Skip to content

Latest commit

 

History

History
33 lines (27 loc) · 1.76 KB

README.md

File metadata and controls

33 lines (27 loc) · 1.76 KB
Raw

Markov Speaking （马尔科夫链随机文本生成）

This project is an package generating random sentences by markov chain. If you have any problems/ideas, please email me, or open your PR. I feel honored to learn from your help.

Platform

The markov_speaking.py is written in Python 3.6, using jieba, codecs, random and re. You need to install jieba by pip2 install jieba.

Usage

  • The markov_speaking.py provides a class Markov, the init of Markov is __init__(self, filepath = None, mode = 0, coding="utf8"). filepath is the file you want to parse, and the sentences the class build will base on this file. mode is 0 if you want to parse English, and 1 if Chinese. coding assigns the codec, default is UTF-8.
  • Let p be a instance of Markov, you can use p.train(self, filepath = '', mode = 0, coding="utf8") to regenerate the instance.
  • After you have built p and trained, you can use p.say(length) to generate a random sentence. The length is the max length of sentence to generate, default is 10.

Examples For Use

with all the material included

>>> import markov_speaking
>>> p = markov_speaking.Markov('swords.txt', 1)
Building prefix dict from the default dictionary ...
Loading model from cache /home/forec/cache
Dumping model to file cache /home/forec/cache
Loading model cost 1.578 seconds.
Prefix dict has been built succesfully.
>>> p.say(5)
忽然想到一计说道师伯令狐师兄行侠仗义。

Update-logs

  • 2019-5-12: change it to py3
  • 2016-10-10: Add project and build repository.
  • 2016-10-11: Fix problems in English part: Not split words by sentences.
  • 2016-10-12: Fix train function.
  • 2016-10-13: Remove useless chinese upper condition.