python对大文件进行乱序
背景
数据处理,文件太大,几十个G的。
思想
逐行读取并记录每行开头的位置,将位置列表乱序,再按乱序后的位置逐行读取并写入新文件。
代码
'''
xxx.py input_file output_file
'''
import os, sys
import random
def get_line_offset(f):
lines_start_offset=list()
f.seek(0)
lines_start_offset.append(f.tell())
line = f.readline()
while line:
line=line.strip()
lines_start_offset.append(f.tell())
line = f.readline()
return lines_start_offset
def rewrite_file(f_in, f_out, lines_start_offset):
for i in range(len(lines_start_offset)):
f_in.seek(lines_start_offset[i], 0)
line=f_in.readline()
f_out.write(line)
if __name__ == "__main__":
path1 = os.path.abspath(sys.argv[1])
path2 = os.path.abspath(sys.argv[3])
f = open(path1, 'r', encoding='utf-8')
f_out = open(path2, 'w', encoding='utf-8')
lines_start_offset = get_line_offset(f)
random.shuffle(lines_start_offset)
rewrite_file(f, f_out, lines_start_offset)
f.close()
f_out.close()