Pyte无法在python3中生成正确的输出

我有一段代码在Python2.7.5中运行良好，但在Python3中不起作用

主要问题是tee.write，它无法写入文件

这段代码假设将20个字母a写入文件/tmp/tee-test-1和/tmp/tee-test-2，但事实并非如此，这两个文件是空的

谁能给我一些建议吗

import sys
import os
import subprocess
#from  netsa.util.shell import *
from string import Template

__author__ = 'Brandon Sandrowicz <brandon@sandrowicz.org>'
__version__ = '0.1'

valid_modes = ['a','w']

def create_tee(files, mode, buffer_size=128):


        if mode not in valid_modes:
            raise IOError("Only valid modes to create_tee() are: %s" % ', '.join(valid_modes))

        tee_list = []
        for file in files:
            if type(file) == str:
                fp = open(file, mode)
                tee_list.append(fp)
            else:
                tee_list.append(file)

        pipe_read, pipe_write = os.pipe()
        pid = os.fork()
        if pid == 0:
            # Child -- Read bytes from the pipe and write them to the specified
            # files.
            try:
                # Close parent's end of the pipe
                os.close(pipe_write)

                bytes = os.read(pipe_read, buffer_size)
                print (bytes)
                while(bytes):
                    for file in tee_list:
                        file.write(bytes)
                        file.flush()
                        # TODO maybe add in fsync() here if the fileno() method
                        # exists on file

                    bytes = os.read(pipe_read, buffer_size)
            except:
                pass
            finally:
                os._exit(255)
        else:
            # Parent -- Return a file object wrapper around the pipe to the
            # child.
            return os.fdopen(pipe_write,'w')



if __name__ == '__main__':

    files = [ '/tmp/tee-test-1', '/tmp/tee-test-2' ]
    num_chars = 100000

    print("Writing %d chars to files (using create_tee):" % num_chars)
    for file in files:
        print(" %s" % file)
    print()

    tee = create_tee(files,mode='a')
    #print("a" * num_chars, end=' ', file=tee)
    tee.write("a" * 20)
    tee.close()
    os.wait()

    for filename in files:
        with open(filename, 'r') as fh:
            chars = len(fh.read())
            print("File '%s' has %d chars" % (filename, chars))

---

好的，我发现这个问题很有趣，也很有挑战性，最后我发现了问题所在，如：

一个常见的问题是文件以错误的模式打开。确保打开带有“t”标志的文本文件和带有“b”标志的二进制文件，这样就解决了许多问题

因此，当您以b数据类型写入数据时，我尝试了以下方法：

    for file in files:
        if type(file) == str:
            fp = open(file, mode+'b')
            tee_list.append(fp)
        else:
            tee_list.append(file)

而且效果很好：

File '/tmp/tee-test-1' has 20 chars
File '/tmp/tee-test-2' has 20 chars

---

定义不起作用。你能编辑你得到的错误吗？这段代码假设在文件'/tmp/tee-test-1''/tmp/tee-test-2'中写入20个字母a，但是它没有，这两个文件是空的，只是测试了它，实际上，它在py2和py3中的行为不同…