def return_html_code(url,use_proxy):
dcap = dict(webdriver.DesiredCapabilities.PHANTOMJS)
dcap["phantomjs.page.settings.userAgent"] = "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_8_4) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/29.0.1547.57 Safari/537.36"
if use_proxy==True:
proxy_address=random.choice(get_proxy_fastest())
proxy_type='https'
print proxy_address,proxy_type
service_args = [
'--proxy='+proxy_address,
'--proxy-type='+proxy_type,
]
driver = webdriver.PhantomJS(desired_capabilities=dcap,service_args=service_args)
else:
driver = webdriver.PhantomJS(desired_capabilities=dcap)
driver.maximize_window()
driver.get(url)
print 'Loading initial page'
# initial wait for the tweets to load
wait = WebDriverWait(driver, 10)
try:
wait.until(EC.visibility_of_element_located((By.CSS_SELECTOR, "li[data-item-id]")))
except TimeoutException:
print 'No tweets here'
driver.quit()
return False
# scroll down to the last tweet until there is no more tweets loaded
print 'Scrolling tweets'
while True:
tweets = driver.find_elements_by_css_selector("li[data-item-id]")
number_of_tweets = len(tweets)
print(number_of_tweets),
# move to the top and then to the bottom 5 times in a row
for _ in range(5):
driver.execute_script("window.scrollTo(0, 0)")
driver.execute_script("arguments[0].scrollIntoView(true);", tweets[-1])
time.sleep(0.5)
try:
wait.until(wait_for_more_than_n_elements_to_be_present((By.CSS_SELECTOR, "li[data-item-id]"), number_of_tweets))
except TimeoutException:
break
html_full_source=driver.page_source
driver.quit()
print "_"*15
#with open("check.html",'w') as f: f.write(html_full_source)
return html_full_source
def return_html_code(url, proxy_use):
vdisplay = Xvfb()
vdisplay.start()
proxy_address_list = get_proxy_fastest()
if proxy_address_list != False:
proxy_address = random.choice(proxy_address_list)
ip, port = proxy_address.split(":")
print ip, port
profile = webdriver.FirefoxProfile()
profile.set_preference("network.proxy.http", ip)
profile.set_preference("network.proxy.http_port", port)
profile.set_preference("network.proxy_type", 1)
driver = webdriver.Firefox(firefox_profile=profile)
else:
print "Using localhost, unable to get proxy"
driver = webdriver.Firefox()
driver.maximize_window()
driver.get(url)
# initial wait for the tweets to load
# initial wait for the tweets to load
wait = WebDriverWait(driver, 30)
try:
wait.until(EC.visibility_of_element_located((By.CSS_SELECTOR, "li[data-item-id]")))
except TimeoutException:
driver.quit()
return False # scroll down to the last tweet until there is no more tweets loaded
while True:
tweets = driver.find_elements_by_css_selector("li[data-item-id]")
print len(tweets) # added in edit 1
number_of_tweets = len(tweets)
driver.execute_script("arguments[0].scrollIntoView(true);", tweets[-1])
try:
wait.until(
wait_for_more_than_n_elements_to_be_present((By.CSS_SELECTOR, "li[data-item-id]"), number_of_tweets)
)
except TimeoutException:
break
html_full_source = driver.page_source
driver.close()
vdisplay.stop()
return html_full_source
