如何从Sephora网站的aria-label标签中提取产品评分？

Hey there! I see you're stuck on grabbing the product rating from Sephora's pages — let's get that sorted out. The main issue with your current code is that soup.find('a', {'aria-label'}) only fetches the HTML element itself, not the actual rating value stored in the aria-label attribute. Plus, Sephora's rating elements have a specific structure we need to target properly.

Let's Break Down the Problem #

Sephora's product rating lives in an <a> tag with an aria-label that looks like this: aria-label="4.7 stars, 1000+ reviews". Your original code wasn't extracting the numeric rating from this attribute, which is why the extraction failed.

Updated Code with Fixes #

Here's your revised code with two key improvements:

1. Properly extracting the rating from the aria-label attribute
2. Adding a user-agent header to avoid being blocked by Sephora's anti-scraping measures

import requests
from bs4 import BeautifulSoup

final_products = [] # Empty list to append product dictionaries before DataFrame conversion
headers = {
    'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/114.0.0.0 Safari/537.36'
}

for current_url in link_list:  # Simplified the loop (no need for index range)
    try:
        resource = requests.get(current_url, headers=headers)
        resource.raise_for_status()  # Raise error for bad HTTP responses (e.g., 403, 404)
        current_data = resource.text
        soup = BeautifulSoup(current_data, 'html.parser')
        
        product = {}
        # Extract product name
        name_elem = soup.find('span', {'class': 'css-0'})
        product['Name'] = name_elem.text.strip() if name_elem else None
        
        # Extract price
        price_elem = soup.find('div', {'class': 'css-slwsq8'})
        product['Price'] = price_elem.text.strip() if price_elem else None
        
        # Extract number of reviews
        review_elems = soup.find_all('span', {'class': 'css-2rg6q7'})
        product['Num_of_reviews'] = review_elems[0].text.strip() if review_elems else None
        
        # Extract number of likes
        likes_elem = soup.find('span', {'data-at': 'product_love_count'})
        product['Num_of_Likes'] = likes_elem.text.strip() if likes_elem else None
        
        # Extract rating (fixed part!)
        rating_elem = soup.find('a', {'aria-label': lambda attr: attr and 'stars' in attr.lower()})
        if rating_elem:
            aria_label = rating_elem['aria-label']
            # Split the label to get the numeric rating (e.g., "4.7" from "4.7 stars...")
            product['Rating'] = aria_label.split()[0]
        else:
            product['Rating'] = None  # Mark as None if no rating found
        
        final_products.append(product)
        
    except requests.exceptions.RequestException as e:
        print(f"Error fetching {current_url}: {e}")
    except Exception as e:
        print(f"Error parsing {current_url}: {e}")

Key Changes Explained #

• User-Agent Header: Sephora blocks requests without a proper user-agent, so adding this mimics a real browser visit to avoid being flagged as a crawler.
• Targeted Rating Extraction: We use a lambda function in find() to locate the <a> tag with an aria-label containing "stars", then extract the first part of that attribute (the numeric rating).
• Graceful Error Handling: Instead of a broad except: pass, we now catch specific request errors and general parsing errors, so you can debug issues without silently skipping pages.
• None Fallbacks: For each field, we check if the element exists before accessing .text, preventing AttributeErrors if a field is missing on a product page.

Quick Notes #

• Sephora's HTML classes might change over time, so if this breaks later, inspect the product page again to update the class names or attributes.
• Consider adding small delays between requests (time.sleep(2)) to avoid overwhelming Sephora's servers and getting blocked.

内容的提问来源于stack exchange，提问作者DJ-coding