2018-07-08 23:28:29 +08:00
|
|
|
''' mbinary
|
|
|
|
|
#########################################################################
|
|
|
|
|
# File : markov.py
|
|
|
|
|
# Author: mbinary
|
|
|
|
|
# Mail: zhuheqin1@gmail.com
|
2019-01-31 12:09:46 +08:00
|
|
|
# Blog: https://mbinary.xyz
|
2018-07-08 23:28:29 +08:00
|
|
|
# Github: https://github.com/mbinary
|
|
|
|
|
# Created Time: 2018-07-06 15:57
|
|
|
|
|
# Description:
|
|
|
|
|
#########################################################################
|
|
|
|
|
'''
|
|
|
|
|
|
|
|
|
|
from random import randint
|
|
|
|
|
import re
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
class markov:
|
|
|
|
|
def __init__(self,txt):
|
|
|
|
|
self.words= self.clean(txt)
|
|
|
|
|
self.dic = self.getDic(self.words)
|
|
|
|
|
def clean(self,text):
|
|
|
|
|
text = text.replace("\n", " ");
|
|
|
|
|
text = text.replace("\"", "");
|
|
|
|
|
|
|
|
|
|
# 保证每个标点符号都和前面的单词在一起
|
|
|
|
|
# 这样不会被剔除，保留在马尔可夫链中
|
|
|
|
|
punctuation = [',', '.', ';',':']
|
|
|
|
|
for symbol in punctuation:
|
|
|
|
|
text = text.replace(symbol, symbol+" ");
|
|
|
|
|
|
|
|
|
|
return re.split(' +',text)
|
|
|
|
|
|
|
|
|
|
def getDic(self,words):
|
|
|
|
|
dic = {}
|
|
|
|
|
end = len(words)
|
|
|
|
|
for i in range(1,end):
|
|
|
|
|
if words[i-1] not in dic:
|
|
|
|
|
dic[words[i-1]] = {words[i]:1}
|
|
|
|
|
elif words[i] not in dic[words[i-1]]:
|
|
|
|
|
dic[words[i-1]][words[i]] = 1
|
|
|
|
|
else: dic[words[i-1]][words[i]] +=1
|
|
|
|
|
return dic
|
|
|
|
|
def getSum(self,dic):
|
|
|
|
|
if '%size' not in dic:
|
|
|
|
|
dic['%size'] = sum(list(dic.values()))
|
|
|
|
|
return dic['%size']
|
|
|
|
|
def nextWord(self,word):
|
|
|
|
|
k = randint(1,self.getSum(self.dic[word]))
|
|
|
|
|
for i,j in self.dic[word].items():
|
|
|
|
|
k-=j
|
|
|
|
|
if k<=0:return i
|
|
|
|
|
def genSentence(self,begin = 'I',length = 30):
|
|
|
|
|
li = [begin]
|
|
|
|
|
nextWord= begin
|
|
|
|
|
for i in range(1,length):
|
|
|
|
|
nextWord= self.nextWord(nextWord)
|
|
|
|
|
li.append(nextWord)
|
|
|
|
|
return ' '.join(li)
|
