pluto-src 0.1.1+0.10.4

#include <cstdint>
#include <memory.h>


//-----------------------------------------------------------------------------
// MurmurHash1Aligned, by Austin Appleby
// Same algorithm as MurmurHash1, but only does aligned reads - should be safer
// on certain platforms. 
// Performance should be equal to or better than the simple version.
// MurmurHash was written by Austin Appleby, and is placed in the public domain
// The author hereby disclaims copyright to this source code.
//-----------------------------------------------------------------------------
unsigned int MurmurHash1Aligned(const void* key, int len, unsigned int seed)
{
  const unsigned int m = 0xc6a4a793;
  const int r = 16;

  const unsigned char * data = (const unsigned char *)key;

  unsigned int h = seed ^ (len * m);

  int align = (uint64_t)data & 3;

  if(align && (len >= 4))
  {
    unsigned int t = 0, d = 0;

    switch(align)
    {
      case 1: 
        t |= data[2] << 16;
        [[fallthrough]];
      case 2:
        t |= data[1] << 8;
        [[fallthrough]];
      case 3:
        t |= data[0];
    }

    t <<= (8 * align);

    data += 4-align;
    len -= 4-align;

    int sl = 8 * (4-align);
    int sr = 8 * align;

    while(len >= 4)
    {
      d = *(unsigned int *)data;
      t = (t >> sr) | (d << sl);
      h += t;
      h *= m;
      h ^= h >> r;
      t = d;

      data += 4;
      len -= 4;
    }

    int pack = len < align ? len : align;

    d = 0;

    switch(pack)
    {
      case 3:
        d |= data[2] << 16;
        [[fallthrough]];
      case 2:
        d |= data[1] << 8;
        [[fallthrough]];
      case 1: 
        d |= data[0];
        [[fallthrough]];
      case 0:
        h += (t >> sr) | (d << sl);
        h *= m;
        h ^= h >> r;
    }

    data += pack;
    len -= pack;
  }
  else
  {
    while(len >= 4)
    {
      h += *(unsigned int *)data;
      h *= m;
      h ^= h >> r;

      data += 4;
      len -= 4;
    }
  }

  switch(len)
  {
    case 3: 
      h += data[2] << 16;
      [[fallthrough]];
    case 2: 
      h += data[1] << 8;
      [[fallthrough]];
    case 1:
      h += data[0];
      h *= m;
      h ^= h >> r;
  };

  h *= m;
  h ^= h >> 10;
  h *= m;
  h ^= h >> 17;

  return h;
}


#if defined(_MSC_VER)

#define BIG_CONSTANT(x) (x)

// Other compilers

#else	// defined(_MSC_VER)

#define BIG_CONSTANT(x) (x##LLU)

#endif // !defined(_MSC_VER)

//-----------------------------------------------------------------------------

uint32_t MurmurHash2 ( const void * key, int len, uint32_t seed )
{
  // 'm' and 'r' are mixing constants generated offline.
  // They're not really 'magic', they just happen to work well.

  const uint32_t m = 0x5bd1e995;
  const int r = 24;

  // Initialize the hash to a 'random' value

  uint32_t h = seed ^ len;

  // Mix 4 bytes at a time into the hash

  const unsigned char * data = (const unsigned char *)key;

  while(len >= 4)
  {
    uint32_t k = *(uint32_t*)data;

    k *= m;
    k ^= k >> r;
    k *= m;

    h *= m;
    h ^= k;

    data += 4;
    len -= 4;
  }

  // Handle the last few bytes of the input array

  switch(len)
  {
    case 3:
      h ^= data[2] << 16;
      [[fallthrough]];
    case 2:
      h ^= data[1] << 8;
      [[fallthrough]];
    case 1:
      h ^= data[0];
      h *= m;
  };

  // Do a few final mixes of the hash to ensure the last few
  // bytes are well-incorporated.

  h ^= h >> 13;
  h *= m;
  h ^= h >> 15;

  return h;
} 

//-----------------------------------------------------------------------------
// MurmurHash2, 64-bit versions, by Austin Appleby

// The same caveats as 32-bit MurmurHash2 apply here - beware of alignment 
// and endian-ness issues if used across multiple platforms.

// 64-bit hash for 64-bit platforms

uint64_t MurmurHash64A ( const void * key, int len, uint64_t seed )
{
  const uint64_t m = BIG_CONSTANT(0xc6a4a7935bd1e995);
  const int r = 47;

  uint64_t h = seed ^ (len * m);

  const uint64_t * data = (const uint64_t *)key;
  const uint64_t * end = data + (len/8);

  while(data != end)
  {
    uint64_t k = *data++;

    k *= m; 
    k ^= k >> r; 
    k *= m; 
    
    h ^= k;
    h *= m; 
  }

  const unsigned char * data2 = (const unsigned char*)data;

  switch(len & 7)
  {
    case 7:
      h ^= uint64_t(data2[6]) << 48;
      [[fallthrough]];
    case 6:
      h ^= uint64_t(data2[5]) << 40;
      [[fallthrough]];
    case 5:
      h ^= uint64_t(data2[4]) << 32;
      [[fallthrough]];
    case 4:
      h ^= uint64_t(data2[3]) << 24;
      [[fallthrough]];
    case 3:
      h ^= uint64_t(data2[2]) << 16;
      [[fallthrough]];
    case 2:
      h ^= uint64_t(data2[1]) << 8;
      [[fallthrough]];
    case 1:
      h ^= uint64_t(data2[0]);
      h *= m;
  };
 
  h ^= h >> r;
  h *= m;
  h ^= h >> r;

  return h;
} 


// 64-bit hash for 32-bit platforms

uint64_t MurmurHash64B ( const void * key, int len, uint64_t seed )
{
  const uint32_t m = 0x5bd1e995;
  const int r = 24;

  uint32_t h1 = uint32_t(seed) ^ len;
  uint32_t h2 = uint32_t(seed >> 32);

  const uint32_t * data = (const uint32_t *)key;

  while(len >= 8)
  {
    uint32_t k1 = *data++;
    k1 *= m; k1 ^= k1 >> r; k1 *= m;
    h1 *= m; h1 ^= k1;
    len -= 4;

    uint32_t k2 = *data++;
    k2 *= m; k2 ^= k2 >> r; k2 *= m;
    h2 *= m; h2 ^= k2;
    len -= 4;
  }

  if(len >= 4)
  {
    uint32_t k1 = *data++;
    k1 *= m; k1 ^= k1 >> r; k1 *= m;
    h1 *= m; h1 ^= k1;
    len -= 4;
  }

  switch(len)
  {
    case 3:
      h2 ^= ((unsigned char*)data)[2] << 16;
      [[fallthrough]];
    case 2:
      h2 ^= ((unsigned char*)data)[1] << 8;
      [[fallthrough]];
    case 1:
      h2 ^= ((unsigned char*)data)[0];
      h2 *= m;
  };

  h1 ^= h2 >> 18; h1 *= m;
  h2 ^= h1 >> 22; h2 *= m;
  h1 ^= h2 >> 17; h1 *= m;
  h2 ^= h1 >> 19; h2 *= m;

  uint64_t h = h1;

  h = (h << 32) | h2;

  return h;
} 

//-----------------------------------------------------------------------------
// MurmurHash2A, by Austin Appleby

// This is a variant of MurmurHash2 modified to use the Merkle-Damgard 
// construction. Bulk speed should be identical to Murmur2, small-key speed 
// will be 10%-20% slower due to the added overhead at the end of the hash.

// This variant fixes a minor issue where null keys were more likely to
// collide with each other than expected, and also makes the function
// more amenable to incremental implementations.

#define mmix(h,k) { k *= m; k ^= k >> r; k *= m; h *= m; h ^= k; }

uint32_t MurmurHash2A ( const void * key, int len, uint32_t seed )
{
  const uint32_t m = 0x5bd1e995;
  const int r = 24;
  uint32_t l = len;

  const unsigned char * data = (const unsigned char *)key;

  uint32_t h = seed;

  while(len >= 4)
  {
    uint32_t k = *(uint32_t*)data;

    mmix(h,k);

    data += 4;
    len -= 4;
  }

  uint32_t t = 0;

  switch(len)
  {
    case 3:
      t ^= data[2] << 16;
      [[fallthrough]];
    case 2:
      t ^= data[1] << 8;
      [[fallthrough]];
    case 1: t ^= data[0];
  };

  mmix(h,t);
  mmix(h,l);

  h ^= h >> 13;
  h *= m;
  h ^= h >> 15;

  return h;
}


//-----------------------------------------------------------------------------
// MurmurHashNeutral2, by Austin Appleby

// Same as MurmurHash2, but endian- and alignment-neutral.
// Half the speed though, alas.

uint32_t MurmurHashNeutral2 ( const void * key, int len, uint32_t seed )
{
  const uint32_t m = 0x5bd1e995;
  const int r = 24;

  uint32_t h = seed ^ len;

  const unsigned char * data = (const unsigned char *)key;

  while(len >= 4)
  {
    uint32_t k;

    k  = data[0];
    k |= data[1] << 8;
    k |= data[2] << 16;
    k |= data[3] << 24;

    k *= m; 
    k ^= k >> r; 
    k *= m;

    h *= m;
    h ^= k;

    data += 4;
    len -= 4;
  }
  
  switch(len)
  {
    case 3:
      h ^= data[2] << 16;
      [[fallthrough]];
    case 2:
      h ^= data[1] << 8;
      [[fallthrough]];
    case 1:
      h ^= data[0];
      h *= m;
  };

  h ^= h >> 13;
  h *= m;
  h ^= h >> 15;

  return h;
} 

//-----------------------------------------------------------------------------
// MurmurHashAligned2, by Austin Appleby

// Same algorithm as MurmurHash2, but only does aligned reads - should be safer
// on certain platforms. 

// Performance will be lower than MurmurHash2

#define MIX(h,k,m) { k *= m; k ^= k >> r; k *= m; h *= m; h ^= k; }


uint32_t MurmurHashAligned2 ( const void * key, int len, uint32_t seed )
{
  const uint32_t m = 0x5bd1e995;
  const int r = 24;

  const unsigned char * data = (const unsigned char *)key;

  uint32_t h = seed ^ len;

  int align = (uint64_t)data & 3;

  if(align && (len >= 4))
  {
    // Pre-load the temp registers

    uint32_t t = 0, d = 0;

    switch(align)
    {
      case 1:
        t |= data[2] << 16;
        [[fallthrough]];
      case 2: 
        t |= data[1] << 8;
        [[fallthrough]];
      case 3:
        t |= data[0];
    }

    t <<= (8 * align);

    data += 4-align;
    len -= 4-align;

    int sl = 8 * (4-align);
    int sr = 8 * align;

    // Mix

    while(len >= 4)
    {
      d = *(uint32_t *)data;
      t = (t >> sr) | (d << sl);

      uint32_t k = t;

      MIX(h,k,m);

      t = d;

      data += 4;
      len -= 4;
    }

    // Handle leftover data in temp registers

    d = 0;

    if(len >= align)
    {
      switch(align)
      {
        case 3:
          d |= data[2] << 16;
          [[fallthrough]];
        case 2:
          d |= data[1] << 8;
          [[fallthrough]];
        case 1:
          d |= data[0];
      }

      uint32_t k = (t >> sr) | (d << sl);
      MIX(h,k,m);

      data += align;
      len -= align;

      //----------
      // Handle tail bytes

      switch(len)
      {
        case 3:
          h ^= data[2] << 16;
          [[fallthrough]];
        case 2:
          h ^= data[1] << 8;
          [[fallthrough]];
        case 1:
          h ^= data[0];
          h *= m;
      };
    }
    else
    {
      switch(len)
      {
        case 3:
          d |= data[2] << 16;
          [[fallthrough]];
        case 2:
          d |= data[1] << 8;
          [[fallthrough]];
        case 1:
          d |= data[0];
          [[fallthrough]];
        case 0:
            h ^= (t >> sr) | (d << sl);
            h *= m;
      }
    }

    h ^= h >> 13;
    h *= m;
    h ^= h >> 15;

    return h;
  }
  else
  {
    while(len >= 4)
    {
      uint32_t k = *(uint32_t *)data;

      MIX(h,k,m);

      data += 4;
      len -= 4;
    }

    //----------
    // Handle tail bytes

    switch(len)
    {
      case 3:
        h ^= data[2] << 16;
        [[fallthrough]];
      case 2:
        h ^= data[1] << 8;
        [[fallthrough]];
      case 1:
        h ^= data[0];
        h *= m;
    };

    h ^= h >> 13;
    h *= m;
    h ^= h >> 15;

    return h;
  }
}

//-----------------------------------------------------------------------------


/*
  SuperFastHash By Paul Hsieh (C) 2004, 2005.  Covered under the Paul Hsieh derivative 
  license. See: 
  http://www.azillionmonkeys.com/qed/weblicense.html for license details.
  http://www.azillionmonkeys.com/qed/hash.html
*/

inline uint16_t get16bits ( const void * p )
{
  return *(const uint16_t*)p;
}

uint32_t SuperFastHash (const signed char * data, int len) {
uint32_t hash = 0, tmp;
int rem;

  if (len <= 0 || data == NULL) return 0;

  rem = len & 3;
  len >>= 2;

  /* Main loop */
  for (;len > 0; len--) {
    hash  += get16bits (data);
    tmp    = (get16bits (data+2) << 11) ^ hash;
    hash   = (hash << 16) ^ tmp;
    data  += 2*sizeof (uint16_t);
    hash  += hash >> 11;
  }

  /* Handle end cases */
  switch (rem) {
    case 3:	hash += get16bits (data);
        hash ^= hash << 16;
        hash ^= data[sizeof (uint16_t)] << 18;
        hash += hash >> 11;
        break;
    case 2:	hash += get16bits (data);
        hash ^= hash << 11;
        hash += hash >> 17;
        break;
    case 1: hash += *data;
        hash ^= hash << 10;
        hash += hash >> 1;
  }

  /* Force "avalanching" of final 127 bits */
  hash ^= hash << 3;
  hash += hash >> 5;
  hash ^= hash << 4;
  hash += hash >> 17;
  hash ^= hash << 25;
  hash += hash >> 6;

  return hash;
}


// "Derived from the RSA Data Security, Inc. MD5 Message Digest Algorithm"

/**
 * \brief          MD5 context structure
 */
typedef struct
{
    unsigned long total[2];     /*!< number of bytes processed  */
    unsigned long state[4];     /*!< intermediate digest state  */
    unsigned char buffer[64];   /*!< data block being processed */
}
md5_context;

/**
 * \brief          MD5 context setup
 *
 * \param ctx      context to be initialized
 */
void md5_starts( md5_context *ctx );

/**
 * \brief          MD5 process buffer
 *
 * \param ctx      MD5 context
 * \param input    buffer holding the  data
 * \param ilen     length of the input data
 */
void md5_update( md5_context *ctx, unsigned char *input, int ilen );

/**
 * \brief          MD5 final digest
 *
 * \param ctx      MD5 context
 * \param output   MD5 checksum result
 */
void md5_finish( md5_context *ctx, unsigned char output[16] );

/**
 * \brief          Output = MD5( input buffer )
 *
 * \param input    buffer holding the  data
 * \param ilen     length of the input data
 * \param output   MD5 checksum result
 */
void md5( unsigned char *input, int ilen, unsigned char output[16] );

/*
 * 32-bit integer manipulation macros (little endian)
 */
#ifndef GET_ULONG_LE
#define GET_ULONG_LE(n,b,i)                             \
{                                                       \
    (n) = ( (unsigned long) (b)[(i)    ]       )        \
        | ( (unsigned long) (b)[(i) + 1] <<  8 )        \
        | ( (unsigned long) (b)[(i) + 2] << 16 )        \
        | ( (unsigned long) (b)[(i) + 3] << 24 );       \
}
#endif

#ifndef PUT_ULONG_LE
#define PUT_ULONG_LE(n,b,i)                             \
{                                                       \
    (b)[(i)    ] = (unsigned char) ( (n)       );       \
    (b)[(i) + 1] = (unsigned char) ( (n) >>  8 );       \
    (b)[(i) + 2] = (unsigned char) ( (n) >> 16 );       \
    (b)[(i) + 3] = (unsigned char) ( (n) >> 24 );       \
}
#endif

/*
 * MD5 context setup
 */
void md5_starts( md5_context *ctx )
{
    ctx->total[0] = 0;
    ctx->total[1] = 0;

    ctx->state[0] = 0x67452301;
    ctx->state[1] = 0xEFCDAB89;
    ctx->state[2] = 0x98BADCFE;
    ctx->state[3] = 0x10325476;
}

static void md5_process( md5_context *ctx, unsigned char data[64] )
{
    unsigned long X[16], A, B, C, D;

    GET_ULONG_LE( X[ 0], data,  0 );
    GET_ULONG_LE( X[ 1], data,  4 );
    GET_ULONG_LE( X[ 2], data,  8 );
    GET_ULONG_LE( X[ 3], data, 12 );
    GET_ULONG_LE( X[ 4], data, 16 );
    GET_ULONG_LE( X[ 5], data, 20 );
    GET_ULONG_LE( X[ 6], data, 24 );
    GET_ULONG_LE( X[ 7], data, 28 );
    GET_ULONG_LE( X[ 8], data, 32 );
    GET_ULONG_LE( X[ 9], data, 36 );
    GET_ULONG_LE( X[10], data, 40 );
    GET_ULONG_LE( X[11], data, 44 );
    GET_ULONG_LE( X[12], data, 48 );
    GET_ULONG_LE( X[13], data, 52 );
    GET_ULONG_LE( X[14], data, 56 );
    GET_ULONG_LE( X[15], data, 60 );

#define S(x,n) ((x << n) | ((x & 0xFFFFFFFF) >> (32 - n)))

#define P(a,b,c,d,k,s,t)                                \
{                                                       \
    a += F(b,c,d) + X[k] + t; a = S(a,s) + b;           \
}

    A = ctx->state[0];
    B = ctx->state[1];
    C = ctx->state[2];
    D = ctx->state[3];

#define F(x,y,z) (z ^ (x & (y ^ z)))

    P( A, B, C, D,  0,  7, 0xD76AA478 );
    P( D, A, B, C,  1, 12, 0xE8C7B756 );
    P( C, D, A, B,  2, 17, 0x242070DB );
    P( B, C, D, A,  3, 22, 0xC1BDCEEE );
    P( A, B, C, D,  4,  7, 0xF57C0FAF );
    P( D, A, B, C,  5, 12, 0x4787C62A );
    P( C, D, A, B,  6, 17, 0xA8304613 );
    P( B, C, D, A,  7, 22, 0xFD469501 );
    P( A, B, C, D,  8,  7, 0x698098D8 );
    P( D, A, B, C,  9, 12, 0x8B44F7AF );
    P( C, D, A, B, 10, 17, 0xFFFF5BB1 );
    P( B, C, D, A, 11, 22, 0x895CD7BE );
    P( A, B, C, D, 12,  7, 0x6B901122 );
    P( D, A, B, C, 13, 12, 0xFD987193 );
    P( C, D, A, B, 14, 17, 0xA679438E );
    P( B, C, D, A, 15, 22, 0x49B40821 );

#undef F

#define F(x,y,z) (y ^ (z & (x ^ y)))

    P( A, B, C, D,  1,  5, 0xF61E2562 );
    P( D, A, B, C,  6,  9, 0xC040B340 );
    P( C, D, A, B, 11, 14, 0x265E5A51 );
    P( B, C, D, A,  0, 20, 0xE9B6C7AA );
    P( A, B, C, D,  5,  5, 0xD62F105D );
    P( D, A, B, C, 10,  9, 0x02441453 );
    P( C, D, A, B, 15, 14, 0xD8A1E681 );
    P( B, C, D, A,  4, 20, 0xE7D3FBC8 );
    P( A, B, C, D,  9,  5, 0x21E1CDE6 );
    P( D, A, B, C, 14,  9, 0xC33707D6 );
    P( C, D, A, B,  3, 14, 0xF4D50D87 );
    P( B, C, D, A,  8, 20, 0x455A14ED );
    P( A, B, C, D, 13,  5, 0xA9E3E905 );
    P( D, A, B, C,  2,  9, 0xFCEFA3F8 );
    P( C, D, A, B,  7, 14, 0x676F02D9 );
    P( B, C, D, A, 12, 20, 0x8D2A4C8A );

#undef F
    
#define F(x,y,z) (x ^ y ^ z)

    P( A, B, C, D,  5,  4, 0xFFFA3942 );
    P( D, A, B, C,  8, 11, 0x8771F681 );
    P( C, D, A, B, 11, 16, 0x6D9D6122 );
    P( B, C, D, A, 14, 23, 0xFDE5380C );
    P( A, B, C, D,  1,  4, 0xA4BEEA44 );
    P( D, A, B, C,  4, 11, 0x4BDECFA9 );
    P( C, D, A, B,  7, 16, 0xF6BB4B60 );
    P( B, C, D, A, 10, 23, 0xBEBFBC70 );
    P( A, B, C, D, 13,  4, 0x289B7EC6 );
    P( D, A, B, C,  0, 11, 0xEAA127FA );
    P( C, D, A, B,  3, 16, 0xD4EF3085 );
    P( B, C, D, A,  6, 23, 0x04881D05 );
    P( A, B, C, D,  9,  4, 0xD9D4D039 );
    P( D, A, B, C, 12, 11, 0xE6DB99E5 );
    P( C, D, A, B, 15, 16, 0x1FA27CF8 );
    P( B, C, D, A,  2, 23, 0xC4AC5665 );

#undef F

#define F(x,y,z) (y ^ (x | ~z))

    P( A, B, C, D,  0,  6, 0xF4292244 );
    P( D, A, B, C,  7, 10, 0x432AFF97 );
    P( C, D, A, B, 14, 15, 0xAB9423A7 );
    P( B, C, D, A,  5, 21, 0xFC93A039 );
    P( A, B, C, D, 12,  6, 0x655B59C3 );
    P( D, A, B, C,  3, 10, 0x8F0CCC92 );
    P( C, D, A, B, 10, 15, 0xFFEFF47D );
    P( B, C, D, A,  1, 21, 0x85845DD1 );
    P( A, B, C, D,  8,  6, 0x6FA87E4F );
    P( D, A, B, C, 15, 10, 0xFE2CE6E0 );
    P( C, D, A, B,  6, 15, 0xA3014314 );
    P( B, C, D, A, 13, 21, 0x4E0811A1 );
    P( A, B, C, D,  4,  6, 0xF7537E82 );
    P( D, A, B, C, 11, 10, 0xBD3AF235 );
    P( C, D, A, B,  2, 15, 0x2AD7D2BB );
    P( B, C, D, A,  9, 21, 0xEB86D391 );

#undef F

    ctx->state[0] += A;
    ctx->state[1] += B;
    ctx->state[2] += C;
    ctx->state[3] += D;
}

/*
 * MD5 process buffer
 */
void md5_update( md5_context *ctx, unsigned char *input, int ilen )
{
    int fill;
    unsigned long left;

    if( ilen <= 0 )
        return;

    left = ctx->total[0] & 0x3F;
    fill = 64 - left;

    ctx->total[0] += ilen;
    ctx->total[0] &= 0xFFFFFFFF;

    if( ctx->total[0] < (unsigned long) ilen )
        ctx->total[1]++;

    if( left && ilen >= fill )
    {
        memcpy( (void *) (ctx->buffer + left),
                (void *) input, fill );
        md5_process( ctx, ctx->buffer );
        input += fill;
        ilen  -= fill;
        left = 0;
    }

    while( ilen >= 64 )
    {
        md5_process( ctx, input );
        input += 64;
        ilen  -= 64;
    }

    if( ilen > 0 )
    {
        memcpy( (void *) (ctx->buffer + left),
                (void *) input, ilen );
    }
}

static const unsigned char md5_padding[64] =
{
 0x80, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
};

/*
 * MD5 final digest
 */
void md5_finish( md5_context *ctx, unsigned char output[16] )
{
    unsigned long last, padn;
    unsigned long high, low;
    unsigned char msglen[8];

    high = ( ctx->total[0] >> 29 )
         | ( ctx->total[1] <<  3 );
    low  = ( ctx->total[0] <<  3 );

    PUT_ULONG_LE( low,  msglen, 0 );
    PUT_ULONG_LE( high, msglen, 4 );

    last = ctx->total[0] & 0x3F;
    padn = ( last < 56 ) ? ( 56 - last ) : ( 120 - last );

    md5_update( ctx, (unsigned char *) md5_padding, padn );
    md5_update( ctx, msglen, 8 );

    PUT_ULONG_LE( ctx->state[0], output,  0 );
    PUT_ULONG_LE( ctx->state[1], output,  4 );
    PUT_ULONG_LE( ctx->state[2], output,  8 );
    PUT_ULONG_LE( ctx->state[3], output, 12 );
}

/*
 * output = MD5( input buffer )
 */
void md5_fn( unsigned char *input, int ilen, unsigned char output[16] )
{
    md5_context ctx;

    md5_starts( &ctx );
    md5_update( &ctx, input, ilen );
    md5_finish( &ctx, output );

    memset( &ctx, 0, sizeof( md5_context ) );
}

// lookup3 by Bob Jekins, code is public domain.
#define rot(x,k) (((x)<<(k)) | ((x)>>(32-(k))))

#define mix(a,b,c) \
{ \
  a -= c;  a ^= rot(c, 4);  c += b; \
  b -= a;  b ^= rot(a, 6);  a += c; \
  c -= b;  c ^= rot(b, 8);  b += a; \
  a -= c;  a ^= rot(c,16);  c += b; \
  b -= a;  b ^= rot(a,19);  a += c; \
  c -= b;  c ^= rot(b, 4);  b += a; \
}

#define final(a,b,c) \
{ \
  c ^= b; c -= rot(b,14); \
  a ^= c; a -= rot(c,11); \
  b ^= a; b -= rot(a,25); \
  c ^= b; c -= rot(b,16); \
  a ^= c; a -= rot(c,4);  \
  b ^= a; b -= rot(a,14); \
  c ^= b; c -= rot(b,24); \
}

uint32_t lookup3_impl ( const void * key, int length, uint32_t initval )
{
  uint32_t a,b,c;                                          /* internal state */

  a = b = c = 0xdeadbeef + ((uint32_t)length) + initval;

  const uint32_t *k = (const uint32_t *)key;         /* read 32-bit chunks */

  /*------ all but last block: aligned reads and affect 32 bits of (a,b,c) */
  while (length > 12)
  {
    a += k[0];
    b += k[1];
    c += k[2];
    mix(a,b,c);
    length -= 12;
    k += 3;
  }

  switch(length)
  {
    case 12:
      c += k[2];
      b += k[1];
      a += k[0];
      break;
    case 11:
      c += k[2] & 0xffffff;
      b += k[1];
      a += k[0];
      break;
    case 10:
      c += k[2] & 0xffff;
      b += k[1];
      a += k[0];
      break;
    case 9:
      c += k[2] & 0xff;
      b += k[1];
      a += k[0];
      break;
    case 8:
      b += k[1];
      a += k[0];
      break;
    case 7:
      b += k[1] & 0xffffff;
      a += k[0];
      break;
    case 6:
      b += k[1] & 0xffff;
      a += k[0];
      break;
    case 5:
      b += k[1] & 0xff;
      a += k[0];
      break;
    case 4:
      a += k[0];
      break;
    case 3:
      a += k[0] & 0xffffff;
      break;
    case 2:
      a += k[0] & 0xffff;
      break;
    case 1:
      a += k[0] & 0xff;
      break;
    case 0: return c; 
  }

  final(a,b,c);

  return c;
}