2021-03-01 17:57:17 +11:00
|
|
|
import re
|
2021-04-11 19:43:49 +10:00
|
|
|
import argparse
|
2021-03-01 17:57:17 +11:00
|
|
|
|
2021-04-11 19:43:49 +10:00
|
|
|
# parse input arguments
|
|
|
|
|
parser = argparse.ArgumentParser()
|
|
|
|
|
parser.add_argument('path_to_file', help = 'path to the csv file')
|
|
|
|
|
args = parser.parse_args()
|
|
|
|
|
|
|
|
|
|
# open file, add all lines to a single string
|
|
|
|
|
file_given = open(args.path_to_file)
|
|
|
|
|
f = ""
|
|
|
|
|
for line in file_given:
|
|
|
|
|
f += line + " "
|
|
|
|
|
file_given.close()
|
|
|
|
|
|
|
|
|
|
# remove non-alphabetic characters, replace all whitespace characters with a
|
|
|
|
|
# single whitespace, and change all uppercase characters to lowercase
|
|
|
|
|
f = re.sub(r'[^a-zA-Z\s]', r'', f)
|
|
|
|
|
f = re.sub(r'\s+', r' ', f)
|
|
|
|
|
f = f.lower()
|
|
|
|
|
|
|
|
|
|
print(f)
|
