kw_tools/hhh/SEO/hhh_phantom.py

from selenium import webdriver
import time
import random
import codecs



fpath=__file__
fpath=fpath.replace('hhh_phantom.py','urls.csv')

print(fpath)

lst=[]
fr=codecs.open(fpath,'r','utf-8')
lines=fr.readlines()
for l in lines:
    elmts=l.split(',')
    lst.append('https://www.hhh.com.tw'+elmts[0])
fr.close()

#lst=['13781','3649','2116']
#for i in range(9999):
if True:
    options = webdriver.ChromeOptions()
    options.add_argument('--headless')
    options.add_argument('--disable-gpu')  # Last I checked this was necessary.
    options.add_argument('--incognito')
    options.add_argument('--no-sandbox')
    options.add_argument('--disable-dev-shm-usage')
    driver = webdriver.Chrome(
    desired_capabilities=options.to_capabilities())
    driver.set_window_size(1400,1000)
    for i in range(3):
        l=random.choice(lst)
    #    driver.get("https://www.hhh.com.tw/cases/detail/"+l+"/index.php")
        driver.get(l)
        print(driver.current_url)
        driver.execute_script("window.scrollTo(0, window.scrollY + 200)")
#        driver.implicitly_wait (2)
        time.sleep(6)
    driver.quit()