Python script not returning correct error status

Question

I am new to python,Need some help!

PYTHON VERSION: 2.7

SCRIPT

import subprocess
import sys

HOST="user@machine"
COMMAND="ps -fu user | grep \"XYZ\" |grep -v grep |wc -l"  #Command to be run

ssh = subprocess.Popen(["ssh", "%s" % HOST, COMMAND],shell=False,stdout=subprocess.PIPE,stderr=subprocess.PIPE)

result = ssh.stdout.readlines()

if result == []:
        error = ssh.stderr.readlines()
        print >>sys.stderr, "ERROR: %s" % error
else:
        result = [qaz.strip('\n') for qaz in result]
        print result

This script results in this output:

['0']

PROBLEM

Even If change the COMMAND to be executed on other machine to some irrelevant string,I am not getting the error message ,say I change the command to below value and run the script again.

COMMAND="adsasdadsps -fu user | grep \"XYZ\" |grep -v grep |wc -l"

The script now also results in same output.

['0']

But When I changed the command as below.

COMMAND="asdadasd"

Then I got the output.

ERROR: ['ksh: line 1: asdadasd: not found\n']

Can anyone help with me understand why I am not getting error message for

COMMAND="adsasdadsps -fu user | grep \"XYZ\" |grep -v grep |wc -l"

SethMMorton · Accepted Answer · 2013-08-14 22:57:44Z

1

You might want to change your logic so that you check if the stderr is empty or not:

result = ssh.stdout.readlines()
error = ssh.stderr.readlines()

if error:
    print >>sys.stderr, "ERROR: {}".format(''.join(error))
else:
    print ''.join(result)

Note that I have changed the error and result messages so they get printed without the brackets. Also, it's better to test if a list is empty by using the fact that an empty list is False instead of comparing to [].

edited Aug 14, 2013 at 22:57

answered Aug 14, 2013 at 18:51

SethMMorton

49.5k13 gold badges72 silver badges90 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

g4ur4v Over a year ago

Hi,This is working great but you need to change print >>sys.stderr, "ERROR: %s".format('\n'.join(error)) to print >>sys.stderr, "ERROR: %s" % format('\n'.join(error))

g4ur4v Over a year ago

print >>sys.stderr, "ERROR: %s".format(''.join(error)) outputs ERROR: %s

SethMMorton Over a year ago

Got it, changed "%s" to "{}"

tdelaney · Accepted Answer · 2013-08-14 19:06:53Z

1

Its a shell thing. The shell sets up the full pipeline even though the first command errors. wc -l reports that it counted '0' lines and that's where the ['0'] comes from. You could add a 'bash' tag to this post for more details, but I think that its a fork/exec that emits the error after the pipeline is setup (that is, the top level shell doesn't know the program didn't exist, its a forked copy that got that bit of news).

The return code isn't going to help (the shell gives you the return code of the last program in the pipeline). You can read stderr but be careful of programs that emit chatty traffic on stderr even though its not a real error.

answered Aug 14, 2013 at 19:06

tdelaney

77.9k6 gold badges91 silver badges129 bronze badges

Collectives™ on Stack Overflow

Python script not returning correct error status

2 Answers 2

3 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

3 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related