DEBUGGING_PYTHON.md

Debugging Python

This tutorial focuses on using pdb to debug python.

Create your buggy script

echo '#!/usr/bin/env python
import sys

def foo():
    a = 1
    b = 2
    c = 1+b
    d = 4-c
    e = 1 - d/0.25
    f = 13.23 / (e + 3)
    g = f / 34
    r = g + 2 + 4 + + 100
    sys.stdout.write("And the answer is {}\n".format(r))

try:
    foo()
except:
    sys.stdout.write("*** Uh uh, this script failed, and the reason is hidden by this try/except ***\n")
' > ./buggy.py
chmod a+x ./buggy.py

Run it and debug with pdb

./buggy.py

You'll see that it fails. To debug, you can use the pdb module. We'll step into the code right before the call to foo(), lmport pdb and add pdb.set_trace() to to the line above.

....
try:
    import pdb;
    pdb.set_trace()
    foo()
...

Now run it again:

./buggy

You'll get the following prompt:

> /Users/jdubowy/tmp/tutorials/debuggingex/buggy.py(18)<module>()
-> foo()
(Pdb)

typing h <return> will give you

Documented commands (type help <topic>):
========================================
EOF    bt         cont      enable  jump  pp       run      unt
a      c          continue  exit    l     q        s        until
alias  cl         d         h       list  quit     step     up
args   clear      debug     help    n     r        tbreak   w
b      commands   disable   ignore  next  restart  u        whatis
break  condition  down      j       p     return   unalias  where

Miscellaneous help topics:
==========================
exec  pdb

Undocumented commands:
======================
retval  rv

We'll use the step/s, next/n, and p, commands. Use s to step into foo. Then, use n to step over each of the commands in foo until you reach the failure. You should see the following:

(Pdb) s
--Call--
> /Users/jdubowy/tmp/tutorials/debuggingex/buggy.py(4)foo()
-> def foo():
(Pdb) n
> /Users/jdubowy/tmp/tutorials/debuggingex/buggy.py(5)foo()
-> a = 1
(Pdb) n
> /Users/jdubowy/tmp/tutorials/debuggingex/buggy.py(6)foo()
-> b = 2
(Pdb) n
> /Users/jdubowy/tmp/tutorials/debuggingex/buggy.py(7)foo()
-> c = 1+b
(Pdb) n
> /Users/jdubowy/tmp/tutorials/debuggingex/buggy.py(8)foo()
-> d = 4-c
(Pdb) n
> /Users/jdubowy/tmp/tutorials/debuggingex/buggy.py(9)foo()
-> e = 1 - d/0.25
(Pdb) n
> /Users/jdubowy/tmp/tutorials/debuggingex/buggy.py(10)foo()
-> f = 13.23 / (e + 3)
(Pdb) n
ZeroDivisionError: 'float division by zero'
> /Users/jdubowy/tmp/tutorials/debuggingex/buggy.py(10)foo()
-> f = 13.23 / (e + 3)
(Pdb) n
--Return--
> /Users/jdubowy/tmp/tutorials/debuggingex/buggy.py(10)foo()->None
-> f = 13.23 / (e + 3)
(Pdb) n
ZeroDivisionError: 'float division by zero'
> /Users/jdubowy/tmp/tutorials/debuggingex/buggy.py(18)<module>()
-> foo()
(Pdb)

The code failed at line 10, f = 13.23 / (e + 3). The script is failing because it's deviding by zero - i.e. e + 3 == 0 and thus e == -3. We've figured out the problem, but to try variable inspection, Kill the script (ctrl-c) and repeat the above steps, this time stopping at line 10 to look at e:

...
> /Users/jdubowy/tmp/tutorials/debuggingex/buggy.py(10)foo()
-> f = 13.23 / (e + 3)
(Pdb) p e
-3.0

As we suspected, e == -3.

Other commands

You can do a lot with just s, n, p. Other useful commands are continue/c for continuing execution, and b for setting breakpoints.