Decentrala/chatbot
12
0
Fork 1
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Compare commits

base: Decentrala:5b4ae05582b700260190f9a6114638b7d3c5e7a5
Branches Tags
Decentrala:master
Decentrala:dev
Decentrala:task_manager
Decentrala:scraper_functions
...
compare: Decentrala:9704474c2929fae0915e2690803949b7641fbc19
Branches Tags
Decentrala:dev
Decentrala:master
Decentrala:task_manager
Decentrala:scraper_functions

2 Commits
5b4ae05582 ... 9704474c29

Author SHA1 Message Date
texhno
9704474c29 Merge branch 'master' of ssh://gitea.dmz.rs:2222/Decentrala/chatbot 2024-02-06 02:38:46 +01:00
texhno
9c63ada6e5 Added try catch for scraper functions 2024-02-06 02:37:19 +01:00
1 changed files with 8 additions and 5 deletions
Show Stats Expand all files Collapse all files

3
scraper_functions.py
View File

@ -2,8 +2,11 @@ import requests
from bs4 import BeautifulSoup from bs4 import BeautifulSoup
def query_external_website(base_url, query): def query_external_website(base_url, query):
try:
page = requests.get(base_url + query) page = requests.get(base_url + query)
soup = BeautifulSoup(page.content, "html.parser") soup = BeautifulSoup(page.content, "html.parser")
title = soup.find("span", class_="mw-page-title-main").text title = soup.find("span", class_="mw-page-title-main").text
content = soup.find(id="mw-content-text").select("p")[2].text content = soup.find(id="mw-content-text").select("p")[2].text
return "\nTITLE:\n" + title + "\n\nCONTENT:\n" + content + "\n\nFULL LINK:\n" + base_url + query return "\nTITLE:\n" + title + "\n\nCONTENT:\n" + content + "\n\nFULL LINK:\n" + base_url + query
except:
return "Can't parse search result :("