// SPDX-License-Identifier: Apache-2.0
// Copyright 2024-2026, Amlal El Mahrouss (amlal@nekernel.org)
// Licensed under the Apache License, Version 2.0 (see LICENSE file)
// Official repository: https://github.com/ne-foss-org/nekernel

#ifndef QR_H
#define QR_H

#include <BootKit/QrPrelude.h>
#include <BootKit/Shared/base.h>
#include <BootKit/Shared/bit.h>
#include <BootKit/Support.h>
#include <CompilerKit/Detail.h>
#include <NeKit/KernelPanic.h>
#include <modules/CoreGfx/CoreGfx.h>

/// @note the QR code is still code 128, it utilizes the same concept of having it's own character
/// set.

namespace qr {
inline uint8_t min_poly = 0b11101, /* Minimal polynomial x^8 + x^4 + x^3 + x^2 + 1 */
    generator           = 0b10;    /* Generator of Galois field */

/// @brief galois finite field multiplication.
inline uint8_t gf_mul(uint8_t a, uint8_t b) {
  uint8_t res = 0;

  for (; b; b >>= 1) {
    if (b & 1) res ^= a;
    if (a & 0x80)
      a = (a << 1) ^ min_poly;
    else
      a <<= 1;
  }

  return res;
}

// Size of Ecc block with respect to level and version. 0 version is for
// padding.
constexpr int ECC_CODEWORDS_PER_BLOCK[4][41] = {
    {0,  7,  10, 15, 20, 26, 18, 20, 24, 30, 18, 20, 24, 26, 30, 22, 24, 28, 30, 28, 28,
     28, 28, 30, 30, 26, 28, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30},
    {0,  10, 16, 26, 18, 24, 16, 18, 22, 22, 26, 30, 22, 22, 24, 24, 28, 28, 26, 26, 26,
     26, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28, 28},
    {0,  13, 22, 18, 26, 18, 24, 18, 22, 20, 24, 28, 26, 24, 20, 30, 24, 28, 28, 26, 30,
     28, 30, 30, 30, 30, 28, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30},
    {0,  17, 28, 22, 16, 22, 28, 26, 26, 24, 28, 24, 28, 22, 24, 24, 30, 28, 28, 26, 28,
     30, 24, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30, 30},
};

// Number of Ecc blocks with respect to level and version. 0 version is for
// padding.
constexpr int N_ECC_BLOCKS[4][41] = {
    {0, 1, 1, 1,  1,  1,  2,  2,  2,  2,  4,  4,  4,  4,  4,  6,  6,  6,  6,  7, 8,
     8, 9, 9, 10, 12, 12, 12, 13, 14, 15, 16, 17, 18, 19, 19, 20, 21, 22, 24, 25},
    {0,  1,  1,  1,  2,  2,  4,  4,  4,  5,  5,  5,  8,  9,  9,  10, 10, 11, 13, 14, 16,
     17, 17, 18, 20, 21, 23, 25, 26, 28, 29, 31, 33, 35, 37, 38, 40, 43, 45, 47, 49},
    {0,  1,  1,  2,  2,  4,  4,  6,  6,  8,  8,  8,  10, 12, 16, 12, 17, 16, 18, 21, 20,
     23, 23, 25, 27, 29, 34, 34, 35, 38, 40, 43, 45, 48, 51, 53, 56, 59, 62, 65, 68},
    {0,  1,  1,  2,  4,  4,  4,  5,  6,  8,  8,  11, 11, 16, 16, 18, 16, 19, 21, 25, 25,
     25, 34, 30, 32, 35, 37, 40, 42, 45, 48, 51, 54, 57, 60, 63, 66, 70, 74, 77, 81},
};

// Positions of central modules of alignment patterns according to version. 0
// version is for padding.
constexpr int ALIGN_POS[41][7] = {
    {},
    {0},
    {6, 18},
    {6, 22},
    {6, 26},
    {6, 30},
    {6, 34},
    {6, 22, 38},
    {6, 24, 42},
    {6, 26, 46},
    {6, 28, 50},
    {6, 30, 54},
    {6, 32, 58},
    {6, 34, 62},
    {6, 26, 46, 66},
    {6, 26, 48, 70},
    {6, 26, 50, 74},
    {6, 30, 54, 78},
    {6, 30, 56, 82},
    {6, 30, 58, 86},
    {6, 34, 62, 90},
    {6, 28, 50, 72, 94},
    {6, 26, 50, 74, 98},
    {6, 30, 54, 78, 102},
    {6, 28, 54, 80, 106},
    {6, 32, 58, 84, 110},
    {6, 30, 58, 86, 114},
    {6, 34, 62, 90, 118},
    {6, 26, 50, 74, 98, 122},
    {6, 30, 54, 78, 102, 126},
    {6, 26, 52, 78, 104, 130},
    {6, 30, 56, 82, 108, 134},
    {6, 34, 60, 86, 112, 138},
    {6, 30, 58, 86, 114, 142},
    {6, 34, 62, 90, 118, 146},
    {6, 30, 54, 78, 102, 126, 150},
    {6, 24, 50, 76, 102, 128, 154},
    {6, 28, 54, 80, 106, 132, 158},
    {6, 32, 58, 84, 110, 136, 162},
    {6, 26, 54, 82, 110, 138, 166},
    {6, 30, 58, 86, 114, 142, 170},
};

// Return n-th bit of arr starting from MSB.
constexpr uint8_t get_bit_r(uint8_t* arr, int n) {
  return (arr[n >> 3] >> (7 - (n & 7))) & 1;
}

// Add up to 16 bits to arr. Data starts from MSB as well as each byte of an
// array.
constexpr void add_bits(uint16_t data, int n, uint8_t* arr, size_t& pos) {
  while (n--) {
    arr[pos >> 3] |= ((data >> n) & 1) << (7 - (pos & 7));
    ++pos;
  }
}

// Translate char to alphanumeric encoding value,
constexpr int alphanumeric(char c) {
  if (c >= '0' && c <= '9') return c - '0';

  if (c >= 'A' && c <= 'Z') return c - 'A' + 10;

  switch (c) {
    case ' ':
      return 36;
    case '$':
      return 37;
    case '%':
      return 38;
    case '*':
      return 39;
    case '+':
      return 40;
    case '-':
      return 41;
    case '.':
      return 42;
    case '/':
      return 43;
    case ':':
      return 44;
  }
  return -1;
}

// Check if string can be encoded in alphanumeric mode.
constexpr bool is_alphanumeric(const char* str, size_t len) {
  for (size_t i = 0; i < len; ++i)
    if (alphanumeric(str[i]) == -1) return false;
  return true;
}

// Check if string can be encoded in numeric mode.
constexpr bool is_numeric(const char* str, size_t len) {
  for (size_t i = 0; i < len; ++i)
    if (str[i] < '0' || str[i] > '9') return false;
  return true;
}

// Check if string can be encoded in kanji mode.
constexpr bool is_kanji(const char* str, size_t len) {
  for (size_t i = 0; i < len; i += 2) {
    uint16_t val = uint16_t(str[i]) | (uint16_t(str[i + 1]) << 8);
    if (val < 0x8140 || val > 0xebbf || (val > 0x9ffc && val < 0xe040)) return false;
  }
  return true;
}

// Reed-Solomon Ecc generator polynomial for the given degree.
constexpr void gf_gen_poly(int degree, uint8_t* poly) {
  SetMem(poly, 0, degree);

  uint8_t root = poly[degree - 1] = 1;

  for (int i = 0; i < degree; ++i) {
    for (int j = 0; j < degree - 1; ++j) poly[j] = gf_mul(poly[j], root) ^ poly[j + 1];
    poly[degree - 1] = gf_mul(poly[degree - 1], root);
    root             = (root << 1) ^ ((root >> 7) * 0x11d);
  }
}

// Polynomial division if Galois Field.
constexpr void gf_poly_div(uint8_t* dividend, size_t len, uint8_t* divisor, int degree,
                           uint8_t* result) {
  SetMem(result, 0, degree);

  for (size_t i = 0; i < len; ++i) {
    uint8_t factor = dividend[i] ^ result[0];
    MoveMem(&result[0], &result[1], degree - 1);
    result[degree - 1] = 0;
    for (int j = 0; j < degree; ++j) result[j] ^= gf_mul(divisor[j], factor);
  }
}

enum Ecc {
  L,
  M,
  Q,
  H,
};

enum Mode {
  M_NUMERIC,
  M_ALPHANUMERIC,
  M_BYTE,
  M_KANJI,
};

// Select appropriate encoding mode for string.
constexpr Mode select_mode(const char* str, size_t len) {
  if (is_numeric(str, len)) return M_NUMERIC;
  if (is_alphanumeric(str, len)) return M_ALPHANUMERIC;
  if (is_kanji(str, len)) return M_KANJI;
  return M_BYTE;
}

// Return size of Character Control Indicator in bits for given version and
// mode.
constexpr int cci(int ver, Mode mode) {
  constexpr int cnt[4][3] = {
      {10, 12, 14},
      {9, 11, 13},
      {8, 16, 16},
      {8, 10, 12},
  };
  if (ver < 10) return cnt[mode][0];
  if (ver < 27) return cnt[mode][1];
  return cnt[mode][2];
}

template <int V>
struct Qr {
 private:
  friend class QrDelegate;
  bool draw(int x, int y);

 public:
  constexpr auto side_size() const { return SIDE; }

  bool module(int x, int y);
  bool encode(const char* str, size_t len, Ecc ecc, int mask = -1);

 private:
  bool encode_data(const char* data, size_t len, Ecc ecc, uint8_t* out);
  void encode_ecc(uint8_t* data, Ecc ecc, uint8_t* out);

  void add_data(uint8_t* data, uint8_t* patterns);
  void add_patterns();
  void add_version();
  void add_format(Ecc ecc, int mask);
  void reserve_patterns(uint8_t* out);

  template <bool Black>
  void draw_rect(int y, int x, int height, int width, uint8_t* out);
  template <bool Black>
  void draw_bound(int y, int x, int height, int width, uint8_t* out);

  template <bool Horizontal>
  int  rule_1_3_score();
  int  penalty_score();
  int  select_mask(Ecc ecc, uint8_t* patterns);
  void apply_mask(int mask, uint8_t* patterns);

 private:
  STATIC_PASS(V >= 1 && V <= 40, "invalid version");
  static constexpr int SIDE          = 17 + V * 4;
  static constexpr int N_BITS        = SIDE * SIDE;
  static constexpr int N_ALIGN       = V == 1 ? 0 : V / 7 + 2;
  static constexpr int N_ALIGN_BITS  = V > 1 ? (N_ALIGN* N_ALIGN - 3) * 25 : 0;
  static constexpr int N_TIMING_BITS = (SIDE - 16) * 2 - (10 * (V > 1 ? N_ALIGN - 2 : 0));
  static constexpr int N_VER_BITS    = V > 6 ? 36 : 0;
  static constexpr int N_DAT_BITS = N_BITS - (192 + N_ALIGN_BITS + N_TIMING_BITS + 31 + N_VER_BITS);
  static constexpr int N_BYTES    = utl::bytes_in_bits(N_BITS);  // Actual number of bytes_in_bits
                                                                 // required to store whole Qr code
  static constexpr int N_DAT_BYTES =
      utl::bytes_in_bits(N_DAT_BITS);  // Actual number of bytes_in_bits required to store
                                       // [data + ecc]
  static constexpr int N_DAT_CAPACITY =
      N_DAT_BITS >> 3;  // Capacity of [data + ecc] without remainder bits
 private:
  /// @brief internal function to retrieve bit from a bitset.
  uint8_t get_arr_bit(uint8_t* arr, unsigned bit) const { return utl::get_arr_bit(arr, bit); }

  /// @brief internal function to set bit from a bitset.
  void set_arr_bit(uint8_t* arr, unsigned bit) { utl::set_arr_bit(arr, bit); }

  /// @brief internal function to clear bit from a bitset.
  void clr_arr_bit(uint8_t* arr, unsigned bit) { utl::clr_arr_bit(arr, bit); }

  uint8_t code[N_BYTES] = {};

  bool status = false;
};

// Get color of a module from left-to-right and top-to-bottom. Black is true.
template <int V>
bool Qr<V>::module(int x, int y) {
  return get_arr_bit(code, y * SIDE + x);
}

/// @brief draw a new QR code.
template <int V>
bool Qr<V>::draw(int whereX, int whereY) {
  if (!this->status) return false;  // it may be invalid.

  cg_init();

  for (int y = 0; y < (this->side_size()); ++y) {
    for (int x = 0; x < (this->side_size()); ++x) {
      FBDrawInRegion((this->module(x, y) ? RGB(00, 00, 00) : RGB(0xFF, 0xFF, 0xFF)), 1, 1,
                     x + whereX, y + whereY);
    }
  }

  cg_clear();

  return false;
}

// Create Qr code with given error correction level. If mask == -1,
// then best mask selected automatically. NOTE: Automatic mask is the
// most expensive operation. Takes about 95 % of all computation time.
template <int V>
bool Qr<V>::encode(const char* str, size_t len, Ecc ecc, int mask) {
  uint8_t data[N_DAT_BYTES]          = {};
  uint8_t data_with_ecc[N_DAT_BYTES] = {};
  uint8_t patterns[N_BYTES]          = {};

  if (!encode_data(str, len, ecc, data)) {
    return status = false;
  }

  encode_ecc(data, ecc, data_with_ecc);

  reserve_patterns(patterns);
  CopyMem(code, patterns, N_BYTES);

  add_data(data_with_ecc, patterns);
  add_patterns();
  add_version();

  mask = mask != -1 ? mask & 7 : select_mask(ecc, patterns);

  add_format(ecc, mask);
  apply_mask(mask, patterns);

  return status = true;
}

template <int V>
bool Qr<V>::encode_data(const char* data, size_t len, Ecc ecc, uint8_t* out) {
  Mode mode = select_mode(data, len);

  size_t n_bits = (N_DAT_CAPACITY - ECC_CODEWORDS_PER_BLOCK[ecc][V] * N_ECC_BLOCKS[ecc][V]) << 3;
  size_t pos    = 0;

  add_bits(1 << mode, 4, out, pos);
  add_bits(len, cci(V, mode), out, pos);

  if (mode == M_NUMERIC) {
    const size_t triplets      = len / 3;
    const size_t triplets_size = triplets * 3;
    const size_t rem           = len % 3;
    const size_t rem_bits      = rem == 2 ? 7 : rem == 1 ? 4 : 0;
    const size_t total_size    = 10 * triplets + rem_bits;

    if (pos + total_size > n_bits) return false;

    char buf[4] = {};

    for (size_t i = 0; i < triplets_size; i += 3) {
      buf[0] = data[i];
      buf[1] = data[i + 1];
      buf[2] = data[i + 2];

      uint16_t num = StringToLong(buf, NULL, 10);
      add_bits(num, 10, out, pos);
    }

    if (rem) {
      buf[0]   = data[triplets_size];
      buf[1]   = data[triplets_size + 1];
      buf[rem] = 0;

      uint16_t num = StringToLong(buf, NULL, 10);
      add_bits(num, rem_bits, out, pos);
    }
  } else if (mode == M_ALPHANUMERIC) {
    if (pos + 11 * (int(len & ~1ul) / 2) > n_bits) return false;

    for (int i = 0; i < int(len & ~1ul); i += 2) {
      uint16_t num = alphanumeric(data[i]) * 45 + alphanumeric(data[i + 1]);
      add_bits(num, 11, out, pos);
    }
    if (len & 1) {
      if (pos + 6 > n_bits) return false;

      add_bits(alphanumeric(data[len - 1]), 6, out, pos);
    }
  } else if (mode == M_BYTE) {
    if (pos + len * 8 > n_bits) return false;

    for (size_t i = 0; i < len; ++i) add_bits(data[i], 8, out, pos);
  } else {
    if (pos + 13 * (len / 2) > n_bits) return false;

    for (size_t i = 0; i < len; i += 2) {
      uint16_t val = ((uint8_t) data[i]) | (((uint8_t) data[i + 1]) << 8);
      uint16_t res = 0;
      val -= val < 0x9FFC ? 0x8140 : 0xC140;
      res += val & 0xff;
      res += (val >> 8) * 0xc0;
      add_bits(res, 13, out, pos);
    }
  }

  size_t padding = n_bits - pos;
  size_t i       = 0;

  add_bits(0, padding > 4 ? 4 : padding, out, pos);

  if (pos & 7) add_bits(0, (8 - pos) & 7, out, pos);

  while (pos < n_bits) add_bits(++i & 1 ? 0xec : 0x11, 8, out, pos);

  return true;
}

template <int V>
void Qr<V>::encode_ecc(uint8_t* data, Ecc ecc, uint8_t* out) {
  int n_blocks = N_ECC_BLOCKS[ecc][V];
  int ecc_len  = ECC_CODEWORDS_PER_BLOCK[ecc][V];

  int n_data_bytes = N_DAT_CAPACITY - ecc_len * n_blocks;

  int n_short_blocks = n_blocks - N_DAT_CAPACITY % n_blocks;
  int short_len      = N_DAT_CAPACITY / n_blocks - ecc_len;

  uint8_t gen_poly[30];
  uint8_t ecc_buf[30];

  gf_gen_poly(ecc_len, gen_poly);

  uint8_t* data_ptr = data;

  for (int i = 0; i < n_blocks; ++i) {
    int data_len = short_len;

    if (i >= n_short_blocks) ++data_len;

    gf_poly_div(data_ptr, data_len, gen_poly, ecc_len, ecc_buf);

    for (int j = 0, k = i; j < data_len; ++j, k += n_blocks) {
      if (j == short_len) k -= n_short_blocks;
      out[k] = data_ptr[j];
    }
    for (int j = 0, k = n_data_bytes + i; j < ecc_len; ++j, k += n_blocks) out[k] = ecc_buf[j];

    data_ptr += data_len;
  }
}

template <int V>
void Qr<V>::add_data(uint8_t* data, uint8_t* patterns) {
  int data_pos = 0;

  for (int x = SIDE - 1; x >= 1; x -= 2) {
    if (x == 6) x = 5;

    for (int i = 0; i < SIDE; ++i) {
      int y     = !((x + 1) & 2) ? SIDE - 1 - i : i;
      int coord = y * SIDE + x;

      if (!get_arr_bit(patterns, coord)) {
        if (get_bit_r(data, data_pos)) set_arr_bit(code, coord);

        ++data_pos;
      }

      if (!get_arr_bit(patterns, coord - 1)) {
        if (get_bit_r(data, data_pos)) set_arr_bit(code, coord - 1);

        ++data_pos;
      }
    }
  }
}

template <int V>
void Qr<V>::add_patterns() {
  // White bounds inside finders
  draw_bound<false>(1, 1, 5, 5, code);
  draw_bound<false>(1, SIDE - 6, 5, 5, code);
  draw_bound<false>(SIDE - 6, 1, 5, 5, code);

  // Finish alignment patterns
  for (int i = 0; i < N_ALIGN; ++i) {
    for (int j = 0; j < N_ALIGN; ++j) {
      if ((!i && !j) || (!i && j == N_ALIGN - 1) || (!j && i == N_ALIGN - 1)) continue;
      draw_bound<false>(ALIGN_POS[V][i] - 1, ALIGN_POS[V][j] - 1, 3, 3, code);
    }
  }

  // Draw white separators
  draw_rect<false>(7, 0, 1, 8, code);
  draw_rect<false>(0, 7, 8, 1, code);
  draw_rect<false>(SIDE - 8, 0, 1, 8, code);
  draw_rect<false>(SIDE - 8, 7, 8, 1, code);
  draw_rect<false>(7, SIDE - 8, 1, 8, code);
  draw_rect<false>(0, SIDE - 8, 8, 1, code);

  // Perforate timing patterns
  for (int i = 7; i < SIDE - 7; i += 2) {
    clr_arr_bit(code, 6 * SIDE + i);
    clr_arr_bit(code, i * SIDE + 6);
  }
}

template <int V>
void Qr<V>::add_version() {
  if (V < 7) return;

  uint32_t rem = V;

  for (uint8_t i = 0; i < 12; ++i) rem = (rem << 1) ^ ((rem >> 11) * 0x1F25);

  uint32_t data = V << 12 | rem;

  for (int x = 0; x < 6; ++x) {
    for (int j = 0; j < 3; ++j) {
      int y = SIDE - 11 + j;

      bool black = (data >> (x * 3 + j)) & 1;

      if (!black) {
        clr_arr_bit(code, y * SIDE + x);
        clr_arr_bit(code, y + SIDE * x);
      }
    }
  }
}

template <int V>
void Qr<V>::add_format(Ecc ecc, int mask) {
  int data = (ecc ^ 1) << 3 | mask;
  int rem  = data;

  for (int i = 0; i < 10; i++) rem = (rem << 1) ^ ((rem >> 9) * 0b10100110111);

  int res = (data << 10 | rem) ^ 0b101010000010010;

  for (int i = 0; i < 6; ++i) {
    if ((res >> i) & 1) {
      set_arr_bit(code, SIDE * 8 + SIDE - 1 - i);
      set_arr_bit(code, SIDE * i + 8);
    } else {
      clr_arr_bit(code, SIDE * 8 + SIDE - 1 - i);
      clr_arr_bit(code, SIDE * i + 8);
    }
  }

  for (int i = 6; i < 8; ++i) {
    if ((res >> i) & 1) {
      set_arr_bit(code, SIDE * 8 + SIDE - 1 - i);
      set_arr_bit(code, SIDE * (i + 1) + 8);
    } else {
      clr_arr_bit(code, SIDE * 8 + SIDE - 1 - i);
      clr_arr_bit(code, SIDE * (i + 1) + 8);
    }
  }

  if ((res >> 8) & 1) {
    set_arr_bit(code, SIDE * 8 + 7);
    set_arr_bit(code, SIDE * (SIDE - 7) + 8);
  } else {
    clr_arr_bit(code, SIDE * 8 + 7);
    clr_arr_bit(code, SIDE * (SIDE - 7) + 8);
  }

  for (int i = 9, j = 5; i < 15; ++i, --j) {
    if ((res >> i) & 1) {
      set_arr_bit(code, SIDE * 8 + j);
      set_arr_bit(code, SIDE * (SIDE - 1 - j) + 8);
    } else {
      clr_arr_bit(code, SIDE * 8 + j);
      clr_arr_bit(code, SIDE * (SIDE - 1 - j) + 8);
    }
  }
}

template <int V>
template <bool B>
void Qr<V>::draw_rect(int y, int x, int height, int width, uint8_t* out) {
  if (B) {
    for (int dy = y * SIDE; dy < (y + height) * SIDE; dy += SIDE)
      for (int dx = x; dx < x + width; ++dx) set_arr_bit(out, dy + dx);
  } else {
    for (int dy = y * SIDE; dy < (y + height) * SIDE; dy += SIDE)
      for (int dx = x; dx < x + width; ++dx) clr_arr_bit(out, dy + dx);
  }
}

template <int V>
template <bool B>
void Qr<V>::draw_bound(int y, int x, int height, int width, uint8_t* out) {
  if (B) {
    for (int i = y * SIDE + x; i < y * SIDE + x + width; ++i) set_arr_bit(out, i);
    for (int i = (y + height - 1) * SIDE + x; i < (y + height - 1) * SIDE + x + width; ++i)
      set_arr_bit(out, i);
    for (int i = (y + 1) * SIDE + x; i < (y + height - 1) * SIDE + x; i += SIDE)
      set_arr_bit(out, i);
    for (int i = (y + 1) * SIDE + x + width - 1; i < (y + height - 1) * SIDE + x + width - 1;
         i += SIDE)
      set_arr_bit(out, i);
  } else {
    for (int i = y * SIDE + x; i < y * SIDE + x + width; ++i) clr_arr_bit(out, i);
    for (int i = (y + height - 1) * SIDE + x; i < (y + height - 1) * SIDE + x + width; ++i)
      clr_arr_bit(out, i);
    for (int i = (y + 1) * SIDE + x; i < (y + height - 1) * SIDE + x; i += SIDE)
      clr_arr_bit(out, i);
    for (int i = (y + 1) * SIDE + x + width - 1; i < (y + height - 1) * SIDE + x + width - 1;
         i += SIDE)
      clr_arr_bit(out, i);
  }
}

template <int V>
void Qr<V>::reserve_patterns(uint8_t* out) {
  draw_rect<true>(0, 6, SIDE, 1, out);
  draw_rect<true>(6, 0, 1, SIDE, out);

  draw_rect<true>(0, 0, 9, 9, out);
  draw_rect<true>(SIDE - 8, 0, 8, 9, out);
  draw_rect<true>(0, SIDE - 8, 9, 8, out);

  for (int i = 0; i < N_ALIGN; ++i) {
    for (int j = 0; j < N_ALIGN; ++j) {
      if ((!i && !j) || (!i && j == N_ALIGN - 1) || (!j && i == N_ALIGN - 1)) continue;
      draw_rect<true>(ALIGN_POS[V][i] - 2, ALIGN_POS[V][j] - 2, 5, 5, out);
    }
  }

  if (V >= 7) {
    draw_rect<true>(SIDE - 11, 0, 3, 6, out);
    draw_rect<true>(0, SIDE - 11, 6, 3, out);
  }
}

template <int V>
template <bool H>
int Qr<V>::rule_1_3_score() {
  constexpr int y_max  = H ? N_BITS : SIDE;
  constexpr int x_max  = H ? SIDE : N_BITS;
  constexpr int y_step = H ? SIDE : 1;
  constexpr int x_step = H ? 1 : SIDE;

  int res = 0;

  for (int y = 0; y < y_max; y += y_step) {
    bool color  = get_arr_bit(code, y);
    int  finder = color;
    int  cnt    = 1;
    for (int x = 1; x < x_max; x += x_step) {
      if (get_arr_bit(code, y + x) == color) {
        ++cnt;
        if (cnt == 5) res += 3;
        if (cnt > 5) ++res;
      } else {
        color = !color;
        cnt   = 1;
      }
      // Finder-like
      finder = ((finder << 1) & 0x7ff) | color;
      if (x >= x_step * 10) {
        if (finder == 0x05d || finder == 0x5d0) res += 40;
      }
    }
  }
  return res;
}

template <int V>
int Qr<V>::penalty_score() {
  int res = 0;

  res += rule_1_3_score<true>();
  res += rule_1_3_score<false>();

  for (int y = 0; y < N_BITS - SIDE; y += SIDE) {
    for (int x = 0; x < SIDE - 1; ++x) {
      bool c = get_arr_bit(code, y + x);

      if (c == get_arr_bit(code, y + x + 1) && c == get_arr_bit(code, y + x + SIDE) &&
          c == get_arr_bit(code, y + x + SIDE + 1))
        res += 3;
    }
  }

  int black = 0;
  for (int y = 0; y < N_BITS; y += SIDE) {
    for (int x = 0; x < SIDE; ++x) black += get_arr_bit(code, y + x);
  }
  res += abs((black * 100) / N_BITS - 50) / 5 * 10;

  return res;
}

template <int V>
int Qr<V>::select_mask(Ecc ecc, uint8_t* patterns) {
  unsigned min_score = -1;
  unsigned score     = 0;
  uint8_t  mask      = 0;

  for (int i = 0; i < 8; ++i) {
    add_format(ecc, i);
    apply_mask(i, patterns);
    score = penalty_score();
    if (score < min_score) {
      mask      = i;
      min_score = score;
    }
    apply_mask(i, patterns);
  }
  return mask;
}

template <int V>
void Qr<V>::apply_mask(int mask, uint8_t* patterns) {
  for (int y = 0, dy = 0; y < SIDE; ++y, dy += SIDE) {
    for (int x = 0; x < SIDE; ++x) {
      int coord = dy + x;

      if (get_arr_bit(patterns, coord)) continue;

      bool keep = true;

      switch (mask) {
        case 0:
          keep = (x + y) & 1;
          break;
        case 1:
          keep = y & 1;
          break;
        case 2:
          keep = x % 3;
          break;
        case 3:
          keep = (x + y) % 3;
          break;
        case 4:
          keep = (y / 2 + x / 3) & 1;
          break;
        case 5:
          keep = x * y % 2 + x * y % 3;
          break;
        case 6:
          keep = (x * y % 2 + x * y % 3) & 1;
          break;
        case 7:
          keep = ((x + y) % 2 + x * y % 3) & 1;
          break;
      }

      if (!keep) {
        if (get_arr_bit(code, coord))
          clr_arr_bit(code, coord);
        else
          set_arr_bit(code, coord);
      }
    }
  }
}

/// @brief QR code encoder class.
class QrDelegate final {
 public:
  explicit QrDelegate() = default;
  ~QrDelegate()         = default;

  NE_COPY_DEFAULT(QrDelegate)

  /// @brief Draw method delegate.
  template <Int32 V>
  bool draw(Qr<V>& subject, Int32 x, Int32 y) {
    return subject.draw(x, y);
  }
};
}  // namespace qr

namespace Kernel::Qr {
using namespace qr;
}  // namespace Kernel::Qr

#endif  // QR_H