VideoX-Fun/validation_samples/rename_missing.py

import os
import sys
import shutil

# --- 脚本正文 (通常不需要修改) ---

def sanitize_filename(name):
    """
    移除或替换文件名中非法的字符。
    您可以根据需要添加更多非法字符。
    """
    # 移除首尾空白字符
    name = name.strip()
    
    # 替换 Windows 和 Unix 系统中的路径分隔符
    name = name.replace('/', '_').replace('\\', '_')
    # 替换其他一些在文件名中可能有问题的字符
    name = name.replace(':', '-').replace('?', '').replace('*', '')
    name = name.replace('"', '').replace('<', '').replace('>', '').replace('|', '')
    
    # 如果处理后名称为空，返回一个默认名
    if not name:
        return "renamed_file"
    return name

def main(dry_run, video_folder_path, target_folder_path, txt_file_path, index, max_len=1800):
    """
    读取TXT文件中的新名称，并将源视频文件复制到目标文件夹，
    并以新名称加上指定的索引后缀命名。

    参数:
    dry_run (bool): 如果为 True，则只打印计划的操作，不实际执行复制。
    video_folder_path (str): 源视频文件所在的文件夹路径。
    target_folder_path (str): 复制文件后的目标文件夹路径。
    txt_file_path (str): 包含新文件名的 TXT 文件路径。
    index (str): 用于附加到新文件名末尾的索引字符串 (例如: "0", "1")。
    """
    
    print("--- 视频批量复制与重命名脚本 ---")
    if dry_run:
        print("*** 警告: 当前处于“演练” (dry_run) 模式。***")
        print("*** 不会实际复制任何文件。只显示计划的操作。***")
    else:
        print("*** 正在执行复制操作... ***")
        
    print("---")
    
    # --- 0. 准备目标文件夹 ---
    if not dry_run:
        try:
            os.makedirs(target_folder_path, exist_ok=True)
            print(f"目标文件夹 '{target_folder_path}' 已确认或创建。")
        except Exception as e:
            print(f"错误: 无法创建目标文件夹 '{target_folder_path}': {e}")
            sys.exit(1)


    # --- 1. 读取并处理新名称列表 ---
    try:
        with open(txt_file_path, 'r', encoding='utf-8') as f:
            new_names = f.readlines()
            new_names = [line.strip() for line in new_names]
            new_names = [name for name in new_names if name]
        
        print(f"成功从 '{txt_file_path}' 读取了 {len(new_names)} 个新名称。")
        
    except FileNotFoundError:
        print(f"错误: 找不到 TXT 文件: {txt_file_path}")
        return
    except Exception as e:
        print(f"读取 TXT 文件时出错: {e}")
        return

    # --- 2. 读取并处理视频文件列表 ---
    try:
        all_files = os.listdir(video_folder_path)
        video_files = [f for f in all_files if f.lower().endswith('.mp4')]
        
        # 关键：按文件名排序，以确保顺序对应
        video_files.sort()
        
        print(f"成功从 '{video_folder_path}' 找到了 {len(video_files)} 个 .mp4 文件。")
        
    except FileNotFoundError:
        print(f"错误: 找不到源视频文件夹: {video_folder_path}")
        return
    except Exception as e:
        print(f"读取源视频文件夹时出错: {e}")
        return
        
    # --- 3. 检查数量是否匹配 ---
    if len(video_files) != len(new_names):
        print(f"!!! 严重错误: 文件数量 ({len(video_files)}) 与 TXT 文件行数 ({len(new_names)}) 不匹配！")
        print("请检查您的文件和 TXT 文件。")
        return

    print("文件数量与名称数量匹配。准备开始复制...")
    print("---")

    # --- 4. 循环并执行复制和重命名 ---
    copy_count = 0
    skipped_count = 0
    
    for old_filename, new_base_name in zip(video_files, new_names):
        
        try:
            # 4a. 获取原始文件的扩展名 (例如: .mp4)
            file_extension = os.path.splitext(old_filename)[1]
            
            # 4b. 清理新名称中的非法字符
            safe_new_base_name = sanitize_filename(new_base_name)

            if len(safe_new_base_name) > max_len:
                # 截断处理
                safe_new_base_name = safe_new_base_name[:max_len]
                print(f"  [警告] 名称 '{new_base_name}' 过长，已截断至前 {max_len} 个字符。")
            
            # 4c. 组合成新的完整文件名 (新名称 + 索引 + 扩展名)
            new_filename = f"{safe_new_base_name}-{index}{file_extension}"
            
            # 4d. 构建完整的新旧路径 (旧路径仍在源文件夹，新路径在目标文件夹)
            old_path = os.path.join(video_folder_path, old_filename)
            new_path = os.path.join(target_folder_path, new_filename)
            
            # 4e. 打印计划
            print(f"[计划] 复制 {old_filename}  --->  到目标文件夹并命名为 {new_filename}")
            
            # 4f. 检查目标文件夹中是否已存在该文件，避免覆盖
            if os.path.exists(new_path):
                print(f"  [跳过] 错误: 目标文件 '{new_filename}' 已存在于目标文件夹。")
                skipped_count += 1
                continue
                
            # 4g. 如果不是 dry_run 模式，则执行复制
            if not dry_run:
                # 使用 copy2 保留文件的元数据（如创建和修改时间）
                shutil.copy2(old_path, new_path)
                copy_count += 1
                
        except Exception as e:
            print(f"  [失败] 处理 {old_filename} 时发生意外错误: {e}")
            skipped_count += 1

    print("\n--- 操作总结 ---")
    if not dry_run:
        print(f"成功复制并重命名文件数量: {copy_count}")
        print(f"跳过或失败文件数量: {skipped_count}")
    else:
        print("当前处于演练模式，未执行实际操作。")
        
if __name__ == "__main__":
    # --- 1. 请修改这里的配置 ---

    # “演练”模式开关
    # True = 只打印会做什么，不实际重命名文件（推荐先运行一次）
    # False = 实际执行重命名操作
    dry_run = False


    # 视频文件夹的路径
    # (例如: '/path/to/your/videos')
    #seeds = [0, 84, 128, 1000]
    seeds = ['r1', 'r2', 'r3', 'r4', 'r5']
    max_len = 180

    index = [0, 1, 2, 3, 4]
    folder_sub = 'all_dimension_longer'
    # folder_sub = 'all_category'
    source_root = '/nfs/ywang29/Reward_finetuning/VideoX-Fun/validation_samples/samples_vbench2_all_wan_aug/missing/ps1'
    #source_root = '/nfs/ywang29/Reward_finetuning/VideoX-Fun/validation_samples/samples_vbench2_all'
    model_name = 'Nov2_1'
    steps = 4000
    model_path = f'output_{model_name}/steps_{steps}'
    # model_path = f'output_pretrain'

    # video_folder_path_list = [f'/nfs/ywang29/Reward_finetuning/VideoX-Fun/validation_samples/samples_vbench_{folder_sub}/{i}/wan-videos-t2v-512*288-30-seed-1/output_Oct16_1/steps_2000' for i in seeds]
    video_folder_path_list = [f'/nfs/ywang29/Reward_finetuning/VideoX-Fun/validation_samples/samples_vbench_{folder_sub}/{i}/wan-videos-t2v-512*288-30-seed-1/{model_path}' for i in seeds]

    # 包含新名称的 TXT 文件的路径
    # (例如: '/path/to/your/names.txt')
    # /nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/prompts/VBench2_full_text.txt
    # txt_file_path_list = [f'/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/prompts/{folder_sub}.txt' for i in seeds]
    txt_file_path_list = f'/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/missing_prompts.txt'
    target_folder = f'/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/sampled_videos/wan1.3b_rwft_{model_name}_{steps}_wan_aug'
    # target_folder = f'/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/sampled_videos/wan1.3b_base_longer'

    
    vbench2 = True
    if vbench2:
        index = [0, 1, 2]
        seeds = ['r1', 'r2', 'r3']
        # index = [0,0,1,1,2,2]
        # seeds = ['r1_p1','r1_p2','r2_p1','r2_p2','r3_p1','r3_p2']
        # index = [2, 2]
        # seeds = ['r3_p1', 'r3_p2']
        video_folder_path_list = [f'{source_root}/{i}/wan-videos-t2v-512*288-30-seed-1/{model_path}' for i in seeds]
        # video_folder_path_list.append(f'{source_root}/wan-videos-t2v-512*288-30/{model_path}')
        txt_file_path_list = f'/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/missing_prompts.txt'

        #txt_file_path_list = '/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/prompts/prompt_aug/Wanx_full_text_aug.txt'
        target_folder = f'/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/sampled_videos/control/wan1.3b_rwft_{model_name}_{steps}_wan_aug'
        #target_folder = f'/nfs/ywang29/Reward_finetuning/VideoX-Fun/VBench/VBench-2.0/sampled_videos/wan1.3b_base_wan_aug'

        max_len = 180

    for i in range(len(video_folder_path_list)):
        if 'p1' in seeds[0]:
            main(dry_run, video_folder_path_list[i], target_folder, txt_file_path_list[i%2], index[i], max_len = max_len)
        else:
            main(dry_run, video_folder_path_list[i], target_folder, txt_file_path_list, index[i], max_len = max_len)

    # main(dry_run, f'{source_root}/wan-videos-t2v-512*288-30/output_Oct16_1/steps_2000', txt_file_path_list[0], 4)

    # seeds = ['r10', 'r11', 'r12']
    # base_video_folder_path_list = [f'/nfs/ywang29/Reward_finetuning/VideoX-Fun/validation_samples/samples_vbench_all_dimension_aug/{i}/wan-videos-t2v-512*288-30-seed-1/output_pretrain' for i in seeds]
    
    # # 包含新名称的 TXT 文件的路径
    # # (例如: '/path/to/your/names.txt')

    # for i in range(len(base_video_folder_path_list)):
    #     main(dry_run, base_video_folder_path_list[i], txt_file_path_list[i], index[i])


    # main(dry_run, f'/nfs/ywang29/Reward_finetuning/VideoX-Fun/validation_samples/samples_vbench_{folder_sub}/wan-videos-t2v-512*288-30/output_Oct16_1/steps_2000', txt_file_path_list[0], 4)
    # main(dry_run, '/nfs/ywang29/Reward_finetuning/VideoX-Fun/validation_samples/samples_vbench_all_category/wan-videos-t2v-512*288-30/output_Oct14_6/steps_2000', txt_file_path_list[0], 4)