This commit is contained in:
ada-dmitry
2023-04-05 09:47:13 +03:00
parent f5579babc2
commit 3497a6f849
3 changed files with 16 additions and 8 deletions
+12 -8
View File
@@ -3,8 +3,11 @@ import wget
from bs4 import BeautifulSoup from bs4 import BeautifulSoup
from selenium import webdriver from selenium import webdriver
from selenium.webdriver.chrome.service import Service from selenium.webdriver.chrome.service import Service
import config
# Парсер и загрузчик в БД//Антипенко Дмитрий # Парсер и загрузчик в БД//Антипенко Дмитрий
connection = psycopg2.connect(host='localhost', dbname='PythonDB', user='postgres', password='Q1w2e3r4') connection = psycopg2.connect(
host=hostname, dbname=databname, user=username, password=passw)
cursor = connection.cursor() cursor = connection.cursor()
@@ -14,19 +17,22 @@ creat_qwery = """ create table Parser
cursor.execute(creat_qwery) cursor.execute(creat_qwery)
connection.commit() connection.commit()
driver = Service('D:\\teach\Prog\chromedriver.exe') driver = Service('D:\\teach\Prog\chromedriver.exe')
browser = webdriver.Chrome(service=driver) browser = webdriver.Chrome(service=driver)
browser.get('https://amwine.ru/catalog/igristoe_vino_i_shampanskoe/igristoe_vino/') browser.get(
'https://amwine.ru/catalog/igristoe_vino_i_shampanskoe/igristoe_vino/')
html_code = browser.page_source html_code = browser.page_source
b_soup = BeautifulSoup(html_code, 'lxml') b_soup = BeautifulSoup(html_code, 'lxml')
name = b_soup.find_all('a', class_="catalog-list-item__title js-product-detail-link") name = b_soup.find_all(
'a', class_="catalog-list-item__title js-product-detail-link")
price = b_soup.find_all('span', class_="middle_price") price = b_soup.find_all('span', class_="middle_price")
priceDis = b_soup.find_all('span', class_="baseoldprice") priceDis = b_soup.find_all('span', class_="baseoldprice")
mark = b_soup.find_all('span', class_="product-rating__rating") mark = b_soup.find_all('span', class_="product-rating__rating")
pictures = b_soup.find_all('div', class_="catalog-list-item__img-wrapper") pictures = b_soup.find_all('div', class_="catalog-list-item__img-wrapper")
for i in range(15): for i in range(15):
url = 'https://amwine.ru'+pictures[i].find('a').find('img').attrs['data-src'] url = 'https://amwine.ru' + \
pictures[i].find('a').find('img').attrs['data-src']
filename = f"Programming\\23.03\img\{i}.jpg" filename = f"Programming\\23.03\img\{i}.jpg"
print(filename) print(filename)
wget.download(url, filename) wget.download(url, filename)
@@ -35,8 +41,6 @@ for i in range(15):
connection.commit() connection.commit()
cursor.close() cursor.close()
connection.close() connection.close()
Binary file not shown.
+4
View File
@@ -0,0 +1,4 @@
hostname = "localhost"
databname = "PythonDB"
username = "postgres"
passw = "Q1w2e3r4"