python执行shell命令

1 os.system

可以返回运行shell命令状态，同时会在终端输出运行结果

例如 ipython中运行如下命令，返回运行状态status

os.system('cat /etc/passwdqc.conf')
min=disabled,24,11,8,7
max=40
passphrase=3
match=4
similar=deny
random=47
enforce=everyone
retry=3
Out[6]: 0

2 os.popen()

可以返回运行结果

popen(command [, mode='r' [, bufsize]]) -> pipe
Open a pipe to/from a command returning a file object.

运行返回结果

In [20]: output = os.popen('cat /proc/cpuinfo')

In [21]: lineLen = []

In [22]: for line in output.readlines():
    lineLen.append(len(line))
   ....:     

In [23]: line
line     lineLen  

In [23]: lineLen
Out[23]: 
[14,
 25,
...

3 如何同时返回结果和运行状态，commands模块:

#String form: <module 'commands' from '/usr/lib64/python2.7/commands.pyc'>
File: /usr/lib64/python2.7/commands.py
Docstring:
Execute shell commands via os.popen() and return status, output.

Interface summary:

import commands

outtext = commands.getoutput(cmd)
(exitstatus, outtext) = commands.getstatusoutput(cmd)
outtext = commands.getstatus(file) # returns output of "ls -ld file"

A trailing newline is removed from the output string.

Encapsulates the basic operation:

pipe = os.popen('{ ' + cmd + '; } 2>&1', 'r')
text = pipe.read()
sts = pipe.close()

commands示例如下:

In [24]: (status, output) = commands.getstatusoutput('cat /proc/cpuinfo')

In [25]: status
Out[25]: 0

In [26]: len(output)
Out[26]: 3859

4 使用模块subprocess

通常项目中经常使用方法为subporcess.Popen, 我们可以在Popen()建立子进程的时候改变标准输入、标准输出和标准错误，并可以利用subprocess.PIPE将多个子进程的输入和输出连接在一起，构成管道(pipe):

import subprocess
child1 = subprocess.Popen(["ls","-l"], stdout=subprocess.PIPE)
child2 = subprocess.Popen(["wc"], stdin=child1.stdout,stdout=subprocess.PIPE)
out = child2.communicate()
print(out)

 在例如使用lsblk查看swap分区的uuid：

import subprocess

child = subprocess.Popen(["lsblk", "-f"], stdout=subprocess.PIPE)
out = child.stdout.readlines()

swap_uuid = None
for item in out:
    line = item.strip().split()
    if len(line) == 4:
        if(line[1] == 'swap'):
            swap_uuid = line[2]
print(swap_uuid)

ipython 中运行"?subprocess"可以发现subprocess是python用来替换os.popen()等管道操作命令的新模块

A more real-world example would look like this:

try:
    retcode = call("mycmd" + " myarg", shell=True)
    if retcode < 0:
        print >>sys.stderr, "Child was terminated by signal", -retcode
    else:
        print >>sys.stderr, "Child returned", retcode
except OSError, e:
    print >>sys.stderr, "Execution failed:", e

相对于上面几种方式，subprocess便于控制和监控进程运行结果,subprocess提供多种函数便于应对父进程对子进程不同要求:

4.1.1 subprocess.call()

父进程父进程等待子进程完成，返回exit code

4.1.2 subprocess.check_call()

父进程等待子进程完成,返回0，如果returncode不为0，则举出错误subprocess.CalledProcessError，该对象包含有returncode属性，可用try...except...来检查

4.1.3 subprocess.check_output()

父进程等待子进程完成

返回子进程向标准输出的输出结果

检查退出信息，如果returncode不为0，则举出错误subprocess.CalledProcessError，该对象包含有returncode属性和output属性，output属性为标准输出的输出结果，可用try...except...来检查

例如:

In [32]: out = subprocess.call("ls -l", shell=True)
total 42244
-rw-rw-r--.  1 *** ***     366 May 26 09:10 ChangeLog

4.2.1 

上面三个函数都是源于Popen()函数的wapper(封装),如果需要更加个性化应用，那么就需要使用popen()函数

Popen对象创建后，主程序不会自动等待子进程完成。我们必须调用对象的wait()方法，父进程才会等待 (也就是阻塞block)

[wenwt@localhost syntax]$ rm subprocess.pyc 
[wenwt@localhost syntax]$ python process.py 
parent process
[wenwt@localhost syntax]$ PING www.google.com (173.194.219.99) 56(84) bytes of data.
^C
[wenwt@localhost syntax]$ 
--- www.google.com ping statistics ---
5 packets transmitted, 0 received, 100% packet loss, time 3999ms

加上wait方法:

[wenwt@localhost syntax]$ python process.py 
PING www.google.com (173.194.219.103) 56(84) bytes of data.

--- www.google.com ping statistics ---
5 packets transmitted, 0 received, 100% packet loss, time 3999ms

parent process

参考文章:Python标准库06 子进程 (subprocess包)