buckbanzai/kuow-mastodon-bot
1
0
Fork 0
Code Issues 4 Pull requests Projects Releases Packages Wiki Activity

Fix formatting

Browse source
This commit is contained in:
Liam Steckler 2024-01-14 14:48:09 -08:00
parent abd9ff4630
commit bb1aa38721
1 changed files with 29 additions and 15 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

44
kuow_fetcher.py
View file

@ -1,14 +1,18 @@
import requests import requests
from bs4 import BeautifulSoup from bs4 import BeautifulSoup
from mastodon import Mastodon from mastodon import Mastodon
kuow_base_url="https://www.kuow.org"
url=kuow_base_url+"/fragments?name=story_grid&source=homepage&view_id=1&page=1&per_page=12" kuow_base_url = "https://www.kuow.org"
url = (
kuow_base_url
+ "/fragments?name=story_grid&source=homepage&view_id=1&page=1&per_page=12"
)
log_file = "kuow_bot_logfile.txt" log_file = "kuow_bot_logfile.txt"
mastodon = Mastodon(access_token = 'kuow_bot_mastodon.secret') mastodon = Mastodon(access_token="kuow_bot_mastodon.secret")
r = requests.get(url) r = requests.get(url)
soup = BeautifulSoup(r.content, 'html.parser') soup = BeautifulSoup(r.content, "html.parser")
articles = soup.find_all("span", class_ = "txt") articles = soup.find_all("span", class_="txt")
# Reverse articles, so that if multiple new ones have been found, they'll be posted in order of when published # Reverse articles, so that if multiple new ones have been found, they'll be posted in order of when published
articles.reverse() articles.reverse()
@ -16,22 +20,32 @@ for article in articles:
article_link = article.find("a").attrs["href"] article_link = article.find("a").attrs["href"]
is_new_article = True is_new_article = True
with open(log_file, 'r') as fp: with open(log_file, "r") as fp:
lines = fp.readlines() lines = fp.readlines()
for row in lines: for row in lines:
if row == article_link+"\n": if row == article_link + "\n":
print("Article "+article_link+" has already been seen") print("Article " + article_link + " has already been seen")
is_new_article = False is_new_article = False
if is_new_article: if is_new_article:
print(article_link+" has not been seen, posting") print(article_link + " has not been seen, posting")
article_lookup = requests.get(kuow_base_url+article_link) article_lookup = requests.get(kuow_base_url + article_link)
article_soup = BeautifulSoup(article_lookup.content, 'html.parser') article_soup = BeautifulSoup(article_lookup.content, "html.parser")
try: try:
article_description = (article_soup.find("meta", attrs={"property": "description"})).attrs["content"].strip() article_description = (
mastodon.status_post(status=article_description+"\n"+kuow_base_url+article_link+"\n#KUOW #News", visibility="unlisted") (article_soup.find("meta", attrs={"property": "description"}))
.attrs["content"]
.strip()
)
mastodon.status_post(
status=article_description
+ "\n"
+ kuow_base_url
+ article_link
+ "\n#KUOW #News",
visibility="unlisted",
)
with open(log_file, "a") as fp: with open(log_file, "a") as fp:
fp.write(article_link+"\n") fp.write(article_link + "\n")
except: except:
print("Could not load a description/post this article") print("Could not load a description/post this article")