平常需要用到这个功能，于是自己写了个
2025.12.15
更新，之前的一堆bug

import os
import re


def path_to_original(vtt_file_path=""):
    '''
        从可能包含多层扩展名的文件路径中提取原始文件名。
        会递归移除指定的媒体文件扩展名(.mp3, .wav, .vtt等)。
    :param vtt_file_path: /home/direct.ory/file.wav.vtt 转换成 /home/direct.ory/file
    :return: /home/direct.ory/file
    '''
    dic_name = os.path.dirname(vtt_file_path)
    file_name = os.path.basename(vtt_file_path)

    # 把文件名分成真实名称和扩展名
    real_file_name, ext_name = os.path.splitext(file_name)
    not_allow_ext = ['.mp3', '.wav', '.vtt']
    # 如果扩展名还包含在真实名称里，继续分离
    while ext_name and ext_name in not_allow_ext:
        real_file_name, ext_name = os.path.splitext(real_file_name)

    if ext_name not in not_allow_ext:
        real_file_name = real_file_name + ext_name
    # 返回无后缀名的路径
    return os.path.join(dic_name, real_file_name)


# 处理时间格式
def deal_time(timeStamp=''):
    # if len(timeStamp) < 1:
    #     return None
    #
    # startTime = timeStamp.split("-->")[0].strip()
    # times = startTime.split(":")
    # times[-1] = times[-1][0: 5]
    # # print(startTime)
    # # print(times)
    # ans = "[{}:{}]".format(times[1], times[2])
    # # print(ans)
    # return ans
    if not timeStamp:
        return None

        # 使用正则表达式匹配时间格式
    time_match = re.match(r'(\d{2}):(\d{2}):(\d{2})\.(\d{3})', timeStamp.split("-->")[0].strip())
    if time_match:
        minutes, seconds = time_match.group(2), time_match.group(3)
        # 将毫秒转换为百分之一秒（LRC格式）
        centiseconds = str(round(int(time_match.group(4)) / 10)).zfill(2)
        return f"[{minutes}:{seconds}.{centiseconds}]"

    return None


# 进行转换
def convert_vtt_to_lrc(vtt_file_path):
    lines = []
    # 打开vtt文件进行读取
    with open(vtt_file_path, 'r', encoding='utf-8') as vtt_file:
        # 读取所有行
        lines = vtt_file.readlines()

    # 创建对应的lrc文件名
    lrc_file_path = path_to_original(vtt_file_path) + '.lrc'
    print(lrc_file_path)
    # lrc_file_path = "{}{}".format(lrc_file_path.split(".")[0], ".lrc")
    # print(lrc_file_path)
    # return
    timeStamp = ''
    # 标记句子的开始和结束

    contents = []
    # 打开lrc文件进行写入

    for line in lines:
        if line.strip().isnumeric():
            # 上一句添加换行
            if contents:  # 确保列表不为空
                contents[-1] = contents[-1] + "\n"
            continue

        # 如果行以时间戳格式开始，则跳过（vtt文件中的时间戳行不需要复制到lrc文件）
        if line.strip().startswith(('00:', '01:', '02:', '03:', '04:', '05:', '06:', '07:', '08:', '09:')):
            timeStamp = deal_time(line.strip())
            # print(timeStamp)
            continue

        # line = line.replace("\n", "")
        # 将其他行写入lrc文件
        contents.append(timeStamp + line)
        # print(timeStamp + line)
        timeStamp = ""

    with open(lrc_file_path, 'w', encoding='utf-8') as lrc_file:
        lrc_file.writelines(contents)

    # remove_blank_lines(vtt_file_path)


def convert_vtt_files_to_lrc(directory):
    # 获取目录下的所有文件和文件夹
    for root, dirs, files in os.walk(directory):
        for file in files:
            # 拼接文件的完整路径
            file_path = os.path.join(root, file)
            # 如果文件是vtt文件，则进行转换
            if file.endswith('.vtt'):
                print(f"Converting {file} to LRC...")
                convert_vtt_to_lrc(file_path)
            # elif file.endswith('.lrc'):
            #     cut(file_path)


# 测试程序
if __name__ == "__main__":
    # 输入要处理的目录路径
    directory_path = input("请输入要处理的目录路径：")
    # 调用函数将vtt文件转换为lrc文件
    convert_vtt_files_to_lrc(directory_path)
    print("Conversion completed!")