Python：将目录中的所有文件写入一个cdv文件

Question

I am trying to create a bimodal graph of a collection of texts such that I can project a network of either texts by words or words by texts. A colleague of mine has indicated that if I can get all my files in a single csv file of the format below, then has a workflow that will take care of the rest:

textfile1, words words words
textfile2, words words words

I have written the following script:

#! /usr/bin/env python

# a script to convert all text files in a directory to the format:
# filename, words from file (no punctuation)

import glob
import re

files = {}
for fpath in glob.glob("*.txt"):
    with open(fpath) as f:
         just_words = re.sub("[^a-zA-Z'-]"," ",f.read())

with open("mastertext.csv", "w") as f:
    for fname in files:
        print >> f , "%s,%s"%(fname,just_words)

This script will run and produce the output file, but the output file is blank and I get no error response -- the source of much learning for me as a Python newbie. Am I on the right track here, and if so what am I missing?

Answer 1

You need to save the data in just_words to files . In this case, I use a list of tuples instead of a dictionary but you can still use a dictionary if you prefer. :-)

files = []
for fpath in glob.glob("*.txt"):
    with open(fpath) as f:
        just_words = re.sub("[^a-zA-Z'-]"," ",f.read())
        files.append((fpath, just_words))

with open("mastertext.csv", "w") as f:
    for fname, just_words in files:
        print >> f , "%s,%s"%(fname,just_words)