Website Logo Scraper

A Python tool for automatically scraping and extracting logo images from websites.

Features

• Extracts logo images from websites using various detection methods
• Processes multiple websites concurrently to improve efficiency
• Uses intelligent heuristics to identify the most likely logo candidate
• Provides comprehensive logging and progress tracking
• Randomized delays and user agents to avoid rate limiting

Installation

1. Clone this repository
2. Install the required dependencies:

pip install -r requirements.txt

Usage

The script requires an input CSV file with a column named website containing the URLs to scrape.

Basic Usage

python logo_scraper.py --input websites.csv --output logos.csv

Command-line Arguments

• --input, -i: Input CSV file with websites (default: websites.csv)
• --output, -o: Output CSV file for logo URLs (default: logos.csv)
• --workers, -w: Number of worker threads (default: 5)
• --delay, -d: Delay between requests in seconds (default: 1.0)
• --timeout, -t: Request timeout in seconds (default: 10)

Example

python logo_scraper.py --input companies.csv --output company_logos.csv --workers 10 --delay 2 --timeout 15

Input Format

The input CSV file should contain a column named website with the URLs to scrape:

website
example.com
google.com
github.com

Output Format

The script outputs a CSV file with the following columns:

• website: The original website URL
• logo_url: The URL of the extracted logo (if found)
• status: The status of the extraction (success, no logo found, or an error message)

How It Works

The logo scraper uses multiple methods to identify logo images:

1. Looks for images with "logo" in their URL, class name, ID, or alt text
2. Checks for images positioned in header elements or as home page links
3. Examines SVG elements with "logo" in their class names
4. Checks for meta tags with OpenGraph images
5. Looks for favicon and apple-touch-icon links

For each website, the script:

1. Sends an HTTP request with a random user agent
2. Parses the HTML content
3. Applies logo detection heuristics
4. Scores potential logo candidates
5. Returns the highest-scoring logo URL

License

MIT