Huffman Encode 哈弗曼编码（python实现）

# coding: utf-8
#Tree-Node Type
class Node:
    def __init__(self,freq):
        self.left = None
        self.right = None
        self.father = None
        self.freq = freq
    def __repr__(self):  #change the string representation of instances,see cookbook  8.1
        return ‘Node({0.freq!r})‘.format(self)
    def isLeft(self):
        return self.father.left == self

#create nodes 创建叶子节点
def createNodes(freqs):
    return [Node(freq) for freq in freqs]

#create Huffman-Tree 创建Huffman树
def createHuffmanTree(nodes):
    queue = nodes[:]  #copy of the nodes
    while len(queue) > 1:
        queue.sort(key=lambda item:item.freq)  #sort the objects by certain attribute
        node_left = queue.pop(0)
        node_right = queue.pop(0)
        node_father = Node(node_left.freq + node_right.freq)
        node_father.left = node_left
        node_father.right = node_right
        node_left.father = node_father
        node_right.father = node_father
        queue.append(node_father)
    queue[0].father = None  #self.father = None,可省去
    return queue[0]   #root node

#Huffman编码
def huffmanEncoding(nodes,root):
    codes = [‘‘] * len(nodes)  #叶子节点数即为编码数
    for i in range(len(nodes)):
        node_tmp = nodes[i]
        while node_tmp != root:
            if node_tmp.isLeft():
                codes[i] = ‘0‘ + codes[i]
            else:
                codes[i] = ‘1‘ + codes[i]
            node_tmp = node_tmp.father
    return codes

if __name__ == ‘__main__‘:
    chars = [‘A‘,‘B‘,‘C‘,‘D‘,‘E‘,‘F‘,‘G‘,‘H‘,‘I‘,‘J‘,‘K‘,‘L‘,‘M‘,‘N‘]
    freqs = [10,4,2,5,3,4,2,6,4,4,3,7,9,6]
    chars_freqs = zip(chars,freqs)
    nodes = createNodes([item[1] for item in chars_freqs])
    root = createHuffmanTree(nodes)
    codes = huffmanEncoding(nodes,root)
    chars_codes = zip(chars_freqs,codes)
    chars_codes.sort(key=lambda item:item[0][1])  #sort the result by the freqs
    for item in chars_codes:
        print ‘Character:%s freq:%-2d   encoding: %s‘ % (item[0][0],item[0][1],item[1])

Character:C freq:2    encoding: 10100
Character:G freq:2    encoding: 10101
Character:E freq:3    encoding: 0000
Character:K freq:3    encoding: 0001
Character:B freq:4    encoding: 0100
Character:F freq:4    encoding: 0101
Character:I freq:4    encoding: 0110
Character:J freq:4    encoding: 0111
Character:D freq:5    encoding: 1011
Character:H freq:6    encoding: 1110
Character:N freq:6    encoding: 1111
Character:L freq:7    encoding: 001
Character:M freq:9    encoding: 100
Character:A freq:10   encoding: 110

注：

1）__repr__修改instance的string representation，方便调试，见python cookbook 8.1节

2）line22和line55 根据提供的key参数对列表进行排序

3）构建Node类，体会OO编程

4）其他实现:Rosettacode

5）Ipython notebook