scossa/scrapeje
1
1
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Update 'ScrapeJE.py'

Browse Source
This commit is contained in:
scossa
2022-12-01 23:26:10 +01:00
parent aaf4818dbc
commit da3522aa76
1 changed files with 17 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
ScrapeJE.py
View File

@@ -2,6 +2,9 @@ import requests
from bs4 import BeautifulSoup from bs4 import BeautifulSoup
#import cloudscraper #import cloudscraper
#import json #import json
import colorama
from colorama import Fore
from colorama import Style
nome=[] nome=[]
desc=[] desc=[]
@@ -11,8 +14,14 @@ scripto=[]
#prende l url della pagina justeat del ristorante in input #prende l url della pagina justeat del ristorante in input
url = "https://www.justeat.it/restaurants-pizzeria-la-garganica-bologna/menu" #url = "https://www.justeat.it/restaurants-pizzeria-girasole-bologna/menu"
#url = input('link della pagina justeat del ristorante: ') #url = input('link della pagina justeat del ristorante: ')
#input ('inserisci il link della pagina justeat del ristorante: ')
print("\n")
url = print(Fore.WHITE + Style.DIM + "es https://www.justeat.it/NOME_RISTORANTE/menu" + Style.RESET_ALL)
url = input(Fore.RED + Style.BRIGHT + "inserisci il link della pagina justeat del ristorante: " + Style.RESET_ALL)
#scrape html scavalcando cloudflare #scrape html scavalcando cloudflare
# scraper = cloudscraper.create_scraper(browser={'browser': 'firefox','platform': 'windows','mobile': False}) # scraper = cloudscraper.create_scraper(browser={'browser': 'firefox','platform': 'windows','mobile': False})
@@ -56,8 +65,13 @@ for menu in soup.find_all(attrs={"data-test-id": "menu-item"}):
att=menu att=menu
#riempie la lista "nome" #riempie la lista "nome"
for att in menu.find(attrs={"data-test-id": "menu-item-name"}): for att in menu.find(attrs={"data-test-id": "menu-item-name"}):
if att != type(None):
if att != " ": if att != " ":
nome.append(att.lstrip().splitlines()[0]) nome.append(att.lstrip().splitlines()[0])
break
else:
continue
#riempie la lista "desc" #riempie la lista "desc"
att=menu.find("p", class_="c-menuItems-description") att=menu.find("p", class_="c-menuItems-description")