#!/usr/bin/env python3 # -*- coding: utf-8 -*- """ 视频处理整合工具 先进行视频切割，再转换为音频 """ import os import sys import argparse import subprocess from pathlib import Path import logging from typing import List, Tuple # 导入自定义模块 from video_splitter import get_video_files, split_video, batch_split_videos from video2audio import Video2AudioExtractor # 设置日志 logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s') logger = logging.getLogger(__name__) class VideoProcessor: """视频处理整合类""" def __init__(self, input_dir: str, output_dir: str, segment_duration: int = 30, audio_format: str = "wav"): """ 初始化视频处理器 Args: input_dir: 输入视频目录 output_dir: 输出目录 segment_duration: 每个片段的时长（秒） audio_format: 音频格式 """ self.input_dir = Path(input_dir) self.output_dir = Path(output_dir) self.segment_duration = segment_duration self.audio_format = audio_format # 创建主输出目录 self.output_dir.mkdir(parents=True, exist_ok=True) logger.info(f"视频处理器初始化完成") logger.info(f"输入目录: {self.input_dir}") logger.info(f"输出目录: {self.output_dir}") def check_dependencies(self) -> bool: """检查依赖是否可用""" try: # 检查ffmpeg subprocess.run(['ffmpeg', '-version'], capture_output=True, check=True) logger.info("✅ ffmpeg 可用") # 检查Python依赖 import decord import numpy import tqdm import soundfile logger.info("✅ Python依赖包可用") return True except (subprocess.CalledProcessError, FileNotFoundError): logger.error("❌ ffmpeg 未安装或不可用") logger.info("请安装 ffmpeg: sudo apt install ffmpeg") return False except ImportError as e: logger.error(f"❌ Python依赖包缺失: {e}") logger.info("请安装依赖: pip install -r requirements.txt") return False def process_videos(self) -> Tuple[int, int, int]: """ 处理视频：先切割，再提取音频 Returns: Tuple[int, int, int]: (成功处理的视频数, 切割的片段数, 提取的音频数) """ logger.info("=" * 60) logger.info("开始视频处理流程") logger.info("=" * 60) # 检查依赖 if not self.check_dependencies(): return 0, 0, 0 # 获取视频文件列表 logger.info("步骤1: 扫描视频文件") video_files = get_video_files(str(self.input_dir)) if not video_files: logger.warning("没有找到视频文件") return 0, 0, 0 logger.info(f"找到 {len(video_files)} 个视频文件") # 统计信息 success_videos = 0 total_segments = 0 total_audio_files = 0 # 处理每个视频 for i, video_file in enumerate(video_files, 1): logger.info(f"\n{'='*40}") logger.info(f"处理第 {i}/{len(video_files)} 个视频: {video_file.name}") logger.info(f"{'='*40}") try: # 步骤2: 切割视频 logger.info("步骤2: 切割视频") if self._split_single_video(video_file): success_videos += 1 # 步骤3: 提取音频 logger.info("步骤3: 提取音频") segments_count, audio_count = self._extract_audio_from_segments(video_file) total_segments += segments_count total_audio_files += audio_count else: logger.error(f"视频 {video_file.name} 处理失败") except Exception as e: logger.error(f"处理视频 {video_file.name} 时出错: {e}") # 输出统计信息 logger.info(f"\n{'='*60}") logger.info("处理完成!") logger.info(f"{'='*60}") logger.info(f"成功处理的视频数: {success_videos}/{len(video_files)}") logger.info(f"总切割片段数: {total_segments}") logger.info(f"总音频文件数: {total_audio_files}") logger.info(f"输出目录: {self.output_dir}") return success_videos, total_segments, total_audio_files def _split_single_video(self, video_file: Path) -> bool: """ 切割单个视频 Args: video_file: 视频文件路径 Returns: bool: 是否成功 """ try: # 为每个视频创建独立的输出目录结构 video_name = video_file.stem video_output_dir = self.output_dir / video_name video_split_dir = video_output_dir / "video_split" # 创建目录 video_split_dir.mkdir(parents=True, exist_ok=True) # 调用split_video，但传入video_split_dir作为输出目录 success = split_video(video_file, str(video_split_dir), self.segment_duration) # 如果视频不足30秒，split_video会将原视频复制到video_split_dir # 如果视频超过30秒，split_video会将片段保存到video_split_dir # 两种情况都符合我们的期望 return success except Exception as e: logger.error(f"切割视频失败: {e}") return False def _extract_audio_from_segments(self, original_video: Path) -> Tuple[int, int]: """ 从视频片段中提取音频 Args: original_video: 原始视频文件路径 Returns: Tuple[int, int]: (片段数, 音频文件数) """ video_name = original_video.stem video_output_dir = self.output_dir / video_name video_split_dir = video_output_dir / "video_split" audio_split_dir = video_output_dir / "audio_split" if not video_split_dir.exists(): logger.warning(f"视频片段目录不存在: {video_split_dir}") return 0, 0 # 获取所有视频片段 video_extensions = ['.mp4', '.avi', '.mov', '.mkv', '.flv', '.wmv', '.webm'] segment_files = [] for ext in video_extensions: segment_files.extend(video_split_dir.glob(f"*{ext}")) if not segment_files: logger.warning(f"没有找到视频片段: {video_split_dir}") return 0, 0 logger.info(f"找到 {len(segment_files)} 个视频片段") # 创建音频输出目录 audio_split_dir.mkdir(exist_ok=True) # 提取音频 audio_count = 0 for segment_file in segment_files: try: logger.info(f"提取音频: {segment_file.name}") # 创建音频提取器，输出到audio_split目录 segment_audio_extractor = Video2AudioExtractor(str(audio_split_dir)) video_out, audio_out = segment_audio_extractor.extract_audio_from_video( str(segment_file), self.audio_format ) if audio_out: audio_count += 1 logger.info(f"音频提取成功: {Path(audio_out).name}") else: logger.warning(f"音频提取失败: {segment_file.name}") except Exception as e: logger.error(f"提取音频时出错: {e}") logger.info(f"音频提取完成: {audio_count}/{len(segment_files)} 个片段") return len(segment_files), audio_count def get_output_structure(self) -> dict: """ 获取输出目录结构信息 Returns: dict: 目录结构信息 """ structure = { "output_dir": str(self.output_dir), "videos": [] } # 统计每个视频的目录结构 if self.output_dir.exists(): for video_dir in self.output_dir.iterdir(): if video_dir.is_dir(): video_info = { "name": video_dir.name, "video_split_count": 0, "audio_split_count": 0 } # 检查video_split目录 video_split_dir = video_dir / "video_split" if video_split_dir.exists(): segments = list(video_split_dir.glob("*.mp4")) video_info["video_split_count"] = len(segments) # 检查audio_split目录 audio_split_dir = video_dir / "audio_split" if audio_split_dir.exists(): audio_files = list(audio_split_dir.glob(f"*.{self.audio_format}")) video_info["audio_split_count"] = len(audio_files) structure["videos"].append(video_info) return structure def main(): """主函数""" parser = argparse.ArgumentParser(description="视频处理整合工具 - 先切割再提取音频") parser.add_argument("--input_dir", "-i", default="/root/autodl-tmp/video", help="输入视频目录 (默认: /root/autodl-tmp/video)") parser.add_argument("--output_dir", "-o", default="/root/autodl-tmp/video_processed", help="输出目录 (默认: /root/autodl-tmp/video_processed)") parser.add_argument("--duration", "-d", type=int, default=30, help="每个片段的时长（秒），默认30秒") parser.add_argument("--audio_format", "-f", default="wav", choices=["wav", "flac", "ogg"], help="音频格式 (默认: wav)") parser.add_argument("--verbose", "-v", action="store_true", help="显示详细日志") parser.add_argument("--check_only", action="store_true", help="仅检查输出目录结构，不进行处理") args = parser.parse_args() if args.verbose: logging.getLogger().setLevel(logging.DEBUG) # 创建视频处理器 processor = VideoProcessor( input_dir=args.input_dir, output_dir=args.output_dir, segment_duration=args.duration, audio_format=args.audio_format ) if args.check_only: # 仅检查输出结构 structure = processor.get_output_structure() print("\n📁 输出目录结构:") print(f"主输出目录: {structure['output_dir']}") if structure['videos']: print(f"\n📹 视频处理统计:") for video in structure['videos']: print(f" {video['name']}:") print(f" 📹 video_split: {video['video_split_count']} 个片段") print(f" 🎵 audio_split: {video['audio_split_count']} 个音频文件") else: print("📁 输出目录为空，没有处理过的视频") else: # 执行完整的处理流程 success_videos, total_segments, total_audio = processor.process_videos() if success_videos > 0: print(f"\n✅ 处理完成!") print(f"成功处理: {success_videos} 个视频") print(f"切割片段: {total_segments} 个") print(f"音频文件: {total_audio} 个") print(f"输出目录: {args.output_dir}") else: print(f"\n❌ 处理失败，请检查输入目录和依赖") if __name__ == "__main__": main()