15 lines
407 B
Python
15 lines
407 B
Python
import sys, re, string
|
|
|
|
def extract_words(path_to_file):
|
|
with open(path_to_file) as f:
|
|
str_data = f.read()
|
|
pattern = re.compile('[\W_]+')
|
|
word_list = pattern.sub(' ', str_data).lower().split()
|
|
|
|
with open('../stop_words.txt') as f:
|
|
stop_words = f.read().split(',')
|
|
stop_words.extend(list(string.ascii_lowercase))
|
|
|
|
return [w for w in word_list if not w in stop_words]
|
|
|