效果图：

私聊效果图.png

引用库：

os 用于查找当前文件夹内的文件
re 聊天记录提取
jieba 中文分词工具
pyecharts 用于绘制文字云（WordCloud）

源码：

# !usr/bin/env python
# -*-coding:utf-8 -*-
import os
import re
import jieba
from pyecharts import WordCloud


def file_name():
    name = []
    for root, dirs, files in os.walk(os.getcwd()):
        for file in files:
            this_file = os.path.splitext(file)
            if this_file[1].lower() == '.txt':
                name.append(this_file[0])
        return name


def add_dict(dic, words):
    for word in words:
        if len(word.strip()):
            if word in dic:
                dic[word] += 1
            else:
                dic[word] = 1


def remove_point(dic):
    apl = list(r'\/|{}[]@#$^&*~`;:"<>?,，。、？“”')
    chara = list(r'的了吗吧的么呢啊哪')
    remove_key = list(chr(i) for i in range(0, 128)) + apl + chara
    for each in remove_key:
        try:
            dic.pop(each)
        except:
            pass


def dict_sort(dic):
    d = list(zip(dic.keys(), dic.values()))
    dd = sorted(d, key=lambda x: x[1], reverse=True)
    return list(zip(*dd))


def wordCloud(title, key, value):
    wordcloud = WordCloud(width=1000, height=700)
    wordcloud.add('', key, value, word_size_range=[17, 100], rotate_step=20, shape='circle')
    # 'circle', 'cardioid', 'diamond', 'triangle-forward', 'triangle', 'pentagon', 'star'
    wordcloud.render(title + '.html')


def main(name):
    with open('%s.txt' % name, 'r', encoding='utf-8') as f:
        file = f.read()
    reg = re.compile(r'(\d{4}-\d\d-\d\d).(\d{1,2}:\d\d:\d\d).(.*)\D([^`]*?)' + '\n\n')
    res = reg.findall(file)
    print(len(res))
    ttl = {}
    count = 0
    for each in res:
        count += 1
        print('当前进度：%s/%s' %(count, len(res)), end='\r')
        words = list(jieba.cut(each[3], cut_all=True))
        add_dict(ttl, words)
    remove_point(ttl)
    ttl_key, ttl_value = dict_sort(ttl)
    wordCloud(name, ttl_key[:300], ttl_value[:300])


if __name__ == '__main__':
    files = file_name()
    for file in files:
        print('当前文件：', file, '\n提取聊天信息数量：', end='')
        main(file)
    input('已完成，按回车键结束……')

代码解释：

前两行，定义解析器和编码方式
import 引入我们需要使用的库
file_name 获取当前文件夹下的聊天记录，返回对应路径
add_dict 使用字典统计词频
remove_point 移除分词结果无用词汇及标点符号(对于不想体现的词汇，均可以在此添加)
dict_sort 按照词频对字典排序
wordCloud 使用pyecharts的文字云工具绘制文字云
main 主函数，控制每一步的进行

注：

由QQ直接导出的群聊聊天记录与私聊聊天记录格式略有区别，请勿直接使用此代码分析，关于群聊，请查看另外一篇 https://www.jianshu.com/p/50acede72d52

最后编辑于：2018.06.25 10:24:03