Temp_Type X001 = mixin(AT!(0, 1));
        immutable Temp_Type X002 = mixin(AT!(0, 2));
        immutable Temp_Type X003 = mixin(AT!(0, 3));
        immutable Temp_Type X004 = mixin(AT!(0, 4));
        immutable Temp_Type X005 = mixin(AT!(0, 5));
        immutable Temp_Type X006 = mixin(AT!(0, 6));
        immutable Temp_Type X007 = mixin(AT!(0, 7));
        immutable Temp_Type X010 = D(F!(0.415735f) * mixin(AT!(1, 0)) + F!(0.791065f) * mixin(AT!(3, 0)) + F!(-0.352443f) * mixin(AT!(5, 0)) + F!(0.277785f) * mixin(AT!(7, 0)));
        immutable Temp_Type X011 = D(F!(0.415735f) * mixin(AT!(1, 1)) + F!(0.791065f) * mixin(AT!(3, 1)) + F!(-0.352443f) * mixin(AT!(5, 1)) + F!(0.277785f) * mixin(AT!(7, 1)));
        immutable Temp_Type X012 = D(F!(0.415735f) * mixin(AT!(1, 2)) + F!(0.791065f) * mixin(AT!(3, 2)) + F!(-0.352443f) * mixin(AT!(5, 2)) + F!(0.277785f) * mixin(AT!(7, 2)));
        immutable Temp_Type X013 = D(F!(0.415735f) * mixin(AT!(1, 3)) + F!(0.791065f) * mixin(AT!(3, 3)) + F!(-0.352443f) * mixin(AT!(5, 3)) + F!(0.277785f) * mixin(AT!(7, 3)));
        immutable Temp_Type X014 = D(F!(0.415735f) * mixin(AT!(1, 4)) + F!(0.791065f) * mixin(AT!(3, 4)) + F!(-0.352443f) * mixin(AT!(5, 4)) + F!(0.277785f) * mixin(AT!(7, 4)));
        immutable Temp_Type X015 = D(F!(0.415735f) * mixin(AT!(1, 5)) + F!(0.791065f) * mixin(AT!(3, 5)) + F!(-0.352443f) * mixin(AT!(5, 5)) + F!(0.277785f) * mixin(AT!(7, 5)));
        immutable Temp_Type X016 = D(F!(0.415735f) * mixin(AT!(1, 6)) + F!(0.791065f) * mixin(AT!(3, 6)) + F!(-0.352443f) * mixin(AT!(5, 6)) + F!(0.277785f) * mixin(AT!(7, 6)));
        immutable Temp_Type X017 = D(F!(0.415735f) * mixin(AT!(1, 7)) + F!(0.791065f) * mixin(AT!(3, 7)) + F!(-0.352443f) * mixin(AT!(5, 7)) + F!(0.277785f) * mixin(AT!(7, 7)));
        immutable Temp_Type X020 = mixin(AT!(4, 0));
        immutable Temp_Type X021 = mixin(AT!(4, 1));
        immutable Temp_Type X022 = mixin(AT!(4, 2));
        immutable Temp_Type X023 = mixin(AT!(4, 3));
        immutable Temp_Type X024 = mixin(AT!(4, 4));
        immutable Temp_Type X025 = mixin(AT!(4, 5));
        immutable Temp_Type X026 = mixin(AT!(4, 6));
        immutable Temp_Type X027 = mixin(AT!(4, 7));
        immutable Temp_Type X030 = D(F!(0.022887f) * mixin(AT!(1, 0)) + F!(-0.097545f) * mixin(AT!(3, 0)) + F!(0.490393f) * mixin(AT!(5, 0)) + F!(0.865723f) * mixin(AT!(7, 0)));
        immutable Temp_Type X031 = D(F!(0.022887f) * mixin(AT!(1, 1)) + F!(-0.097545f) * mixin(AT!(3, 1)) + F!(0.490393f) * mixin(AT!(5, 1)) + F!(0.865723f) * mixin(AT!(7, 1)));
        immutable Temp_Type X032 = D(F!(0.022887f) * mixin(AT!(1, 2)) + F!(-0.097545f) * mixin(AT!(3, 2)) + F!(0.490393f) * mixin(AT!(5, 2)) + F!(0.865723f) * mixin(AT!(7, 2)));
        immutable Temp_Type X033 = D(F!(0.022887f) * mixin(AT!(1, 3)) + F!(-0.097545f) * mixin(AT!(3, 3)) + F!(0.490393f) * mixin(AT!(5, 3)) + F!(0.865723f) * mixin(AT!(7, 3)));
        immutable Temp_Type X034 = D(F!(0.022887f) * mixin(AT!(1, 4)) + F!(-0.097545f) * mixin(AT!(3, 4)) + F!(0.490393f) * mixin(AT!(5, 4)) + F!(0.865723f) * mixin(AT!(7, 4)));
        immutable Temp_Type X035 = D(F!(0.022887f) * mixin(AT!(1, 5)) + F!(-0.097545f) * mixin(AT!(3, 5)) + F!(0.490393f) * mixin(AT!(5, 5)) + F!(0.865723f) * mixin(AT!(7, 5)));
        immutable Temp_Type X036 = D(F!(0.022887f) * mixin(AT!(1, 6)) + F!(-0.097545f) * mixin(AT!(3, 6)) + F!(0.490393f) * mixin(AT!(5, 6)) + F!(0.865723f) * mixin(AT!(7, 6)));
        immutable Temp_Type X037 = D(F!(0.022887f) * mixin(AT!(1, 7)) + F!(-0.097545f) * mixin(AT!(3, 7)) + F!(0.490393f) * mixin(AT!(5, 7)) + F!(0.865723f) * mixin(AT!(7, 7)));

        // 4x4 = 4x8 times 8x4, matrix 1 is constant
        P.at(0, 0) = X000;
        P.at(0, 1) = D(X001 * F!(0.415735f) + X003 * F!(0.791065f) + X005 * F!(-0.352443f) + X007 * F!(0.277785f));
        P.at(0, 2) = X004;
        P.at(0, 3) = D(X001 * F!(0.022887f) + X003 * F!(-0.097545f) + X005 * F!(0.490393f) + X007 * F!(0.865723f));
        P.at(1, 0) = X010;
        P.at(1, 1) = D(X011 * F!(0.415735f) + X013 * F!(0.791065f) + X015 * F!(-0.352443f) + X017 * F!(0.277785f));
        P.at(1, 2) = X014;
        P.at(1, 3) = D(X011 * F!(0.022887f) + X013 * F!(-0.097545f) + X015 * F!(0.490393f) + X017 * F!(0.865723f));
        P.at(2, 0) = X020;
        P.at(2, 1) = D(X021 * F!(0.415735f) + X023 * F!(0.791065f) + X025 * F!(-0.352443f) + X027 * F!(0.277785f));
        P.at(2, 2) = X024;
        P.at(2, 3) = D(X021 * F!(0.022887f) + X023 * F!(-0.097545f) + X025 * F!(0.490393f) + X027 * F!(0.865723f));
        P.at(3, 0) = X030;
        P.at(3, 1) = D(X031 * F!(0.415735f) + X033 * F!(0.791065f) + X035 * F!(-0.352443f) + X037 * F!(0.277785f));
        P.at(3, 2) = X034;
        P.at(3, 3) = D(X031 * F!(0.022887f) + X033 * F!(-0.097545f) + X035 * F!(0.490393f) + X037 * F!(0.865723f));
        // 40 muls 24 adds

        // 4x4 = 4x8 times 8x4, matrix 1 is constant
        Q.at(0, 0) = D(X001 * F!(0.906127f) + X003 * F!(-0.318190f) + X005 * F!(0.212608f) + X007 * F!(-0.180240f));
        Q.at(0, 1) = X002;
        Q.at(0, 2) = D(X001 * F!(-0.074658f) + X003 * F!(0.513280f) + X005 * F!(0.768178f) + X007 * F!(-0.375330f));
        Q.at(0, 3) = X006;
        Q.at(1, 0) = D(X011 * F!(0.906127f) + X013 * F!(-0.318190f) + X015 * F!(0.212608f) + X017 * F!(-0.180240f));
        Q.at(1, 1) = X012;
        Q.at(1, 2) = D(X011 * F!(-0.074658f) + X013 * F!(0.513280f) + X015 * F!(0.768178f) + X017 * F!(-0.375330f));
        Q.at(1, 3) = X016;
        Q.at(2, 0) = D(X021 * F!(0.906127f) + X023 * F!(-0.318190f) + X025 * F!(0.212608f) + X027 * F!(-0.180240f));
        Q.at(2, 1) = X022;
        Q.at(2, 2) = D(X021 * F!(-0.074658f) + X023 * F!(0.513280f) + X025 * F!(0.768178f) + X027 * F!(-0.375330f));
        Q.at(2, 3) = X026;
        Q.at(3, 0) = D(X031 * F!(0.906127f) + X033 * F!(-0.318190f) + X035 * F!(0.212608f) + X037 * F!(-0.180240f));
        Q.at(3, 1) = X032;
        Q.at(3, 2) = D(X031 * F!(-0.074658f) + X033 * F!(0.513280f) + X035 * F!(0.768178f) + X037 * F!(-0.375330f));
        Q.at(3, 3) = X036;
        // 40 muls 24 adds
      }
    }

    static struct R_S(int NUM_ROWS, int NUM_COLS) {
      static void calc(ref Matrix44 R, ref Matrix44 S, const(jpgd_block_t)* pSrc) {
        //auto AT (int c, int r) nothrow @trusted @nogc { return (c >= NUM_COLS || r >= NUM_ROWS ? 0 : pSrc[c+r*8]); }
        template AT(int c, int r) {
          static if (c >= NUM_COLS || r >= NUM_ROWS) enum AT = "0"; else enum AT = "pSrc["~c.stringof~"+"~r.stringof~"*8]";
        }
        // 4x8 = 4x8 times 8x8, matrix 0 is constant
        immutable Temp_Type X100 = D(F!(0.906127f) * mixin(AT!(1, 0)) + F!(-0.318190f) * mixin(AT!(3, 0)) + F!(0.212608f) * mixin(AT!(5, 0)) + F!(-0.180240f) * mixin(AT!(7, 0)));
        immutable Temp_Type X101 = D(F!(0.906127f) * mixin(AT!(1, 1)) + F!(-0.318190f) * mixin(AT!(3, 1)) + F!(0.212608f) * mixin(AT!(5, 1)) + F!(-0.180240f) * mixin(AT!(7, 1)));
        immutable Temp_Type X102 = D(F!(0.906127f) * mixin(AT!(1, 2)) + F!(-0.318190f) * mixin(AT!(3, 2)) + F!(0.212608f) * mixin(AT!(5, 2)) + F!(-0.180240f) * mixin(AT!(7, 2)));
        immutable Temp_Type X103 = D(F!(0.906127f) * mixin(AT!(1, 3)) + F!(-0.318190f) * mixin(AT!(3, 3)) + F!(0.212608f) * mixin(AT!(5, 3)) + F!(-0.180240f) * mixin(AT!(7, 3)));
        immutable Temp_Type X104 = D(F!(0.906127f) * mixin(AT!(1, 4)) + F!(-0.318190f) * mixin(AT!(3, 4)) + F!(0.212608f) * mixin(AT!(5, 4)) + F!(-0.180240f) * mixin(AT!(7, 4)));
        immutable Temp_Type X105 = D(F!(0.906127f) * mixin(AT!(1, 5)) + F!(-0.318190f) * mixin(AT!(3, 5)) + F!(0.212608f) * mixin(AT!(5, 5)) + F!(-0.180240f) * mixin(AT!(7, 5)));
        immutable Temp_Type X106 = D(F!(0.906127f) * mixin(AT!(1, 6)) + F!(-0.318190f) * mixin(AT!(3, 6)) + F!(0.212608f) * mixin(AT!(5, 6)) + F!(-0.180240f) * mixin(AT!(7, 6)));
        immutable Temp_Type X107 = D(F!(0.906127f) * mixin(AT!(1, 7)) + F!(-0.318190f) * mixin(AT!(3, 7)) + F!(0.212608f) * mixin(AT!(5, 7)) + F!(-0.180240f) * mixin(AT!(7, 7)));
        immutable Temp_Type X110 = mixin(AT!(2, 0));
        immutable Temp_Type X111 = mixin(AT!(2, 1));
        immutable Temp_Type X112 = mixin(AT!(2, 2));
        immutable Temp_Type X113 = mixin(AT!(2, 3));
        immutable Temp_Type X114 = mixin(AT!(2, 4));
        immutable Temp_Type X115 = mixin(AT!(2, 5));
        immutable Temp_Type X116 = mixin(AT!(2, 6));
        immutable Temp_Type X117 = mixin(AT!(2, 7));
        immutable Temp_Type X120 = D(F!(-0.074658f) * mixin(AT!(1, 0)) + F!(0.513280f) * mixin(AT!(3, 0)) + F!(0.768178f) * mixin(AT!(5, 0)) + F!(-0.375330f) * mixin(AT!(7, 0)));
        immutable Temp_Type X121 = D(F!(-0.074658f) * mixin(AT!(1, 1)) + F!(0.513280f) * mixin(AT!(3, 1)) + F!(0.768178f) * mixin(AT!(5, 1)) + F!(-0.375330f) * mixin(AT!(7, 1)));
        immutable Temp_Type X122 = D(F!(-0.074658f) * mixin(AT!(1, 2)) + F!(0.513280f) * mixin(AT!(3, 2)) + F!(0.768178f) * mixin(AT!(5, 2)) + F!(-0.375330f) * mixin(AT!(7, 2)));
        immutable Temp_Type X123 = D(F!(-0.074658f) * mixin(AT!(1, 3)) + F!(0.513280f) * mixin(AT!(3, 3)) + F!(0.768178f) * mixin(AT!(5, 3)) + F!(-0.375330f) * mixin(AT!(7, 3)));
        immutable Temp_Type X124 = D(F!(-0.074658f) * mixin(AT!(1, 4)) + F!(0.513280f) * mixin(AT!(3, 4)) + F!(0.768178f) * mixin(AT!(5, 4)) + F!(-0.375330f) * mixin(AT!(7, 4)));
        immutable Temp_Type X125 = D(F!(-0.074658f) * mixin(AT!(1, 5)) + F!(0.513280f) * mixin(AT!(3, 5)) + F!(0.768178f) * mixin(AT!(5, 5)) + F!(-0.375330f) * mixin(AT!(7, 5)));
        immutable Temp_Type X126 = D(F!(-0.074658f) * mixin(AT!(1, 6)) + F!(0.513280f) * mixin(AT!(3, 6)) + F!(0.768178f) * mixin(AT!(5, 6)) + F!(-0.375330f) * mixin(AT!(7, 6)));
        immutable Temp_Type X127 = D(F!(-0.074658f) * mixin(AT!(1, 7)) + F!(0.513280f) * mixin(AT!(3, 7)) + F!(0.768178f) * mixin(AT!(5, 7)) + F!(-0.375330f) * mixin(AT!(7, 7)));
        immutable Temp_Type X130 = mixin(AT!(6, 0));
        immutable Temp_Type X131 = mixin(AT!(6, 1));
        immutable Temp_Type X132 = mixin(AT!(6, 2));
        immutable Temp_Type X133 = mixin(AT!(6, 3));
        immutable Temp_Type X134 = mixin(AT!(6, 4));
        immutable Temp_Type X135 = mixin(AT!(6, 5));
        immutable Temp_Type X136 = mixin(AT!(6, 6));
        immutable Temp_Type X137 = mixin(AT!(6, 7));
        // 80 muls 48 adds

        // 4x4 = 4x8 times 8x4, matrix 1 is constant
        R.at(0, 0) = X100;
        R.at(0, 1) = D(X101 * F!(0.415735f) + X103 * F!(0.791065f) + X105 * F!(-0.352443f) + X107 * F!(0.277785f));
        R.at(0, 2) = X104;
        R.at(0, 3) = D(X101 * F!(0.022887f) + X103 * F!(-0.097545f) + X105 * F!(0.490393f) + X107 * F!(0.865723f));
        R.at(1, 0) = X110;
        R.at(1, 1) = D(X111 * F!(0.415735f) + X113 * F!(0.791065f) + X115 * F!(-0.352443f) + X117 * F!(0.277785f));
        R.at(1, 2) = X114;
        R.at(1, 3) = D(X111 * F!(0.022887f) + X113 * F!(-0.097545f) + X115 * F!(0.490393f) + X117 * F!(0.865723f));
        R.at(2, 0) = X120;
        R.at(2, 1) = D(X121 * F!(0.415735f) + X123 * F!(0.791065f) + X125 * F!(-0.352443f) + X127 * F!(0.277785f));
        R.at(2, 2) = X124;
        R.at(2, 3) = D(X121 * F!(0.022887f) + X123 * F!(-0.097545f) + X125 * F!(0.490393f) + X127 * F!(0.865723f));
        R.at(3, 0) = X130;
        R.at(3, 1) = D(X131 * F!(0.415735f) + X133 * F!(0.791065f) + X135 * F!(-0.352443f) + X137 * F!(0.277785f));
        R.at(3, 2) = X134;
        R.at(3, 3) = D(X131 * F!(0.022887f) + X133 * F!(-0.097545f) + X135 * F!(0.490393f) + X137 * F!(0.865723f));
        // 40 muls 24 adds
        // 4x4 = 4x8 times 8x4, matrix 1 is constant
        S.at(0, 0) = D(X101 * F!(0.906127f) + X103 * F!(-0.318190f) + X105 * F!(0.212608f) + X107 * F!(-0.180240f));
        S.at(0, 1) = X102;
        S.at(0, 2) = D(X101 * F!(-0.074658f) + X103 * F!(0.513280f) + X105 * F!(0.768178f) + X107 * F!(-0.375330f));
        S.at(0, 3) = X106;
        S.at(1, 0) = D(X111 * F!(0.906127f) + X113 * F!(-0.318190f) + X115 * F!(0.212608f) + X117 * F!(-0.180240f));
        S.at(1, 1) = X112;
        S.at(1, 2) = D(X111 * F!(-0.074658f) + X113 * F!(0.513280f) + X115 * F!(0.768178f) + X117 * F!(-0.375330f));
        S.at(1, 3) = X116;
        S.at(2, 0) = D(X121 * F!(0.906127f) + X123 * F!(-0.318190f) + X125 * F!(0.212608f) + X127 * F!(-0.180240f));
        S.at(2, 1) = X122;
        S.at(2, 2) = D(X121 * F!(-0.074658f) + X123 * F!(0.513280f) + X125 * F!(0.768178f) + X127 * F!(-0.375330f));
        S.at(2, 3) = X126;
        S.at(3, 0) = D(X131 * F!(0.906127f) + X133 * F!(-0.318190f) + X135 * F!(0.212608f) + X137 * F!(-0.180240f));
        S.at(3, 1) = X132;
        S.at(3, 2) = D(X131 * F!(-0.074658f) + X133 * F!(0.513280f) + X135 * F!(0.768178f) + X137 * F!(-0.375330f));
        S.at(3, 3) = X136;
        // 40 muls 24 adds
      }
    }
  } // end namespace DCT_Upsample

  // Unconditionally frees all allocated m_blocks.
  void free_all_blocks () {
    //m_pStream = null;
    readfn = null;
    for (mem_block *b = m_pMem_blocks; b; ) {
      mem_block* n = b.m_pNext;
      jpgd_free(b);
      b = n;
    }
    m_pMem_blocks = null;
  }

  // This method handles all errors. It will never return.
  // It could easily be changed to use C++ exceptions.
  /*JPGD_NORETURN*/ void stop_decoding (jpgd_status status) {
    m_error_code = status;
    free_all_blocks();
    //longjmp(m_jmp_state, status);
    assert(false, "jpeg decoding error");
  }

  void* alloc (size_t nSize, bool zero=false) {
    nSize = (JPGD_MAX(nSize, 1) + 3) & ~3;
    char *rv = null;
    for (mem_block *b = m_pMem_blocks; b; b = b.m_pNext)
    {
      if ((b.m_used_count + nSize) <= b.m_size)
      {
        rv = b.m_data.ptr + b.m_used_count;
        b.m_used_count += nSize;
        break;
      }
    }
    if (!rv)
    {
      int capacity = cast(int) JPGD_MAX(32768 - 256, (nSize + 2047) & ~2047);
      mem_block *b = cast(mem_block*)jpgd_malloc(mem_block.sizeof + capacity);
      if (!b) { stop_decoding(JPGD_NOTENOUGHMEM); }
      b.m_pNext = m_pMem_blocks; m_pMem_blocks = b;
      b.m_used_count = nSize;
      b.m_size = capacity;
      rv = b.m_data.ptr;
    }
    if (zero) memset(rv, 0, nSize);
    return rv;
  }

  void word_clear (void *p, ushort c, uint n) {
    ubyte *pD = cast(ubyte*)p;
    immutable ubyte l = c & 0xFF, h = (c >> 8) & 0xFF;
    while (n)
    {
      pD[0] = l; pD[1] = h; pD += 2;
      n--;
    }
  }

  // Refill the input buffer.
  // This method will sit in a loop until (A) the buffer is full or (B)
  // the stream's read() method reports and end of file condition.
  void prep_in_buffer () {
    m_in_buf_left = 0;
    m_pIn_buf_ofs = m_in_buf.ptr;

    if (m_eof_flag)
      return;

    do
    {
      int bytes_read = readfn(m_in_buf.ptr + m_in_buf_left, JPGD_IN_BUF_SIZE - m_in_buf_left, &m_eof_flag, userData);
      if (bytes_read == -1)
        stop_decoding(JPGD_STREAM_READ);

      m_in_buf_left += bytes_read;
    } while ((m_in_buf_left < JPGD_IN_BUF_SIZE) && (!m_eof_flag));

    m_total_bytes_read += m_in_buf_left;

    // Pad the end of the block with M_EOI (prevents the decompressor from going off the rails if the stream is invalid).
    // (This dates way back to when this decompressor was written in C/asm, and the all-asm Huffman decoder did some fancy things to increase perf.)
    word_clear(m_pIn_buf_ofs + m_in_buf_left, 0xD9FF, 64);
  }

  // Read a Huffman code table.
  void read_dht_marker () {
    int i, index, count;
    ubyte[17] huff_num;
    ubyte[256] huff_val;

    uint num_left = get_bits(16);

    if (num_left < 2)
      stop_decoding(JPGD_BAD_DHT_MARKER);

    num_left -= 2;

    while (num_left)
    {
      index = get_bits(8);

      huff_num.ptr[0] = 0;

      count = 0;

      for (i = 1; i <= 16; i++)
      {
        huff_num.ptr[i] = cast(ubyte)(get_bits(8));
        count += huff_num.ptr[i];
      }

      if (count > 255)
        stop_decoding(JPGD_BAD_DHT_COUNTS);

      for (i = 0; i < count; i++)
        huff_val.ptr[i] = cast(ubyte)(get_bits(8));

      i = 1 + 16 + count;

      if (num_left < cast(uint)i)
        stop_decoding(JPGD_BAD_DHT_MARKER);

      num_left -= i;

      if ((index & 0x10) > 0x10)
        stop_decoding(JPGD_BAD_DHT_INDEX);

      index = (index & 0x0F) + ((index & 0x10) >> 4) * (JPGD_MAX_HUFF_TABLES >> 1);

      if (index >= JPGD_MAX_HUFF_TABLES)
        stop_decoding(JPGD_BAD_DHT_INDEX);

      if (!m_huff_num.ptr[index])
        m_huff_num.ptr[index] = cast(ubyte*)alloc(17);

      if (!m_huff_val.ptr[index])
        m_huff_val.ptr[index] = cast(ubyte*)alloc(256);

      m_huff_ac.ptr[index] = (index & 0x10) != 0;
      memcpy(m_huff_num.ptr[index], huff_num.ptr, 17);
      memcpy(m_huff_val.ptr[index], huff_val.ptr, 256);
    }
  }

  // Read a quantization table.
  void read_dqt_marker () {
    int n, i, prec;
    uint num_left;
    uint temp;

    num_left = get_bits(16);

    if (num_left < 2)
      stop_decoding(JPGD_BAD_DQT_MARKER);

    num_left -= 2;

    while (num_left)
    {
      n = get_bits(8);
      prec = n >> 4;
      n &= 0x0F;

      if (n >= JPGD_MAX_QUANT_TABLES)
        stop_decoding(JPGD_BAD_DQT_TABLE);

      if (!m_quant.ptr[n])
        m_quant.ptr[n] = cast(jpgd_quant_t*)alloc(64 * jpgd_quant_t.sizeof);

      // read quantization entries, in zag order
      for (i = 0; i < 64; i++)
      {
        temp = get_bits(8);

        if (prec)
          temp = (temp << 8) + get_bits(8);

        m_quant.ptr[n][i] = cast(jpgd_quant_t)(temp);
      }

      i = 64 + 1;

      if (prec)
        i += 64;

      if (num_left < cast(uint)i)
        stop_decoding(JPGD_BAD_DQT_LENGTH);

      num_left -= i;
    }
  }

  // Read the start of frame (SOF) marker.
  void read_sof_marker () {
    int i;
    uint num_left;

    num_left = get_bits(16);

    if (get_bits(8) != 8)   /* precision: sorry, only 8-bit precision is supported right now */
      stop_decoding(JPGD_BAD_PRECISION);

    m_image_y_size = get_bits(16);

    if ((m_image_y_size < 1) || (m_image_y_size > JPGD_MAX_HEIGHT))
      stop_decoding(JPGD_BAD_HEIGHT);

    m_image_x_size = get_bits(16);

    if ((m_image_x_size < 1) || (m_image_x_size > JPGD_MAX_WIDTH))
      stop_decoding(JPGD_BAD_WIDTH);

    m_comps_in_frame = get_bits(8);

    if (m_comps_in_frame > JPGD_MAX_COMPONENTS)
      stop_decoding(JPGD_TOO_MANY_COMPONENTS);

    if (num_left != cast(uint)(m_comps_in_frame * 3 + 8))
      stop_decoding(JPGD_BAD_SOF_LENGTH);

    for (i = 0; i < m_comps_in_frame; i++)
    {
      m_comp_ident.ptr[i]  = get_bits(8);
      m_comp_h_samp.ptr[i] = get_bits(4);
      m_comp_v_samp.ptr[i] = get_bits(4);
      m_comp_quant.ptr[i]  = get_bits(8);
    }
  }

  // Used to skip unrecognized markers.
  void skip_variable_marker () {
    uint num_left;

    num_left = get_bits(16);

    if (num_left < 2)
      stop_decoding(JPGD_BAD_VARIABLE_MARKER);

    num_left -= 2;

    while (num_left)
    {
      get_bits(8);
      num_left--;
    }
  }

  // Read a define restart interval (DRI) marker.
  void read_dri_marker () {
    if (get_bits(16) != 4)
      stop_decoding(JPGD_BAD_DRI_LENGTH);

    m_restart_interval = get_bits(16);
  }

  // Read a start of scan (SOS) marker.
  void read_sos_marker () {
    uint num_left;
    int i, ci, n, c, cc;

    num_left = get_bits(16);

    n = get_bits(8);

    m_comps_in_scan = n;

    num_left -= 3;

    if ( (num_left != cast(uint)(n * 2 + 3)) || (n < 1) || (n > JPGD_MAX_COMPS_IN_SCAN) )
      stop_decoding(JPGD_BAD_SOS_LENGTH);

    for (i = 0; i < n; i++)
    {
      cc = get_bits(8);
      c = get_bits(8);
      num_left -= 2;

      for (ci = 0; ci < m_comps_in_frame; ci++)
        if (cc == m_comp_ident.ptr[ci])
          break;

      if (ci >= m_comps_in_frame)
        stop_decoding(JPGD_BAD_SOS_COMP_ID);

      m_comp_list.ptr[i]    = ci;
      m_comp_dc_tab.ptr[ci] = (c >> 4) & 15;
      m_comp_ac_tab.ptr[ci] = (c & 15) + (JPGD_MAX_HUFF_TABLES >> 1);
    }

    m_spectral_start  = get_bits(8);
    m_spectral_end    = get_bits(8);
    m_successive_high = get_bits(4);
    m_successive_low  = get_bits(4);

    if (!m_progressive_flag)
    {
      m_spectral_start = 0;
      m_spectral_end = 63;
    }

    num_left -= 3;

    /* read past whatever is num_left */
    while (num_left)
    {
      get_bits(8);
      num_left--;
    }
  }

  // Finds the next marker.
  int next_marker () {
    uint c, bytes;

    bytes = 0;

    do
    {
      do
      {
        bytes++;
        c = get_bits(8);
      } while (c != 0xFF);

      do
      {
        c = get_bits(8);
      } while (c == 0xFF);

    } while (c == 0);

    // If bytes > 0 here, there where extra bytes before the marker (not good).

    return c;
  }

  // Process markers. Returns when an SOFx, SOI, EOI, or SOS marker is
  // encountered.
  int process_markers () {
    int c;

    for ( ; ; ) {
      c = next_marker();

      switch (c)
      {
        case M_SOF0:
        case M_SOF1:
        case M_SOF2:
        case M_SOF3:
        case M_SOF5:
        case M_SOF6:
        case M_SOF7:
        //case M_JPG:
        case M_SOF9:
        case M_SOF10:
        case M_SOF11:
        case M_SOF13:
        case M_SOF14:
        case M_SOF15:
        case M_SOI:
        case M_EOI:
        case M_SOS:
          return c;
        case M_DHT:
          read_dht_marker();
          break;
        // No arithmitic support - dumb patents!
        case M_DAC:
          stop_decoding(JPGD_NO_ARITHMITIC_SUPPORT);
          break;
        case M_DQT:
          read_dqt_marker();
          break;
        case M_DRI:
          read_dri_marker();
          break;

        case M_APP0:
            uint num_left;

            num_left = get_bits(16);

            if (num_left < 7)
                stop_decoding(JPGD_BAD_VARIABLE_MARKER);

            num_left -= 2;

            ubyte[5] jfif_id;
            foreach(i; 0..5)
                jfif_id[i] = cast(ubyte) get_bits(8);

            num_left -= 5;
            static immutable ubyte[5] JFIF = [0x4A, 0x46, 0x49, 0x46, 0x00];
            if (jfif_id == JFIF && num_left >= 7)
            {
                // skip version
                get_bits(16);
                uint units = get_bits(8);
                int Xdensity = get_bits(16);
                int Ydensity = get_bits(16);
                num_left -= 7;

                m_pixelAspectRatio = (Xdensity/cast(double)Ydensity);

                switch (units)
                {
                    case 0: // no units, just a ratio
                        m_pixelsPerInchX = -1;
                        m_pixelsPerInchY = -1;
                        break;

                    case 1: // dot per inch
                        m_pixelsPerInchX = Xdensity;
                        m_pixelsPerInchY = Ydensity;
                        break;

                    case 2: // dot per cm
                        m_pixelsPerInchX = convertInchesToMeters(Xdensity * 100.0f);
                        m_pixelsPerInchY = convertInchesToMeters(Ydensity * 100.0f);
                        break;
                    default:
                }
            }

            // skip rests of chunk

            while (num_left)
            {
                get_bits(8);
                num_left--;
            }
            break;

        case M_APP0+1: // possibly EXIF data
         
            uint num_left;
            num_left = get_bits(16);

            if (num_left < 2)
                stop_decoding(JPGD_BAD_VARIABLE_MARKER);
            num_left -= 2;

            ubyte[] exifData = (cast(ubyte*) malloc(num_left))[0..num_left];
            scope(exit) free(exifData.ptr);

            foreach(i; 0..num_left)
                exifData[i] = cast(ubyte)(get_bits(8));

            const(ubyte)* s = exifData.ptr;

            ubyte[6] exif_id;
            foreach(i; 0..6)
                exif_id[i] = read_ubyte(s);

            const(ubyte)* remainExifData = s;

            static immutable ubyte[6] ExifIdentifierCode = [0x45, 0x78, 0x69, 0x66, 0x00, 0x00]; // "Exif\0\0"
            if (exif_id == ExifIdentifierCode)
            {
                // See EXIF specification: http://www.cipa.jp/std/documents/e/DC-008-2012_E.pdf

                const(ubyte)* tiffFile = s; // save exif chunk from "start of TIFF file"

                ushort byteOrder = read_ushort_BE(s);
                if (byteOrder != 0x4949 && byteOrder != 0x4D4D)
                    stop_decoding(JPGD_DECODE_ERROR);
                bool littleEndian = (byteOrder == 0x4949);

                ushort version_ = littleEndian ? read_ushort_LE(s) : read_ushort_BE(s);
                if (version_ != 42)
                    stop_decoding(JPGD_DECODE_ERROR);

                uint offset = littleEndian ? read_uint_LE(s) : read_uint_BE(s);

                double resolutionX = 72;
                double resolutionY = 72;
                int unit = 2;

                // parse all IFDs
                while(offset != 0)
                {
                    if (offset > exifData.length)
                        stop_decoding(JPGD_DECODE_ERROR);
                    const(ubyte)* pIFD = tiffFile + offset;
                    ushort numEntries = littleEndian ? read_ushort_LE(pIFD) : read_ushort_BE(pIFD);

                    foreach(entry; 0..numEntries)
                    {
                        ushort tag = littleEndian ? read_ushort_LE(pIFD) : read_ushort_BE(pIFD);
                        ushort type = littleEndian ? read_ushort_LE(pIFD) : read_ushort_BE(pIFD);
                        uint count = littleEndian ? read_uint_LE(pIFD) : read_uint_BE(pIFD);
                        uint valueOffset = littleEndian ? read_uint_LE(pIFD) : read_uint_BE(pIFD);

                        if (tag == 282 || tag == 283) // XResolution
                        {
                            const(ubyte)* tagData = tiffFile + valueOffset;
                            double num = littleEndian ? read_uint_LE(tagData) : read_uint_BE(tagData);
                            double denom = littleEndian ? read_uint_LE(tagData) : read_uint_BE(tagData);
                            double frac = num / denom;
                            if (tag == 282)
                                resolutionX = frac;
                            else
                                resolutionY = frac;
                        }

                        if (tag == 296) // unit
                            unit = valueOffset;
                    }
                    offset = littleEndian ? read_uint_LE(pIFD) : read_uint_BE(pIFD);
                }

                if (unit == 2) // inches
                {
                    m_pixelsPerInchX = resolutionX;
                    m_pixelsPerInchY = resolutionY;
                    m_pixelAspectRatio = resolutionX / resolutionY;
                }
                else if (unit == 3) // dots per cm
                {
                    m_pixelsPerInchX = convertInchesToMeters(resolutionX * 100);
                    m_pixelsPerInchY = convertInchesToMeters(resolutionY * 100);
                    m_pixelAspectRatio = resolutionX / resolutionY;
                }
            }
            break;

        case M_JPG:
        case M_RST0:    /* no parameters */
        case M_RST1:
        case M_RST2:
        case M_RST3:
        case M_RST4:
        case M_RST5:
        case M_RST6:
        case M_RST7:
        case M_TEM:
          stop_decoding(JPGD_UNEXPECTED_MARKER);
          break;
        default:    /* must be DNL, DHP, EXP, APPn, JPGn, COM, or RESn or APP0 */
          skip_variable_marker();
          break;
      }
    }
  }

  // Finds the start of image (SOI) marker.
  // This code is rather defensive: it only checks the first 512 bytes to avoid
  // false positives.
  void locate_soi_marker () {
    uint lastchar, thischar;
    uint bytesleft;

    lastchar = get_bits(8);

    thischar = get_bits(8);

    /* ok if it's a normal JPEG file without a special header */

    if ((lastchar == 0xFF) && (thischar == M_SOI))
      return;

    bytesleft = 4096; //512;

    for ( ; ; )
    {
      if (--bytesleft == 0)
        stop_decoding(JPGD_NOT_JPEG);

      lastchar = thischar;

      thischar = get_bits(8);

      if (lastchar == 0xFF)
      {
        if (thischar == M_SOI)
          break;
        else if (thischar == M_EOI) // get_bits will keep returning M_EOI if we read past the end
          stop_decoding(JPGD_NOT_JPEG);
      }
    }

    // Check the next character after marker: if it's not 0xFF, it can't be the start of the next marker, so the file is bad.
    thischar = (m_bit_buf >> 24) & 0xFF;

    if (thischar != 0xFF)
      stop_decoding(JPGD_NOT_JPEG);
  }

  // Find a start of frame (SOF) marker.
  void locate_sof_marker () {
    locate_soi_marker();

    int c = process_markers();

    switch (c)
    {
      case M_SOF2:
        m_progressive_flag = true;
        goto case;
      case M_SOF0:  /* baseline DCT */
      case M_SOF1:  /* extended sequential DCT */
        read_sof_marker();
        break;
      case M_SOF9:  /* Arithmitic coding */
        stop_decoding(JPGD_NO_ARITHMITIC_SUPPORT);
        break;
      default:
        stop_decoding(JPGD_UNSUPPORTED_MARKER);
        break;
    }
  }

  // Find a start of scan (SOS) marker.
  int locate_sos_marker () {
    int c;

    c = process_markers();

    if (c == M_EOI)
      return false;
    else if (c != M_SOS)
      stop_decoding(JPGD_UNEXPECTED_MARKER);

    read_sos_marker();

    return true;
  }

  // Reset everything to default/uninitialized state.
  void initit (JpegStreamReadFunc rfn, void* userData) 
  {
    m_pMem_blocks = null;
    m_error_code = JPGD_SUCCESS;
    m_ready_flag = false;
    m_image_x_size = m_image_y_size = 0;
    readfn = rfn;
    this.userData = userData;
    m_progressive_flag = false;

    memset(m_huff_ac.ptr, 0, m_huff_ac.sizeof);
    memset(m_huff_num.ptr, 0, m_huff_num.sizeof);
    memset(m_huff_val.ptr, 0, m_huff_val.sizeof);
    memset(m_quant.ptr, 0, m_quant.sizeof);

    m_scan_type = 0;
    m_comps_in_frame = 0;

    memset(m_comp_h_samp.ptr, 0, m_comp_h_samp.sizeof);
    memset(m_comp_v_samp.ptr, 0, m_comp_v_samp.sizeof);
    memset(m_comp_quant.ptr, 0, m_comp_quant.sizeof);
    memset(m_comp_ident.ptr, 0, m_comp_ident.sizeof);
    memset(m_comp_h_blocks.ptr, 0, m_comp_h_blocks.sizeof);
    memset(m_comp_v_blocks.ptr, 0, m_comp_v_blocks.sizeof);

    m_comps_in_scan = 0;
    memset(m_comp_list.ptr, 0, m_comp_list.sizeof);
    memset(m_comp_dc_tab.ptr, 0, m_comp_dc_tab.sizeof);
    memset(m_comp_ac_tab.ptr, 0, m_comp_ac_tab.sizeof);

    m_spectral_start = 0;
    m_spectral_end = 0;
    m_successive_low = 0;
    m_successive_high = 0;
    m_max_mcu_x_size = 0;
    m_max_mcu_y_size = 0;
    m_blocks_per_mcu = 0;
    m_max_blocks_per_row = 0;
    m_mcus_per_row = 0;
    m_mcus_per_col = 0;
    m_expanded_blocks_per_component = 0;
    m_expanded_blocks_per_mcu = 0;
    m_expanded_blocks_per_row = 0;
    m_freq_domain_chroma_upsample = false;

    memset(m_mcu_org.ptr, 0, m_mcu_org.sizeof);

    m_total_lines_left = 0;
    m_mcu_lines_left = 0;
    m_real_dest_bytes_per_scan_line = 0;
    m_dest_bytes_per_scan_line = 0;
    m_dest_bytes_per_pixel = 0;

    memset(m_pHuff_tabs.ptr, 0, m_pHuff_tabs.sizeof);

    memset(m_dc_coeffs.ptr, 0, m_dc_coeffs.sizeof);
    memset(m_ac_coeffs.ptr, 0, m_ac_coeffs.sizeof);
    memset(m_block_y_mcu.ptr, 0, m_block_y_mcu.sizeof);

    m_eob_run = 0;

    memset(m_block_y_mcu.ptr, 0, m_block_y_mcu.sizeof);

    m_pIn_buf_ofs = m_in_buf.ptr;
    m_in_buf_left = 0;
    m_eof_flag = false;
    m_tem_flag = 0;

    memset(m_in_buf_pad_start.ptr, 0, m_in_buf_pad_start.sizeof);
    memset(m_in_buf.ptr, 0, m_in_buf.sizeof);
    memset(m_in_buf_pad_end.ptr, 0, m_in_buf_pad_end.sizeof);

    m_restart_interval = 0;
    m_restarts_left    = 0;
    m_next_restart_num = 0;

    m_max_mcus_per_row = 0;
    m_max_blocks_per_mcu = 0;
    m_max_mcus_per_col = 0;

    memset(m_last_dc_val.ptr, 0, m_last_dc_val.sizeof);
    m_pMCU_coefficients = null;
    m_pSample_buf = null;

    m_total_bytes_read = 0;

    m_pScan_line_0 = null;
    m_pScan_line_1 = null;

    // Ready the input buffer.
    prep_in_buffer();

    // Prime the bit buffer.
    m_bits_left = 16;
    m_bit_buf = 0;

    get_bits(16);
    get_bits(16);

    for (int i = 0; i < JPGD_MAX_BLOCKS_PER_MCU; i++)
      m_mcu_block_max_zag.ptr[i] = 64;
  }

  enum SCALEBITS = 16;
  enum ONE_HALF = (cast(int) 1 << (SCALEBITS-1));
  enum FIX(float x) = (cast(int)((x) * (1L<<SCALEBITS) + 0.5f));

  // Create a few tables that allow us to quickly convert YCbCr to RGB.
  void create_look_ups () {
    for (int i = 0; i <= 255; i++)
    {
      int k = i - 128;
      m_crr.ptr[i] = ( FIX!(1.40200f)  * k + ONE_HALF) >> SCALEBITS;
      m_cbb.ptr[i] = ( FIX!(1.77200f)  * k + ONE_HALF) >> SCALEBITS;
      m_crg.ptr[i] = (-FIX!(0.71414f)) * k;
      m_cbg.ptr[i] = (-FIX!(0.34414f)) * k + ONE_HALF;
    }
  }

  // This method throws back into the stream any bytes that where read
  // into the bit buffer during initial marker scanning.
  void fix_in_buffer () {
    // In case any 0xFF's where pulled into the buffer during marker scanning.
    assert((m_bits_left & 7) == 0);

    if (m_bits_left == 16)
      stuff_char(cast(ubyte)(m_bit_buf & 0xFF));

    if (m_bits_left >= 8)
      stuff_char(cast(ubyte)((m_bit_buf >> 8) & 0xFF));

    stuff_char(cast(ubyte)((m_bit_buf >> 16) & 0xFF));
    stuff_char(cast(ubyte)((m_bit_buf >> 24) & 0xFF));

    m_bits_left = 16;
    get_bits_no_markers(16);
    get_bits_no_markers(16);
  }

  void transform_mcu (int mcu_row) {
    jpgd_block_t* pSrc_ptr = m_pMCU_coefficients;
    ubyte* pDst_ptr = m_pSample_buf + mcu_row * m_blocks_per_mcu * 64;

    for (int mcu_block = 0; mcu_block < m_blocks_per_mcu; mcu_block++)
    {
      idct(pSrc_ptr, pDst_ptr, m_mcu_block_max_zag.ptr[mcu_block]);
      pSrc_ptr += 64;
      pDst_ptr += 64;
    }
  }

  static immutable ubyte[64] s_max_rc = [
    17, 18, 34, 50, 50, 51, 52, 52, 52, 68, 84, 84, 84, 84, 85, 86, 86, 86, 86, 86,
    102, 118, 118, 118, 118, 118, 118, 119, 120, 120, 120, 120, 120, 120, 120, 136,
    136, 136, 136, 136, 136, 136, 136, 136, 136, 136, 136, 136, 136, 136, 136, 136,
    136, 136, 136, 136, 136, 136, 136, 136, 136, 136, 136, 136
  ];

  void transform_mcu_expand (int mcu_row) {
    jpgd_block_t* pSrc_ptr = m_pMCU_coefficients;
    ubyte* pDst_ptr = m_pSample_buf + mcu_row * m_expanded_blocks_per_mcu * 64;

    // Y IDCT
    int mcu_block;
    for (mcu_block = 0; mcu_block < m_expanded_blocks_per_component; mcu_block++)
    {
      idct(pSrc_ptr, pDst_ptr, m_mcu_block_max_zag.ptr[mcu_block]);
      pSrc_ptr += 64;
      pDst_ptr += 64;
    }

    // Chroma IDCT, with upsampling
    jpgd_block_t[64] temp_block;

    for (int i = 0; i < 2; i++)
    {
      DCT_Upsample.Matrix44 P, Q, R, S;

      assert(m_mcu_block_max_zag.ptr[mcu_block] >= 1);
      assert(m_mcu_block_max_zag.ptr[mcu_block] <= 64);

      int max_zag = m_mcu_block_max_zag.ptr[mcu_block++] - 1;
      if (max_zag <= 0) max_zag = 0; // should never happen, only here to shut up static analysis
      switch (s_max_rc.ptr[max_zag])
      {
      case 1*16+1:
        DCT_Upsample.P_Q!(1, 1).calc(P, Q, pSrc_ptr);
        DCT_Upsample.R_S!(1, 1).calc(R, S, pSrc_ptr);
        break;
      case 1*16+2:
        DCT_Upsample.P_Q!(1, 2).calc(P, Q, pSrc_ptr);
        DCT_Upsample.R_S!(1, 2).calc(R, S, pSrc_ptr);
        break;
      case 2*16+2:
        DCT_Upsample.P_Q!(2, 2).calc(P, Q, pSrc_ptr);
        DCT_Upsample.R_S!(2, 2).calc(R, S, pSrc_ptr);
        break;
      case 3*16+2:
        DCT_Upsample.P_Q!(3, 2).calc(P, Q, pSrc_ptr);
        DCT_Upsample.R_S!(3, 2).calc(R, S, pSrc_ptr);
        break;
      case 3*16+3:
        DCT_Upsample.P_Q!(3, 3).calc(P, Q, pSrc_ptr);
        DCT_Upsample.R_S!(3, 3).calc(R, S, pSrc_ptr);
        break;
      case 3*16+4:
        DCT_Upsample.P_Q!(3, 4).calc(P, Q, pSrc_ptr);
        DCT_Upsample.R_S!(3, 4).calc(R, S, pSrc_ptr);
        break;
      case 4*16+4:
        DCT_Upsample.P_Q!(4, 4).calc(P, Q, pSrc_ptr);
        DCT_Upsample.R_S!(4, 4).calc(R, S, pSrc_ptr);
        break;
      case 5*16+4:
        DCT_Upsample.P_Q!(5, 4).calc(P, Q, pSrc_ptr);
        DCT_Upsample.R_S!(5, 4).calc(R, S, pSrc_ptr);
        break;
      case 5*16+5:
        DCT_Upsample.P_Q!(5, 5).calc(P, Q, pSrc_ptr);
        DCT_Upsample.R_S!(5, 5).calc(R, S, pSrc_ptr);
        break;
      case 5*16+6:
        DCT_Upsample.P_Q!(5, 6).calc(P, Q, pSrc_ptr);
        DCT_Upsample.R_S!(5, 6).calc(R, S, pSrc_ptr);
        break;
      case 6*16+6:
        DCT_Upsample.P_Q!(6, 6).calc(P, Q, pSrc_ptr);
        DCT_Upsample.R_S!(6, 6).calc(R, S, pSrc_ptr);
        break;
      case 7*16+6:
        DCT_Upsample.P_Q!(7, 6).calc(P, Q, pSrc_ptr);
        DCT_Upsample.R_S!(7, 6).calc(R, S, pSrc_ptr);
        break;
      case 7*16+7:
        DCT_Upsample.P_Q!(7, 7).calc(P, Q, pSrc_ptr);
        DCT_Upsample.R_S!(7, 7).calc(R, S, pSrc_ptr);
        break;
      case 7*16+8:
        DCT_Upsample.P_Q!(7, 8).calc(P, Q, pSrc_ptr);
        DCT_Upsample.R_S!(7, 8).calc(R, S, pSrc_ptr);
        break;
      case 8*16+8:
        DCT_Upsample.P_Q!(8, 8).calc(P, Q, pSrc_ptr);
        DCT_Upsample.R_S!(8, 8).calc(R, S, pSrc_ptr);
        break;
      default:
        assert(false);
      }

      auto a = DCT_Upsample.Matrix44(P + Q);
      P -= Q;
      DCT_Upsample.Matrix44* b = &P;
      auto c = DCT_Upsample.Matrix44(R + S);
      R -= S;
      DCT_Upsample.Matrix44* d = &R;

      DCT_Upsample.Matrix44.add_and_store(temp_block.ptr, a, c);
      idct_4x4(temp_block.ptr, pDst_ptr);
      pDst_ptr += 64;

      DCT_Upsample.Matrix44.sub_and_store(temp_block.ptr, a, c);
      idct_4x4(temp_block.ptr, pDst_ptr);
      pDst_ptr += 64;

      DCT_Upsample.Matrix44.add_and_store(temp_block.ptr, *b, *d);
      idct_4x4(temp_block.ptr, pDst_ptr);
      pDst_ptr += 64;

      DCT_Upsample.Matrix44.sub_and_store(temp_block.ptr, *b, *d);
      idct_4x4(temp_block.ptr, pDst_ptr);
      pDst_ptr += 64;

      pSrc_ptr += 64;
    }
  }

  // Loads and dequantizes the next row of (already decoded) coefficients.
  // Progressive images only.
  void load_next_row () {
    int i;
    jpgd_block_t *p;
    jpgd_quant_t *q;
    int mcu_row, mcu_block, row_block = 0;
    int component_num, component_id;
    int[JPGD_MAX_COMPONENTS] block_x_mcu;

    memset(block_x_mcu.ptr, 0, JPGD_MAX_COMPONENTS * int.sizeof);

    for (mcu_row = 0; mcu_row < m_mcus_per_row; mcu_row++)
    {
      int block_x_mcu_ofs = 0, block_y_mcu_ofs = 0;

      for (mcu_block = 0; mcu_block < m_blocks_per_mcu; mcu_block++)
      {
        component_id = m_mcu_org.ptr[mcu_block];
        q = m_quant.ptr[m_comp_quant.ptr[component_id]];

        p = m_pMCU_coefficients + 64 * mcu_block;

        jpgd_block_t* pAC = coeff_buf_getp(m_ac_coeffs.ptr[component_id], block_x_mcu.ptr[component_id] + block_x_mcu_ofs, m_block_y_mcu.ptr[component_id] + block_y_mcu_ofs);
        jpgd_block_t* pDC = coeff_buf_getp(m_dc_coeffs.ptr[component_id], block_x_mcu.ptr[component_id] + block_x_mcu_ofs, m_block_y_mcu.ptr[component_id] + block_y_mcu_ofs);
        p[0] = pDC[0];
        memcpy(&p[1], &pAC[1], 63 * jpgd_block_t.sizeof);

        for (i = 63; i > 0; i--)
          if (p[g_ZAG[i]])
            break;

        m_mcu_block_max_zag.ptr[mcu_block] = i + 1;

        for ( ; i >= 0; i--)
          if (p[g_ZAG[i]])
            p[g_ZAG[i]] = cast(jpgd_block_t)(p[g_ZAG[i]] * q[i]);

        row_block++;

        if (m_comps_in_scan == 1)
          block_x_mcu.ptr[component_id]++;
        else
        {
          if (++block_x_mcu_ofs == m_comp_h_samp.ptr[component_id])
          {
            block_x_mcu_ofs = 0;

            if (++block_y_mcu_ofs == m_comp_v_samp.ptr[component_id])
            {
              block_y_mcu_ofs = 0;

              block_x_mcu.ptr[component_id] += m_comp_h_samp.ptr[component_id];
            }
          }
        }
      }

      if (m_freq_domain_chroma_upsample)
        transform_mcu_expand(mcu_row);
      else
        transform_mcu(mcu_row);
    }

    if (m_comps_in_scan == 1)
      m_block_y_mcu.ptr[m_comp_list.ptr[0]]++;
    else
    {
      for (component_num = 0; component_num < m_comps_in_scan; component_num++)
      {
        component_id = m_comp_list.ptr[component_num];

        m_block_y_mcu.ptr[component_id] += m_comp_v_samp.ptr[component_id];
      }
    }
  }

  // Restart interval processing.
  void process_restart () {
    int i;
    int c = 0;

    // Align to a byte boundry
    // FIXME: Is this really necessary? get_bits_no_markers() never reads in markers!
    //get_bits_no_markers(m_bits_left & 7);

    // Let's scan a little bit to find the marker, but not _too_ far.
    // 1536 is a "fudge factor" that determines how much to scan.
    for (i = 1536; i > 0; i--)
      if (get_char() == 0xFF)
        break;

    if (i == 0)
      stop_decoding(JPGD_BAD_RESTART_MARKER);

    for ( ; i > 0; i--)
      if ((c = get_char()) != 0xFF)
        break;

    if (i == 0)
      stop_decoding(JPGD_BAD_RESTART_MARKER);

    // Is it the expected marker? If not, something bad happened.
    if (c != (m_next_restart_num + M_RST0))
      stop_decoding(JPGD_BAD_RESTART_MARKER);

    // Reset each component's DC prediction values.
    memset(&m_last_dc_val, 0, m_comps_in_frame * uint.sizeof);

    m_eob_run = 0;

    m_restarts_left = m_restart_interval;

    m_next_restart_num = (m_next_restart_num + 1) & 7;

    // Get the bit buffer going again...

    m_bits_left = 16;
    get_bits_no_markers(16);
    get_bits_no_markers(16);
  }

  // Decodes and dequantizes the next row of coefficients.
  void decode_next_row () {
    int row_block = 0;

    for (int mcu_row = 0; mcu_row < m_mcus_per_row; mcu_row++)
    {
      if ((m_restart_interval) && (m_restarts_left == 0))
        process_restart();

      jpgd_block_t* p = m_pMCU_coefficients;
      for (int mcu_block = 0; mcu_block < m_blocks_per_mcu; mcu_block++, p += 64)
      {
        int component_id = m_mcu_org.ptr[mcu_block];
        jpgd_quant_t* q = m_quant.ptr[m_comp_quant.ptr[component_id]];

        int r, s;
        s = huff_decode(m_pHuff_tabs.ptr[m_comp_dc_tab.ptr[component_id]], r);
        s = JPGD_HUFF_EXTEND(r, s);

        m_last_dc_val.ptr[component_id] = (s += m_last_dc_val.ptr[component_id]);

        p[0] = cast(jpgd_block_t)(s * q[0]);

        int prev_num_set = m_mcu_block_max_zag.ptr[mcu_block];

        huff_tables *pH = m_pHuff_tabs.ptr[m_comp_ac_tab.ptr[component_id]];

        int k;
        for (k = 1; k < 64; k++)
        {
          int extra_bits;
          s = huff_decode(pH, extra_bits);

          r = s >> 4;
          s &= 15;

          if (s)
          {
            if (r)
            {
              if ((k + r) > 63)
                stop_decoding(JPGD_DECODE_ERROR);

              if (k < prev_num_set)
              {
                int n = JPGD_MIN(r, prev_num_set - k);
                int kt = k;
                while (n--)
                  p[g_ZAG[kt++]] = 0;
              }

              k += r;
            }

            s = JPGD_HUFF_EXTEND(extra_bits, s);

            assert(k < 64);

            p[g_ZAG[k]] = cast(jpgd_block_t)( s * q[k] ); // dequantize
          }
          else
          {
            if (r == 15)
            {
              if ((k + 16) > 64)
                stop_decoding(JPGD_DECODE_ERROR);

              if (k < prev_num_set)
              {
                int n = JPGD_MIN(16, prev_num_set - k);
                int kt = k;
                while (n--)
                {
                  assert(kt <= 63);
                  p[g_ZAG[kt++]] = 0;
                }
              }

              k += 16 - 1; // - 1 because the loop counter is k
              assert(p[g_ZAG[k]] == 0);
            }
            else
              break;
          }
        }

        if (k < prev_num_set)
        {
          int kt = k;
          while (kt < prev_num_set)
            p[g_ZAG[kt++]] = 0;
        }

        m_mcu_block_max_zag.ptr[mcu_block] = k;

        row_block++;
      }

      if (m_freq_domain_chroma_upsample)
        transform_mcu_expand(mcu_row);
      else
        transform_mcu(mcu_row);

      m_restarts_left--;
    }
  }

  // YCbCr H1V1 (1x1:1:1, 3 m_blocks per MCU) to RGB
  void H1V1Convert () {
    int row = m_max_mcu_y_size - m_mcu_lines_left;
    ubyte *d = m_pScan_line_0;
    ubyte *s = m_pSample_buf + row * 8;

    for (int i = m_max_mcus_per_row; i > 0; i--)
    {
      for (int j = 0; j < 8; j++)
      {
        int y = s[j];
        int cb = s[64+j];
        int cr = s[128+j];


        __m128i zero = _mm_setzero_si128();
        __m128i A = _mm_setr_epi32(y + m_crr.ptr[cr], 
                                   y + ((m_crg.ptr[cr] + m_cbg.ptr[cb]) >> 16),
                                   y + m_cbb.ptr[cb],
                                   255);
        A = _mm_packs_epi32(A, zero);
        A = _mm_packus_epi16(A, zero);
        _mm_storeu_si32(&d[0], A);
        d += 4;
      }

      s += 64*3;
    }
  }

  // YCbCr H2V1 (2x1:1:1, 4 m_blocks per MCU) to RGB
  void H2V1Convert () 
  {
    int row = m_max_mcu_y_size - m_mcu_lines_left;
    ubyte *d0 = m_pScan_line_0;
    ubyte *y = m_pSample_buf + row * 8;
    ubyte *c = m_pSample_buf + 2*64 + row * 8;

    for (int i = m_max_mcus_per_row; i > 0; i--)
    {
      for (int l = 0; l < 2; l++)
      {
        for (int j = 0; j < 4; j++)
        {
          int cb = c[0];
          int cr = c[64];

          int rc = m_crr.ptr[cr];
          int gc = ((m_crg.ptr[cr] + m_cbg.ptr[cb]) >> 16);
          int bc = m_cbb.ptr[cb];

          int yy = y[j<<1];
          d0[0] = clamp(yy+rc);
          d0[1] = clamp(yy+gc);
          d0[2] = clamp(yy+bc);
          d0[3] = 255;

          yy = y[(j<<1)+1];
          d0[4] = clamp(yy+rc);
          d0[5] = clamp(yy+gc);
          d0[6] = clamp(yy+bc);
          d0[7] = 255;

          d0 += 8;

          c++;
        }
        y += 64;
      }

      y += 64*4 - 64*2;
      c += 64*4 - 8;
    }
  }

  // YCbCr H2V1 (1x2:1:1, 4 m_blocks per MCU) to RGB
  void H1V2Convert () {
    int row = m_max_mcu_y_size - m_mcu_lines_left;
    ubyte *d0 = m_pScan_line_0;
    ubyte *d1 = m_pScan_line_1;
    ubyte *y;
    ubyte *c;

    if (row < 8)
      y = m_pSample_buf + row * 8;
    else
      y = m_pSample_buf + 64*1 + (row & 7) * 8;

    c = m_pSample_buf + 64*2 + (row >> 1) * 8;

    for (int i = m_max_mcus_per_row; i > 0; i--)
    {
      for (int j = 0; j < 8; j++)
      {
        int cb = c[0+j];
        int cr = c[64+j];

        int rc = m_crr.ptr[cr];
        int gc = ((m_crg.ptr[cr] + m_cbg.ptr[cb]) >> 16);
        int bc = m_cbb.ptr[cb];

        int yy = y[j];
        d0[0] = clamp(yy+rc);
        d0[1] = clamp(yy+gc);
        d0[2] = clamp(yy+bc);
        d0[3] = 255;

        yy = y[8+j];
        d1[0] = clamp(yy+rc);
        d1[1] = clamp(yy+gc);
        d1[2] = clamp(yy+bc);
        d1[3] = 255;

        d0 += 4;
        d1 += 4;
      }

      y += 64*4;
      c += 64*4;
    }
  }

  // YCbCr H2V2 (2x2:1:1, 6 m_blocks per MCU) to RGB
  void H2V2Convert () {
    int row = m_max_mcu_y_size - m_mcu_lines_left;
    ubyte *d0 = m_pScan_line_0;
    ubyte *d1 = m_pScan_line_1;
    ubyte *y;
    ubyte *c;

    if (row < 8)
      y = m_pSample_buf + row * 8;
    else
      y = m_pSample_buf + 64*2 + (row & 7) * 8;

    c = m_pSample_buf + 64*4 + (row >> 1) * 8;

    for (int i = m_max_mcus_per_row; i > 0; i--)
    {
      for (int l = 0; l < 2; l++)
      {
        for (int j = 0; j < 8; j += 2)
        {
          int cb = c[0];
          int cr = c[64];

          int rc = m_crr.ptr[cr];
          int gc = ((m_crg.ptr[cr] + m_cbg.ptr[cb]) >> 16);
          int bc = m_cbb.ptr[cb];

          int yy = y[j];
          d0[0] = clamp(yy+rc);
          d0[1] = clamp(yy+gc);
          d0[2] = clamp(yy+bc);
          d0[3] = 255;

          yy = y[j+1];
          d0[4] = clamp(yy+rc);
          d0[5] = clamp(yy+gc);
          d0[6] = clamp(yy+bc);
          d0[7] = 255;

          yy = y[j+8];
          d1[0] = clamp(yy+rc);
          d1[1] = clamp(yy+gc);
          d1[2] = clamp(yy+bc);
          d1[3] = 255;

          yy = y[j+8+1];
          d1[4] = clamp(yy+rc);
          d1[5] = clamp(yy+gc);
          d1[6] = clamp(yy+bc);
          d1[7] = 255;

          d0 += 8;
          d1 += 8;

          c++;
        }
        y += 64;
      }

      y += 64*6 - 64*2;
      c += 64*6 - 8;
    }
  }

  // Y (1 block per MCU) to 8-bit grayscale
  void gray_convert () {
    int row = m_max_mcu_y_size - m_mcu_lines_left;
    ubyte *d = m_pScan_line_0;
    ubyte *s = m_pSample_buf + row * 8;

    for (int i = m_max_mcus_per_row; i > 0; i--)
    {
      *cast(uint*)d = *cast(uint*)s;
      *cast(uint*)(&d[4]) = *cast(uint*)(&s[4]);

      s += 64;
      d += 8;
    }
  }


  void expanded_convert () 
  {
    int row = m_max_mcu_y_size - m_mcu_lines_left;

    ubyte* Py = m_pSample_buf + (row / 8) * 64 * m_comp_h_samp.ptr[0] + (row & 7) * 8;

    ubyte* d = m_pScan_line_0;

    for (int i = m_max_mcus_per_row; i > 0; i--)
    {
      for (int k = 0; k < m_max_mcu_x_size; k += 8)
      {
        immutable int Y_ofs = k * 8;
        immutable int Cb_ofs = Y_ofs + 64 * m_expanded_blocks_per_component;
        immutable int Cr_ofs = Y_ofs + 64 * m_expanded_blocks_per_component * 2;
        for (int j = 0; j < 8; j++)
        {
          int y = Py[Y_ofs + j];
          int cb = Py[Cb_ofs + j];
          int cr = Py[Cr_ofs + j];

          __m128i zero = _mm_setzero_si128();
          __m128i A = _mm_setr_epi32(y + m_crr.ptr[cr], 
                                     y + ((m_crg.ptr[cr] + m_cbg.ptr[cb]) >> 16),
                                     y + m_cbb.ptr[cb],
                                     255);
          A = _mm_packs_epi32(A, zero);
          A = _mm_packus_epi16(A, zero);
          _mm_storeu_si32(&d[0], A);
          d += 4;
        }
      }

      Py += 64 * m_expanded_blocks_per_mcu;
    }
  }

  // Find end of image (EOI) marker, so we can return to the user the exact size of the input stream.
  void find_eoi () {
    if (!m_progressive_flag)
    {
      // Attempt to read the EOI marker.
      //get_bits_no_markers(m_bits_left & 7);

      // Prime the bit buffer
      m_bits_left = 16;
      get_bits(16);
      get_bits(16);

      // The next marker _should_ be EOI
      process_markers();
    }

    m_total_bytes_read -= m_in_buf_left;
  }

  // Creates the tables needed for efficient Huffman decoding.
  void make_huff_table (int index, huff_tables *pH) {
    int p, i, l, si;
    ubyte[257] huffsize;
    uint[257] huffcode;
    uint code;
    uint subtree;
    int code_size;
    int lastp;
    int nextfreeentry;
    int currententry;

    pH.ac_table = m_huff_ac.ptr[index] != 0;

    p = 0;

    for (l = 1; l <= 16; l++)
    {
      for (i = 1; i <= m_huff_num.ptr[index][l]; i++)
        huffsize.ptr[p++] = cast(ubyte)(l);
    }

    huffsize.ptr[p] = 0;

    lastp = p;

    code = 0;
    si = huffsize.ptr[0];
    p = 0;

    while (huffsize.ptr[p])
    {
      while (huffsize.ptr[p] == si)
      {
        huffcode.ptr[p++] = code;
        code++;
      }

      code <<= 1;
      si++;
    }

    memset(pH.look_up.ptr, 0, pH.look_up.sizeof);
    memset(pH.look_up2.ptr, 0, pH.look_up2.sizeof);
    memset(pH.tree.ptr, 0, pH.tree.sizeof);
    memset(pH.code_size.ptr, 0, pH.code_size.sizeof);

    nextfreeentry = -1;

    p = 0;

    while (p < lastp)
    {
      i = m_huff_val.ptr[index][p];
      code = huffcode.ptr[p];
      code_size = huffsize.ptr[p];

      pH.code_size.ptr[i] = cast(ubyte)(code_size);

      if (code_size <= 8)
      {
        code <<= (8 - code_size);

        for (l = 1 << (8 - code_size); l > 0; l--)
        {
          assert(i < 256);

          pH.look_up.ptr[code] = i;

          bool has_extrabits = false;
          int extra_bits = 0;
          int num_extra_bits = i & 15;

          int bits_to_fetch = code_size;
          if (num_extra_bits)
          {
            int total_codesize = code_size + num_extra_bits;
            if (total_codesize <= 8)
            {
              has_extrabits = true;
              extra_bits = ((1 << num_extra_bits) - 1) & (code >> (8 - total_codesize));
              assert(extra_bits <= 0x7FFF);
              bits_to_fetch += num_extra_bits;
            }
          }

          if (!has_extrabits)
            pH.look_up2.ptr[code] = i | (bits_to_fetch << 8);
          else
            pH.look_up2.ptr[code] = i | 0x8000 | (extra_bits << 16) | (bits_to_fetch << 8);

          code++;
        }
      }
      else
      {
        subtree = (code >> (code_size - 8)) & 0xFF;

        currententry = pH.look_up.ptr[subtree];

        if (currententry == 0)
        {
          pH.look_up.ptr[subtree] = currententry = nextfreeentry;
          pH.look_up2.ptr[subtree] = currententry = nextfreeentry;

          nextfreeentry -= 2;
        }

        code <<= (16 - (code_size - 8));

        for (l = code_size; l > 9; l--)
        {
          if ((code & 0x8000) == 0)
            currententry--;

          if (pH.tree.ptr[-currententry - 1] == 0)
          {
            pH.tree.ptr[-currententry - 1] = nextfreeentry;

            currententry = nextfreeentry;

            nextfreeentry -= 2;
          }
          else
            currententry = pH.tree.ptr[-currententry - 1];

          code <<= 1;
        }

        if ((code & 0x8000) == 0)
          currententry--;

        pH.tree.ptr[-currententry - 1] = i;
      }

      p++;
    }
  }

  // Verifies the quantization tables needed for this scan are available.
  void check_quant_tables () {
    for (int i = 0; i < m_comps_in_scan; i++)
      if (m_quant.ptr[m_comp_quant.ptr[m_comp_list.ptr[i]]] == null)
        stop_decoding(JPGD_UNDEFINED_QUANT_TABLE);
  }

  // Verifies that all the Huffman tables needed for this scan are available.
  void check_huff_tables () {
    for (int i = 0; i < m_comps_in_scan; i++)
    {
      if ((m_spectral_start == 0) && (m_huff_num.ptr[m_comp_dc_tab.ptr[m_comp_list.ptr[i]]] == null))
        stop_decoding(JPGD_UNDEFINED_HUFF_TABLE);

      if ((m_spectral_end > 0) && (m_huff_num.ptr[m_comp_ac_tab.ptr[m_comp_list.ptr[i]]] == null))
        stop_decoding(JPGD_UNDEFINED_HUFF_TABLE);
    }

    for (int i = 0; i < JPGD_MAX_HUFF_TABLES; i++)
      if (m_huff_num.ptr[i])
      {
        if (!m_pHuff_tabs.ptr[i])
          m_pHuff_tabs.ptr[i] = cast(huff_tables*)alloc(huff_tables.sizeof);

        make_huff_table(i, m_pHuff_tabs.ptr[i]);
      }
  }

  // Determines the component order inside each MCU.
  // Also calcs how many MCU's are on each row, etc.
  void calc_mcu_block_order () {
    int component_num, component_id;
    int max_h_samp = 0, max_v_samp = 0;

    for (component_id = 0; component_id < m_comps_in_frame; component_id++)
    {
      if (m_comp_h_samp.ptr[component_id] > max_h_samp)
        max_h_samp = m_comp_h_samp.ptr[component_id];

      if (m_comp_v_samp.ptr[component_id] > max_v_samp)
        max_v_samp = m_comp_v_samp.ptr[component_id];
    }

    for (component_id = 0; component_id < m_comps_in_frame; component_id++)
    {
      m_comp_h_blocks.ptr[component_id] = ((((m_image_x_size * m_comp_h_samp.ptr[component_id]) + (max_h_samp - 1)) / max_h_samp) + 7) / 8;
      m_comp_v_blocks.ptr[component_id] = ((((m_image_y_size * m_comp_v_samp.ptr[component_id]) + (max_v_samp - 1)) / max_v_samp) + 7) / 8;
    }

    if (m_comps_in_scan == 1)
    {
      m_mcus_per_row = m_comp_h_blocks.ptr[m_comp_list.ptr[0]];
      m_mcus_per_col = m_comp_v_blocks.ptr[m_comp_list.ptr[0]];
    }
    else
    {
      m_mcus_per_row = (((m_image_x_size + 7) / 8) + (max_h_samp - 1)) / max_h_samp;
      m_mcus_per_col = (((m_image_y_size + 7) / 8) + (max_v_samp - 1)) / max_v_samp;
    }

    if (m_comps_in_scan == 1)
    {
      m_mcu_org.ptr[0] = m_comp_list.ptr[0];

      m_blocks_per_mcu = 1;
    }
    else
    {
      m_blocks_per_mcu = 0;

      for (component_num = 0; component_num < m_comps_in_scan; component_num++)
      {
        int num_blocks;

        component_id = m_comp_list.ptr[component_num];

        num_blocks = m_comp_h_samp.ptr[component_id] * m_comp_v_samp.ptr[component_id];

        while (num_blocks--)
          m_mcu_org.ptr[m_blocks_per_mcu++] = component_id;
      }
    }
  }

  // Starts a new scan.
  int init_scan () {
    if (!locate_sos_marker())
      return false;

    calc_mcu_block_order();

    check_huff_tables();

    check_quant_tables();

    memset(m_last_dc_val.ptr, 0, m_comps_in_frame * uint.sizeof);

    m_eob_run = 0;

    if (m_restart_interval)
    {
      m_restarts_left = m_restart_interval;
      m_next_restart_num = 0;
    }

    fix_in_buffer();

    return true;
  }

  // Starts a frame. Determines if the number of components or sampling factors
  // are supported.
  void init_frame () {
    int i;

    if (m_comps_in_frame == 1)
    {
      if ((m_comp_h_samp.ptr[0] != 1) || (m_comp_v_samp.ptr[0] != 1))
        stop_decoding(JPGD_UNSUPPORTED_SAMP_FACTORS);

      m_scan_type = JPGD_GRAYSCALE;
      m_max_blocks_per_mcu = 1;
      m_max_mcu_x_size = 8;
      m_max_mcu_y_size = 8;
    }
    else if (m_comps_in_frame == 3)
    {
      if ( ((m_comp_h_samp.ptr[1] != 1) || (m_comp_v_samp.ptr[1] != 1)) ||
           ((m_comp_h_samp.ptr[2] != 1) || (m_comp_v_samp.ptr[2] != 1)) )
        stop_decoding(JPGD_UNSUPPORTED_SAMP_FACTORS);

      if ((m_comp_h_samp.ptr[0] == 1) && (m_comp_v_samp.ptr[0] == 1))
      {
        m_scan_type = JPGD_YH1V1;

        m_max_blocks_per_mcu = 3;
        m_max_mcu_x_size = 8;
        m_max_mcu_y_size = 8;
      }
      else if ((m_comp_h_samp.ptr[0] == 2) && (m_comp_v_samp.ptr[0] == 1))
      {
        m_scan_type = JPGD_YH2V1;
        m_max_blocks_per_mcu = 4;
        m_max_mcu_x_size = 16;
        m_max_mcu_y_size = 8;
      }
      else if ((m_comp_h_samp.ptr[0] == 1) && (m_comp_v_samp.ptr[0] == 2))
      {
        m_scan_type = JPGD_YH1V2;
        m_max_blocks_per_mcu = 4;
        m_max_mcu_x_size = 8;
        m_max_mcu_y_size = 16;
      }
      else if ((m_comp_h_samp.ptr[0] == 2) && (m_comp_v_samp.ptr[0] == 2))
      {
        m_scan_type = JPGD_YH2V2;
        m_max_blocks_per_mcu = 6;
        m_max_mcu_x_size = 16;
        m_max_mcu_y_size = 16;
      }
      else
        stop_decoding(JPGD_UNSUPPORTED_SAMP_FACTORS);
    }
    else
      stop_decoding(JPGD_UNSUPPORTED_COLORSPACE);

    m_max_mcus_per_row = (m_image_x_size + (m_max_mcu_x_size - 1)) / m_max_mcu_x_size;
    m_max_mcus_per_col = (m_image_y_size + (m_max_mcu_y_size - 1)) / m_max_mcu_y_size;

    // These values are for the *destination* pixels: after conversion.
    if (m_scan_type == JPGD_GRAYSCALE)
      m_dest_bytes_per_pixel = 1;
    else
      m_dest_bytes_per_pixel = 4;

    m_dest_bytes_per_scan_line = ((m_image_x_size + 15) & 0xFFF0) * m_dest_bytes_per_pixel;

    m_real_dest_bytes_per_scan_line = (m_image_x_size * m_dest_bytes_per_pixel);

    // Initialize two scan line buffers.
    m_pScan_line_0 = cast(ubyte*)alloc(m_dest_bytes_per_scan_line, true);
    if ((m_scan_type == JPGD_YH1V2) || (m_scan_type == JPGD_YH2V2))
      m_pScan_line_1 = cast(ubyte*)alloc(m_dest_bytes_per_scan_line, true);

    m_max_blocks_per_row = m_max_mcus_per_row * m_max_blocks_per_mcu;

    // Should never happen
    if (m_max_blocks_per_row > JPGD_MAX_BLOCKS_PER_ROW)
      stop_decoding(JPGD_ASSERTION_ERROR);

    // Allocate the coefficient buffer, enough for one MCU
    m_pMCU_coefficients = cast(jpgd_block_t*)alloc(m_max_blocks_per_mcu * 64 * jpgd_block_t.sizeof);

    for (i = 0; i < m_max_blocks_per_mcu; i++)
      m_mcu_block_max_zag.ptr[i] = 64;

    m_expanded_blocks_per_component = m_comp_h_samp.ptr[0] * m_comp_v_samp.ptr[0];
    m_expanded_blocks_per_mcu = m_expanded_blocks_per_component * m_comps_in_frame;
    m_expanded_blocks_per_row = m_max_mcus_per_row * m_expanded_blocks_per_mcu;
    // Freq. domain chroma upsampling is only supported for H2V2 subsampling factor (the most common one I've seen).
    m_freq_domain_chroma_upsample = false;
    version(JPGD_SUPPORT_FREQ_DOMAIN_UPSAMPLING) {
      m_freq_domain_chroma_upsample = (m_expanded_blocks_per_mcu == 4*3);
    }

    if (m_freq_domain_chroma_upsample)
      m_pSample_buf = cast(ubyte*)alloc(m_expanded_blocks_per_row * 64);
    else
      m_pSample_buf = cast(ubyte*)alloc(m_max_blocks_per_row * 64);

    m_total_lines_left = m_image_y_size;

    m_mcu_lines_left = 0;

    create_look_ups();
  }

  // The coeff_buf series of methods originally stored the coefficients
  // into a "virtual" file which was located in EMS, XMS, or a disk file. A cache
  // was used to make this process more efficient. Now, we can store the entire
  // thing in RAM.
  coeff_buf* coeff_buf_open(int block_num_x, int block_num_y, int block_len_x, int block_len_y) {
    coeff_buf* cb = cast(coeff_buf*)alloc(coeff_buf.sizeof);

    cb.block_num_x = block_num_x;
    cb.block_num_y = block_num_y;
    cb.block_len_x = block_len_x;
    cb.block_len_y = block_len_y;
    cb.block_size = (block_len_x * block_len_y) * cast(int)(jpgd_block_t.sizeof);
    cb.pData = cast(ubyte*)alloc(cb.block_size * block_num_x * block_num_y, true);
    return cb;
  }

  jpgd_block_t* coeff_buf_getp (coeff_buf *cb, int block_x, int block_y) {
    assert((block_x < cb.block_num_x) && (block_y < cb.block_num_y));
    return cast(jpgd_block_t*)(cb.pData + block_x * cb.block_size + block_y * (cb.block_size * cb.block_num_x));
  }

  // The following methods decode the various types of m_blocks encountered
  // in progressively encoded images.
  static void decode_block_dc_first (ref jpeg_decoder pD, int component_id, int block_x, int block_y) {
    int s, r;
    jpgd_block_t *p = pD.coeff_buf_getp(pD.m_dc_coeffs.ptr[component_id], block_x, block_y);

    if ((s = pD.huff_decode(pD.m_pHuff_tabs.ptr[pD.m_comp_dc_tab.ptr[component_id]])) != 0)
    {
      r = pD.get_bits_no_markers(s);
      s = JPGD_HUFF_EXTEND(r, s);
    }

    pD.m_last_dc_val.ptr[component_id] = (s += pD.m_last_dc_val.ptr[component_id]);

    p[0] = cast(jpgd_block_t)(s << pD.m_successive_low);
  }

  static void decode_block_dc_refine (ref jpeg_decoder pD, int component_id, int block_x, int block_y) {
    if (pD.get_bits_no_markers(1))
    {
      jpgd_block_t *p = pD.coeff_buf_getp(pD.m_dc_coeffs.ptr[component_id], block_x, block_y);

      p[0] |= (1 << pD.m_successive_low);
    }
  }

  static void decode_block_ac_first (ref jpeg_decoder pD, int component_id, int block_x, int block_y) {
    int k, s, r;

    if (pD.m_eob_run)
    {
      pD.m_eob_run--;
      return;
    }

    jpgd_block_t *p = pD.coeff_buf_getp(pD.m_ac_coeffs.ptr[component_id], block_x, block_y);

    for (k = pD.m_spectral_start; k <= pD.m_spectral_end; k++)
    {
      s = pD.huff_decode(pD.m_pHuff_tabs.ptr[pD.m_comp_ac_tab.ptr[component_id]]);

      r = s >> 4;
      s &= 15;

      if (s)
      {
        if ((k += r) > 63)
          pD.stop_decoding(JPGD_DECODE_ERROR);

        r = pD.get_bits_no_markers(s);
        s = JPGD_HUFF_EXTEND(r, s);

        p[g_ZAG[k]] = cast(jpgd_block_t)(s << pD.m_successive_low);
      }
      else
      {
        if (r == 15)
        {
          if ((k += 15) > 63)
            pD.stop_decoding(JPGD_DECODE_ERROR);
        }
        else
        {
          pD.m_eob_run = 1 << r;

          if (r)
            pD.m_eob_run += pD.get_bits_no_markers(r);

          pD.m_eob_run--;

          break;
        }
      }
    }
  }

  static void decode_block_ac_refine (ref jpeg_decoder pD, int component_id, int block_x, int block_y) {
    int s, k, r;
    int p1 = 1 << pD.m_successive_low;
    int m1 = (-1) << pD.m_successive_low;
    jpgd_block_t *p = pD.coeff_buf_getp(pD.m_ac_coeffs.ptr[component_id], block_x, block_y);

    assert(pD.m_spectral_end <= 63);

    k = pD.m_spectral_start;

    if (pD.m_eob_run == 0)
    {
      for ( ; k <= pD.m_spectral_end; k++)
      {
        s = pD.huff_decode(pD.m_pHuff_tabs.ptr[pD.m_comp_ac_tab.ptr[component_id]]);

        r = s >> 4;
        s &= 15;

        if (s)
        {
          if (s != 1)
            pD.stop_decoding(JPGD_DECODE_ERROR);

          if (pD.get_bits_no_markers(1))
            s = p1;
          else
            s = m1;
        }
        else
        {
          if (r != 15)
          {
            pD.m_eob_run = 1 << r;

            if (r)
              pD.m_eob_run += pD.get_bits_no_markers(r);

            break;
          }
        }

        do
        {
          jpgd_block_t *this_coef = p + g_ZAG[k & 63];

          if (*this_coef != 0)
          {
            if (pD.get_bits_no_markers(1))
            {
              if ((*this_coef & p1) == 0)
              {
                if (*this_coef >= 0)
                  *this_coef = cast(jpgd_block_t)(*this_coef + p1);
                else
                  *this_coef = cast(jpgd_block_t)(*this_coef + m1);
              }
            }
          }
          else
          {
            if (--r < 0)
              break;
          }

          k++;

        } while (k <= pD.m_spectral_end);

        if ((s) && (k < 64))
        {
          p[g_ZAG[k]] = cast(jpgd_block_t)(s);
        }
      }
    }

    if (pD.m_eob_run > 0)
    {
      for ( ; k <= pD.m_spectral_end; k++)
      {
        jpgd_block_t *this_coef = p + g_ZAG[k & 63]; // logical AND to shut up static code analysis

        if (*this_coef != 0)
        {
          if (pD.get_bits_no_markers(1))
          {
            if ((*this_coef & p1) == 0)
            {
              if (*this_coef >= 0)
                *this_coef = cast(jpgd_block_t)(*this_coef + p1);
              else
                *this_coef = cast(jpgd_block_t)(*this_coef + m1);
            }
          }
        }
      }

      pD.m_eob_run--;
    }
  }

  // Decode a scan in a progressively encoded image.
  void decode_scan (pDecode_block_func decode_block_func) {
    int mcu_row, mcu_col, mcu_block;
    int[JPGD_MAX_COMPONENTS] block_x_mcu;
    int[JPGD_MAX_COMPONENTS] m_block_y_mcu;

    memset(m_block_y_mcu.ptr, 0, m_block_y_mcu.sizeof);

    for (mcu_col = 0; mcu_col < m_mcus_per_col; mcu_col++)
    {
      int component_num, component_id;

      memset(block_x_mcu.ptr, 0, block_x_mcu.sizeof);

      for (mcu_row = 0; mcu_row < m_mcus_per_row; mcu_row++)
      {
        int block_x_mcu_ofs = 0, block_y_mcu_ofs = 0;

        if ((m_restart_interval) && (m_restarts_left == 0))
          process_restart();

        for (mcu_block = 0; mcu_block < m_blocks_per_mcu; mcu_block++)
        {
          component_id = m_mcu_org.ptr[mcu_block];

          decode_block_func(this, component_id, block_x_mcu.ptr[component_id] + block_x_mcu_ofs, m_block_y_mcu.ptr[component_id] + block_y_mcu_ofs);

          if (m_comps_in_scan == 1)
            block_x_mcu.ptr[component_id]++;
          else
          {
            if (++block_x_mcu_ofs == m_comp_h_samp.ptr[component_id])
            {
              block_x_mcu_ofs = 0;

              if (++block_y_mcu_ofs == m_comp_v_samp.ptr[component_id])
              {
                block_y_mcu_ofs = 0;
                block_x_mcu.ptr[component_id] += m_comp_h_samp.ptr[component_id];
              }
            }
          }
        }

        m_restarts_left--;
      }

      if (m_comps_in_scan == 1)
        m_block_y_mcu.ptr[m_comp_list.ptr[0]]++;
      else
      {
        for (component_num = 0; component_num < m_comps_in_scan; component_num++)
        {
          component_id = m_comp_list.ptr[component_num];
          m_block_y_mcu.ptr[component_id] += m_comp_v_samp.ptr[component_id];
        }
      }
    }
  }

  // Decode a progressively encoded image.
  void init_progressive () {
    int i;

    if (m_comps_in_frame == 4)
      stop_decoding(JPGD_UNSUPPORTED_COLORSPACE);

    // Allocate the coefficient buffers.
    for (i = 0; i < m_comps_in_frame; i++)
    {
      m_dc_coeffs.ptr[i] = coeff_buf_open(m_max_mcus_per_row * m_comp_h_samp.ptr[i], m_max_mcus_per_col * m_comp_v_samp.ptr[i], 1, 1);
      m_ac_coeffs.ptr[i] = coeff_buf_open(m_max_mcus_per_row * m_comp_h_samp.ptr[i], m_max_mcus_per_col * m_comp_v_samp.ptr[i], 8, 8);
    }

    for ( ; ; )
    {
      int dc_only_scan, refinement_scan;
      pDecode_block_func decode_block_func;

      if (!init_scan())
        break;

      dc_only_scan = (m_spectral_start == 0);
      refinement_scan = (m_successive_high != 0);

      if ((m_spectral_start > m_spectral_end) || (m_spectral_end > 63))
        stop_decoding(JPGD_BAD_SOS_SPECTRAL);

      if (dc_only_scan)
      {
        if (m_spectral_end)
          stop_decoding(JPGD_BAD_SOS_SPECTRAL);
      }
      else if (m_comps_in_scan != 1)  /* AC scans can only contain one component */
        stop_decoding(JPGD_BAD_SOS_SPECTRAL);

      if ((refinement_scan) && (m_successive_low != m_successive_high - 1))
        stop_decoding(JPGD_BAD_SOS_SUCCESSIVE);

      if (dc_only_scan)
      {
        if (refinement_scan)
          decode_block_func = &decode_block_dc_refine;
        else
          decode_block_func = &decode_block_dc_first;
      }
      else
      {
        if (refinement_scan)
          decode_block_func = &decode_block_ac_refine;
        else
          decode_block_func = &decode_block_ac_first;
      }

      decode_scan(decode_block_func);

      m_bits_left = 16;
      get_bits(16);
      get_bits(16);
    }

    m_comps_in_scan = m_comps_in_frame;

    for (i = 0; i < m_comps_in_frame; i++)
      m_comp_list.ptr[i] = i;

    calc_mcu_block_order();
  }

  void init_sequential () {
    if (!init_scan())
      stop_decoding(JPGD_UNEXPECTED_MARKER);
  }

  void decode_start () {
    init_frame();

    if (m_progressive_flag)
      init_progressive();
    else
      init_sequential();
  }

  void decode_init (JpegStreamReadFunc rfn, void* userData) {
    initit(rfn, userData);
    locate_sof_marker();
  }
}

// ////////////////////////////////////////////////////////////////////////// //
/// decompress JPEG image, what else?
/// you can specify required color components in `req_comps` (3 for RGB or 4 for RGBA), or leave it as is to use image value.
/// Returns pixelAspectRatio and dotsPerInchY, -1 if not available.
public ubyte[] decompress_jpeg_image_from_stream(scope JpegStreamReadFunc rfn, void* userData,
                                                 out int width, out int height, out int actual_comps, 
                                                 out float pixelAspectRatio, out float dotsPerInchY,
                                                 int req_comps=-1) {

  //actual_comps = 0;
  if (rfn is null) return null;
  if (req_comps != -1 && req_comps != 1 && req_comps != 3 && req_comps != 4) return null;

  auto decoder = jpeg_decoder(rfn, userData);
  if (decoder.error_code != JPGD_SUCCESS) return null;
  version(jpegd_test) scope(exit) { import core.stdc.stdio : printf; printf("%u bytes read.\n", cast(uint)decoder.total_bytes_read); }

  immutable int image_width = decoder.width;
  immutable int image_height = decoder.height;
  width = image_width;
  height = image_height;
  pixelAspectRatio = -1;
  dotsPerInchY = -1;
  actual_comps = decoder.num_components;
  if (req_comps < 0) req_comps = decoder.num_components;

  if (decoder.begin_decoding() != JPGD_SUCCESS) return null;

  immutable int dst_bpl = image_width*req_comps;

   ubyte* pImage_data = cast(ubyte*)jpgd_malloc(dst_bpl*image_height);
   if (pImage_data is null) return null;
   auto idata = pImage_data[0..dst_bpl*image_height];

  for (int y = 0; y < image_height; ++y) {
    const(ubyte)* pScan_line;
    uint scan_line_len;
    if (decoder.decode(/*(const void**)*/cast(void**)&pScan_line, &scan_line_len) != JPGD_SUCCESS) {
      jpgd_free(pImage_data);
      return null;
    }

    ubyte* pDst = pImage_data+y*dst_bpl;

    if ((req_comps == 1 && decoder.num_components == 1) || (req_comps == 4 && decoder.num_components == 3)) {
      memcpy(pDst, pScan_line, dst_bpl);
    } else if (decoder.num_components == 1) {
      if (req_comps == 3) {
        for (int x = 0; x < image_width; ++x) {
          ubyte luma = pScan_line[x];
          pDst[0] = luma;
          pDst[1] = luma;
          pDst[2] = luma;
          pDst += 3;
        }
      } else {
        for (int x = 0; x < image_width; ++x) {
          ubyte luma = pScan_line[x];
          pDst[0] = luma;
          pDst[1] = luma;
          pDst[2] = luma;
          pDst[3] = 255;
          pDst += 4;
        }
      }
    } else if (decoder.num_components == 3) {
      if (req_comps == 1) {
        immutable int YR = 19595, YG = 38470, YB = 7471;
        for (int x = 0; x < image_width; ++x) {
          int r = pScan_line[x*4+0];
          int g = pScan_line[x*4+1];
          int b = pScan_line[x*4+2];
          *pDst++ = cast(ubyte)((r * YR + g * YG + b * YB + 32768) >> 16);
        }
      } else {
        for (int x = 0; x < image_width; ++x) {
          pDst[0] = pScan_line[x*4+0];
          pDst[1] = pScan_line[x*4+1];
          pDst[2] = pScan_line[x*4+2];
          pDst += 3;
        }
      }
    }
  }

  pixelAspectRatio = decoder.m_pixelAspectRatio;
  dotsPerInchY = decoder.m_pixelsPerInchY;

  return idata;
}