Advertisement
Not a member of Pastebin yet?
Sign Up,
it unlocks many cool features!
- import requests
- import base64
- from bs4 import BeautifulSoup
- import json
- from base64 import b64encode
- kategoria= input("Podaj kategorie ktora chcesz przeszukac:")
- url = 'https://www.morele.net/wyszukiwarka/0/0/,,,,,,,,,,,,/1/?q='+kategoria
- print(url)
- response = requests.get(url)
- soup = BeautifulSoup(response.text, "html5lib")
- body = soup.body
- tytuly = body.find_all('h2', {'class':'cat-product-name'})
- ceny = body.find_all('div', {'class' : 'price-new'})
- parametry = body.find_all('div', {'class': 'cat-product-feature'})
- zdjecia = body.find_all('a',{'class': 'cat_product_image'})
- class Base64Encoder(json.JSONEncoder):
- def default(self,o):
- if isinstance(o,bytes):
- return b64encode(o).decode()
- return json.JSONEncoder.default(self,0)
- def get_as_base64(url):
- return base64.b64encode(requests.get(url).content)
- tablicaCen = []
- for cena in ceny:
- tablicaCen.append(cena)
- tablicaTytulow = []
- for tytul in tytuly:
- tablicaTytulow.append(tytul)
- tablicaParametrow = []
- for parametr in parametry:
- tablicaParametrow.append(parametr)
- tablicaZdjec = []
- for zdjecie in zdjecia:
- tablicaZdjec.append(zdjecie)
- jsony = [{"Tytul": t, "Cena": c, "parametr": p, "Zdjecie": get_as_base64(z)} for t,c,p,z in zip(tablicaTytulow,tablicaCen,tablicaParametrow,tablicaZdjec)]
- with open('data.json','w',encoding='utf-8') as f:
- json.dump(jsony,f,ensure_ascii=False,indent=4,cls=Base64Encoder)
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement