使用验证码处理进行新闻和媒体聚合

新闻媒体和媒体平台使用验证码来保护内容免遭自动聚合。媒体监察员、公关公司和研究组织需要以编程方式收集文章。 CaptchaAI 处理跨新闻来源的验证码挑战。

新闻平台验证码

新闻聚合器

import requests
import time
import re
from bs4 import BeautifulSoup
from datetime import datetime
import json

CAPTCHAAI_KEY = "YOUR_API_KEY"
CAPTCHAAI_URL = "https://ocr.captchaai.com"


def solve_captcha(method, sitekey, pageurl, **kwargs):
    data = {
        "key": CAPTCHAAI_KEY, "method": method,
        "googlekey": sitekey, "pageurl": pageurl, "json": 1,
    }
    data.update(kwargs)
    resp = requests.post(f"{CAPTCHAAI_URL}/in.php", data=data)
    task_id = resp.json()["request"]
    for _ in range(60):
        time.sleep(5)
        result = requests.get(f"{CAPTCHAAI_URL}/res.php", params={
            "key": CAPTCHAAI_KEY, "action": "get",
            "id": task_id, "json": 1,
        })
        r = result.json()
        if r["request"] != "CAPCHA_NOT_READY":
            return r["request"]
    raise TimeoutError("Timeout")


class NewsAggregator:
    def __init__(self, proxy=None):
        self.session = requests.Session()
        if proxy:
            self.session.proxies = {"http": proxy, "https": proxy}
        self.session.headers.update({
            "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) "
            "AppleWebKit/537.36 Chrome/126.0.0.0 Safari/537.36",
            "Accept-Language": "en-US,en;q=0.9",
        })

    def collect_headlines(self, source_url, section=None):
        """Collect headlines from a news source."""
        url = f"{source_url}/{section}" if section else source_url
        resp = self.session.get(url, timeout=30)

        if self._has_captcha(resp.text):
            resp = self._solve_and_retry(resp.text, url)

        soup = BeautifulSoup(resp.text, "html.parser")
        articles = []

        for item in soup.select("article, .story, .headline-item, h2 a, h3 a"):
            link = item if item.name == "a" else item.select_one("a")
            if link:
                articles.append({
                    "title": link.get_text(strip=True),
                    "url": self._abs_url(source_url, link.get("href", "")),
                    "source": source_url,
                    "collected_at": datetime.now().isoformat(),
                })

        return articles

    def get_article(self, article_url):
        """Fetch full article content."""
        resp = self.session.get(article_url, timeout=30)

        if self._has_captcha(resp.text):
            resp = self._solve_and_retry(resp.text, article_url)

        soup = BeautifulSoup(resp.text, "html.parser")

        # Remove unwanted elements
        for tag in soup.select("script, style, nav, footer, .ad, .sidebar"):
            tag.decompose()

        content_el = soup.select_one(
            "article, .article-body, .story-body, .entry-content"
        )

        return {
            "url": article_url,
            "title": self._text(soup, "h1, .article-title"),
            "author": self._text(soup, ".author, .byline, [rel='author']"),
            "date": self._text(soup, "time, .publish-date, .article-date"),
            "content": content_el.get_text(separator="\n", strip=True) if content_el else "",
            "word_count": len(content_el.get_text().split()) if content_el else 0,
        }

    def aggregate_sources(self, sources, max_articles_per=20):
        """Aggregate headlines across multiple sources."""
        all_articles = []

        for source in sources:
            try:
                articles = self.collect_headlines(source["url"], source.get("section"))
                all_articles.extend(articles[:max_articles_per])
                print(f"{source['name']}: {len(articles)} headlines")
            except Exception as e:
                print(f"{source['name']}: Error - {e}")
            time.sleep(3)

        return all_articles

    def _has_captcha(self, html):
        return any(tag in html.lower() for tag in [
            'data-sitekey', 'g-recaptcha', 'cf-turnstile',
        ])

    def _solve_and_retry(self, html, url):
        match = re.search(r'data-sitekey="([^"]+)"', html)
        if not match:
            return self.session.get(url)
        sitekey = match.group(1)
        if 'cf-turnstile' in html:
            token = solve_captcha("turnstile", sitekey, url)
            return self.session.post(url, data={"cf-turnstile-response": token})
        token = solve_captcha("userrecaptcha", sitekey, url)
        return self.session.post(url, data={"g-recaptcha-response": token})

    def _text(self, soup, selector):
        el = soup.select_one(selector)
        return el.get_text(strip=True) if el else ""

    def _abs_url(self, base, href):
        if href.startswith("http"):
            return href
        return base.rstrip("/") + "/" + href.lstrip("/")


# Usage
aggregator = NewsAggregator(
    proxy="http://user:pass@residential.proxy.com:5000"
)

sources = [
    {"name": "Tech News A", "url": "https://technews-a.example.com", "section": "latest"},
    {"name": "Business B", "url": "https://business-b.example.com", "section": "tech"},
    {"name": "Industry C", "url": "https://industry-c.example.com"},
]

headlines = aggregator.aggregate_sources(sources)
print(f"Total: {len(headlines)} headlines collected")

基于关键词的新闻监控

class NewsMonitor:
    def __init__(self, keywords, sources, proxy=None):
        self.keywords = [kw.lower() for kw in keywords]
        self.aggregator = NewsAggregator(proxy=proxy)
        self.sources = sources
        self.seen_urls = set()

    def scan(self):
        """Scan for articles matching keywords."""
        headlines = self.aggregator.aggregate_sources(self.sources)
        matches = []

        for article in headlines:
            if article["url"] in self.seen_urls:
                continue

            title_lower = article["title"].lower()
            matched_kws = [kw for kw in self.keywords if kw in title_lower]

            if matched_kws:
                article["matched_keywords"] = matched_kws
                matches.append(article)
                self.seen_urls.add(article["url"])

        return matches

    def continuous_monitor(self, interval_min=30):
        """Run continuous monitoring with alerts."""
        while True:
            matches = self.scan()
            if matches:
                print(f"\n=== {len(matches)} new matches found ===")
                for m in matches:
                    print(f"  [{', '.join(m['matched_keywords'])}] {m['title']}")
                    print(f"    {m['url']}")
            else:
                print(f"No new matches at {datetime.now().strftime('%H:%M')}")

            time.sleep(interval_min * 60)


# Monitor for specific topics
monitor = NewsMonitor(
    keywords=["captcha", "bot detection", "web scraping", "automation"],
    sources=sources,
    proxy="http://user:pass@residential.proxy.com:5000",
)
matches = monitor.scan()

域预算计划

• 根据新鲜度需求、CAPTCHA 压力和内容流的价值为每个源分配爬网预算。
• 首先减慢敏感域的速度，而不是在所有发布商中应用相同的重试模式。
• 跟踪每个域的解决、错过和获取节奏，以便可以根据实际信号调整预算。

故障排除

常问问题

新闻聚合合法吗？

收集头条新闻和元数据用于研究或监测是常见的做法。未经许可复制完整受版权保护的文章是不可以的。使用片段并链接回原始来源。

如何处理付费内容？

检测付费墙（查找付费墙 CSS 类或有限的内容长度）并对其进行标记。仅访问您有权查看的内容。

哪种代理类型最适合新闻网站？

轮换自有服务器基础设施效果最好。主要新闻媒体使用 Cloudflare，它会积极阻止数据中心 IP。

相关指南

• 轮换自有服务器基础设施
• 社交媒体研究

来自任何来源的汇总新闻 -获取您的 CaptchaAI 密钥并自动化内容收集。