This fixes issue #1 #2

Open
j wants to merge 123 commits from Branch_Karen into main
2 changed files with 39 additions and 5 deletions
Showing only changes of commit 72167507f9 - Show all commits
+6 -5
View File
@@ -1,5 +1,6 @@
from selenium import webdriver 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
import json import json
# Configure the Selenium driver options # Configure the Selenium driver options
@@ -9,13 +10,13 @@ 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/Pete-and-Gerry-s-Organic-Large-Brown-Eggs/65902?category=dairy02' url = 'https://www.sayweee.com/en/product/Want-Want-Senbei-Rice-Crackers/17119'
driver.get(url) driver.get(url)
next_data_element = None next_data_element = None
for script_element in driver.find_elements_by_xpath('//script'): try:
if '__NEXT_DATA__' in script_element.get_attribute('innerHTML'): next_data_element = driver.find_element(By.XPATH, '//script[contains(text(), "__NEXT_DATA__")]')
next_data_element = script_element except:
break print('__NEXT_DATA__ element not found')
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')
+33
View File
@@ -0,0 +1,33 @@
import re
import json
from selenium import webdriver
from selenium.webdriver.chrome.options import Options
from selenium.webdriver.common.by import By
# 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
url = 'https://www.sayweee.com/en/product/Want-Want-Senbei-Rice-Crackers/17119'
driver.get(url)
next_data_element = None
for script_element in driver.find_elements(By.XPATH, '//script'):
script_text = script_element.get_attribute('innerHTML')
match = re.search(r'__NEXT_DATA__ = (\{.*\});', script_text)
if match:
next_data_json = match.group(1)
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)
break
if next_data_element is None:
print('__NEXT_DATA__ element not found')
# Close the Selenium driver
driver.quit()