python小程序-0005

第5题：有一个目录，放了你一个月的日记，都是 txt，为了避免分词的问题，假设内容都是英文，请统计出你认为每篇日记最重要的词。

#!/usr/bin/env python3
# -*- coding : utf-8 -*-

import os
import re
from collections import Counter

ignore_words = ['I','am','is','be','that','the','and','maybe','it','a','not']


def getWordNum(filesource):
    with open(filesource) as f:
        r = re.findall('\w+',f.read())
        return Counter(r)

def getImportantWord(dirpath):
    for file in os.listdir(dirpath):
        filepath = os.path.join(dirpath,file)
        totalCnt = getWordNum(filepath)
        for word in ignore_words:
            totalCnt[word] = 0
        print('The most import word in %s is %s.' % (filepath,totalCnt.most_common()[0][0]))


if __name__ == '__main__':
    dirpath = input('Please input dirpath: ')
    getImportantWord(dirpath)