我写了一封信来处理错误,并重新引导输出和其他东西。

import shlex
import psutil
import subprocess

def call_cmd(cmd, stdout=sys.stdout, quiet=False, shell=False, raise_exceptions=True, use_shlex=True, timeout=None):
    """Exec command by command line like 'ln -ls "/var/log"'
    """
    if not quiet:
        print("Run %s", str(cmd))
    if use_shlex and isinstance(cmd, (str, unicode)):
        cmd = shlex.split(cmd)
    if timeout is None:
        process = subprocess.Popen(cmd, stdout=stdout, stderr=sys.stderr, shell=shell)
        retcode = process.wait()
    else:
        process = subprocess.Popen(cmd, stdout=stdout, stderr=sys.stderr, shell=shell)
        p = psutil.Process(process.pid)
        finish, alive = psutil.wait_procs([p], timeout)
        if len(alive) > 0:
            ps = p.children()
            ps.insert(0, p)
            print('waiting for timeout again due to child process check')
            finish, alive = psutil.wait_procs(ps, 0)
        if len(alive) > 0:
            print('process {} will be killed'.format([p.pid for p in alive]))
            for p in alive:
                p.kill()
            if raise_exceptions:
                print('External program timeout at {} {}'.format(timeout, cmd))
                raise CalledProcessTimeout(1, cmd)
        retcode = process.wait()
    if retcode and raise_exceptions:
        print("External program failed %s", str(cmd))
        raise subprocess.CalledProcessError(retcode, cmd)

你可以这样称之为:

cmd = 'ln -ls "/var/log"'
stdout = 'out.txt'
call_cmd(cmd, stdout)

我会建议使用子过程模块而不是os.system,因为它会让您逃脱,因此更安全。

subprocess.call(['ping', 'localhost'])

这里有很多答案,但没有一个满足了我所有的需求。

我需要运行命令并捕获输出和输出代码. 我需要时间out 执行的程序,并强迫它出去,如果时间out 达到,并杀死它的所有儿童过程. 我需要它在 Windows XP 和以后, Cygwin 和 Linux 工作。

def _run(command, timeout_s=False, shell=False):
    ### run a process, capture the output and wait for it to finish. if timeout is specified then Kill the subprocess and its children when the timeout is reached (if parent did not detach)
    ## usage: _run(arg1, arg2, arg3)
        # arg1: command + arguments. Always pass a string; the function will split it when needed
        # arg2: (optional) timeout in seconds before force killing
        # arg3: (optional) shell usage. default shell=False
    ## return: a list containing: exit code, output, and if timeout was reached or not

    # - Tested on Python 2 and 3 on Windows XP, Windows 7, Cygwin and Linux.
    # - preexec_fn=os.setsid (py2) is equivalent to start_new_session (py3) (works on Linux only), in Windows and Cygwin we use TASKKILL
    # - we use stderr=subprocess.STDOUT to merge standard error and standard output
    import sys, subprocess, os, signal, shlex, time

    def _runPY3(command, timeout_s=None, shell=False):
        # py3.3+ because: timeout was added to communicate() in py3.3.
        new_session=False
        if sys.platform.startswith('linux'): new_session=True
        p = subprocess.Popen(command, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, start_new_session=new_session, shell=shell)

        try:
            out = p.communicate(timeout=timeout_s)[0].decode('utf-8')
            is_timeout_reached = False
        except subprocess.TimeoutExpired:
            print('Timeout reached: Killing the whole process group...')
            killAll(p.pid)
            out = p.communicate()[0].decode('utf-8')
            is_timeout_reached = True
        return p.returncode, out, is_timeout_reached

    def _runPY2(command, timeout_s=0, shell=False):
        preexec=None
        if sys.platform.startswith('linux'): preexec=os.setsid
        p = subprocess.Popen(command, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, preexec_fn=preexec, shell=shell)

        start_time = time.time()
        is_timeout_reached = False
        while timeout_s and p.poll() == None:
            if time.time()-start_time >= timeout_s:
                print('Timeout reached: Killing the whole process group...')
                killAll(p.pid)
                is_timeout_reached = True
                break
            time.sleep(1)
        out = p.communicate()[0].decode('utf-8')
        return p.returncode, out, is_timeout_reached

    def killAll(ParentPid):
        if sys.platform.startswith('linux'):
            os.killpg(os.getpgid(ParentPid), signal.SIGTERM)
        elif sys.platform.startswith('cygwin'):
            # subprocess.Popen(shlex.split('bash -c "TASKKILL /F /PID $(</proc/{pid}/winpid) /T"'.format(pid=ParentPid)))
            winpid=int(open("/proc/{pid}/winpid".format(pid=ParentPid)).read())
            subprocess.Popen(['TASKKILL', '/F', '/PID', str(winpid), '/T'])
        elif sys.platform.startswith('win32'):
            subprocess.Popen(['TASKKILL', '/F', '/PID', str(ParentPid), '/T'])

    # - In Windows, we never need to split the command, but in Cygwin and Linux we need to split if shell=False (default), shlex will split the command for us
    if shell==False and (sys.platform.startswith('cygwin') or sys.platform.startswith('linux')):
        command=shlex.split(command)

    if sys.version_info >= (3, 3): # py3.3+
        if timeout_s==False:
            returnCode, output, is_timeout_reached = _runPY3(command, timeout_s=None, shell=shell)
        else:
            returnCode, output, is_timeout_reached = _runPY3(command, timeout_s=timeout_s, shell=shell)
    else:  # Python 2 and up to 3.2
        if timeout_s==False:
            returnCode, output, is_timeout_reached = _runPY2(command, timeout_s=0, shell=shell)
        else:
            returnCode, output, is_timeout_reached = _runPY2(command, timeout_s=timeout_s, shell=shell)

    return returnCode, output, is_timeout_reached

然后用它如下:

总是将命令作为一个行(它更容易)。你不需要分开它;函数在需要时将分开它。

所以我们可以用这个类似的时间:

a=_run('cmd /c echo 11111 & echo 22222 & calc',3)
for i in a[1].splitlines(): print(i)

或者没有时间:

b=_run('cmd /c echo 11111 & echo 22222 & calc')

更多例子:

b=_run('''wmic nic where 'NetConnectionID="Local Area Connection"' get NetConnectionStatus /value''')
print(b)

c=_run('cmd /C netsh interface ip show address "Local Area Connection"')
print(c)

d=_run('printf "<%s>\n" "{foo}"')
print(d)

你也可以指定Shell=True,但在大多数情况下,这个功能是无用的。我宁愿自己选择我想要的Shell,但这里是,如果你也需要它:

# windows
e=_run('echo 11111 & echo 22222 & calc',3, shell=True)
print(e)
# Cygwin/Linux:
f=_run('printf "<%s>\n" "{foo}"', shell=True)
print(f)

為什麼我沒有使用更簡單的新方法 subprocess.run()?

因为它支持在Python 3.7+,但最后支持的Python版本在Windows XP是3.4. 而且由于这个功能的Timeout论点在Windows中是无用的,它不会杀死执行命令的儿童过程. 如果你使用 capture_output + Timeout论点,它会挂在如果有一个儿童过程仍然运行。

我总是使用纺织品来做这些事情,这里是一个演示代码:

from fabric.operations import local
result = local('ls', capture=True)
print "Content:/n%s" % (result, )

但这似乎是一个很好的工具:sh(Python子处理界面)。

看看一个例子:

from sh import vgdisplay
print vgdisplay()
print vgdisplay('-v')
print vgdisplay(v=True)

2015 年更新: Python 3.5 添加了 subprocess.run 比 subprocess.Popen 更容易使用。

>>> subprocess.run(["ls", "-l"])  # doesn't capture output
CompletedProcess(args=['ls', '-l'], returncode=0)

>>> subprocess.run("exit 1", shell=True, check=True)
Traceback (most recent call last):
  ...
subprocess.CalledProcessError: Command 'exit 1' returned non-zero exit status 1

>>> subprocess.run(["ls", "-l", "/dev/null"], capture_output=True)
CompletedProcess(args=['ls', '-l', '/dev/null'], returncode=0,
stdout=b'crw-rw-rw- 1 root root 1, 3 Jan 23 16:23 /dev/null\n', stderr=b'')

典型的实施:

import subprocess

p = subprocess.Popen('ls', shell=True, stdout=subprocess.PIPE, stderr=subprocess.STDOUT)
for line in p.stdout.readlines():
    print line,
retval = p.wait()

事实上,你可以简单地忽略这些参数(stdout=和stderr=)并将像os.system()一样行事。