sidski/helseveileder_webscraper
Archived
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

✨ export info to csv-file

Browse source 
Co-authored-by: haraldnilsen <harald_998@hotmail.com>
Signed-off-by: Sindre Kjelsrud <kjelsrudsindre@gmail.com>
This commit is contained in:
Sindre Kjelsrud 2024-01-08 15:03:42 +01:00
parent 1a1b6c9737
commit 709937da9d
Signed by untrusted user who does not match committer: sidski
GPG key ID: D2BBDF3EDE6BA9A6
1 changed files with 14 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

17
main.py
View file

@ -1,6 +1,7 @@
import httpx import httpx
from selectolax.parser import HTMLParser from selectolax.parser import HTMLParser
import re import re
import csv
def fetch_question_url(url): def fetch_question_url(url):
# Define headers # Define headers
@ -15,7 +16,7 @@ def fetch_question_url(url):
"list": "frequently_asked_questions", "list": "frequently_asked_questions",
"idCategory": "8288", # Id for category "idCategory": "8288", # Id for category
"from": "0", "from": "0",
"size": "10", # Adjust this to get more results "size": "100", # Adjust this to get more results
"gender": "false", "gender": "false",
"age": "false", "age": "false",
"zone": "default", "zone": "default",
@ -77,5 +78,15 @@ def fetch_all_info(url):
studenterspor_url = "https://www.studenterspor.no/ajax_handler.php" studenterspor_url = "https://www.studenterspor.no/ajax_handler.php"
question_urls = fetch_question_url(studenterspor_url) question_urls = fetch_question_url(studenterspor_url)
for url in question_urls: # Open a new CSV file to write the data
print(fetch_all_info(url)) with open('studenterspor.csv', mode='w', newline='', encoding='utf-8') as file:
writer = csv.writer(file)
# Write the header
writer.writerow(['Question', 'Answer', 'Signature'])
# Fetch info for each URL and write it to the CSV file
for url in question_urls:
info = fetch_all_info(url)
writer.writerow(info) # Write the question, answer, and signature as a new row
print("Done! Your data is now in 'studenterspor.csv'")