feat: 添加 Cross 模式统计报告功能

2026-03-09 16:53:24 +08:00
parent 3850bc33dc
commit 8e101491d2
3 changed files with 433 additions and 0 deletions
@@ -0,0 +1,280 @@
+"""
+Cross 模式统计报告功能
+
+提供事件记录和统计报告生成功能，支持 Cross 和 Normal 两种模式。
+"""
+
+import json
+import os
+import fcntl
+from datetime import datetime, timezone
+from pathlib import Path
+from typing import Optional
+
+# 统计数据文件路径
+STATS_FILE = Path(__file__).parent / ".stats_data.json"
+
+# 敏感字段列表（用于过滤）
+SENSITIVE_FIELDS = {"password", "token", "secret", "api_key", "apikey", "credential"}
+
+
+def _filter_sensitive_data(data: dict) -> dict:
+    """过滤敏感数据
+    
+    递归过滤字典中的敏感字段，将值替换为 [REDACTED]
+    """
+    if not isinstance(data, dict):
+        return data
+    
+    filtered = {}
+    for key, value in data.items():
+        key_lower = key.lower().replace("-", "_").replace(" ", "_")
+        
+        # 检查是否为敏感字段
+        is_sensitive = any(s in key_lower for s in SENSITIVE_FIELDS)
+        
+        if is_sensitive:
+            filtered[key] = "[REDACTED]"
+        elif isinstance(value, dict):
+            filtered[key] = _filter_sensitive_data(value)
+        else:
+            filtered[key] = value
+    
+    return filtered
+
+
+def _ensure_stats_file():
+    """确保统计数据文件存在，并设置正确权限"""
+    if not STATS_FILE.exists():
+        STATS_FILE.write_text(json.dumps({"events": []}, ensure_ascii=False, indent=2))
+        os.chmod(STATS_FILE, 0o600)
+
+
+def _load_stats() -> dict:
+    """加载统计数据（带文件锁）"""
+    _ensure_stats_file()
+    
+    with open(STATS_FILE, "r") as f:
+        fcntl.flock(f.fileno(), fcntl.LOCK_SH)  # 共享锁
+        try:
+            data = json.load(f)
+        finally:
+            fcntl.flock(f.fileno(), fcntl.LOCK_UN)
+    
+    return data
+
+
+def _save_stats(data: dict):
+    """保存统计数据（带文件锁）"""
+    _ensure_stats_file()
+    
+    with open(STATS_FILE, "w") as f:
+        fcntl.flock(f.fileno(), fcntl.LOCK_EX)  # 排他锁
+        try:
+            json.dump(data, f, ensure_ascii=False, indent=2)
+        finally:
+            fcntl.flock(f.fileno(), fcntl.LOCK_UN)
+    
+    # 确保文件权限为 600
+    os.chmod(STATS_FILE, 0o600)
+
+
+def record_event(
+    event_type: str,
+    mode: str,
+    metadata: Optional[dict] = None,
+    status: str = "success"
+) -> bool:
+    """记录一个事件
+    
+    Args:
+        event_type: 事件类型 (debate, review, audit)
+        mode: 运行模式 (cross, normal)
+        metadata: 事件元数据（可选）
+        status: 事件状态 (success, failed)
+    
+    Returns:
+        bool: 记录是否成功
+    """
+    # 验证参数
+    valid_types = {"debate", "review", "audit"}
+    valid_modes = {"cross", "normal"}
+    valid_status = {"success", "failed"}
+    
+    if event_type not in valid_types:
+        print(f"错误: 无效的事件类型 '{event_type}'，有效值: {valid_types}")
+        return False
+    
+    if mode not in valid_modes:
+        print(f"错误: 无效的模式 '{mode}'，有效值: {valid_modes}")
+        return False
+    
+    if status not in valid_status:
+        print(f"错误: 无效的状态 '{status}'，有效值: {valid_status}")
+        return False
+    
+    # 过滤敏感数据
+    if metadata:
+        metadata = _filter_sensitive_data(metadata)
+    else:
+        metadata = {}
+    
+    # 创建事件记录
+    event = {
+        "timestamp": datetime.now(timezone.utc).isoformat(),
+        "event_type": event_type,
+        "mode": mode,
+        "status": status,
+        "metadata": metadata
+    }
+    
+    try:
+        data = _load_stats()
+        data["events"].append(event)
+        _save_stats(data)
+        return True
+    except Exception as e:
+        print(f"记录事件失败: {e}")
+        return False
+
+
+def get_stats() -> dict:
+    """获取统计汇总数据
+    
+    Returns:
+        dict: 统计汇总数据
+    """
+    data = _load_stats()
+    events = data.get("events", [])
+    
+    # 初始化统计结构
+    stats = {
+        "total_events": len(events),
+        "by_type": {
+            "debate": {"total": 0, "success": 0, "failed": 0},
+            "review": {"total": 0, "success": 0, "failed": 0},
+            "audit": {"total": 0, "success": 0, "failed": 0}
+        },
+        "by_mode": {
+            "cross": {"total": 0, "success": 0, "failed": 0},
+            "normal": {"total": 0, "success": 0, "failed": 0}
+        },
+        "cross_mode_breakdown": {
+            "debate": {"total": 0, "success": 0, "failed": 0},
+            "review": {"total": 0, "success": 0, "failed": 0},
+            "audit": {"total": 0, "success": 0, "failed": 0}
+        },
+        "normal_mode_breakdown": {
+            "debate": {"total": 0, "success": 0, "failed": 0},
+            "review": {"total": 0, "success": 0, "failed": 0},
+            "audit": {"total": 0, "success": 0, "failed": 0}
+        },
+        "success_rate": 0.0,
+        "first_event": None,
+        "last_event": None
+    }
+    
+    if not events:
+        return stats
+    
+    # 统计各类数据
+    for event in events:
+        event_type = event["event_type"]
+        mode = event["mode"]
+        status = event["status"]
+        
+        # 按类型统计
+        stats["by_type"][event_type]["total"] += 1
+        stats["by_type"][event_type][status] += 1
+        
+        # 按模式统计
+        stats["by_mode"][mode]["total"] += 1
+        stats["by_mode"][mode][status] += 1
+        
+        # 按模式+类型交叉统计
+        breakdown_key = f"{mode}_mode_breakdown"
+        stats[breakdown_key][event_type]["total"] += 1
+        stats[breakdown_key][event_type][status] += 1
+    
+    # 计算成功率
+    total = stats["total_events"]
+    success_count = sum(t["success"] for t in stats["by_type"].values())
+    stats["success_rate"] = round(success_count / total * 100, 2) if total > 0 else 0.0
+    
+    # 时间范围
+    stats["first_event"] = events[0]["timestamp"]
+    stats["last_event"] = events[-1]["timestamp"]
+    
+    return stats
+
+
+def format_stats_report(stats: dict) -> str:
+    """格式化统计报告为可读字符串
+    
+    Args:
+        stats: get_stats() 返回的统计数据
+    
+    Returns:
+        str: 格式化的统计报告
+    """
+    lines = []
+    lines.append("=" * 50)
+    lines.append("       Cross 模式统计报告")
+    lines.append("=" * 50)
+    
+    # 总览
+    lines.append(f"\n📊 总览")
+    lines.append(f"  总事件数: {stats['total_events']}")
+    lines.append(f"  成功率: {stats['success_rate']}%")
+    if stats['first_event']:
+        lines.append(f"  时间范围: {stats['first_event']} ~ {stats['last_event']}")
+    
+    # 按模式统计
+    lines.append(f"\n📈 按模式统计")
+    for mode, data in stats["by_mode"].items():
+        mode_name = "Cross 模式" if mode == "cross" else "Normal 模式"
+        rate = round(data["success"] / data["total"] * 100, 2) if data["total"] > 0 else 0
+        lines.append(f"  {mode_name}:")
+        lines.append(f"    总数: {data['total']} | 成功: {data['success']} | 失败: {data['failed']} | 成功率: {rate}%")
+    
+    # 按事件类型统计
+    lines.append(f"\n📋 按事件类型统计")
+    for event_type, data in stats["by_type"].items():
+        type_name = {"debate": "辩论", "review": "评审", "audit": "审计"}.get(event_type, event_type)
+        rate = round(data["success"] / data["total"] * 100, 2) if data["total"] > 0 else 0
+        lines.append(f"  {type_name}:")
+        lines.append(f"    总数: {data['total']} | 成功: {data['success']} | 失败: {data['failed']} | 成功率: {rate}%")
+    
+    # Cross 模式详细分解
+    lines.append(f"\n🔍 Cross 模式详细分解")
+    for event_type, data in stats["cross_mode_breakdown"].items():
+        type_name = {"debate": "辩论", "review": "评审", "audit": "审计"}.get(event_type, event_type)
+        rate = round(data["success"] / data["total"] * 100, 2) if data["total"] > 0 else 0
+        lines.append(f"  {type_name}:")
+        lines.append(f"    总数: {data['total']} | 成功: {data['success']} | 失败: {data['failed']} | 成功率: {rate}%")
+    
+    # Normal 模式详细分解
+    lines.append(f"\n🔍 Normal 模式详细分解")
+    for event_type, data in stats["normal_mode_breakdown"].items():
+        type_name = {"debate": "辩论", "review": "评审", "audit": "审计"}.get(event_type, event_type)
+        rate = round(data["success"] / data["total"] * 100, 2) if data["total"] > 0 else 0
+        lines.append(f"  {type_name}:")
+        lines.append(f"    总数: {data['total']} | 成功: {data['success']} | 失败: {data['failed']} | 成功率: {rate}%")
+    
+    lines.append("\n" + "=" * 50)
+    
+    return "\n".join(lines)
+
+
+if __name__ == "__main__":
+    # 测试代码
+    print("测试 stats.py 模块...")
+    
+    # 记录一些测试事件
+    record_event("debate", "cross", {"topic": "AI 安全"}, "success")
+    record_event("review", "normal", {"pr_id": 123}, "success")
+    record_event("audit", "cross", {"password": "secret123"}, "failed")  # 敏感数据会被过滤
+    
+    # 获取并打印统计
+    stats = get_stats()
+    print(format_stats_report(stats))