ada/c22712
1
0
Fork 0
mirror of https://github.com/KUlishevgeniy/c22712.git synced 2026-09-24 08:00:22 +00:00
Code Issues Packages Projects Releases Wiki Activity

Update parcing.py

Browse Source
This commit is contained in:
eauskova
2023-05-09 02:58:15 +03:00
committed by GitHub
parent abb924fb6c
commit f985f9b2d5
1 changed files with 32 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files
Задания/task1/Uskova/parcing.py
+32 -11
View File
@@ -1,14 +1,35 @@
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 time import psycopg2
s = Service('C:\\Users\\Yekaterina\\Downloads\\chromedriver_win32\\chromedriver.exe') import wget
browser = webdriver.Chrome(service=s)
browser.get('https://www.citilink.ru/catalog/noutbuki/') browser = webdriver.Chrome(service=Service('C:\\Users\\Yekaterina\\Downloads\\chromedriver_win32.zip\\chromedriver.exe'))
html_text = browser.page_source browser.get('https://msk.sushi-market.com/menu/rolly')
soup = BeautifulSoup(html_text, 'lxml') soup = BeautifulSoup(browser.page_source, "lxml")
name=soup.find_all('div', class_='app-catalog-1tp0ino e1an64qs0')
print (name[0].text) Name = soup.find_all(attrs={"class": "goodsBlockName device-show"})
description=soup.find_all('div', class_='app-catalog-1o4umte eevw8x70') Info = soup.find_all(attrs={"class": "goodsBlockWeight"})
print (description[0].text) Des = soup.find_all(attrs={"class": "cardsList_text"})
Price = soup.find_all(attrs={"class": "goodsBlockPrice"})
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()