Scraping Images

Introduction

You've definitely started to hone your skills at scraping now! With that, let's look at another data format you're apt to want to pull from the web: images! In this lesson, you'll see how to save images from the web as well as display them in a Pandas DataFrame for easy perusal!

Objectives

You will be able to:

Select specific elements from HTML using Beautiful Soup
Identify and scrape images from a web page

Grabbing an HTML Page

Start with the same page that you've been working with: books.toscrape.com.

from bs4 import BeautifulSoup
import requests

html_page = requests.get('http://books.toscrape.com/') # Make a get request to retrieve the page
soup = BeautifulSoup(html_page.content, 'html.parser') # Pass the page contents to beautiful soup for parsing
warning = soup.find('div', class_="alert alert-warning")
book_container = warning.nextSibling.nextSibling

Finding Images

First, simply retrieve a list of images by searching for img tags with beautiful soup:

images = book_container.findAll('img')
ex_img = images[0] # Preview an entry
ex_img

<img alt="A Light in the Attic" class="thumbnail" src="media/cache/2c/da/2cdad67c44b002e7ead0cc35693c0e8b.jpg"/>

# Use tab complete to preview what types of methods are available for the entry
# ex_img.

# While there's plenty of other methods to explore, simply select the url for the image for now.
ex_img.attrs['src']

'media/cache/2c/da/2cdad67c44b002e7ead0cc35693c0e8b.jpg'

Saving Images

Great! Now that you have a URL (well, a URL extension to be more precise) you can download the image locally!

import shutil

url_base = "http://books.toscrape.com/"
url_ext = ex_img.attrs['src']
full_url = url_base + url_ext
r = requests.get(full_url, stream=True)
if r.status_code == 200:
    with open("images/book1.jpg", 'wb') as f:
        r.raw.decode_content = True
        shutil.copyfileobj(r.raw, f)

Showing Images in the File Directory

You can also run a simple bash command in a standalone cell to preview that the image is indeed there:

ls images/

book-section.png  book14.jpg        book2.jpg         book7.jpg
book1.jpg         book15.jpg        book20.jpg        book8.jpg
book10.jpg        book16.jpg        book3.jpg         book9.jpg
book11.jpg        book17.jpg        book4.jpg
book12.jpg        book18.jpg        book5.jpg
book13.jpg        book19.jpg        book6.jpg

Previewing an Individual Image

import matplotlib.pyplot as plt
import matplotlib.image as mpimg

img = mpimg.imread('images/book1.jpg')
imgplot = plt.imshow(img)
plt.show()

Displaying Images in Pandas DataFrames

You can even display images within a pandas DataFrame by using a little HTML yourself!

import pandas as pd
from IPython.display import Image, HTML

row1 = [ex_img.attrs['alt'], '<img src="images/book1.jpg"/>']
df = pd.DataFrame(row1).transpose()
df.columns = ['title', 'cover']
HTML(df.to_html(escape=False))

	title	cover
0	A Light in the Attic

All Together Now

data = []
for n, img in enumerate(images):
    url_base = "http://books.toscrape.com/"
    url_ext = img.attrs['src']
    full_url = url_base + url_ext
    r = requests.get(full_url, stream=True)
    path = "images/book{}.jpg".format(n+1)
    title = img.attrs['alt']
    if r.status_code == 200:
        with open(path, 'wb') as f:
            r.raw.decode_content = True
            shutil.copyfileobj(r.raw, f)
        row = [title, '<img src="{}"/>'.format(path)]
        data.append(row)
df = pd.DataFrame(data)
print('Number of rows: ', len(df))
df.columns = ['title', 'cover']
HTML(df.to_html(escape=False))

Number of rows:  20

	title	cover
0	A Light in the Attic
1	Tipping the Velvet
2	Soumission
3	Sharp Objects
4	Sapiens: A Brief History of Humankind
5	The Requiem Red
6	The Dirty Little Secrets of Getting Your Dream Job
7	The Coming Woman: A Novel Based on the Life of the Infamous Feminist, Victoria Woodhull
8	The Boys in the Boat: Nine Americans and Their Epic Quest for Gold at the 1936 Berlin Olympics
9	The Black Maria
10	Starving Hearts (Triangular Trade Trilogy, #1)
11	Shakespeare's Sonnets
12	Set Me Free
13	Scott Pilgrim's Precious Little Life (Scott Pilgrim #1)
14	Rip it Up and Start Again
15	Our Band Could Be Your Life: Scenes from the American Indie Underground, 1981-1991
16	Olio
17	Mesaerion: The Best Science Fiction Stories 1800-1849
18	Libertarianism for Beginners
19	It's Only the Himalayas

Summary

Voila! You now know how to use your knowledge of HTML and Beautiful Soup to scrape images. You really are turning into a scraping champion! Now, go get scraping!

sbaltodano/dsc-scraping-images