SVIG/python-bootcamp
7
0
Fork 1
Code Issues Pull Requests 1 Releases Wiki Activity

fp2_4 update

Browse Source
This commit is contained in:
Karenzz
2023-05-01 12:52:50 -04:00
parent 72167507f9
commit 7baaac206f
1 changed files with 14 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files
fp2_4.py
+14 -8
View File
@@ -2,36 +2,42 @@ from selenium import webdriver
from selenium.webdriver.chrome.options import Options from selenium.webdriver.chrome.options import Options
from selenium.webdriver.common.by import By from selenium.webdriver.common.by import By
import json import json
import time
# Configure the Selenium driver options # Configure the Selenium driver options
options = Options() options = Options()
options.add_argument('--headless') #options.add_argument('--headless')
options.add_argument('--disable-gpu') options.add_argument('--disable-gpu')
driver = webdriver.Chrome(options=options) driver = webdriver.Chrome(options=options)
# Load the page and extract the __NEXT_DATA__ data # Load the page and extract the __NEXT_DATA__ data
url = 'https://www.sayweee.com/en/product/Want-Want-Senbei-Rice-Crackers/17119' url = 'https://www.sayweee.com/en/product/Want-Want-Senbei-Rice-Crackers/17119'
driver.get(url) driver.get(url)
time.sleep(5)
next_data_element = None next_data_element = None
try: #try:
next_data_element = driver.find_element(By.XPATH, '//script[contains(text(), "__NEXT_DATA__")]') next_data_element = driver.find_element(By.XPATH, "//script[@type='application/ld+json']")
except: #except:
print('__NEXT_DATA__ element not found') # print('__NEXT_DATA__ element not found')
#time.sleep(1000)
if next_data_element is not None: if next_data_element is not None:
next_data_text = next_data_element.get_attribute('innerHTML') next_data_text = next_data_element.get_attribute('innerHTML')
print(next_data_text)
next_data_start_index = next_data_text.index('{') next_data_start_index = next_data_text.index('{')
next_data_end_index = next_data_text.rindex('}') + 1 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 = next_data_text[next_data_start_index:next_data_end_index]
next_data = json.loads(next_data_json) next_data = json.loads(next_data_json)
# Extract the price from the __NEXT_DATA__ data # Extract the price from the __NEXT_DATA__ data
price = next_data['props']['pageProps']['product']['price'] # price = next_data['props']['pageProps']['product']['price']
print(next_data)
price = next_data['offers']['price']
# Print the price # Print the price
print(price) print(price)
else: else:
print('__NEXT_DATA__ element not found') print('__NEXT_DATA__ element not found')
time.sleep(1000)
# Close the Selenium driver # Close the Selenium driver
driver.quit() #driver.quit()