I want to scrape pictures from a public Instagram account. I'm pretty familiar with bs4 so I started with that. Using the element inspector on Chrome, I noted the pictures are in an unordered list and li has class 'photo', so I figure, what the hell -- can't be that hard to scrape with findAll, right?
Wrong: it doesn't return anything (code below) and I soon notice that the code shown in element inspector and the code that I drew from requests were not the same AKA no unordered list in the code I pulled from requests.
Any idea how I can get the code that shows up in element inspector?
Just for the record, this was my code to start, which didn't work because the unordered list was not there:
from bs4 import BeautifulSoup
import requests
import re
r = requests.get('http://instagram.com/umnpics/')
soup = BeautifulSoup(r.text)
for x in soup.findAll('li', {'class':'photo'}):
print x
Thank you for your help.
If you look at the source code for the page, you'll see that some javascript generates the webpage. What you see in the element browser is the webpage after the script has been run, and beautifulsoup just gets the html file. In order to parse the rendered webpage you'll need to use something like Selenium to render the webpage for you.
So, for example, this is how it would look with Selenium:
from bs4 import BeautifulSoup
import selenium.webdriver as webdriver
url = 'http://instagram.com/umnpics/'
driver = webdriver.Firefox()
driver.get(url)
soup = BeautifulSoup(driver.page_source)
for x in soup.findAll('li', {'class':'photo'}):
print x
Now the soup should be what you are expecting.
Related
Hi I am trying to scrape different websites but for some of them it outputs a cloudflare url .
As the code below, I want the href values of the anchor tags for every tags that have a value(which is the href/a url.) and print each one of them in new lines.
For example in this code:
some text
its should return the https://www.google.com in output which for some websites it doesn't work at all or return the output that I mentioned below.
import requests
from bs4 import BeautifulSoup
from selenium import webdriver
from webdriver_manager.chrome import ChromeDriverManager
page = requests.get("https://www.swappa.com")
soup = BeautifulSoup(page.content, "html.parser")
links = soup.find_all("a", href=True)
for link in links:
print(link['href'])
and vs code output this:
https://www.cloudflare.com/?utm_source=challenge&utm_campaign=j
this is not a problem of the beautiful soup but yeah this site first detects your IP and many data like which kinds of browser you are using etc and then forward it to the main site. this is a new type of blocking the web scrapping.
I am trying to use Beautiful Soup 4 to help me download an image from Imgur, although I doubt the Imgur part is relevant. As an example, I'm using the webpage here: https://imgur.com/t/lenovo/mLwnorj
My code is as follows:
import webbrowser, time, sys, requests, os, bs4 # Not all libraries are used in this code snippet
from selenium import webdriver
browser = webdriver.Firefox()
browser.get("https://imgur.com/t/lenovo/mLwnorj")
res = requests.get(https://imgur.com/t/lenovo/mLwnorj)
res.raise_for_status()
soup = bs4.BeautifulSoup(res.text, features="html.parser")
imageElement = soup.findAll('img', {'class': 'post-image-placeholder'})
print(imageElement)
The HTML code on the Imgur link contains a part that reads as:
<img alt="" src="//i.imgur.com/JfLsH5y.jpg" class="post-image-placeholder" style="max-width: 100%; min-height: 546px;" original-title="">
which I found by picking the first image element on the page using the point and click tool in Inspect Element.
The problem is that I would expect there to be two items in imageElement, one for each image, however, the print function shows []. I have also tried other forms of soup.findAll('img', {'class': 'post-image-placeholder'}) such as soup.findall("img[class='post-image-placeholder']") but that made no difference.
Furthermore, when I used
imageElement = soup.select("h1[class='post-title']")
,just to test, the print function did return a match, which made me wonder if it had something to do with the tag.
[<h1 class="post-title">Cable management increases performance. </h1>]
Thank you for your time and effort
The fundamental problem here seems to be that the actual <img ...> element is not present when the page is first loaded. The best solution to this, in my opinion, would be to take advantage of the selenium webdriver that you already have available to grab the image. Selenium will allow the page to properly render (with JavaScript and all), and then locate whatever elements you care about.
For example:
import webbrowser, time, sys, requests, os, bs4 # Not all libraries are used in this code snippet
from selenium import webdriver
# For pretty debugging output
import pprint
browser = webdriver.Firefox()
browser.get("https://imgur.com/t/lenovo/mLwnorj")
# Give the page up to 10 seconds of a grace period to finish rendering
# before complaining about images not being found.
browser.implicitly_wait(10)
# Find elements via Selenium's search
selenium_image_elements = browser.find_elements_by_css_selector('img.post-image-placeholder')
pprint.pprint(selenium_image_elements)
# Use page source to attempt to find them with BeautifulSoup 4
soup = bs4.BeautifulSoup(browser.page_source, features="html.parser")
soup_image_elements = soup.findAll('img', {'class': 'post-image-placeholder'})
pprint.pprint(soup_image_elements)
I cannot say that I have tested this code yet on my side, but the general concept should work.
Update:
I went ahead and tested this on my side, fixed some errors in the code, and I then got the results I was hoping to see:
If a website will insert objects after page load you will need to use Selenium instead of requests.
from bs4 import BeautifulSoup
from selenium import webdriver
url = 'https://imgur.com/t/lenovo/mLwnorj'
browser = webdriver.Firefox()
browser.get(url)
soup = BeautifulSoup(browser.page_source, 'html.parser')
images = soup.find_all('img', {'class': 'post-image-placeholder'})
[print(image['src']) for image in images]
# //i.imgur.com/JfLsH5yr.jpg
# //i.imgur.com/lLcKMBzr.jpg
I am new to web scraping/coding, and I am trying to use Python requests/BeautifulSoup to parse through the html code in order to get some physical and chemical properties.
For some reason, although I have used the following script for other websites successfully, BeautifulSoup has only printed a few lines from the header and footer, and then pages of HTML code that doesn't really make sense. This is the code I have been using:
import requests
from bs4 import BeautifulSoup
url='https://comptox.epa.gov/dashboard/dsstoxdb/results?search=ammonia#properties'
response = requests.get(url).text
soup=BeautifulSoup(response,'lxml')
print(soup.prettify())
When I try to find the table or even a row, it gives no output. Is there something I haven't accounted for? Any help would be greatly appreciated!
It is present in one of the attributes. You can extract as follows (there is a lot more info there but I subset to physical properties
import requests
from bs4 import BeautifulSoup as bs
import json
url = "https://comptox.epa.gov/dashboard/dsstoxdb/results?search=ammonia#properties"
r = requests.get(url)
soup = bs(r.content, 'lxml')
soup.select_one('[data-result]')['data-result']
data = json.loads(soup.select_one('[data-result]')['data-result'])
properties = data['physprop']
print(properties)
It's pretty common that if a page is populated by JavaScript after you load it requests and BeautifulSoup will not process the page correctly. The best thing to do is likely switch to the selenium module which allows your program to dynamically access the page and interact with elements. After loading (and maybe clicking on a couple elements) you can feed the HTML to BeautifulSoup and process it how you wish. The basic framework I recommend you start with would look like:
from selenium import webdriver
browser = webdriver.Chrome() # You'll need to download drivers from link above
browser.implicitly_wait(10) # probably unnecessary, just makes sure all pages you visit fully load
browser.get('https://stips.co.il/explore')
while True:
input('Press Enter to print HTML')
HTML = browser.page_source
print(HTML)
Just click around in the browser and when you want to see if the HTML is correct, click back to your prompt and press ENTER. This is how you would locate elements automatically, so you don't have to manually interact with the page every time
I want to load the list of images in this page in Python. However, when I opened the page in my browser (Chrome or Safari) and opened the dev tools, the inspector returned the list of images as <img class="grid-item--image">....
However, when I tried to parse it in Python, the result seemed different. Specifically, I got the list of images as <img class="carousel--image"...>, whereas the soup.findAll("img", "grid-item--image") did return an empty list. Also, I tried saving those images using its srcset tag, most of the saved images are NOT those that were listed on the web.
I think the web page used some sort of technics when rendering. How can I parse the web pages successfully?
I used BeautifulSoup 4 on Python 3.5. I loaded the page as follows:
import requests
from bs4 import BeautifulSoup
html = requests.get(url).text
soup = BeautifulSoup(html, "html.parser", from_encoding="utf-8")
return soup
You would do better to use something like selenium for this as follows:
from bs4 import BeautifulSoup
from selenium import webdriver
browser = webdriver.Firefox()
browser.get("http://www.vogue.com/fashion-shows/fall-2016-menswear/fendi#collection")
html_source = browser.page_source
soup = BeautifulSoup(html_source, "html.parser")
for item in soup.find_all("img", {"class":"grid-item--image"}):
print(item.get('srcset'))
This would display the following kind of output:
http://assets.vogue.com/photos/569d37e434324c316bd70f04/master/w_195/_FEN0016.jpg
http://assets.vogue.com/photos/569d37e5d928983d20a78e4f/master/w_195/_FEN0027.jpg
http://assets.vogue.com/photos/569d37e834324c316bd70f0a/master/w_195/_FEN0041.jpg
http://assets.vogue.com/photos/569d37e334324c316bd70efe/master/w_195/_FEN0049.jpg
http://assets.vogue.com/photos/569d37e702e08d8957a11e32/master/w_195/_FEN0059.jpg
...
...
...
http://assets.vogue.com/photos/569d3836486d6d3e20ae9625/master/w_195/_FEN0616.jpg
http://assets.vogue.com/photos/569d381834324c316bd70f3b/master/w_195/_FEN0634.jpg
http://assets.vogue.com/photos/569d3829fa6d6c9057f91d2a/master/w_195/_FEN0649.jpg
http://assets.vogue.com/photos/569d382234324c316bd70f41/master/w_195/_FEN0663.jpg
http://assets.vogue.com/photos/569d382b7dcd2a8a57748d05/master/w_195/_FEN0678.jpg
http://assets.vogue.com/photos/569d381334324c316bd70f2f/master/w_195/_FEN0690.jpg
http://assets.vogue.com/photos/569d382dd928983d20a78eb1/master/w_195/_FEN0846.jpg
This allows the full rendering of the page to take place inside the browser, and the resulting HTML can then be obtained.
I was trying to do web scraping and was using the following code :
import mechanize
from bs4 import BeautifulSoup
url = "http://www.thehindu.com/archive/web/2010/06/19/"
br = mechanize.Browser()
htmltext = br.open(url).read()
link_dictionary = {}
soup = BeautifulSoup(htmltext)
for tag_li in soup.findAll('li', attrs={"data-section":"Chennai"}):
for link in tag_li.findAll('a'):
link_dictionary[link.string] = link.get('href')
print link_dictionary[link.string]
urlnew = link_dictionary[link.string]
brnew = mechanize.Browser()
htmltextnew = brnew.open(urlnew).read()
articletext = ""
soupnew = BeautifulSoup(htmltextnew)
for tag in soupnew.findAll('p'):
articletext += tag.text
print articletext
I was unable to get any printed values by using this. But on using attrs={"data-section":"Business"} instead of attrs={"data-section":"Chennai"} I was able to get the desired output. Can someone help me?
READ THE TERMS OF SERVICES OF THE WEBSITE BEFORE SCRAPING
If you are using firebug or inspect element in Chrome, you might see some contents that will not be seen if you are using Mechanize or Urllib2.
For example, when you view the source code of the page sent out by you. (Right click view source in Chrome). and search for data-section tag, you won't see any tags which chennai, I am not 100% sure but I will say those contents need to be populated by Javascript ..etc. which requires the functionality of a browser.
If I were you, I will use selenium to open up the page and then get the source page from there, then the HTML collected in that way will be more like what you see in a browser.
Cited here
from selenium import webdriver
from bs4 import BeautifulSoup
import time
driver = webdriver.Firefox()
driver.get("URL GOES HERE")
# I noticed there is an ad here, sleep til page fully loaded.
time.sleep(10)
soup = BeautifulSoup(driver.page_source)
print len(soup.findAll(...}))
# or you can work directly in selenium
...
driver.close()
And the output for me is 8