python NLP数据集分割大文件

码狂☆2024-04-19 19:03

python NLP数据集分割大文件

NLP数据文件有时候特别大的文件，需要分割成N个小文件来处理

部分提取：可以提取N份，每份K行

全部分割：分割整个文件，每一份K行

python 复制代码

import os

def split_file(filename, outdir,num_lines):
    """ 将文件按行数进行分割 \n
        filename 文件名 \n
        num_lines 每份包含的行数 \n
    """
    file_name_without_path_and_extension = os.path.splitext(os.path.basename(filename))[0]

    with open(filename, 'r') as f:
        current_chunk = 1
        current_line = 0
        current_output = open(f"{outdir}{file_name_without_path_and_extension}{current_chunk}.txt", 'w')
        for line in f:
            current_output.write(line)
            current_line += 1
            if current_line >= num_lines:
                current_output.close()
                current_chunk += 1
                current_line = 0
                current_output = open(f"{outdir}{file_name_without_path_and_extension}{current_chunk}.txt", 'w')
        current_output.close()

def split_file_max_chunks(filename,outdir, num_lines, max_chunks):
    """ 将文件按行数进行分割 \n
        filename 文件名 \n
        num_lines 每份包含的行数 \n
        max_chunks 最大分出多少份 \n
    """
    file_name_without_path_and_extension = os.path.splitext(os.path.basename(filename))[0]

    with open(filename, 'r') as f:
        current_chunk = 1
        current_line = 0
        current_output = open(f"{outdir}{file_name_without_path_and_extension}{current_chunk}.txt", 'w')
        for line in f:
            current_output.write(line)
            current_line += 1
            if current_line >= num_lines:
                current_output.close()
                if current_chunk >= max_chunks:
                    break
                current_chunk += 1
                current_line = 0
                current_output = open(f"{outdir}{file_name_without_path_and_extension}{current_chunk}.txt", 'w')  # 这里更新了current_output
        current_output.close()

def main():
    large_filename = "./data/large_file_1G.txt"
    outdir="./docs/"
    num_lines = 1000  # 每份包含 1000 行
    split_file(large_filename,outdir, num_lines)
    
    # max_chunks = 30   # 最大分出 30 份
    # split_file_max_chunks(large_filename,outdir, num_lines, max_chunks)

   

if __name__ == "__main__":
    main()