最近我一直在iPhone上玩一款名为《Scramble》的游戏。有些人可能知道这个游戏叫拼字游戏。从本质上讲,当游戏开始时,你会得到一个字母矩阵:

F X I E
A M L O
E W B X
A S T U

The goal of the game is to find as many words as you can that can be formed by chaining letters together. You can start with any letter, and all the letters that surround it are fair game, and then once you move on to the next letter, all the letters that surround that letter are fair game, except for any previously used letters. So in the grid above, for example, I could come up with the words LOB, TUX, SEA, FAME, etc. Words must be at least 3 characters, and no more than NxN characters, which would be 16 in this game but can vary in some implementations. While this game is fun and addictive, I am apparently not very good at it and I wanted to cheat a little bit by making a program that would give me the best possible words (the longer the word the more points you get).

(来源:boggled.org)

不幸的是,我不太擅长算法或它们的效率等等。我的第一次尝试使用一个像这样的字典(约2.3MB),并进行线性搜索,试图匹配字典条目的组合。这需要花费很长时间来找到可能的单词,因为你每轮只有2分钟的时间,这是不够的。

我很有兴趣看看是否有任何Stackoverflowers可以提出更有效的解决方案。我主要是在寻找使用三大p的解决方案:Python、PHP和Perl,尽管任何使用Java或c++的东西也很酷,因为速度是至关重要的。

目前的解决方案:

Adam Rosenfield, Python, ~20岁 John Fouhy, Python, ~3秒 Kent Fredric, Perl, ~1s Darius Bacon, Python, ~1s rvarcher, VB。净,~ 1 s Paolo Bergantino, PHP(实时链接),~5s(本地~2s)


当前回答

给定一个有N行M列的Boggle板,让我们假设如下:

N*M基本上大于可能单词的数量 N*M基本上大于可能的最长单词

在这些假设下,该解的复杂度为O(N*M)。

我认为比较这个示例板的运行时间在很多方面都没有重点,但是为了完整性,在我的现代MacBook Pro上,这个解决方案在0.2秒内完成。

这个解决方案将为语料库中的每个单词找到所有可能的路径。

#!/usr/bin/env ruby
# Example usage: ./boggle-solver --board "fxie amlo ewbx astu"

autoload :Matrix, 'matrix'
autoload :OptionParser, 'optparse'

DEFAULT_CORPUS_PATH = '/usr/share/dict/words'.freeze

# Functions

def filter_corpus(matrix, corpus, min_word_length)
  board_char_counts = Hash.new(0)
  matrix.each { |c| board_char_counts[c] += 1 }

  max_word_length = matrix.row_count * matrix.column_count
  boggleable_regex = /^[#{board_char_counts.keys.reduce(:+)}]{#{min_word_length},#{max_word_length}}$/
  corpus.select{ |w| w.match boggleable_regex }.select do |w|
    word_char_counts = Hash.new(0)
    w.each_char { |c| word_char_counts[c] += 1 }
    word_char_counts.all? { |c, count| board_char_counts[c] >= count }
  end
end

def neighbors(point, matrix)
  i, j = point
  ([i-1, 0].max .. [i+1, matrix.row_count-1].min).inject([]) do |r, new_i|
    ([j-1, 0].max .. [j+1, matrix.column_count-1].min).inject(r) do |r, new_j|
      neighbor = [new_i, new_j]
      neighbor.eql?(point) ? r : r << neighbor
    end
  end
end

def expand_path(path, word, matrix)
  return [path] if path.length == word.length

  next_char = word[path.length]
  viable_neighbors = neighbors(path[-1], matrix).select do |point|
    !path.include?(point) && matrix.element(*point).eql?(next_char)
  end

  viable_neighbors.inject([]) do |result, point|
    result + expand_path(path.dup << point, word, matrix)
  end
end

def find_paths(word, matrix)
  result = []
  matrix.each_with_index do |c, i, j|
    result += expand_path([[i, j]], word, matrix) if c.eql?(word[0])
  end
  result
end

def solve(matrix, corpus, min_word_length: 3)
  boggleable_corpus = filter_corpus(matrix, corpus, min_word_length)
  boggleable_corpus.inject({}) do |result, w|
    paths = find_paths(w, matrix)
    result[w] = paths unless paths.empty?
    result
  end
end

# Script

options = { corpus_path: DEFAULT_CORPUS_PATH }
option_parser = OptionParser.new do |opts|
  opts.banner = 'Usage: boggle-solver --board <value> [--corpus <value>]'

  opts.on('--board BOARD', String, 'The board (e.g. "fxi aml ewb ast")') do |b|
    options[:board] = b
  end

  opts.on('--corpus CORPUS_PATH', String, 'Corpus file path') do |c|
    options[:corpus_path] = c
  end

  opts.on_tail('-h', '--help', 'Shows usage') do
    STDOUT.puts opts
    exit
  end
end
option_parser.parse!

unless options[:board]
  STDERR.puts option_parser
  exit false
end

unless File.file? options[:corpus_path]
  STDERR.puts "No corpus exists - #{options[:corpus_path]}"
  exit false
end

rows = options[:board].downcase.scan(/\S+/).map{ |row| row.scan(/./) }

raw_corpus = File.readlines(options[:corpus_path])
corpus = raw_corpus.map{ |w| w.downcase.rstrip }.uniq.sort

solution = solve(Matrix.rows(rows), corpus)
solution.each_pair do |w, paths|
  STDOUT.puts w
  paths.each do |path|
    STDOUT.puts "\t" + path.map{ |point| point.inspect }.join(', ')
  end
end
STDOUT.puts "TOTAL: #{solution.count}"

其他回答

我很快完美地解决了这个问题。我把它放进了一个安卓应用程序。在play store链接中查看视频,看看它是如何运作的。

单词作弊是一个应用程序,“破解”任何矩阵风格的文字游戏。这个应用程序 来帮我在文字混淆器上作弊。它可以用于单词搜索, 沙沙,单词,单词查找器,单词破解,拼字游戏,和更多!

在这里可以看到 https://play.google.com/store/apps/details?id=com.harris.wordcracker

在视频中查看应用程序的操作 https://www.youtube.com/watch?v=DL2974WmNAI

I wrote my solver in C++. I implemented a custom tree structure. I'm not sure it can be considered a trie but it's similar. Each node has 26 branches, 1 for each letter of the alphabet. I traverse the branches of the boggle board in parallel with the branches of my dictionary. If the branch does not exist in the dictionary, I stop searching it on the Boggle board. I convert all the letters on the board to ints. So 'A' = 0. Since it's just arrays, lookup is always O(1). Each node stores if it completes a word and how many words exist in its children. The tree is pruned as words are found to reduce repeatedly searching for the same words. I believe pruning is also O(1).

CPU: Pentium SU2700 1.3GHz 内存:3 gb

在< 1秒内加载178,590个单词的字典。 在4秒内解决100x100 Boggle (Boggle .txt)。约44000字。 解决4x4 Boggle游戏的速度太快,无法提供有意义的基准。:)

快速Boggle求解GitHub回购

该解决方案还提供了在给定的板中搜索的方向

一件事:

1. Uses trie to save all the word in the english to fasten the search
2. The uses DFS to search the words in Boggle

输出:

Found "pic" directions from (4,0)(p) go  → →
Found "pick" directions from (4,0)(p) go  → → ↑
Found "pickman" directions from (4,0)(p) go  → → ↑ ↑ ↖ ↑
Found "picket" directions from (4,0)(p) go  → → ↑ ↗ ↖
Found "picked" directions from (4,0)(p) go  → → ↑ ↗ ↘
Found "pickle" directions from (4,0)(p) go  → → ↑ ↘ →

代码:

from collections import defaultdict
from nltk.corpus import words
from nltk.corpus import stopwords
from nltk.tokenize import word_tokenize

english_words = words.words()

# If you wan to remove stop words
# stop_words = set(stopwords.words('english'))
# english_words = [w for w in english_words if w not in stop_words]

boggle = [
    ['c', 'n', 't', 's', 's'],
    ['d', 'a', 't', 'i', 'n'],
    ['o', 'o', 'm', 'e', 'l'],
    ['s', 'i', 'k', 'n', 'd'],
    ['p', 'i', 'c', 'l', 'e']
]

# Instead of X and Y co-ordinates
# better to use Row and column
lenc = len(boggle[0])
lenr = len(boggle)

# Initialize trie datastructure
trie_node = {'valid': False, 'next': {}}

# lets get the delta to find all the nighbors
neighbors_delta = [
    (-1,-1, "↖"),
    (-1, 0, "↑"),
    (-1, 1, "↗"),
    (0, -1, "←"),
    (0,  1, "→"),
    (1, -1, "↙"),
    (1,  0, "↓"),
    (1,  1, "↘"),
]


def gen_trie(word, node):
    """udpates the trie datastructure using the given word"""
    if not word:
        return

    if word[0] not in node:
        node[word[0]] = {'valid': len(word) == 1, 'next': {}}

    # recursively build trie
    gen_trie(word[1:], node[word[0]])


def build_trie(words, trie):
    """Builds trie data structure from the list of words given"""
    for word in words:
        gen_trie(word, trie)
    return trie


def get_neighbors(r, c):
    """Returns the neighbors for a given co-ordinates"""
    n = []
    for neigh in neighbors_delta:
        new_r = r + neigh[0]
        new_c = c + neigh[1]

        if (new_r >= lenr) or (new_c >= lenc) or (new_r < 0) or (new_c < 0):
            continue
        n.append((new_r, new_c, neigh[2]))
    return n


def dfs(r, c, visited, trie, now_word, direction):
    """Scan the graph using DFS"""
    if (r, c) in visited:
        return

    letter = boggle[r][c]
    visited.append((r, c))

    if letter in trie:
        now_word += letter

        if trie[letter]['valid']:
            print('Found "{}" {}'.format(now_word, direction))

        neighbors = get_neighbors(r, c)
        for n in neighbors:
            dfs(n[0], n[1], visited[::], trie[letter], now_word, direction + " " + n[2])


def main(trie_node):
    """Initiate the search for words in boggle"""
    trie_node = build_trie(english_words, trie_node)

    # print the board
    print("Given board")
    for i in range(lenr):print (boggle[i])
    print ('\n')

    for r in range(lenr):
        for c in range(lenc):
            letter = boggle[r][c]
            dfs(r, c, [], trie_node, '', 'directions from ({},{})({}) go '.format(r, c, letter))


if __name__ == '__main__':
    main(trie_node)

只是为了好玩,我在bash中实现了一个。 它不是超级快,但很合理。

http://dev.xkyle.com/bashboggle/

最快的解决方案可能是将字典存储在一个trie中。然后,创建一个三元组队列(x, y, s),其中队列中的每个元素对应于一个可以在网格中拼写的单词的前缀s,结束于位置(x, y)。初始化队列中有N x N个元素(其中N是网格的大小),网格中的每个正方形都有一个元素。然后,算法进行如下:

While the queue is not empty:
  Dequeue a triple (x, y, s)
  For each square (x', y') with letter c adjacent to (x, y):
    If s+c is a word, output s+c
    If s+c is a prefix of a word, insert (x', y', s+c) into the queue

如果将字典存储在trie中,则可以在常数时间内测试s+c是否是单词或单词的前缀(前提是还在每个队列数据中保留一些额外的元数据,例如指向trie中当前节点的指针),因此此算法的运行时间为O(可拼写的单词数量)。

[编辑]下面是我刚刚编写的Python实现:

#!/usr/bin/python

class TrieNode:
    def __init__(self, parent, value):
        self.parent = parent
        self.children = [None] * 26
        self.isWord = False
        if parent is not None:
            parent.children[ord(value) - 97] = self

def MakeTrie(dictfile):
    dict = open(dictfile)
    root = TrieNode(None, '')
    for word in dict:
        curNode = root
        for letter in word.lower():
            if 97 <= ord(letter) < 123:
                nextNode = curNode.children[ord(letter) - 97]
                if nextNode is None:
                    nextNode = TrieNode(curNode, letter)
                curNode = nextNode
        curNode.isWord = True
    return root

def BoggleWords(grid, dict):
    rows = len(grid)
    cols = len(grid[0])
    queue = []
    words = []
    for y in range(cols):
        for x in range(rows):
            c = grid[y][x]
            node = dict.children[ord(c) - 97]
            if node is not None:
                queue.append((x, y, c, node))
    while queue:
        x, y, s, node = queue[0]
        del queue[0]
        for dx, dy in ((1, 0), (1, -1), (0, -1), (-1, -1), (-1, 0), (-1, 1), (0, 1), (1, 1)):
            x2, y2 = x + dx, y + dy
            if 0 <= x2 < cols and 0 <= y2 < rows:
                s2 = s + grid[y2][x2]
                node2 = node.children[ord(grid[y2][x2]) - 97]
                if node2 is not None:
                    if node2.isWord:
                        words.append(s2)
                    queue.append((x2, y2, s2, node2))

    return words

使用示例:

d = MakeTrie('/usr/share/dict/words')
print(BoggleWords(['fxie','amlo','ewbx','astu'], d))

输出:

['fa', 'xi', 'ie', 'io', 'el', 'am', 'ax', 'ae', 'aw', 'mi', 'ma', 'me', 'lo', 'li', 'oe', 'ox', 'em', 'ea', 'ea', 'es', 'wa', 'we', 'wa', 'bo', 'bu', 'as', 'aw', 'ae', 'st', 'se', 'sa', 'tu', 'ut', 'fam', 'fae', 'imi', 'eli', 'elm', 'elb', 'ami', 'ama', 'ame', 'aes', 'awl', 'awa', 'awe', 'awa', 'mix', 'mim', 'mil', 'mam', 'max', 'mae', 'maw', 'mew', 'mem', 'mes', 'lob', 'lox', 'lei', 'leo', 'lie', 'lim', 'oil', 'olm', 'ewe', 'eme', 'wax', 'waf', 'wae', 'waw', 'wem', 'wea', 'wea', 'was', 'waw', 'wae', 'bob', 'blo', 'bub', 'but', 'ast', 'ase', 'asa', 'awl', 'awa', 'awe', 'awa', 'aes', 'swa', 'swa', 'sew', 'sea', 'sea', 'saw', 'tux', 'tub', 'tut', 'twa', 'twa', 'tst', 'utu', 'fama', 'fame', 'ixil', 'imam', 'amli', 'amil', 'ambo', 'axil', 'axle', 'mimi', 'mima', 'mime', 'milo', 'mile', 'mewl', 'mese', 'mesa', 'lolo', 'lobo', 'lima', 'lime', 'limb', 'lile', 'oime', 'oleo', 'olio', 'oboe', 'obol', 'emim', 'emil', 'east', 'ease', 'wame', 'wawa', 'wawa', 'weam', 'west', 'wese', 'wast', 'wase', 'wawa', 'wawa', 'boil', 'bolo', 'bole', 'bobo', 'blob', 'bleo', 'bubo', 'asem', 'stub', 'stut', 'swam', 'semi', 'seme', 'seam', 'seax', 'sasa', 'sawt', 'tutu', 'tuts', 'twae', 'twas', 'twae', 'ilima', 'amble', 'axile', 'awest', 'mamie', 'mambo', 'maxim', 'mease', 'mesem', 'limax', 'limes', 'limbo', 'limbu', 'obole', 'emesa', 'embox', 'awest', 'swami', 'famble', 'mimble', 'maxima', 'embolo', 'embole', 'wamble', 'semese', 'semble', 'sawbwa', 'sawbwa']

Notes: This program doesn't output 1-letter words, or filter by word length at all. That's easy to add but not really relevant to the problem. It also outputs some words multiple times if they can be spelled in multiple ways. If a given word can be spelled in many different ways (worst case: every letter in the grid is the same (e.g. 'A') and a word like 'aaaaaaaaaa' is in your dictionary), then the running time will get horribly exponential. Filtering out duplicates and sorting is trivial to due after the algorithm has finished.