Rename to hkt.sh

projects/news-bot/summarizer.py

@@ -0,0 +1,97 @@
+"""
+新闻总结生成模块
+使用规则引擎生成中文总结（不调用外部 AI API）
+"""
+import logging
+from datetime import datetime, timedelta
+from collections import defaultdict
+
+logger = logging.getLogger(__name__)
+
+
+def generate_summary(news_list: list, period: str = "全天") -> str:
+    """
+    生成新闻总结
+    period: "最近1小时" / "上午" / "全天" / "昨晚到今早"
+    """
+    if not news_list:
+        return f"📭 {period}暂无重要新闻"
+
+    # 按评分排序，高分在前
+    sorted_news = sorted(news_list, key=lambda x: x.get("score", 0), reverse=True)
+
+    # 按来源分组统计
+    source_count = defaultdict(int)
+    for n in sorted_news:
+        source_count[n.get("source_name", "未知")] += 1
+
+    # 构建总结
+    lines = []
+    lines.append(f"📊 *{period}新闻总结*")
+    lines.append(f"━━━━━━━━━━━━━━━")
+    lines.append(f"📰 共 {len(sorted_news)} 条快讯")
+    lines.append("")
+
+    # 🔴 重磅新闻（评分 >= 8）
+    hot_news = [n for n in sorted_news if n.get("score", 0) >= 8]
+    if hot_news:
+        lines.append("🔴 *重磅快讯*")
+        for n in hot_news[:5]:
+            emoji = _score_emoji(n.get("score", 0))
+            lines.append(f"  {emoji} {n['title']}")
+        lines.append("")
+
+    # 🟡 值得关注（评分 6-7）
+    mid_news = [n for n in sorted_news if 6 <= n.get("score", 0) <= 7]
+    if mid_news:
+        lines.append("🟡 *值得关注*")
+        for n in mid_news[:8]:
+            lines.append(f"  • {n['title']}")
+        lines.append("")
+
+    # 来源统计
+    lines.append("📡 *来源分布*")
+    for src, cnt in sorted(source_count.items(), key=lambda x: x[1], reverse=True):
+        lines.append(f"  {src}: {cnt}条")
+
+    lines.append("")
+    lines.append(f"⏰ 生成时间: {datetime.now().strftime('%H:%M')}")
+
+    return "\n".join(lines)
+
+
+def _score_emoji(score: int) -> str:
+    """根据评分返回 emoji"""
+    if score >= 9:
+        return "🔥"
+    elif score >= 8:
+        return "⚡"
+    elif score >= 7:
+        return "📌"
+    elif score >= 6:
+        return "📎"
+    else:
+        return "•"
+
+
+def get_period_range(period: str) -> tuple:
+    """
+    根据时间段名称返回 (start_ts, end_ts)
+    """
+    now = datetime.now()
+    today_start = now.replace(hour=0, minute=0, second=0, microsecond=0)
+
+    if period == "最近1小时":
+        start = now - timedelta(hours=1)
+    elif period == "上午":
+        start = today_start.replace(hour=6)
+        now = min(now, today_start.replace(hour=12))
+    elif period == "全天":
+        start = today_start
+    elif period == "昨晚到今早":
+        start = (today_start - timedelta(days=1)).replace(hour=20)
+        now = today_start.replace(hour=8)
+    else:
+        start = now - timedelta(hours=3)
+
+    return int(start.timestamp()), int(now.timestamp())