无法使用“in”运算符在未定义中搜索“d”

Question

我正在尝试使用实现特里树的深度优先搜索算法在 JavaScript 中编写一个难以解决的问题。构建 trie 时，我没有收到任何错误；但是，当 trie 到达 dfs 算法时，它给了我“类型：错误，不能使用 'in' 运算符在未定义中搜索 'd'。

我无法确定问题出在我的 generate_trie 函数还是我的 depth_first_search 算法中。您可以在下面找到该程序的所有代码。

// Sample Boggle Dictionary
var boggle_dxctionary = ['apple', 'pickle', 'side',
  'sick', 'mood', 'cat',
  'cats', 'man', 'super',
  'antman', 'godzilla', 'dog',
  'dot', 'sine', 'cos',
  'signal', 'bitcoin', 'cool',
  'kick', 'zapper'
];

// Sample Boggle Board
var boggle_board = [
  ['c', 'n', 't', ],
  ['d', 'a', 't', ],
  ['o', 'o', 'm', ],
];

var column_length = boggle_board[0].length;
var row_length = boggle_board.length;
var trie_node = {
  'valid': false,
  'next': {}
};

var neighbors_delta = [
  [-1, -1],
  [-1, 0],
  [-1, 1],
  [0, -1],
  [0, 1],
  [1, -1],
  [1, 0],
  [1, 1],
];

function generate_trie(word, node) {
  if (!(word)) {
    return;
  }
  if ((word[0] in node) == false) {
    node[word[0]] = {
      'valid': (word.length == 1),
      'next': {}
    };
  }
  generate_trie(word.slice(1, ), node[word[0]]);
}

function build_trie(boggle_dxct, trie) {
  for (word = 0; word < boggle_dxct.length; word++) {
    generate_trie(boggle_dxct[word], trie);
  }
  return trie;
}

function get_neighbors(row, column) {
  var neighbors = [];

  for (neighbor = 0; neighbor < neighbors_delta.length; neighbor++) {
    new_row = row + neighbors_delta[neighbor][0];
    new_column = column + neighbors_delta[neighbor][1];

    if (new_row >= row_length || new_column >= column_length || new_row < 0 || new_column < 0) {
      continue;
    }
    neighbors.push([new_row, new_column]);
  }
  return neighbors;
}


function depth_first_search(row, column, visited, trie, current_word, found_words, board) {
  var row_column_pair = [row, column];
  console.log(row_column_pair);

  if (row_column_pair in visited) {
    return;
  }

  letter = board[row][column];
  visited.push(row_column_pair);

  console.log("Up to here is good3");
  console.log(letter);
  if (letter in trie) {
    current_word = current_word + letter;
    console.log("Up to here is good4");
    if (trie[letter]['valid']) {
      found_words.push(current_word);
    }

    neighbors = get_neighbors(row, column);
    for (n = 0; n < neighbors.length; n++) {
      console.log("Up to here is good5");
      depth_first_search(neighbors[n][0], neighbors[n][1], visited.slice(0, visited.length), trie[letter], current_word, found_words, board);
      console.log("Up to here is good6");
    }
  }
}

function main(trie_node, board) {
  trie_node = build_trie(boggle_dxctionary, trie_node);
  found_words = [];

  for (r = 0; r < row_length; r++) {
    console.log("Up to here is good1");
    for (c = 0; c < column_length; c++) {
      var visited = [];
      var current_word = '';
      depth_first_search(r, c, visited, trie_node, current_word, found_words, board);
      console.log("Up to here is good2");
    }
  }
  console.log(found_words);
}
main(trie_node,boggle_board);

Answer 1

问题是您没有将letter 声明为depth_first_search 中的局部变量。所以当它在for循环中调用自己时，递归调用会改变letter的值，而trie[letter]什么也找不到。

总是使用局部变量，除非你有特定的理由不这样做。我在下面的代码中添加了var 声明。

// Sample Boggle Dictionary
var boggle_dxctionary = ['apple', 'pickle', 'side',
  'sick', 'mood', 'cat',
  'cats', 'man', 'super',
  'antman', 'godzilla', 'dog',
  'dot', 'sine', 'cos',
  'signal', 'bitcoin', 'cool',
  'kick', 'zapper'
];

// Sample Boggle Board
var boggle_board = [
  ['c', 'n', 't', ],
  ['d', 'a', 't', ],
  ['o', 'o', 'm', ],
];

var column_length = boggle_board[0].length;
var row_length = boggle_board.length;
var trie_node = {
  'valid': false,
  'next': {}
};

var neighbors_delta = [
  [-1, -1],
  [-1, 0],
  [-1, 1],
  [0, -1],
  [0, 1],
  [1, -1],
  [1, 0],
  [1, 1],
];

function generate_trie(word, node) {
  if (!(word)) {
    return;
  }
  if ((word[0] in node) == false) {
    node[word[0]] = {
      'valid': (word.length == 1),
      'next': {}
    };
  }
  generate_trie(word.slice(1, ), node[word[0]]);
}

function build_trie(boggle_dxct, trie) {
  for (var word = 0; word < boggle_dxct.length; word++) {
    generate_trie(boggle_dxct[word], trie);
  }
  return trie;
}

function get_neighbors(row, column) {
  var neighbors = [];

  for (var neighbor = 0; neighbor < neighbors_delta.length; neighbor++) {
    var new_row = row + neighbors_delta[neighbor][0];
    var new_column = column + neighbors_delta[neighbor][1];

    if (new_row >= row_length || new_column >= column_length || new_row < 0 || new_column < 0) {
      continue;
    }
    neighbors.push([new_row, new_column]);
  }
  return neighbors;
}


function depth_first_search(row, column, visited, trie, current_word, found_words, board) {
  var row_column_pair = [row, column];
  console.log(row_column_pair);

  if (row_column_pair in visited) {
    return;
  }

  var letter = board[row][column];
  visited.push(row_column_pair);

  console.log("Up to here is good3");
  console.log(letter);
  if (letter in trie) {
    current_word = current_word + letter;
    console.log("Up to here is good4");
    if (trie[letter]['valid']) {
      found_words.push(current_word);
    }

    var neighbors = get_neighbors(row, column);
    for (n = 0; n < neighbors.length; n++) {
      console.log("Up to here is good5");
      depth_first_search(neighbors[n][0], neighbors[n][1], visited.slice(0, visited.length), trie[letter], current_word, found_words, board);
      console.log("Up to here is good6");
    }
  }
}

function main(trie_node, board) {
  trie_node = build_trie(boggle_dxctionary, trie_node);
  var found_words = [];

  for (r = 0; r < row_length; r++) {
    console.log("Up to here is good1");
    for (c = 0; c < column_length; c++) {
      var visited = [];
      var current_word = '';
      depth_first_search(r, c, visited, trie_node, current_word, found_words, board);
      console.log("Up to here is good2");
    }
  }
  console.log(found_words);
}

main(trie_node, boggle_board);