logilab-astng #46273 wrong type inference for subprocess.Popen(...).communicate(...) [validation pending]

Comments

• 2011/05/30 14:36, written by anon

  Test case:
  

  #!/usr/bin/env python
  import subprocess
  a, b = subprocess.Popen('true', stdout=subprocess.PIPE, stderr=subprocess.PIPE).communicate()
  print a.strip()
  

  E1103: 4: Instance of 'list' has no 'strip' member (but some types could not be inferred)
  
• 2011/05/30 14:36, written by anon

  Test case:
  

  #!/usr/bin/env python
  import subprocess
  a, b = subprocess.Popen('true', stdout=subprocess.PIPE, stderr=subprocess.PIPE).communicate()
  print a.strip()
  

  E1103: 4: Instance of 'list' has no 'strip' member (but some types could not be inferred)
  
• 2011/11/23 21:03, written by anon

  I have this issue to
  
• 2012/01/20 20:16, written by anon

  +1
  
• 2012/08/03 21:31, written by anon

  I have this issue as well.
  
• 2012/08/16 19:54, written by anon

  +1
  
• 2012/08/29 09:43, written by sthenault

  for those to which this matters: this is a good opportunity to contribute to the pylint-brain :)

  more info here: http://www.logilab.org/blogentry/100364
• 2012/09/13 12:43, written by alutz

  This bug is corrected by a functionnality implemented in pylint-brain
• 2012/10/12 15:39, written by anon

  Since the bug fix we now get:
  

  #!/usr/bin/env python
  import subprocess
  a = subprocess.Popen('true').wait()
  

  E1101: 3,4: Instance of 'Popen' has no 'wait' member
  
• 2012/11/01 18:09, written by anon

  I have the same problem with Popen - wait, returncode and stdin yield E1101
  
  • 2012/11/15 11:34, written by anon

    Me 2.
    
• 2013/01/22 15:20, written by anon

  The bugfix didn't completely fix the issues, I'm still getting :
  Instance of 'Popen' has no 'wait' member
  Instance of 'Popen' has no 'returncode' member