Fontawesome v6.5.1 scraper

fa-scrape.py

import os
import re
import requests
from urllib.parse import urljoin, urlparse

def download_file(url, directory_path):
    filename = url.split("/")[-1]
    filepath = os.path.join(directory_path, filename)
    with open(filepath, "wb") as f:
        f.write(requests.get(url).content)
    print(f"Downloaded {filename} to {directory_path}")

def fetch_and_download_files(css_url, base_url, download_dir):
    response = requests.get(css_url)
    if response.status_code == 200:
        css_content = response.text
        file_urls = re.findall(r"url\((.*?)\)", css_content)
        for url in file_urls:
            if url.endswith(".woff2") or url.endswith(".ttf"):
                # Join the base_url with the parent directory of css_url and the url from CSS
                url = urljoin(urljoin(base_url, os.path.dirname(css_url)), url.strip("'\""))
                # Include version number
                url = url.replace("releases/", f"releases/v6.5.1/")
                print("Constructed URL:", url)  # Add this line
                parsed_url = urlparse(url)
                filename = os.path.basename(parsed_url.path)
                directory_path = os.path.join(download_dir, "webfonts")
                os.makedirs(directory_path, exist_ok=True)
                download_file(url, directory_path)
    else:
        print(f"Failed to fetch CSS. Status code: {response.status_code}")


def main():
    base_url = "https://site-assets.fontawesome.com/releases/v6.5.1/"
    css_url = urljoin(base_url, "css/all.css")
    download_dir = "./releases/v6.5.1/"

    # Create the directory structure
    os.makedirs(os.path.join(download_dir, "css"), exist_ok=True)

    css_filename = "all.css"
    css_filepath = os.path.join(download_dir, "css", css_filename)

    # Download and save the CSS file
    with open(css_filepath, "wb") as f:
        f.write(requests.get(css_url).content)
    print(f"Downloaded CSS file to {css_filepath}")

    # Fetch and download the .woff2 files
    fetch_and_download_files(css_url, base_url, download_dir)

if __name__ == "__main__":
    main()