Marcov_kokoro

import codecs
import re
import random
from janome.tokenizer import Tokenizer

def buildWordDict(text):
  # 改行や空白、ふりがなを消す
  delete_symbol = ["|", "　"]
  delete_expression = ["《.*》", "［.*\］", "\r", "\n"]
  for ds in delete_symbol:
    text = text.replace(ds, "")
  for de in delete_expression:
    text = re.sub(de, "", text)

# トークンを生成して、テキストを単語ごとに区切る
  t = Tokenizer()
  words = t.tokenize(text, wakati=True)
  
  wordDict = {}
# 単語の辞書を作成
  for i in range(0, len(words)-1):
    previousPhrase = ""
    nextPhrase = ""
    for word in words[i]:
      previousPhrase += word
    for word in words[i+1]:
      nextPhrase += word
    wordDict.setdefault(previousPhrase, [])
    wordDict[previousPhrase].append(nextPhrase)
  return wordDict

def retrieveRandomword(wordList):
  randomIndex = random.randint(0, len(wordList)-1)
  return wordList[randomIndex]

# テキストファイルを読み込む。読み込めないものは無視するためにcodecsモジュールを使い、引数にignoreを指定
text = codecs.open('/Users/yusuke/Downloads/kokoro.txt', 'r', 'shift-jis', 'ignore').read()
wordDict = buildWordDict(text)

# 何回続けるか定義
length = 100
chain =["私"]
for i in range(1, length):
  newWord = retrieveRandomword(wordDict[chain[-1]])
  chain.append(newWord)

print("".join(chain))