带有验证码处理的广告验证工作流程

广告验证需要访问数千个网页来检查广告投放、品牌安全和合规性。许多发布商网站使用验证码来阻止自动检查。 CaptchaAI 让您的验证管道保持运行。

广告验证检查哪些内容

执行

import requests
import time
import re
import json
import os
from datetime import datetime

API_KEY = os.environ["CAPTCHAAI_API_KEY"]


def solve_captcha(method, params):
    params["key"] = API_KEY
    params["method"] = method

    resp = requests.get("https://ocr.captchaai.com/in.php", params=params)
    if not resp.text.startswith("OK|"):
        raise Exception(resp.text)

    task_id = resp.text.split("|")[1]
    for _ in range(60):
        time.sleep(5)
        result = requests.get("https://ocr.captchaai.com/res.php", params={
            "key": API_KEY, "action": "get", "id": task_id,
        })
        if result.text == "CAPCHA_NOT_READY":
            continue
        if result.text.startswith("OK|"):
            return result.text.split("|", 1)[1]
        raise Exception(result.text)
    raise TimeoutError()


def verify_ad_placement(url, session):
    """Verify ad placement on a publisher page."""
    resp = session.get(url)

    # Solve CAPTCHA if present
    match = re.search(r'data-sitekey=["\']([A-Za-z0-9_-]+)["\']', resp.text)
    if match:
        token = solve_captcha("userrecaptcha", {
            "googlekey": match.group(1),
            "pageurl": url,
        })
        resp = session.post(url, data={"g-recaptcha-response": token})

    html = resp.text

    # Check for ad elements
    result = {
        "url": url,
        "timestamp": datetime.utcnow().isoformat(),
        "ads_found": [],
        "brand_safety": True,
        "captcha_solved": match is not None,
    }

    # Detect ad tags
    ad_patterns = [
        (r'googletag\.pubads', "Google Ad Manager"),
        (r'doubleclick\.net', "DFP/DoubleClick"),
        (r'ad\.doubleclick', "DoubleClick"),
        (r'amazon-adsystem', "Amazon Ads"),
        (r'criteo\.com/.*\.js', "Criteo"),
    ]

    for pattern, name in ad_patterns:
        if re.search(pattern, html):
            result["ads_found"].append(name)

    # Brand safety check — flag problematic content
    safety_keywords = [
        "violence", "hate speech", "explicit",
        "gambling", "illegal",
    ]
    page_text = re.sub(r'<[^>]+>', '', html).lower()
    for keyword in safety_keywords:
        if keyword in page_text:
            result["brand_safety"] = False
            break

    return result


def run_verification(urls, output_file="verification_report.json"):
    """Run ad verification across multiple publisher URLs."""
    session = requests.Session()
    session.headers["User-Agent"] = (
        "Mozilla/5.0 (Windows NT 10.0; Win64; x64) "
        "AppleWebKit/537.36 Chrome/120.0.0.0"
    )

    results = []
    for i, url in enumerate(urls):
        try:
            result = verify_ad_placement(url, session)
            results.append(result)
            ads = ", ".join(result["ads_found"]) or "None"
            safe = "SAFE" if result["brand_safety"] else "UNSAFE"
            print(f"  [{i+1}/{len(urls)}] {url}: {ads} [{safe}]")
        except Exception as e:
            results.append({
                "url": url,
                "error": str(e),
                "timestamp": datetime.utcnow().isoformat(),
            })
            print(f"  [{i+1}/{len(urls)}] {url}: ERROR - {e}")

        time.sleep(2)

    with open(output_file, "w") as f:
        json.dump(results, f, indent=2)

    # Summary
    total = len(results)
    safe = sum(1 for r in results if r.get("brand_safety"))
    captchas = sum(1 for r in results if r.get("captcha_solved"))
    errors = sum(1 for r in results if "error" in r)

    print(f"\n  Total: {total} | Safe: {safe} | CAPTCHAs solved: {captchas} | Errors: {errors}")

    return results


# Publisher URLs to verify
publisher_urls = [
    "https://publisher1.com/article/tech-news",
    "https://publisher2.com/sports/latest",
    "https://publisher3.com/finance/markets",
]

run_verification(publisher_urls)

使用受 Cloudflare 保护的发布商进行扩展

许多优质发布商都使用 Cloudflare。应对Turnstile和全面挑战：

def handle_cloudflare(url, session):
    """Handle Cloudflare-protected publisher pages."""
    resp = session.get(url)

    if "cf-turnstile" in resp.text:
        match = re.search(r'data-sitekey=["\']([^"\']+)', resp.text)
        if match:
            token = solve_captcha("turnstile", {
                "sitekey": match.group(1),
                "pageurl": url,
            })
            return session.post(url, data={
                "cf-turnstile-response": token,
            })

    if resp.status_code == 403 and "cf-browser-verification" in resp.text:
        data = solve_captcha("cloudflare_challenge", {
            "pageurl": url,
            "proxy": "user:pass@proxy:port",
            "proxytype": "HTTP",
        })
        # Parse qa_session_cookie and use same proxy
        return data

    return resp

常问问题

我每小时可以验证多少页？

使用 CaptchaAI，您每小时可以验证 200-500 页，具体取决于验证码频率和解决时间。

这适用于视频广告验证吗？

此方法适用于展示广告和原生广告。视频广告验证通常需要使用 Selenium 或 Playwright 进行浏览器渲染。

不同地区如何处理？

使用目标地区的代理。 CaptchaAI 支持代理参数，因此求解上下文与您的地理定位相匹配。

相关指南

• 抓取受保护的网站
• 代理设置最佳实践
• 无头浏览器验证码问题