Skip to main content

Extracting data from HTML with Python

In this lesson we'll finish extracting product data from the downloaded HTML. With help of basic string manipulation we'll focus on cleaning and correctly representing the product price.

Locating the right HTML elements is the first step of a successful data extraction, so it's no surprise that we're already close to having the data in the correct form. The last bit that still requires our attention is the price:

$ python main.py
JBL Flip 4 Waterproof Portable Bluetooth Speaker | $74.95
Sony XBR-950G BRAVIA 4K HDR Ultra HD TV | From $1,398.00
...

Let's summarize what stands in our way if we want to have it in our Python program as a number:

  • A dollar sign precedes the number,
  • the number contains decimal commas for better human readability, and
  • some prices start with From, which reveals there is a certain complexity in how the shop deals with prices.

Representing price​

The last bullet point is the most important to figure out before we start coding. We thought we'll be scraping numbers, but in the middle of our effort, we discovered that the price is actually a range.

It's because some products have variants with different prices. Later in the course we'll get to crawling, i.e. following links and scraping data from more than just one page. That will allow us to get exact prices for all the products, but for now let's extract just what's in the listing.

Ideally we'd go and discuss the problem with those who are about to use the resulting data. For their purposes, is the fact that some prices are just minimum prices important? What would be the most useful representation of the range for them? Maybe they'd tell us that it's okay if we just remove the From prefix?

price_text = product.select_one(".price").contents[-1]
price = price_text.removeprefix("From ")

In other cases, they'd tell us the data must include the range. And in cases when we just don't know, the safest option is to include all the information we have and leave the decision on what's important to later stages. One approach could be having the exact and minimum prices as separate values. If we don't know the exact price, we leave it empty:

price_text = product.select_one(".price").contents[-1]
if price_text.startswith("From "):
    min_price = price_text.removeprefix("From ")
    price = None
else:
    min_price = price_text
    price = min_price
Built-in string methods

If you're not proficient in Python's string methods, .startswith() checks the beginning of a given string, and .removeprefix() removes something from the beginning of a given string.

The whole program would look like this:

import httpx
from bs4 import BeautifulSoup

url = "https://warehouse-theme-metal.myshopify.com/collections/sales"
response = httpx.get(url)
response.raise_for_status()

html_code = response.text
soup = BeautifulSoup(html_code, "html.parser")

for product in soup.select(".product-item"):
    title = product.select_one(".product-item__title").text

    price_text = product.select_one(".price").contents[-1]
    if price_text.startswith("From "):
        min_price = price_text.removeprefix("From ")
        price = None
    else:
        min_price = price_text
        price = min_price

    print(title, min_price, price, sep=" | ")

Removing white space​

Often, the strings we extract from a web page start or end with some amount of whitespace, typically space characters or newline characters, which come from the indentation of the HTML tags.

We call the operation of removing whitespace stripping or trimming, and it's so useful in many applications that programming languages and libraries include ready-made tools for it. Let's add Python's built-in .strip():

title = product.select_one(".product-item__title").text.strip()

price_text = product.select_one(".price").contents[-1].strip()
Handling strings in Beautiful Soup

Beautiful Soup offers several attributes when it comes to working with strings:

These might be useful in some complex scenarios, but in our case, they won't make scraping the title or price any shorter or more elegant.

Removing dollar sign and commas​

We got rid of the From and possible whitespace, but we still can't save the price as a number in our Python program:

>>> price = "$1,998.00"
>>> float(price)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ValueError: could not convert string to float: '$1,998.00'
Interactive Python

The demonstration above is inside the Python's interactive REPL. It's a useful playground where you can try how code behaves before you use it in your program.

We need to remove the dollar sign and the decimal commas. For this type of cleaning, regular expressions are often the best tool for the job, but in this case .replace() is also sufficient:

price_text = (
    product
    .select_one(".price")
    .contents[-1]
    .strip()
    .replace("$", "")
    .replace(",", "")
)

Representing money in programs​

Now we should be able to add float(), so that we have the prices not as a text, but as numbers:

if price_text.startswith("From "):
    min_price = float(price_text.removeprefix("From "))
    price = None
else:
    min_price = float(price_text)
    price = min_price

Great! Only if we didn't overlook an important pitfall called floating-point error. In short, computers save float() numbers in a way which isn't always reliable:

>>> 0.1 + 0.2
0.30000000000000004

These errors are small and usually don't matter, but sometimes they can add up and cause unpleasant discrepancies. That's why it's typically best to avoid float() when working with money. Let's instead use Python's built-in Decimal() type:

from decimal import Decimal
import httpx
from bs4 import BeautifulSoup

url = "https://warehouse-theme-metal.myshopify.com/collections/sales"
response = httpx.get(url)
response.raise_for_status()

html_code = response.text
soup = BeautifulSoup(html_code, "html.parser")

for product in soup.select(".product-item"):
    title = product.select_one(".product-item__title").text.strip()

    price_text = (
        product
        .select_one(".price")
        .contents[-1]
        .strip()
        .replace("$", "")
        .replace(",", "")
    )
    if price_text.startswith("From "):
        min_price = Decimal(price_text.removeprefix("From "))
        price = None
    else:
        min_price = Decimal(price_text)
        price = min_price

    print(title, min_price, price, sep=" | ")

If we run the code above, we have nice, clean data about all the products!

$ python main.py
JBL Flip 4 Waterproof Portable Bluetooth Speaker | 74.95 | 74.95
Sony XBR-950G BRAVIA 4K HDR Ultra HD TV | 1398.00 | None
...

Well, not to spoil the excitement, but in its current form, the data isn't very useful. In the next lesson we'll save the product details to a file which data analysts can use or other programs can read.

Exercises​

These challenges are here to help you test what you’ve learned in this lesson. Try to resist the urge to peek at the solutions right away. Remember, the best learning happens when you dive in and do it yourself!

Real world

You're about to touch the real web, which is practical and exciting! But websites change, so some exercises might break. If you run into any issues, please leave a comment below or file a GitHub Issue.

Scrape units on stock​

Change our scraper so that it extracts how many units of each product are on stock. Your program should print the following. Note the unit amounts at the end of each line:

JBL Flip 4 Waterproof Portable Bluetooth Speaker 672
Sony XBR-950G BRAVIA 4K HDR Ultra HD TV 77
Sony SACS9 10" Active Subwoofer 7
Sony PS-HX500 Hi-Res USB Turntable 15
Klipsch R-120SW Powerful Detailed Home Speaker - Unit 0
Denon AH-C720 In-Ear Headphones 236
...
Solution
import httpx
from bs4 import BeautifulSoup

url = "https://warehouse-theme-metal.myshopify.com/collections/sales"
response = httpx.get(url)
response.raise_for_status()

html_code = response.text
soup = BeautifulSoup(html_code, "html.parser")

for product in soup.select(".product-item"):
    title = product.select_one(".product-item__title").text.strip()

    units_text = (
        product
        .select_one(".product-item__inventory")
        .text
        .removeprefix("In stock,")
        .removeprefix("Only")
        .removesuffix(" left")
        .removesuffix("units")
        .strip()
    )
    if "Sold out" in units_text:
        units = 0
    else:
        units = int(units_text)

    print(title, units)

Use regular expressions​

Simplify the code from previous exercise. Use regular expressions to parse the number of units. You can match digits using a range like [0-9] or by a special sequence \d. To match more characters of the same type you can use +.

Solution
import re
import httpx
from bs4 import BeautifulSoup

url = "https://warehouse-theme-metal.myshopify.com/collections/sales"
response = httpx.get(url)
response.raise_for_status()

html_code = response.text
soup = BeautifulSoup(html_code, "html.parser")

for product in soup.select(".product-item"):
    title = product.select_one(".product-item__title").text.strip()

    units_text = product.select_one(".product-item__inventory").text
    if re_match := re.search(r"\d+", units_text):
        units = int(re_match.group())
    else:
        units = 0

    print(title, units)

Scrape publish dates of F1 news​

Download Guardian's page with the latest F1 news and use Beautiful Soup to parse it. Print titles and publish dates of all the listed articles. This is the URL:

https://www.theguardian.com/sport/formulaone

Your program should print something like the following. Note the dates at the end of each line:

Wolff confident Mercedes are heading to front of grid after Canada improvement 2024-06-10
Frustrated Lando Norris blames McLaren team for missed chance 2024-06-09
Max Verstappen wins Canadian Grand Prix: F1 – as it happened 2024-06-09
...

Hints:

Solution
import httpx
from datetime import datetime
from bs4 import BeautifulSoup

url = "https://www.theguardian.com/sport/formulaone"
response = httpx.get(url)
response.raise_for_status()

html_code = response.text
soup = BeautifulSoup(html_code, "html.parser")

for article in soup.select("#maincontent ul li"):
    title = article.select_one("h3").text.strip()

    time_iso = article.select_one("time")["datetime"].strip()
    published_at = datetime.fromisoformat(time_iso)
    published_on = published_at.date()

    print(title, published_on)