2023-04-26 19:14:54 -04:00
|
|
|
from selenium import webdriver
|
|
|
|
|
from selenium.webdriver.chrome.options import Options
|
2023-04-28 19:30:00 -04:00
|
|
|
from selenium.webdriver.common.by import By
|
2023-04-26 19:14:54 -04:00
|
|
|
import json
|
|
|
|
|
|
|
|
|
|
# Configure the Selenium driver options
|
|
|
|
|
options = Options()
|
|
|
|
|
options.add_argument('--headless')
|
|
|
|
|
options.add_argument('--disable-gpu')
|
|
|
|
|
driver = webdriver.Chrome(options=options)
|
|
|
|
|
|
|
|
|
|
# Load the page and extract the __NEXT_DATA__ data
|
2023-04-28 19:30:00 -04:00
|
|
|
url = 'https://www.sayweee.com/en/product/Want-Want-Senbei-Rice-Crackers/17119'
|
2023-04-26 19:14:54 -04:00
|
|
|
driver.get(url)
|
|
|
|
|
next_data_element = None
|
2023-04-28 19:30:00 -04:00
|
|
|
try:
|
|
|
|
|
next_data_element = driver.find_element(By.XPATH, '//script[contains(text(), "__NEXT_DATA__")]')
|
|
|
|
|
except:
|
|
|
|
|
print('__NEXT_DATA__ element not found')
|
2023-04-26 19:14:54 -04:00
|
|
|
|
|
|
|
|
if next_data_element is not None:
|
|
|
|
|
next_data_text = next_data_element.get_attribute('innerHTML')
|
|
|
|
|
next_data_start_index = next_data_text.index('{')
|
|
|
|
|
next_data_end_index = next_data_text.rindex('}') + 1
|
|
|
|
|
next_data_json = next_data_text[next_data_start_index:next_data_end_index]
|
|
|
|
|
next_data = json.loads(next_data_json)
|
|
|
|
|
|
|
|
|
|
# Extract the price from the __NEXT_DATA__ data
|
|
|
|
|
price = next_data['props']['pageProps']['product']['price']
|
|
|
|
|
|
|
|
|
|
# Print the price
|
|
|
|
|
print(price)
|
|
|
|
|
else:
|
|
|
|
|
print('__NEXT_DATA__ element not found')
|
|
|
|
|
|
|
|
|
|
# Close the Selenium driver
|
|
|
|
|
driver.quit()
|