utils.py

import logging

def handle_scp(scp_path):
    '''
    Read scp file script
    input: 
          scp_path: .scp file's file path
    output: 
          scp_dict: {'key':'wave file path'}
    '''
    scp_dict = dict()
    line = 0
    lines = open(scp_path, 'r').readlines()
    for l in lines:
        scp_parts = l.strip().split()
        line += 1
        if len(scp_parts) != 2:
            raise RuntimeError("For {}, format error in line[{:d}]: {}".format(
                scp_path, line, scp_parts))
        if len(scp_parts) == 2:
            key, value = scp_parts
        if key in scp_dict:
            raise ValueError("Duplicated key \'{0}\' exists in {1}".format(
                key, scp_path))

        scp_dict[key] = value

    return scp_dict


def get_logger(name, format_str="%(asctime)s [%(pathname)s:%(lineno)s - %(levelname)s ] %(message)s",
               date_format='%Y-%m-%d %H:%M:%S', file=False):
    logger = logging.getLogger(name)
    logger.setLevel(logging.INFO)
    # file or console
    handler = logging.StreamHandler() if not file else logging.FileHandler(
        name)
    handler.setLevel(logging.INFO)
    formatter = logging.Formatter(fmt=format_str, datefmt=date_format)
    handler.setFormatter(formatter)
    logger.addHandler(handler)
    return logger


if __name__ == "__main__":
    print(len(handle_scp('./dataset/create_scp/cv_s2.scp')))