ada/c22712
1
0
Fork 0
mirror of https://github.com/KUlishevgeniy/c22712.git synced 2026-09-23 23:50:21 +00:00
Code Issues Packages Projects Releases Wiki Activity

Update parcing.py

Browse Source
This commit is contained in:
eauskova
2023-05-09 02:59:26 +03:00
committed by GitHub
parent 8a4ba475d0
commit 2e9837fdf4
1 changed files with 11 additions and 30 deletions
Download Patch File Download Diff File Expand all files Collapse all files
Задания/task1/Uskova/parcing.py
+11 -30
View File
@@ -1,35 +1,16 @@
from bs4 import BeautifulSoup from bs4 import BeautifulSoup
from selenium.webdriver import Chrome
from selenium import webdriver from selenium import webdriver
from selenium.webdriver.chrome.service import Service from selenium.webdriver.chrome.service import Service
import psycopg2 import time
import wget s = Service('C:\\Users\\Yekaterina\\Downloads\\chromedriver_win32\\chromedriver.exe')
browser = webdriver.Chrome(service=s)
browser.get('https://www.citilink.ru/catalog/noutbuki/')
html_text = browser.page_source
soup = BeautifulSoup(html_text, 'lxml')
browser = webdriver.Chrome(service=Service('C:\\Users\\Yekaterina\\Downloads\\chromedriver_win32.zip\\chromedriver.exe'))
browser.get('https://msk.sushi-market.com/menu/rolly')
soup = BeautifulSoup(browser.page_source, "lxml")
Name = soup.find_all(attrs={"class": "goodsBlockName device-show"}) name=soup.find_all('div', class_='app-catalog-1tp0ino e1an64qs0')
Info = soup.find_all(attrs={"class": "goodsBlockWeight"}) print (name[1].text)
Des = soup.find_all(attrs={"class": "cardsList_text"}) description=soup.find_all('div', class_='app-catalog-1o4umte eevw8x70')
Price = soup.find_all(attrs={"class": "goodsBlockPrice"}) print (description[1].text)
Image = soup.find_all(attrs={"class": "cardsList_img add-to-cart__img"})
connection = psycopg2.connect(host='localhost', dbname='dbdata', user='postgres', password='Q1w2e3r4')
cursor = connection.cursor()
create_q = '''CREATE TABLE Food (ID serial primary key, Name varchar(5000), Info varchar(5000), Des varchar(10000), Price varchar(6000), src varchar(1000))'''
cursor.execute(create_q)
connection.commit()
for j in range(25):
url = Image[j].find('img').attrs['src']
tempf = f"img\\food{j}.jpg"
wget.download(url, tempf)
insert_query = f'''INSERT into public.Food(Name, Info, Des, Price, src) values ('{Name[j].text}','{Info[j].text}','{Des[j].text}', '{Price[j].text}', '{tempf}') '''
cursor.execute(insert_query)
connection.commit()
cursor.execute("SELECT * from Food")
print(cursor.fetchall())
cursor.close()
connection.close()