Using python Requests with javascript pages

Question

I am trying to use the Requests framework with python (http://docs.python-requests.org/en/latest/) but the page I am trying to get to uses javascript to fetch the info that I want.

I have tried to search on the web for a solution but the fact that I am searching with the keyword javascript most of the stuff I am getting is how to scrape with the javascript language.

Is there anyway to use the requests framework with pages that use javascript?

Solution 1

You are going to have to make the same request (using the Requests library) that the javascript is making. You can use any number of tools (including those built into Chrome and Firefox) to inspect the http request that is coming from javascript and simply make this request yourself from Python.

Solution 2

Good news: there is now a requests module that supports javascript: https://pypi.org/project/requests-html/

from requests_html import HTMLSession

session = HTMLSession()

r = session.get('http://www.yourjspage.com')

r.html.render()  # this call executes the js in the page

As a bonus this wraps BeautifulSoup, I think, so you can do things like

r.html.find('#myElementID').text

which returns the content of the HTML element as you'd expect.

Solution 3

While Selenium might seem tempting and useful, it has one main problem that can't be fixed: performance. By calculating every single thing a browser does, you will need a lot more power. Even PhantomJS does not compete with a simple request. I recommend that you will only use Selenium when you really need to click buttons. If you only need javascript, I recommend PyQt (check https://www.youtube.com/watch?v=FSH77vnOGqU to learn it).

However, if you want to use Selenium, I recommend Chrome over PhantomJS. Many users have problems with PhantomJS where a website simply does not work in Phantom. Chrome can be headless (non-graphical) too!

First, make sure you have installed ChromeDriver, which Selenium depends on for using Google Chrome.

Then, make sure you have Google Chrome of version 60 or higher by checking it in the URL chrome://settings/help

Now, all you need to do is the following code:

from selenium.webdriver.chrome.options import Options
from selenium import webdriver

chrome_options = Options()
chrome_options.add_argument("--headless")

driver = webdriver.Chrome(chrome_options=chrome_options)

If you do not know how to use Selenium, here is a quick overview:

driver.get("https://www.google.com") #Browser goes to google.com

Finding elements: Use either the ELEMENTS or ELEMENT method. Examples:

driver.find_element_by_css_selector("div.logo-subtext") #Find your country in Google. (singular)

• driver.find_element(s)_by_css_selector(css_selector) # Every element that matches this CSS selector
• driver.find_element(s)_by_class_name(class_name) # Every element with the following class
• driver.find_element(s)_by_id(id) # Every element with the following ID
• driver.find_element(s)_by_link_text(link_text) # Every with the full link text
• driver.find_element(s)_by_partial_link_text(partial_link_text) # Every with partial link text.
• driver.find_element(s)_by_name(name) # Every element where name=argument
• driver.find_element(s)_by_tag_name(tag_name) # Every element with the tag name argument

Ok! I found an element (or elements list). But what do I do now?

Here are the methods you can do on an element elem:

• elem.tag_name # Could return button in a .
• elem.get_attribute("id") # Returns the ID of an element.
• elem.text # The inner text of an element.
• elem.clear() # Clears a text input.
• elem.is_displayed() # True for visible elements, False for invisible elements.
• elem.is_enabled() # True for an enabled input, False otherwise.
• elem.is_selected() # Is this radio button or checkbox element selected?
• elem.location # A dictionary representing the X and Y location of an element on the screen.
• elem.click() # Click elem.
• elem.send_keys("thelegend27") # Type thelegend27 into elem (useful for text inputs)
• elem.submit() # Submit the form in which elem takes part.

Special commands:

• driver.back() # Click the Back button.
• driver.forward() # Click the Forward button.
• driver.refresh() # Refresh the page.
• driver.quit() # Close the browser including all the tabs.
• foo = driver.execute_script("return 'hello';") # Execute javascript (COULD TAKE RETURN VALUES!)

Comments

[biw]

So there is no way to have requests use javascript.

[sberry]

No, Requests is an http library. It cannot run javascript.

[biw]

I used Chrome tools to debug the website and look for what the Javascript was calling. You can see the results of what I created at github.com/719Ben/myCUinfo-API

[phrogg]

Shouldn't it be r.html.find('#myElementID').text? And also r = session.get('http://www.yourjspage.com')?

[Joshua Stafford]

After fixing the issues that Phil pointed out, I still got "RuntimeError: Cannot use HTMLSession within an existing event loop. Use AsyncHTMLSession instead."

[KubaFYI]

This exists only for python 3 as far as I'm concerned. Is there anything Python 2.7 users can do here?

[Alejandro Braun]

@KubaFYI Yes, they can start moving things over to python3

[Vanessa]

@HuckIt To solve this problem, you'll import AsyncHTMLSession instead of HTMLSession and the render will be called with await session.get(url).result().arender(). I just got this problem and this is how I solved it.

[Sinan Çetinkaya]

As it's written in its doc requests.readthedocs.io/projects/requests-html/en/latest/… requests_html uses Chromium in the background. So it's Chromium browser controlled by a requests-like wrapper.

[cikatomo]

So far this is the best. You can also get nice JSON so its easier to get data

[Eric Nelson]

Unfortunately, requests-html is python 3.6 only, and I'm on 3.8

[Akash Patel]

r.html.render() Is there any way to execute it in chrome browser? I am getting "this browser is not supported. Please reconnect using the Chrome browser.". Which I have set for using browser other than chrome.

[Corsaka]

@EricNelson It still works for me in 3.9. I don't believe there's any real issues with using a module built for 3.6.

[asmaier]

From the documentation: Note, the first time you ever run the render() method, it will download Chromium into your home directory. This only happens once.