python3 scripts/crawl_google_images.py "search keyword" --count 10

python3 scripts/download_images.py urls.txt --output ./images

import re
# Extract from a[href*="imgurl="] links
match = re.search(r'imgurl=([^&]+)', href)
original_url = match.group(1)

# Google sometimes embeds metadata in div.rg_meta
data = json.loads(div.text_content)
original_url = data['ou']  # original URL

from playwright.sync_api import sync_playwright
import re

def crawl_google_images(keyword, count=10):
    with sync_playwright() as p:
        browser = p.chromium.launch(headless=True)
        page = browser.new_page()
        
        # Navigate to Google Images
        page.goto(f"https://www.google.com/search?q={keyword}&tbm=isch")
        page.wait_for_timeout(3000)
        
        # Method 1: Extract from imgurl parameter
        links = page.eval_on_selector_all('a[href*="imgurl="]', 
            'els => els.map(e => e.href)')
        
        original_urls = []
        for link in links:
            match = re.search(r'imgurl=([^&]+)', link)
            if match:
                url = match.group(1)
                # URL decode
                url = url.replace('%3A', ':').replace('%2F', '/')
                if 'gstatic' not in url and 'google' not in url:
                    original_urls.append(url)
        
        browser.close()
        return original_urls[:count]

import requests

def download_image(url, output_path):
    headers = {
        "User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36"
    }
    r = requests.get(url, headers=headers, timeout=30)
    if r.status_code == 200:
        with open(output_path, "wb") as f:
            f.write(r.content)
        return len(r.content)
    return 0