//+------------------------------------------------------------------+
//|                                                         SHA3.mqh |
//|                                   Copyright 2026, Niquel Mendoza |
//|                                             https://www.mql5.com |
//+------------------------------------------------------------------+
#property copyright "Copyright 2026, Niquel Mendoza"
#property link      "https://www.mql5.com"
#property strict

#ifndef CRYPTOBYLEO_SRC_HASH_SHA3_MQH
#define CRYPTOBYLEO_SRC_HASH_SHA3_MQH

//+------------------------------------------------------------------+
//|                                                                  |
//+------------------------------------------------------------------+
#include "Base.mqh"

//+------------------------------------------------------------------+
//|                                                                  |
//+------------------------------------------------------------------+
//---
#define SHA3_DELIM_SHA3  (0x06)
#define SHA3_DELIM_SHAKE (0x1F)

//---
#define SHA3_256_OUT_BSIZE  (32)
#define SHA3_256_BLOCK_SIZE (136)

//---
#define SHA3_512_OUT_BSIZE  (64)
#define SHA3_512_BLOCK_SIZE (72)


//---
#define SHA3_384_OUT_BSIZE   (48)
#define SHA3_384_BLOCK_SIZE  (104)

//---
#define SHA3_SHAKE_128_BLOCK_SIZE (168)
#define SHA3_SHAKE_256_BLOCK_SIZE (136)

//+------------------------------------------------------------------+
//|                                                                  |
//+------------------------------------------------------------------+
namespace TSN
{
//+------------------------------------------------------------------+
//|                                                                  |
//+------------------------------------------------------------------+
#define SHA3_LOADINLINE(idx)                  \
      s[idx] ^= ((uint64_t)in[pos++]) |       \
                ((uint64_t)in[pos++] << 8)  | \
                ((uint64_t)in[pos++] << 16) | \
                ((uint64_t)in[pos++] << 24) | \
                ((uint64_t)in[pos++] << 32) | \
                ((uint64_t)in[pos++] << 40) | \
                ((uint64_t)in[pos++] << 48) | \
                ((uint64_t)in[pos++] << 56);

//+------------------------------------------------------------------+
//|                                                                  |
//+------------------------------------------------------------------+
class CCryptoHashSHA3
 {
private:
  static void        SHA3_KECCAK_F(ulong &s[]);

public:
                     CCryptoHashSHA3(void) {}
                    ~CCryptoHashSHA3(void) {}

  //---
  union SHA3_UN
   {
    uchar block[200];
    ulong uv[25];
   };

  //--- Por limitaciones del compilador de que no se permite tempaltes medio raras
  // en static class lo hare inline
  //---
  // TChunkBytes = tamaño del bloquje en bytes que se procesar en cada it
  // ins= tmaño del inpunt (in)
  // in= input
  // out=salida
  // outl= tamaño de salida
  template <typename TChunkBytes, typename TDelimBytes>
  static void        SHA3_HASH(const uchar &in[], int ins, uchar &out[], int outl, int o_s)
   {
    //--- Iniciales
    ulong s[25] = {0ULL, 0ULL, 0ULL, 0ULL, 0ULL, 0ULL, 0ULL, 0ULL, 0ULL, 0ULL, 0ULL, 0ULL, 0ULL, 0ULL, 0ULL,
                   0ULL, 0ULL, 0ULL, 0ULL, 0ULL, 0ULL, 0ULL, 0ULL, 0ULL, 0ULL
                  };
    const int rho[25] = {0, 1, 62, 28, 27, 36, 44, 6, 55, 20, 3, 10, 43, 25, 39, 41, 45, 15, 21, 8, 18, 2, 61, 56, 14};
    const int pi[25]  = {0, 10, 20, 5, 15, 16, 1, 11, 21, 6, 7, 17, 2, 12, 22, 23, 8, 18, 3, 13, 14, 24, 9, 19, 4};
    const ulong rc[24] =
     {
      0x0000000000000001, 0x0000000000008082, 0x800000000000808a, 0x8000000080008000,
      0x000000000000808b, 0x0000000080000001, 0x8000000080008081, 0x8000000000008009,
      0x000000000000008a, 0x0000000000000088, 0x0000000080008009, 0x000000008000000a,
      0x000000008000808b, 0x800000000000008b, 0x8000000000008089, 0x8000000000008003,
      0x8000000000008002, 0x8000000000000080, 0x000000000000800a, 0x800000008000000a,
      0x8000000080008081, 0x8000000000008080, 0x0000000080000001, 0x8000000080008008
     };


    //--- paso 1 absorcion
    int pos = 0;
    while(ins - pos >= sizeof(TChunkBytes))
     {
      //---- carga inline (72 bytes iniciales el minimo comun entre todos.. de ahi algunos cargan mas..)
      SHA3_LOADINLINE(0)
      SHA3_LOADINLINE(1)
      SHA3_LOADINLINE(2)
      SHA3_LOADINLINE(3)
      SHA3_LOADINLINE(4)
      SHA3_LOADINLINE(5)
      SHA3_LOADINLINE(6)
      SHA3_LOADINLINE(7)
      SHA3_LOADINLINE(8)

      //---- carga manual
      for(int idx = 9; idx < (sizeof(TChunkBytes) >> 3); idx++)
       {
        s[idx] ^= ((uint64_t)in[pos++]) |
                  ((uint64_t)in[pos++] << 8)  |
                  ((uint64_t)in[pos++] << 16) |
                  ((uint64_t)in[pos++] << 24) |
                  ((uint64_t)in[pos++] << 32) |
                  ((uint64_t)in[pos++] << 40) |
                  ((uint64_t)in[pos++] << 48) |
                  ((uint64_t)in[pos++] << 56);
       }

      //---
      // HOT PATH inline aqui mismo
      //------- statart of mixing
      // Nuestro "valor" en el cual operamos..
      /*
      mapa=
      5 filas 5 columnas y cada valor es un lane de 64 bits
      */

      //--- Temp
      ulong c[5], d[5];
      ulong b[25];

      //--- Rondas.
      for(int r = 0; r < 24; r++)
       {
        //--- theta
        /* wiki
        Calcula la paridad de cada una de las 5 w (320, cuando w = 64 )
        columnas de 5 bits y realiza la operación OR exclusiva con dos columnas cercanas siguiendo un patrón regular.
        Para ser precisos, a [ i ][  j ][ k ] ← a [ i ][  j ][ k ] ⊕ paridad(a[0...4][ j -1][ k ]) ⊕ paridad(a[0...4][ j +1][ k −1] )
        */
        // Por cada columna
        /*for(int x = 0; x < 5; x++)
          c[x] = s[x] ^ s[x + 5] ^ s[x + 10] ^ s[x + 15] ^ s[x + 20];
        */ // Denserollado:
        c[0] = s[0] ^ s[5] ^ s[10] ^ s[15] ^ s[20];
        c[1] = s[1] ^ s[6] ^ s[11] ^ s[16] ^ s[21];
        c[2] = s[2] ^ s[7] ^ s[12] ^ s[17] ^ s[22];
        c[3] = s[3] ^ s[8] ^ s[13] ^ s[18] ^ s[23];
        c[4] = s[4] ^ s[9] ^ s[14] ^ s[19] ^ s[24];

        // Final
        //for(int x = 0; x < 5; x++)
        /*{
         const ulong cr = c[(x + 1) % 5];
         d[x] = c[(x + 4) % 5] ^ ((cr << 1) | (cr >> 63));
        }*/
        // Denserollado
         {
          const ulong cr = c[1];
          d[0] = c[4] ^ ((cr << 1) | (cr >> 63));
         }
         {
          const ulong cr = c[2];
          d[1] = c[0] ^ ((cr << 1) | (cr >> 63));
         }
         {
          const ulong cr = c[3];
          d[2] = c[1] ^ ((cr << 1) | (cr >> 63));
         }
         {
          const ulong cr = c[4];
          d[3] = c[2] ^ ((cr << 1) | (cr >> 63));
         }
         {
          const ulong cr = c[0];
          d[4] = c[3] ^ ((cr << 1) | (cr >> 63));
         }

        // unrrolled
        s[0] ^= d[0];
        s[1] ^= d[1];
        s[2] ^= d[2];
        s[3] ^= d[3];
        s[4] ^= d[4];
        s[5] ^= d[0];
        s[6] ^= d[1];
        s[7] ^= d[2];
        s[8] ^= d[3];
        s[9] ^= d[4];
        s[10] ^= d[0];
        s[11] ^= d[1];
        s[12] ^= d[2];
        s[13] ^= d[3];
        s[14] ^= d[4];
        s[15] ^= d[0];
        s[16] ^= d[1];
        s[17] ^= d[2];
        s[18] ^= d[3];
        s[19] ^= d[4];
        s[20] ^= d[0];
        s[21] ^= d[1];
        s[22] ^= d[2];
        s[23] ^= d[3];
        s[24] ^= d[4];

        //--- RHO+PI
        for(int i = 0; i < 25; i++)
         {
          /*
          Gire bit a bit cada una de las 25 palabras por un número triangular diferente 0, 1, 3, 6, 10, 15, .... (rho[])
          Para ser precisos, a [0][0] no se rota, y para todo 0 ≤ t < 24 ,
          a [ i ][  j ][ k ] ← a [ i ][  j ][ k −( t +1)( t +2)/2] , donde
          */
          const int n = rho[i];
          b[pi[i]] = (n == 0) ? s[i] : ((s[i] << n) | (s[i] >> (64 - n)));
         }

        //--- CHI fila a fila
        for(int y = 0; y < 5; y++)
         {
          /*
          Combinación bit a bit a lo largo de las filas, usando x ← x ⊕ (¬ y & z ) . Para ser precisos, a
          [ i ][  j ][ k ] ← a [ i ][  j ][ k ] ⊕ (¬ a [ i ][  j + 1 ][ k ] &a a
          [ i ][  j + 2 ][ k ]) . Esta es la única operación no lineal en SHA-3.
          */
          //---
          ulong row0 = b[5 * y], row1 = b[1 + 5 * y], row2 = b[2 + 5 * y], row3 = b[3 + 5 * y], row4 = b[4 + 5 * y];

          //---
          s[5 * y]   = row0 ^ ((~row1) & row2);
          s[1 + 5 * y] = row1 ^ ((~row2) & row3);
          s[2 + 5 * y] = row2 ^ ((~row3) & row4);
          s[3 + 5 * y] = row3 ^ ((~row4) & row0);
          s[4 + 5 * y] = row4 ^ ((~row0) & row1);
         }

        //--- iota
        s[0] ^= rc[r];
       }



      //-----
      //------- end of mixing
      //pos += sizeof(TChunkBytes);
     }
    // Lo demas ya no inline dado que no da una ganancia sinficativa por lo visto
    // dado qeu se llama poco...

    //--- ultimo bloque parcial + padding (0x06 o 0x1F ... 0x80)
    SHA3_UN v;
    ArrayInitialize(v.block, 0);

    //---
    const int rem = ins - pos;

    //---
    for(int i = 0; i < rem; i++)
      v.block[i] = in[pos + i];

    //---
    v.block[rem] = sizeof(TDelimBytes);
    v.block[sizeof(TChunkBytes) - 1] |= 0x80;

    //--- Ahora lo que sobra..
    const int fin = (sizeof(TChunkBytes) >> 3);
    for(int i = 0; i < fin; i++)
     {
      /*const ulong w = ((ulong)block[i])       | ((ulong)block[i + 1] << 8)  |
                      ((ulong)block[i + 2] << 16) | ((ulong)block[i + 3] << 24) |
                      ((ulong)block[i + 4] << 32) | ((ulong)block[i + 5] << 40) |
                      ((ulong)block[i + 6] << 48) | ((ulong)block[i + 7] << 56);*/
      s[i] ^= v.uv[i];
     }
    SHA3_KECCAK_F(s);

    //---
    //  Print("hola", sizeof(TDelimBytes));
    if(sizeof(TDelimBytes) == SHA3_DELIM_SHA3)
     {
      const int fin = (outl >> 3); // / 8
      for(int k = 0; k < fin;)
       {
        const ulong t = s[k++];
        out[o_s++] = uchar(t);
        out[o_s++] = uchar(t >> 8);
        out[o_s++] = uchar(t >> 16);
        out[o_s++] = uchar(t >> 24);
        out[o_s++] = uchar(t >> 32);
        out[o_s++] = uchar(t >> 40);
        out[o_s++] = uchar(t >> 48);
        out[o_s++] = uchar(t >> 56);
       }
     }
    else // dinamico
     {
      outl += o_s;
      while(o_s < outl)
       {
        //---
        int t = outl - o_s; // tomado
        if(t > sizeof(TChunkBytes))
          t = sizeof(TChunkBytes); // clamp  a lo maximo a tomar..

        //---
        for(int i = 0; i < t; i++)
         {
          // en que lane
          //  const int w = ; // i / 8
          const int sh = (i & 7) << 3; // i % 8 * 8 - cuanto demzpasmroes del lane de 64 desde cuando se mrpiza
          out[o_s + i] = (uchar)(s[i >> 3] >> sh); // escribimos
         }
        o_s += t; // aumentaos lo que ya tenemos

        //---
        if(o_s < outl) // volvemos.. se requier mas de lo normal asi qeu aplicanos denuevo para la nueva salida
          SHA3_KECCAK_F(s);
       }
     }
   }

  //---- Especificos
  static __forceinline void  SHA3_256(const uchar &in[], int ins, uchar &out[], const int o_s);
  static __forceinline void  SHA3_384(const uchar &in[], int ins, uchar &out[], const int o_s);
  static __forceinline void  SHA3_512(const uchar &in[], int ins, uchar &out[], const int o_s);
  static __forceinline void  SHAKE128(const uchar &in[], int ins, uchar &out[], int outl, const int o_s);
  static __forceinline void  SHAKE256(const uchar &in[], int ins, uchar &out[], int outl, const int o_s);
 };

//+------------------------------------------------------------------+
//|                                                                  |
//+------------------------------------------------------------------+
//--- keccak-f[1600] funcion base
// s =arrya de entarad (25) size ya con
static void CCryptoHashSHA3::SHA3_KECCAK_F(ulong &s[])
 {
//--- Constatnes
  static const int rho[25] = {0, 1, 62, 28, 27, 36, 44, 6, 55, 20, 3, 10, 43, 25, 39, 41, 45, 15, 21, 8, 18, 2, 61, 56, 14};
  static const int pi[25]  = {0, 10, 20, 5, 15, 16, 1, 11, 21, 6, 7, 17, 2, 12, 22, 23, 8, 18, 3, 13, 14, 24, 9, 19, 4};
  static const ulong rc[24] =
   {
    0x0000000000000001, 0x0000000000008082, 0x800000000000808a, 0x8000000080008000,
    0x000000000000808b, 0x0000000080000001, 0x8000000080008081, 0x8000000000008009,
    0x000000000000008a, 0x0000000000000088, 0x0000000080008009, 0x000000008000000a,
    0x000000008000808b, 0x800000000000008b, 0x8000000000008089, 0x8000000000008003,
    0x8000000000008002, 0x8000000000000080, 0x000000000000800a, 0x800000008000000a,
    0x8000000080008081, 0x8000000000008080, 0x0000000080000001, 0x8000000080008008
   };
// Nuestro "valor" en el cual operamos..
  /*
  mapa=
  5 filas 5 columnas y cada valor es un lane de 64 bits
  */

//--- Temp
  ulong c[5], d[5];
  ulong b[25];

//--- Rondas.
  for(int r = 0; r < 24; r++)
   {
    //--- theta
    /* wiki
    Calcula la paridad de cada una de las 5 w (320, cuando w = 64 )
    columnas de 5 bits y realiza la operación OR exclusiva con dos columnas cercanas siguiendo un patrón regular.
    Para ser precisos, a [ i ][  j ][ k ] ← a [ i ][  j ][ k ] ⊕ paridad(a[0...4][ j -1][ k ]) ⊕ paridad(a[0...4][ j +1][ k −1] )
    */
    // Por cada columna
    /*for(int x = 0; x < 5; x++)
      c[x] = s[x] ^ s[x + 5] ^ s[x + 10] ^ s[x + 15] ^ s[x + 20];
    */ // Denserollado:
    c[0] = s[0] ^ s[5] ^ s[10] ^ s[15] ^ s[20];
    c[1] = s[1] ^ s[6] ^ s[11] ^ s[16] ^ s[21];
    c[2] = s[2] ^ s[7] ^ s[12] ^ s[17] ^ s[22];
    c[3] = s[3] ^ s[8] ^ s[13] ^ s[18] ^ s[23];
    c[4] = s[4] ^ s[9] ^ s[14] ^ s[19] ^ s[24];

    // Final
    //for(int x = 0; x < 5; x++)
    /*{
     const ulong cr = c[(x + 1) % 5];
     d[x] = c[(x + 4) % 5] ^ ((cr << 1) | (cr >> 63));
    }*/
    // Denserollado
     {
      const ulong cr = c[1];
      d[0] = c[4] ^ ((cr << 1) | (cr >> 63));
     }
     {
      const ulong cr = c[2];
      d[1] = c[0] ^ ((cr << 1) | (cr >> 63));
     }
     {
      const ulong cr = c[3];
      d[2] = c[1] ^ ((cr << 1) | (cr >> 63));
     }
     {
      const ulong cr = c[4];
      d[3] = c[2] ^ ((cr << 1) | (cr >> 63));
     }
     {
      const ulong cr = c[0];
      d[4] = c[3] ^ ((cr << 1) | (cr >> 63));
     }

    // unrrolled
    s[0] ^= d[0];
    s[1] ^= d[1];
    s[2] ^= d[2];
    s[3] ^= d[3];
    s[4] ^= d[4];
    s[5] ^= d[0];
    s[6] ^= d[1];
    s[7] ^= d[2];
    s[8] ^= d[3];
    s[9] ^= d[4];
    s[10] ^= d[0];
    s[11] ^= d[1];
    s[12] ^= d[2];
    s[13] ^= d[3];
    s[14] ^= d[4];
    s[15] ^= d[0];
    s[16] ^= d[1];
    s[17] ^= d[2];
    s[18] ^= d[3];
    s[19] ^= d[4];
    s[20] ^= d[0];
    s[21] ^= d[1];
    s[22] ^= d[2];
    s[23] ^= d[3];
    s[24] ^= d[4];

    //--- RHO+PI
    for(int i = 0; i < 25; i++)
     {
      /*
      Gire bit a bit cada una de las 25 palabras por un número triangular diferente 0, 1, 3, 6, 10, 15, .... (rho[])
      Para ser precisos, a [0][0] no se rota, y para todo 0 ≤ t < 24 ,
      a [ i ][  j ][ k ] ← a [ i ][  j ][ k −( t +1)( t +2)/2] , donde
      */
      const int n = rho[i];
      b[pi[i]] = (n == 0) ? s[i] : ((s[i] << n) | (s[i] >> (64 - n)));
     }

    //--- CHI fila a fila
    for(int y = 0; y < 5; y++)
     {
      /*
      Combinación bit a bit a lo largo de las filas, usando x ← x ⊕ (¬ y & z ) . Para ser precisos, a
      [ i ][  j ][ k ] ← a [ i ][  j ][ k ] ⊕ (¬ a [ i ][  j + 1 ][ k ] &a a
      [ i ][  j + 2 ][ k ]) . Esta es la única operación no lineal en SHA-3.
      */
      //---
      ulong row0 = b[5 * y], row1 = b[1 + 5 * y], row2 = b[2 + 5 * y], row3 = b[3 + 5 * y], row4 = b[4 + 5 * y];

      //---
      s[5 * y]   = row0 ^ ((~row1) & row2);
      s[1 + 5 * y] = row1 ^ ((~row2) & row3);
      s[2 + 5 * y] = row2 ^ ((~row3) & row4);
      s[3 + 5 * y] = row3 ^ ((~row4) & row0);
      s[4 + 5 * y] = row4 ^ ((~row0) & row1);
     }

    //--- iota
    s[0] ^= rc[r];
   }
 }



//+------------------------------------------------------------------+
//|                                                                  |
//+------------------------------------------------------------------+
#define SHA3_CHUNKT_IMP(BYTES, DLIM, P) \
struct SHA3_CHUNKT_BUFFER_##P     \
  {                               \
   const uchar data[BYTES];       \
  };                              \
struct SHA3_DLIM_BUFFER_##P       \
  {                               \
   const uchar data[DLIM];        \
  };
  
//---
SHA3_CHUNKT_IMP(SHA3_256_BLOCK_SIZE, SHA3_DELIM_SHA3, SHA3_256)
SHA3_CHUNKT_IMP(SHA3_384_BLOCK_SIZE, SHA3_DELIM_SHA3, SHA3_384)
SHA3_CHUNKT_IMP(SHA3_512_BLOCK_SIZE, SHA3_DELIM_SHA3, SHA3_512)
SHA3_CHUNKT_IMP(SHA3_SHAKE_128_BLOCK_SIZE, SHA3_DELIM_SHAKE, SHA3_SHAKE_128)
SHA3_CHUNKT_IMP(SHA3_SHAKE_256_BLOCK_SIZE, SHA3_DELIM_SHAKE, SHA3_SHAKE_256)


//+------------------------------------------------------------------+
//| Esepcific implemtantion                                          |
//+------------------------------------------------------------------+
static __forceinline void CCryptoHashSHA3::SHA3_256(const uchar &in[], int ins, uchar &out[], const int o_s)
 {
  CCryptoHashSHA3::SHA3_HASH<SHA3_CHUNKT_BUFFER_SHA3_256, SHA3_DLIM_BUFFER_SHA3_256>(in, ins, out, SHA3_256_OUT_BSIZE, o_s);
 }
static __forceinline void CCryptoHashSHA3::SHA3_384(const uchar &in[], int ins, uchar &out[], const int o_s)
 {
  CCryptoHashSHA3::SHA3_HASH<SHA3_CHUNKT_BUFFER_SHA3_384, SHA3_DLIM_BUFFER_SHA3_384>(in, ins, out, SHA3_384_OUT_BSIZE, o_s);
 }
static __forceinline void CCryptoHashSHA3::SHA3_512(const uchar &in[], int ins, uchar &out[], const int o_s)
 {
  CCryptoHashSHA3::SHA3_HASH<SHA3_CHUNKT_BUFFER_SHA3_512, SHA3_DLIM_BUFFER_SHA3_512>(in, ins, out, SHA3_512_OUT_BSIZE, o_s);
 }
static __forceinline void CCryptoHashSHA3::SHAKE128(const uchar &in[], int ins, uchar &out[], int outl, const int o_s)
 {
  CCryptoHashSHA3::SHA3_HASH<SHA3_CHUNKT_BUFFER_SHA3_SHAKE_128, SHA3_DLIM_BUFFER_SHA3_SHAKE_128>(in, ins, out, outl, o_s);
 }
static __forceinline void CCryptoHashSHA3::SHAKE256(const uchar &in[], int ins, uchar &out[], int outl, const int o_s)
 {
  CCryptoHashSHA3::SHA3_HASH<SHA3_CHUNKT_BUFFER_SHA3_SHAKE_256, SHA3_DLIM_BUFFER_SHA3_SHAKE_256>(in, ins, out, outl, o_s);
 }

//---
}
//---
#endif // CRYPTOBYLEO_SRC_HASH_SHA3_MQH
//+------------------------------------------------------------------+