NAC_Blockchain/nac-lens/src/retry.rs

//! NAC Lens重试机制和日志系统
//! 
//! 实现错误传播、重试机制和日志记录

use std::collections::VecDeque;
use std::sync::{Arc, Mutex};
use std::time::{Duration, Instant};
use serde::{Serialize, Deserialize};
use crate::error::{Nrpc4Error, Result};

/// 重试策略
#[derive(Debug, Clone, Copy, PartialEq, Eq, Serialize, Deserialize)]
pub enum RetryStrategy {
    /// 固定延迟
    FixedDelay,
    /// 指数退避
    ExponentialBackoff,
    /// 线性退避
    LinearBackoff,
}

/// 重试配置
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct RetryConfig {
    /// 最大重试次数
    pub max_retries: u32,
    /// 初始延迟（毫秒）
    pub initial_delay: u64,
    /// 最大延迟（毫秒）
    pub max_delay: u64,
    /// 重试策略
    pub strategy: RetryStrategy,
    /// 退避因子（用于指数/线性退避）
    pub backoff_factor: f64,
    /// 是否启用
    pub enabled: bool,
}

impl Default for RetryConfig {
    fn default() -> Self {
        Self {
            max_retries: 3,
            initial_delay: 1000,
            max_delay: 30000,
            strategy: RetryStrategy::ExponentialBackoff,
            backoff_factor: 2.0,
            enabled: true,
        }
    }
}

/// 重试状态
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct RetryState {
    /// 尝试次数
    pub attempt: u32,
    /// 下次重试延迟（毫秒）
    pub next_delay: u64,
    /// 最后错误
    pub last_error: Option<String>,
    /// 开始时间
    pub start_time: u64,
}

/// 重试管理器
#[derive(Debug)]
pub struct RetryManager {
    /// 配置
    config: RetryConfig,
    /// 重试状态映射
    states: Arc<Mutex<std::collections::HashMap<String, RetryState>>>,
}

impl RetryManager {
    /// 创建新的重试管理器
    pub fn new(config: RetryConfig) -> Self {
        Self {
            config,
            states: Arc::new(Mutex::new(std::collections::HashMap::new())),
        }
    }

    /// 开始重试
    pub fn start_retry(&self, operation_id: String) {
        if !self.config.enabled {
            return;
        }

        let state = RetryState {
            attempt: 0,
            next_delay: self.config.initial_delay,
            last_error: None,
            start_time: Self::current_timestamp(),
        };

        let mut states = self.states.lock().unwrap();
        states.insert(operation_id, state);
    }

    /// 记录失败
    pub fn record_failure(&self, operation_id: &str, error: String) -> bool {
        if !self.config.enabled {
            return false;
        }

        let mut states = self.states.lock().unwrap();
        
        if let Some(state) = states.get_mut(operation_id) {
            state.attempt += 1;
            state.last_error = Some(error);

            // 检查是否达到最大重试次数
            if state.attempt >= self.config.max_retries {
                return false;
            }

            // 计算下次延迟
            state.next_delay = self.calculate_delay(state.attempt);
            true
        } else {
            false
        }
    }

    /// 记录成功
    pub fn record_success(&self, operation_id: &str) {
        let mut states = self.states.lock().unwrap();
        states.remove(operation_id);
    }

    /// 获取重试状态
    pub fn get_state(&self, operation_id: &str) -> Option<RetryState> {
        let states = self.states.lock().unwrap();
        states.get(operation_id).cloned()
    }

    /// 计算延迟
    fn calculate_delay(&self, attempt: u32) -> u64 {
        let delay = match self.config.strategy {
            RetryStrategy::FixedDelay => self.config.initial_delay,
            RetryStrategy::ExponentialBackoff => {
                let delay = self.config.initial_delay as f64
                    * self.config.backoff_factor.powi(attempt as i32 - 1);
                delay as u64
            }
            RetryStrategy::LinearBackoff => {
                self.config.initial_delay + (attempt as u64 - 1) * 1000
            }
        };

        std::cmp::min(delay, self.config.max_delay)
    }

    /// 获取当前时间戳
    fn current_timestamp() -> u64 {
        std::time::SystemTime::now()
            .duration_since(std::time::UNIX_EPOCH)
            .unwrap()
            .as_secs()
    }
}

/// 日志级别
#[derive(Debug, Clone, Copy, PartialEq, Eq, PartialOrd, Ord, Serialize, Deserialize)]
pub enum LogLevel {
    /// 跟踪
    Trace,
    /// 调试
    Debug,
    /// 信息
    Info,
    /// 警告
    Warning,
    /// 错误
    Error,
    /// 致命
    Fatal,
}

/// 日志记录
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct LogRecord {
    /// 日志ID
    pub id: String,
    /// 级别
    pub level: LogLevel,
    /// 消息
    pub message: String,
    /// 模块
    pub module: String,
    /// 时间戳
    pub timestamp: u64,
    /// 额外数据
    pub data: Option<String>,
}

/// 日志配置
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct LogConfig {
    /// 最小日志级别
    pub min_level: LogLevel,
    /// 最大日志数
    pub max_logs: usize,
    /// 是否启用控制台输出
    pub console_output: bool,
    /// 是否启用文件输出
    pub file_output: bool,
    /// 日志文件路径
    pub file_path: Option<String>,
}

impl Default for LogConfig {
    fn default() -> Self {
        Self {
            min_level: LogLevel::Info,
            max_logs: 10000,
            console_output: true,
            file_output: false,
            file_path: None,
        }
    }
}

/// 日志记录器
#[derive(Debug)]
pub struct Logger {
    /// 配置
    config: LogConfig,
    /// 日志队列
    logs: Arc<Mutex<VecDeque<LogRecord>>>,
    /// 下一个日志ID
    next_log_id: Arc<Mutex<u64>>,
}

impl Logger {
    /// 创建新的日志记录器
    pub fn new(config: LogConfig) -> Self {
        Self {
            config,
            logs: Arc::new(Mutex::new(VecDeque::new())),
            next_log_id: Arc::new(Mutex::new(1)),
        }
    }

    /// 记录日志
    pub fn log(
        &self,
        level: LogLevel,
        module: String,
        message: String,
        data: Option<String>,
    ) -> String {
        // 检查日志级别
        if level < self.config.min_level {
            return String::new();
        }

        let mut next_id = self.next_log_id.lock().unwrap();
        let log_id = format!("LOG-{:08}", *next_id);
        *next_id += 1;
        drop(next_id);

        let record = LogRecord {
            id: log_id.clone(),
            level,
            message: message.clone(),
            module: module.clone(),
            timestamp: Self::current_timestamp(),
            data,
        };

        // 控制台输出
        if self.config.console_output {
            self.print_log(&record);
        }

        // 添加到队列
        let mut logs = self.logs.lock().unwrap();
        logs.push_back(record);

        // 限制日志数量
        if logs.len() > self.config.max_logs {
            logs.pop_front();
        }

        log_id
    }

    /// 打印日志
    fn print_log(&self, record: &LogRecord) {
        let level_str = match record.level {
            LogLevel::Trace => "TRACE",
            LogLevel::Debug => "DEBUG",
            LogLevel::Info => "INFO",
            LogLevel::Warning => "WARN",
            LogLevel::Error => "ERROR",
            LogLevel::Fatal => "FATAL",
        };

        println!(
            "[{}] [{}] [{}] {}",
            level_str, record.module, record.timestamp, record.message
        );
    }

    /// Trace日志
    pub fn trace(&self, module: String, message: String) {
        self.log(LogLevel::Trace, module, message, None);
    }

    /// Debug日志
    pub fn debug(&self, module: String, message: String) {
        self.log(LogLevel::Debug, module, message, None);
    }

    /// Info日志
    pub fn info(&self, module: String, message: String) {
        self.log(LogLevel::Info, module, message, None);
    }

    /// Warning日志
    pub fn warning(&self, module: String, message: String) {
        self.log(LogLevel::Warning, module, message, None);
    }

    /// Error日志
    pub fn error(&self, module: String, message: String) {
        self.log(LogLevel::Error, module, message, None);
    }

    /// Fatal日志
    pub fn fatal(&self, module: String, message: String) {
        self.log(LogLevel::Fatal, module, message, None);
    }

    /// 获取日志
    pub fn get_log(&self, log_id: &str) -> Option<LogRecord> {
        let logs = self.logs.lock().unwrap();
        logs.iter().find(|l| l.id == log_id).cloned()
    }

    /// 获取所有日志
    pub fn get_all_logs(&self) -> Vec<LogRecord> {
        let logs = self.logs.lock().unwrap();
        logs.iter().cloned().collect()
    }

    /// 按级别获取日志
    pub fn get_logs_by_level(&self, level: LogLevel) -> Vec<LogRecord> {
        let logs = self.logs.lock().unwrap();
        logs.iter()
            .filter(|l| l.level == level)
            .cloned()
            .collect()
    }

    /// 按模块获取日志
    pub fn get_logs_by_module(&self, module: &str) -> Vec<LogRecord> {
        let logs = self.logs.lock().unwrap();
        logs.iter()
            .filter(|l| l.module == module)
            .cloned()
            .collect()
    }

    /// 清空日志
    pub fn clear_logs(&self) {
        let mut logs = self.logs.lock().unwrap();
        logs.clear();
    }

    /// 获取日志数量
    pub fn get_log_count(&self) -> usize {
        let logs = self.logs.lock().unwrap();
        logs.len()
    }

    /// 获取当前时间戳
    fn current_timestamp() -> u64 {
        std::time::SystemTime::now()
            .duration_since(std::time::UNIX_EPOCH)
            .unwrap()
            .as_secs()
    }
}

/// 错误传播器
#[derive(Debug)]
pub struct ErrorPropagator {
    /// 日志记录器
    logger: Arc<Logger>,
    /// 重试管理器
    retry_manager: Arc<RetryManager>,
}

impl ErrorPropagator {
    /// 创建新的错误传播器
    pub fn new(logger: Arc<Logger>, retry_manager: Arc<RetryManager>) -> Self {
        Self {
            logger,
            retry_manager,
        }
    }

    /// 处理错误
    pub fn handle_error(
        &self,
        operation_id: &str,
        error: &Nrpc4Error,
        module: &str,
    ) -> bool {
        // 记录错误日志
        self.logger.error(
            module.to_string(),
            format!("Operation {} failed: {}", operation_id, error),
        );

        // 记录失败并检查是否应该重试
        let should_retry = self.retry_manager.record_failure(
            operation_id,
            error.to_string(),
        );

        if should_retry {
            if let Some(state) = self.retry_manager.get_state(operation_id) {
                self.logger.info(
                    module.to_string(),
                    format!(
                        "Retrying operation {} (attempt {}/{})",
                        operation_id,
                        state.attempt + 1,
                        self.retry_manager.config.max_retries
                    ),
                );
            }
        } else {
            self.logger.error(
                module.to_string(),
                format!("Operation {} failed after max retries", operation_id),
            );
        }

        should_retry
    }

    /// 处理成功
    pub fn handle_success(&self, operation_id: &str, module: &str) {
        self.retry_manager.record_success(operation_id);
        
        self.logger.info(
            module.to_string(),
            format!("Operation {} succeeded", operation_id),
        );
    }
}

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn test_retry_manager() {
        let config = RetryConfig::default();
        let manager = RetryManager::new(config);
        
        manager.start_retry("op1".to_string());
        
        let should_retry = manager.record_failure("op1", "Error 1".to_string());
        assert!(should_retry);
        
        let state = manager.get_state("op1").unwrap();
        assert_eq!(state.attempt, 1);
    }

    #[test]
    fn test_retry_max_attempts() {
        let mut config = RetryConfig::default();
        config.max_retries = 2;
        let manager = RetryManager::new(config);
        
        manager.start_retry("op1".to_string());
        
        assert!(manager.record_failure("op1", "Error 1".to_string()));
        assert!(!manager.record_failure("op1", "Error 2".to_string()));
    }

    #[test]
    fn test_exponential_backoff() {
        let config = RetryConfig {
            max_retries: 5,
            initial_delay: 1000,
            max_delay: 30000,
            strategy: RetryStrategy::ExponentialBackoff,
            backoff_factor: 2.0,
            enabled: true,
        };
        let manager = RetryManager::new(config);
        
        manager.start_retry("op1".to_string());
        manager.record_failure("op1", "Error 1".to_string());
        
        let state = manager.get_state("op1").unwrap();
        assert_eq!(state.next_delay, 1000);
        
        manager.record_failure("op1", "Error 2".to_string());
        let state = manager.get_state("op1").unwrap();
        assert_eq!(state.next_delay, 2000);
    }

    #[test]
    fn test_logger() {
        let config = LogConfig::default();
        let logger = Logger::new(config);
        
        logger.info("test".to_string(), "Test message".to_string());
        logger.error("test".to_string(), "Error message".to_string());
        
        assert_eq!(logger.get_log_count(), 2);
        
        let info_logs = logger.get_logs_by_level(LogLevel::Info);
        assert_eq!(info_logs.len(), 1);
        
        let test_logs = logger.get_logs_by_module("test");
        assert_eq!(test_logs.len(), 2);
    }

    #[test]
    fn test_logger_level_filter() {
        let mut config = LogConfig::default();
        config.min_level = LogLevel::Warning;
        let logger = Logger::new(config);
        
        logger.info("test".to_string(), "Info message".to_string());
        logger.warning("test".to_string(), "Warning message".to_string());
        logger.error("test".to_string(), "Error message".to_string());
        
        // Info级别的日志应该被过滤掉
        assert_eq!(logger.get_log_count(), 2);
    }

    #[test]
    fn test_error_propagator() {
        let log_config = LogConfig::default();
        let logger = Arc::new(Logger::new(log_config));
        
        let retry_config = RetryConfig::default();
        let retry_manager = Arc::new(RetryManager::new(retry_config));
        
        let propagator = ErrorPropagator::new(logger.clone(), retry_manager.clone());
        
        retry_manager.start_retry("op1".to_string());
        
        let error = Nrpc4Error::NetworkError("Connection failed".to_string());
        let should_retry = propagator.handle_error("op1", &error, "test");
        
        assert!(should_retry);
        assert!(logger.get_log_count() > 0);
    }
}