DEV Community

Cover image for Scrape Google Recipe Results with Python
Dmitriy Zub ☀️
Dmitriy Zub ☀️

Posted on • Edited on

Scrape Google Recipe Results with Python

#python #webscraping #tutorial #datascience

Intro

This blog post is a continuation of Google's web scraping series. Here you'll see how to scrape Google Recipe Results from Organic Results using Python with beautifulsoup, requests, lxml libraries. An alternative API solution will be shown.

Imports 

import requests, lxml from bs4 import BeautifulSoup from serpapi import GoogleSearch
Enter fullscreen mode Exit fullscreen mode

What will be scraped

image

Process

SelectorGadget Chrome extension was used to grab CSS selectors.

Selecting Container, Title, Rating, Reviews, Time to cook, Ingredients, Source

Selecting URL

Code 

import requests, lxml from bs4 import BeautifulSoup headers = { "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/70.0.3538.102 Safari/537.36 Edge/18.19582" } params = {"q": "lasagna recipe", "hl": "en", 'gl': 'us'} response = requests.get("https://www.google.com/search", params=params, headers=headers) soup = BeautifulSoup(response.text, 'lxml') for result in soup.select('.cv2VAd'): try: title = result.select_one('.hfac6d').text except: title = None source = result.select_one('.KuNgxf').text total_time = result.select_one('.z8gr9e').text # returns a list() if need to extract certain ingredient  ingredients = result.select_one('.LDr9cf').text.split(',') rating = result.select_one('.YDIN4c').text reviews = result.select_one('.HypWnf').text.replace('(', '').replace(')', '') print(f'{title}\n{source}\n{total_time}\n{rating}\n{reviews}\n{ingredients}\n') -------------- ''' World's Best Lasagna Allrecipes 3 hr 15 min 4.8 19K ['Sweet italian sausage', ' lean ground beef', ' ricotta cheese', ' tomato sauce', ' lasagna noodles'] '''
Enter fullscreen mode Exit fullscreen mode

Extracting all recipes

If you want to extend this program and extract every recipe that Google provides, then it's time for Selenium.

I believe that there're several ways to make it work, I'll show one approach.

To get actual clicking until there's nothing to click on we need to use a while loop in combination with .is_displayed() method which returns True or False.

The logic is that it clicks on the Show more button and once there's nothing to click on it breaks out of the while loop and begin to scrape all recipes that where loaded during clicking:

While loop snippet explanation:

while True: # locates show more button element  show_more_button = driver.find_element_by_xpath('//*[@id="isl_13"]/div[5]/div[2]').is_displayed() time.sleep(1) try: # clicks on show more button  WebDriverWait(driver, 20).until(EC.element_to_be_clickable((By.XPATH, '//*[@id="isl_13"]/div[5]/div[2]'))).click() except: # pass were used to ignore exception error and continue execution  pass # if show_more_button element becomes False, breaks the loop  if show_more_button == False: break # under the hood it looks like this ''' True True True True True True True True True True True True False # breaks from the loop '''
Enter fullscreen mode Exit fullscreen mode

Full code: 

import time from selenium import webdriver from selenium.webdriver.chrome.options import Options from selenium.webdriver.support.wait import WebDriverWait from selenium.webdriver.support import expected_conditions as EC from selenium.webdriver.common.by import By driver = webdriver.Chrome(executable_path='path/to/chromedriver.exe') driver.get('https://www.google.com/search?q=lasagna+recipe&hl=en') # buffer for everything to load time.sleep(10) while True: # returns True or False. If element is not displayed (False), breaks out of the while loop  # pardon my french (xpath)  show_more_button = driver.find_element_by_xpath('/html/body/div[7]/div/div[10]/div[1]/div/div[2]/div[2]/div/div/div[1]/div/div/g-section-with-header/div[2]/g-expandable-container/div/div/div/div[5]/div[2]').is_displayed() print(show_more_button) # just for debug  time.sleep(1) try: WebDriverWait(driver, 5).until(EC.element_to_be_clickable((By.XPATH, '/html/body/div[7]/div/div[10]/div[1]/div/div[2]/div[2]/div/div/div[1]/div/div/g-section-with-header/div[2]/g-expandable-container/div/div/div/div[5]/div[2]'))).click() except: pass # if show_more_button element becomes False, break the loop  if show_more_button == False: break for index, result in enumerate(driver.find_elements_by_css_selector('.cv2VAd')): try: title = result.find_element_by_css_selector('.hfac6d').text except: title = None link = result.find_element_by_css_selector('.cv2VAd .v1uiFd a').get_attribute('href') source = result.find_element_by_css_selector('.KuNgxf').text try: total_time = result.find_element_by_css_selector('.wHYlTd').text except: total_time = None try: # stays the list if need to extract certain ingredient  ingredients = result.find_element_by_css_selector('.LDr9cf').text.split(',') except: ingredients = None try: rating = result.find_element_by_css_selector('.YDIN4c').text except: rating = None try: reviews = result.find_element_by_css_selector('.HypWnf').text.replace('(', '').replace(')', '') except: reviews = None print(f'{index + 1}\n{title}\n{link}\n{source}\n{total_time}\n{ingredients}\n{rating}\n{reviews}\n') driver.quit() ----------- ''' 1 # first element World's Best Lasagna https://www.allrecipes.com/recipe/23600/worlds-best-lasagna/ Allrecipes 3 hr 15 min ['Sweet italian sausage', ' lean ground beef', ' ricotta cheese', ' tomato sauce', ' lasagna noodles'] 4.8 19K ... 102 # last element Last Minute Red Lasagna https://www.101cookbooks.com/last-minute-red-lasagna-recipe/ 101 Cookbooks 40 min ['Whole wheat no', ' red lentils', ' red pepper flakes', ' fresh pasta sheets', ' mozzarella cheese'] 4.4 29 '''
Enter fullscreen mode Exit fullscreen mode

Note that GIF is sped-up

Using Google Recipes Results API

SerpApi is a paid API with a free plan.

The difference is that you get structured JSON with correctly formatted data, for example, 19K reviews will be 19000 which is useful if you're doing some sort of analysis and don't have to waste time on making conversion. Also, you get access to thumbnails data if you need them.

Note: it scrapes only 3 results, just like with beautifulsoup.

All recipe results is currently under development (check out in the docs if it's already implemented)

import json # just for pretty output from serpapi import GoogleSearch params = { "api_key": "YOUR_API_KEY", "engine": "google", "q": "lasagna recipe", "gl": "us", "hl": "en" } search = GoogleSearch(params) results = search.get_dict() for result in results['recipes_results']: print(json.dumps(result, indent=2, ensure_ascii=False)) ------------- ''' { "title": "World's Best Lasagna", "link": "https://www.allrecipes.com/recipe/23600/worlds-best-lasagna/", "source": "Allrecipes", "rating": 4.8, "reviews": 19000, "total_time": "3 hr 15 min", "ingredients": [ "Sweet italian sausage", "lean ground beef", "ricotta cheese", "tomato sauce", "lasagna noodles" ], "thumbnail": "https://serpapi.com/searches/60e58864ad7fa97ae27832be/images/b6c5f341384d29417e5d9a3b87dba8a0a0c34837ef00120357584c23f98de567.jpeg" } '''
Enter fullscreen mode Exit fullscreen mode

Acknowledgements

Links

Code in the online IDE • Google Recipes Results API

Outro

If you have any questions or something isn't working correctly or you want to write something else, feel free to drop a comment in the comment section or via Twitter at @serp_api.

Yours,
Dimitry, and the rest of SerpApi Team.

Top comments (2)

Collapse
 
mhihasan profile image
Hasanul Islam •

Do you think selecting by css selector is a good idea? Because, it changes frequently.

Collapse
 
dmitryzub profile image
Dmitriy Zub ☀️ •

Hi @mhihasan, thank you for the comment🙂

It is not a particulary good idea unless you want to fix erros in the parser on regular basis when CSS selector is changed🙂

In case of recipe results, there's no other option to parse data from i.e there's no way to parse data from inline <script> tags or from XHR or other type of request. Only from CSS selectors.

We can can use jscontroller/jsname/jsaction attributes if possible instead since they're changing less frequently or not at all.