Question

我有一个适用于Linux的脚本，但与Windows上的Python 3不兼容（Python 3.5.2是经过测试的版本）。该脚本的目标是实时捕获子进程的stdout / stderr。这是一个重现错误的脚本：

import subprocess
import os
import sys

from multiprocessing import Process, Queue
try:
    from Queue import Empty
except ImportError:
    from queue import Empty  # python 3.x

ON_POSIX = 'posix' in sys.builtin_module_names


def enqueue_output(out, q):
    for line in iter(out.readline, b''):
        q.put(line)
    out.close()


def output_monitor_queue(out):
    q = Queue()
    p = Process(target=enqueue_output, args=(out, q))
    p.start()
    return q, p


def update_from_output_queue(q, out):
    try:
        line = q.get_nowait()
    except Empty:
        return out

    out += str(line)
    return out


def runit():
    stdout, stderr = '', ''
    proc = subprocess.Popen(['ipconfig.exe'], stdout=subprocess.PIPE, stderr=subprocess.PIPE)

    # We need to use subprocesses to capture the IO, otherwise they will block one another
    # i.e. a check against stderr will sit waiting on stderr before returning
    # we use Queues to communicate
    qout, pout = output_monitor_queue(proc.stdout)
    qerr, perr = output_monitor_queue(proc.stderr)

    prev_std = None

    # Loop until the process is complete + both stdout/stderr have EOFd
    while proc.poll() is None or pout.is_alive() or perr.is_alive():

        # Check for updates from either (non-blocking)
        stdout = update_from_output_queue(qout, stdout)
        stderr = update_from_output_queue(qerr, stderr)

        # If there are changes, update the db
        if (stdout, stderr) != prev_std:
            print(stdout, stderr)
            prev_std = (stdout, stderr)


runit()

运行它时，会抛出以下错误：

Traceback（最近一次调用最后一次）：文件＆＃34; C：\ Users \ IEUser \ Desktop \ failed.py＆＃34;，第62行，in runit（）文件＆＃34; C：\ Users \ IEUser \ Desktop \ failed.py＆＃34;，第44行，在runit中 qout，pout = output_monitor_queue（proc.stdout）文件＆＃34; C：\ Users \ IEUser \ Desktop \ failed.py＆＃34;，第23行，在output_monitor_queue中 p.start（）文件＆＃34; C：\ Users \ IEUser \ AppData \ Local \ Programs \ Python \ Python35-32 \ lib \ multiprocessing \ process.py＆＃34;，第105行，开始 self._popen = self._Popen（self）File＆＃34; C：\ Users \ IEUser \ AppData \ Local \ Programs \ Python \ Python35-32 \ lib \ multiprocessing \ context.py＆＃34;，第212行，在_Popen return _default_context.get_context（）。Process._Popen（process_obj）文件＆＃34; C：\用户\ IEUser \应用程序数据\本地\程序\的Python \ Python35-32 \ lib中\多\ context.py＆＃34 ;, 第311行，在_Popen 返回Popen（process_obj）文件＆＃34; C：\ Users \ IEUser \ AppData \ Local \ Programs \ Python \ Python35-32 \ lib \ multiprocessing \ popen_spawn_win32.py＆＃34;，第66行，在 init 中 reduction.dump（process_obj，to_child）文件＆＃34; C：\ Users \ IEUser \ AppData \ Local \ Programs \ Python \ Python35-32 \ lib \ multiprocessing \ reduction.py＆＃34;，第59行，在转储中 ForkingPickler（文件，协议）.dump（obj）TypeError：无法序列化＆＃39; _io.BufferedReader＆＃39;对象

是否有人遇到此问题并且知道有效的跨平台解决方案？

无法将subprocess.PIPE传递给Windows上的进程

0 个答案: