MechanicalClerk / redirect.py
Melchior
Update redirect.py
03c0baf
raw
history blame contribute delete
893 Bytes
import requests
import csv
def process_csv(input_file, txt_projectName):
# STEP 1 in processing outputs
# takes the url as supplied by MechanicalClerk via GoogleNews and turns it into the proper article links
def get_redirected_url(url):
response = requests.get("https://" + url)
return response.url
# Read the input .csv file
with open(input_file.name, 'r', encoding="utf-8") as file:
csv_data = list(csv.reader(file, delimiter='\t'))
# Process each row in the .csv file except the header
for row in csv_data[1:]:
row[3] = get_redirected_url(row[3])
# Write the modified .csv data to a new file
with open(txt_projectName+'_processedSearchResults.tsv', 'w', newline='') as file:
writer = csv.writer(file, delimiter='\t')
writer.writerows(csv_data)
return txt_projectName+'_processedSearchResults.tsv'