苹果用"铬合金人"讽刺 Chrome：浏览器隐私战从广告打到代码层

"""
audit_trackers.py — 审计网页中的第三方追踪脚本与 Cookie
用法: python audit_trackers.py <url>
依赖: pip install requests beautifulsoup4
"""

import sys
import re
from urllib.parse import urlparse
from collections import Counter

import requests
from bs4 import BeautifulSoup

# 已知追踪/广告域名片段（可自行扩展）
TRACKER_KEYWORDS = [
    "doubleclick", "googlesyndication", "google-analytics",
    "googletagmanager", "facebook", "fbcdn", "meta",
    "adservice", "adnxs", "amazon-adsystem", "criteo",
    "taboola", "outbrain", "hotjar", "clarity",
    "segment", "amplitude", "mixpanel", "quantserve",
    "scorecardresearch", "moatads", "rubiconproject",
]

def is_tracker(domain: str) -> bool:
    return any(kw in domain.lower() for kw in TRACKER_KEYWORDS)

def audit(url: str):
    # 发起请求，保留 Cookie
    session = requests.Session()
    resp = session.get(url, timeout=15, headers={
        "User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) "
                       "AppleWebKit/605.1.15 (KHTML, like Gecko) "
                       "Version/18.1 Safari/605.1.15"
    })

    page_domain = urlparse(url).netloc
    soup = BeautifulSoup(resp.text, "html.parser")

    # 收集所有 <script src=...> 的域名
    script_domains = []
    for tag in soup.find_all("script", src=True):
        d = urlparse(tag["src"]).netloc
        if d and d != page_domain:
            script_domains.append(d)

    # 收集 <img> 像素追踪（1x1 或无尺寸的第三方图片）
    pixel_domains = []
    for tag in soup.find_all("img", src=True):
        d = urlparse(tag["src"]).netloc
        w = tag.get("width", "")
        h = tag.get("height", "")
        if d and d != page_domain and (str(w) in ("1", "") or str(h) in ("1", "")):
            pixel_domains.append(d)

    # 收集 <link rel=preconnect> 的域名（预连接暗示频繁第三方请求）
    preconnect_domains = []
    for tag in soup.find_all("link", rel="preconnect"):
        href = tag.get("href", "")
        if href:
            preconnect_domains.append(urlparse(href).netloc)

    # 分析 Cookie
    cookies = session.cookies.get_dict()
    third_party_cookies = {
        k: v for k, v in cookies.items()
        if any(is_tracker(d) for d in script_domains)
    }

    # 输出报告
    tracker_scripts = [d for d in script_domains if is_tracker(d)]
    non_tracker_3rd = [d for d in script_domains if not is_tracker(d) and d != page_domain]

    print(f"=== 追踪审计报告: {url} ===")
    print(f"页面域名: {page_domain}")
    print(f"第三方脚本总数: {len(script_domains)}")
    print(f"  已识别追踪脚本: {len(tracker_scripts)}")
    for d, cnt in Counter(tracker_scripts).most_common():
        print(f"    - {d} (出现 {cnt} 次)")
    print(f"  其他第三方脚本: {len(non_tracker_3rd)}")
    for d, cnt in Counter(non_tracker_3rd).most_common():
        print(f"    - {d} (出现 {cnt} 次)")
    print(f"像素追踪请求: {len(pixel_domains)}")
    for d in pixel_domains:
        print(f"    - {d}")
    print(f"预连接域名: {len(preconnect_domains)}")
    for d in preconnect_domains:
        marker = " ⚠️ 追踪域" if is_tracker(d) else ""
        print(f"    - {d}{marker}")
    print(f"会话 Cookie 总数: {len(cookies)}")
    print(f"疑似追踪相关 Cookie: {len(third_party_cookies)}")
    for k, v in third_party_cookies.items():
        print(f"    - {k} = {v[:40]}...")

if __name__ == "__main__":
    if len(sys.argv) < 2:
        print("用法: python audit_trackers.py <url>")
        sys.exit(1)
    audit(sys.argv[1])

pip install requests beautifulsoup4
python audit_trackers.py https://www.example-news-site.com