server.js

// server.js - AI News Backend
const express = require('express');
const cors = require('cors');
const axios = require('axios');
const fs = require('fs');
const path = require('path');
const { Buffer } = require('buffer');

const app = express();
app.use(cors());
app.use(express.json());

// AI 新闻搜索关键词
const AI_KEYWORDS = [
  'AI artificial intelligence',
  'machine learning',
  'deep learning',
  'OpenAI ChatGPT',
  'Google Gemini',
  'Anthropic Claude',
  'Llama LLM',
  'AI regulation'
];

// 缓存文件路径
const CACHE_FILE = path.join(__dirname, 'news_cache.json');
const CACHE_EXPIRY = 24 * 60 * 60 * 1000; // 24小时

// 获取新闻缓存
function getNewsCache() {
  try {
    if (fs.existsSync(CACHE_FILE)) {
      const data = JSON.parse(fs.readFileSync(CACHE_FILE, 'utf8'));
      if (Date.now() - data.timestamp < CACHE_EXPIRY) {
        return data.news;
      }
    }
  } catch (e) {
    console.error('Cache read error:', e);
  }
  return null;
}

// 保存新闻缓存
function saveNewsCache(news) {
  try {
    fs.writeFileSync(CACHE_FILE, JSON.stringify({
      timestamp: Date.now(),
      news: news
    }));
  } catch (e) {
    console.error('Cache write error:', e);
  }
}

// 搜索 AI 新闻 (使用 RSS 源)
async function searchAINews() {
  const news = [];
  const rssUrls = [
    'https://feeds.feedburner.com/TechCrunch/',
    'https://www.theverge.com/rss/index.xml',
    'https://wired.com/feed/rss'
  ];

  for (const url of rssUrls) {
    try {
      const response = await axios.get(url, {
        timeout: 10000,
        headers: {
          'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36'
        }
      });
      
      // 简单解析 RSS - 更强的 description 提取
      const items = response.data.match(/<item[^>]*>[\s\S]*?<\/item>/gi) || [];
      
      for (const item of items.slice(0, 5)) {
        const title = item.match(/<title[^>]*>([^<]+)<\/title>/i)?.[1] || '';
        const link = item.match(/<link[^>]*>([^<]+)<\/link>/i)?.[1] || '';
        // 匹配 description，支持 CDATA 和多行内容
        const descMatch = item.match(/<description[^>]*>([\s\S]*?)<\/description>/i);
        let desc = descMatch ? descMatch[1] : '';
        // 移除 CDATA 包装
        desc = desc.replace(/<!\[CDATA\[([\s\S]*?)\]\]>/i, '$1');
        // 清理并处理描述
        desc = cleanText(desc).slice(0, 500);
        
        // 检查是否与 AI 相关
        const isAIRelated = AI_KEYWORDS.some(kw => 
          title.toLowerCase().includes(kw.toLowerCase()) ||
          desc.toLowerCase().includes(kw.toLowerCase())
        );
        
        if (isAIRelated || news.length < 10) {
          // 去重
          if (!news.find(n => n.title === cleanText(title))) {
            news.push({
              id: Buffer.from(cleanText(title)).toString('base64').slice(0, 16),
              title: cleanText(title).slice(0, 100),
              description: cleanText(desc).slice(0, 200),
              summary: generateSummary(cleanText(desc)),
              content: cleanText(desc),
              link: link,
              source: new URL(url).hostname,
              time: new Date().toLocaleDateString('zh-CN'),
              image: 'https://picsum.photos/300/200?random=' + Math.random()
            });
          }
        }
      }
    } catch (e) {
      console.error(`Failed to fetch ${url}:`, e.message);
    }
  }

  return news.slice(0, 15);
}

// 生成摘要
function generateSummary(text) {
  if (!text) return '暂无摘要';
  // 简单摘要：取前200字
  return text.slice(0, 200) + '...';
}

// 清理文本：移除HTML标签并解码HTML实体
function cleanText(text) {
  if (!text) return '';
  return text
    .replace(/<[^>]+>/g, '')  // 移除HTML标签
    .replace(/&nbsp;/g, ' ')
    .replace(/&amp;/g, '&')
    .replace(/&lt;/g, '<')
    .replace(/&gt;/g, '>')
    .replace(/&quot;/g, '"')
    .replace(/&#39;/g, "'")
    .replace(/&#(\d+);/g, (_, num) => String.fromCharCode(num))
    .replace(/&#x([0-9a-fA-F]+);/g, (_, hex) => String.fromCharCode(parseInt(hex, 16)))
    .trim();
}

// 解码 HTML 实体（不删除标签，用于需要保留格式的地方）
function decodeHTML(html) {
  if (!html) return '';
  return html
    .replace(/&nbsp;/g, ' ')
    .replace(/&amp;/g, '&')
    .replace(/&lt;/g, '<')
    .replace(/&gt;/g, '>')
    .replace(/&quot;/g, '"')
    .replace(/&#39;/g, "'")
    .replace(/&#(\d+);/g, (_, num) => String.fromCharCode(num))
    .replace(/&#x([0-9a-fA-F]+);/g, (_, hex) => String.fromCharCode(parseInt(hex, 16)));
}

// 生成每日摘要
function generateDailySummary(news) {
  if (!news || news.length === 0) return '暂无资讯';
  
  const summary = `今日共收集 ${news.length} 条 AI 资讯。`;
  const topNews = news.slice(0, 3).map((n, i) => `${i + 1}. ${cleanText(n.title)}`).join('；');
  
  return summary + '主要动态：' + topNews;
}

// API 路由
app.get('/api/news', async (req, res) => {
  try {
    // 检查缓存
    let news = getNewsCache();
    
    if (!news) {
      console.log('Fetching fresh news...');
      news = await searchAINews();
      saveNewsCache(news);
    }
    
    const dailySummary = generateDailySummary(news);
    
    res.json({
      success: true,
      data: news,
      summary: dailySummary,
      updateTime: new Date().toISOString()
    });
  } catch (e) {
    console.error('News API error:', e);
    res.json({
      success: false,
      message: e.message
    });
  }
});

// 手动刷新新闻
app.post('/api/refresh', async (req, res) => {
  try {
    const news = await searchAINews();
    saveNewsCache(news);
    
    res.json({
      success: true,
      message: '刷新成功',
      count: news.length
    });
  } catch (e) {
    res.json({
      success: false,
      message: e.message
    });
  }
});

const PORT = process.env.PORT || 3000;
app.listen(PORT, () => {
  console.log(`AI News Server running on port ${PORT}`);
});