Storing Python Program Execution Flow (Function call Flow)

Question

I am developing a project in which I have to store all the function that were called in each request-response cycle and store them. I do not need to store values of the variable, all I need to store is function that we were called with their parameters and their order in execution. I am using mongodb to store this trace.

Answer 1

You could use a function decorator for convenience.

import functools
import logging

def log_me(func):
    @functools.wraps(func)
    def inner(*args, **kwargs):
        logging.debug('name: %s, args: %s, kwargs: %s', func.__name__, args, kwargs)                                        
        return func(*args, **kwargs)
    return inner

Then decorate your function(s) to log.

@log_me
def test(x):
    return x + 2

Test call.

In [10]: test(3)
DEBUG:root:name: test, args: (3,), kwargs: {}
Out[10]: 5

If you wanted to store the entries in MongoDB directly instead of first logging to the logging module you can replace the logging.debug line with code that creates an entry in your database.

Answer 2

sys.settrace traces a function for debugging but can be modified for this problem. Maybe something like this -

import sys

def trace_calls(frame, event, arg):
    if event != 'call':
        return
    co = frame.f_code
    func_name = co.co_name
    if func_name == 'write':
        # Ignore write() calls from print statements
        return
    func_line_no = frame.f_lineno
    func_filename = co.co_filename
    caller = frame.f_back
    caller_line_no = caller.f_lineno
    caller_filename = caller.f_code.co_filename
    print 'Call to %s on line %s of %s from line %s of %s' % \
        (func_name, func_line_no, func_filename,
         caller_line_no, caller_filename)

Also see profiling. It describes how often and for how long various parts of the program executed