The Python Multiprocess Debugger mPdb based on Pdb

This Debugger can be used in both Python2 and Python3.

To use the debugger in its simplest form:

 1 import numpy
 2 import flask
 3 import tensorflow
 4
 5 a = 2
 6 b = 3
 7 c = 7
 8 import mpdb
 9 mpdb.set_trace()
10 for i in range(9): # set breakpoint here, just like pdb.set_trace()
11     a += i
10     b += 2*i
11     c = a + b

running mpdb_cmd.py in another terminal after running the multiprocess code you want to debug, just like below:

python mpdb_cmd.py

or

python3 mpdb_cmd.py

or

python2 mpdb_cmd.py

depend on your own system setting.

Other operations the same as Pdb.

Note: the running order doesn't matter, running mpdb_cmd.py first is also ok, but mpdb_cmd.py must be in the directory where the main program Python file which you want to debug is, so recommend use 'mv' command to move mpdb_cmd.py if you need change main program which you want to debug.

Installation

Running install.py like below:

python3 install.py

or

python install.py

or

python2 install.py

depend on your own system setting. Actually, running install.py is finding in which sys.path pdb.py is, and then copy mpdb.py to the path.

Example

There is a test directory in current directory, a test_code.py is in test directory.

1. copy mpdb_cmd.py to test directory.
2. run test_code.py or mpdb_cmd.py in current terminal, which one doesn't matter, just choose one.
3. open another terminal, change to test directory and run the mpdb_cmd.py if you run the test_code.py in previous step, vice versa.
4. if succeed, you will see (Pdb) prompt in the terminal you run the mpdb_cmd.py.