import time import logging from typing import Tuple, Dict, List, Optional from frame_finder import FrameFinder from frame import FrameFormat from queue import Queue import os from datetime import datetime import csv import numpy as np class FrameProcessor: """帧处理器 - 处理帧数据并统计信息""" # 缓冲区大小设置为64KB，可以根据实际情况调整 BUFFER_SIZE = 64 * 1024 # 64KB def __init__(self, num_channels: int = 8): """ 初始化帧处理器 Args: num_channels: 通道数量，默认为8 """ self.finder = FrameFinder() self.frame_count = 0 self.valid_frames = 0 self.corrupt_frames = 0 self.invalid_bytes = 0 self.start_time = None self.total_bytes_processed = 0 self.buffer = bytearray() # 用于存储未处理完的数据 self.frame_queue = Queue() # 用于存储发现的帧 self.dump_start_time = None # 用于记录导出开始时间 self.num_channels = num_channels # 保存通道数量 def process_buffer(self, data: bytes) -> Tuple[int, int, int, int]: """ 处理数据缓冲区 Args: data: 要处理的数据 Returns: Tuple[int, int, int, int]: (总帧数, 有效帧数, 损坏帧数, 无效字节数) """ if self.start_time is None: self.start_time = time.time() self.total_bytes_processed += len(data) # 将新数据添加到缓冲区 self.buffer.extend(data) # 如果缓冲区过大，可以在这里添加清理逻辑 # if len(self.buffer) > self.BUFFER_SIZE * 2: # self.buffer = self.buffer[-self.BUFFER_SIZE:] while len(self.buffer) > 0: frame_data, next_pos = self.finder.find_frame_in_buffer(self.buffer) if frame_data: self.frame_count += 1 # 使用当前通道数创建 FrameFormat 实例 frame_format = FrameFormat(self.num_channels) if frame_format.parse_frame(frame_data): self.valid_frames += 1 # 将有效帧加入队列 self.frame_queue.put(frame_data) else: self.corrupt_frames += 1 # 更新缓冲区 self.buffer = self.buffer[next_pos:] else: # 如果没有找到帧，next_pos 表示需要保留的数据起始位置 if next_pos == len(self.buffer): break elif next_pos == 0: break else: # 移除无效数据 self.invalid_bytes += next_pos self.buffer = self.buffer[next_pos:] return self.frame_count, self.valid_frames, self.corrupt_frames, self.invalid_bytes def get_frame(self, timeout: float = None) -> bytes: """ 从队列中获取一个帧 Args: timeout: 超时时间（秒），None表示永久等待 Returns: bytes: 帧数据，如果超时则返回None """ try: return self.frame_queue.get(timeout=timeout) except: return None def get_statistics(self) -> Dict: """ 获取处理统计信息 Returns: Dict: 统计信息 """ if self.start_time is None: runtime = 0 else: runtime = time.time() - self.start_time stats = { "total_bytes": self.total_bytes_processed, "total_frames": self.frame_count, "valid_frames": self.valid_frames, "corrupt_frames": self.corrupt_frames, "invalid_bytes": self.invalid_bytes, "runtime": runtime, "bytes_per_second": self.total_bytes_processed/runtime if runtime > 0 else 0, "frames_per_second": self.frame_count/runtime if runtime > 0 else 0, "queue_size": self.frame_queue.qsize() # 添加队列大小信息 } return stats def _parse_frame_data(self, frame_data: bytes) -> Optional[Dict]: """解析帧数据""" try: # 使用当前通道数创建 FrameFormat 实例 frame_format = FrameFormat(self.num_channels) frame = frame_format.parse_frame(frame_data) if not frame: return None # 将通道数据转换为16进制字符串 channels = [] for ch in frame.channels: ch_bytes = ch.to_bytes(4, byteorder='big', signed=True) ch_hex = ch_bytes.hex() channels.append(ch_hex) return { 'index': frame.frame_idx, # 添加帧索引 'timestamp': frame.timestamp, 'channels': channels } except Exception as e: self.logger.error(f"解析帧数据失败: {e}") return None def dump_frames(self, output_dir: str = 'dumps', format: str = 'dat', max_frames: int = None, include_raw: bool = False) -> Dict: """ 将有效帧导出到文件 Args: output_dir: 输出目录 format: 输出格式 ('dat' 或 'npy') max_frames: 最大导出帧数，None表示导出所有帧 include_raw: 是否包含原始数据 Returns: Dict: 导出统计信息 """ # 创建输出目录 if not os.path.exists(output_dir): os.makedirs(output_dir) # 生成输出文件名 timestamp = datetime.now().strftime('%Y%m%d_%H%M%S') if format == 'dat': filename = f'frames_{timestamp}.dat' else: filename = f'frames_{timestamp}.npy' filepath = os.path.join(output_dir, filename) # 初始化统计信息 stats = { 'total_frames': 0, 'exported_frames': 0, 'start_time': None, 'end_time': None, 'file_size': 0, 'format': format } try: self.dump_start_time = time.time() stats['start_time'] = self.dump_start_time if format == 'dat': # DAT格式导出 - 每行一帧数据 with open(filepath, 'w') as f: frame_idx = 0 # 初始化帧索引 while not self.frame_queue.empty(): if max_frames and stats['exported_frames'] >= max_frames: break frame_data = self.get_frame() if not frame_data: break parsed_data = self._parse_frame_data(frame_data) if not parsed_data: continue # 格式化数据行：索引 时间戳 通道1 通道2 ... 通道8 line = f"{parsed_data['index']} {parsed_data['timestamp']:.6f}" for ch in parsed_data['channels']: line += f" {ch}" # 使用16进制字符串 if include_raw: line += f" {parsed_data['raw_data']}" f.write(line + '\n') stats['exported_frames'] += 1 frame_idx += 1 # 更新帧索引 else: # NPY格式导出 frames_data = [] frame_idx = 0 # 初始化帧索引 while not self.frame_queue.empty(): if max_frames and stats['exported_frames'] >= max_frames: break frame_data = self.get_frame() if not frame_data: break parsed_data = self._parse_frame_data(frame_data) if not parsed_data: continue # 准备帧数据 frame_array = np.array([ parsed_data['index'], parsed_data['timestamp'], *[int.from_bytes(bytes.fromhex(ch), byteorder='big', signed=True) for ch in parsed_data['channels']] # 将16进制字符串转回有符号整数 ]) frames_data.append(frame_array) stats['exported_frames'] += 1 frame_idx += 1 # 更新帧索引 if frames_data: frames_array = np.array(frames_data) np.save(filepath, frames_array) # 更新统计信息 stats['end_time'] = time.time() stats['total_frames'] = self.valid_frames stats['file_size'] = os.path.getsize(filepath) return stats except Exception as e: logging.error(f"导出数据时出错: {e}") raise