Friday, 30 December 2016

Testing Python Code

Testing your code is very important to know if your code is having the expected behavior. No one likes a software program which would give a weird result or a program which would break in the middle of a critical task and throw of ugly error codes. For this and many more reasons it’s very important to test you code before making it public, testing is not a “best practice” rather it’s a necessity to deliver clean and delicious code which will obey all the commands that the user provides.
There are many different python modules like unittest that help you to test units of your code and then there are some tools like py.test that helps you achieve the same goal, but I want to point out a simple & fast way to quickly test your piece of code without importing any separate module other than the one that is included in your project or which you have created for testing your code.
Testing code in isolation
In every type of code you must first design your code in modularised way to make debugging and maintenance easier . Let me show you one example -
Consider the following code -
  1. def func_example1 (a,b):
  2.   return a+b
  3. def func_example2 (a,b):
  4.   return a*b
  5. def func_main(func_example1(a,b), func_example2(a,b)):
  6.   result = func_example1 + func_example2
  7.   return result
  8. func_main(func_example1(1,2), func_example2(3,4))   
The above pseudo code is modularised meaning that a specific feature is divided into different functions like func_example1, func_example2 & func_main. Now since the code is divided into different functions so we can test each function separately and to do this it’s better if a separate “test module” is created where we can invoke a particular function under test with mock objects
to check that the function is really performing the intended action. Let’s consider the above pseudo code example, I will place the above code in a module named as “main_module” (you can name it as you like). I will create another module and name it as “test_module” & import the “main_module” into that -
  1. import main_module
  2. # todo - call function to be tested
We have our module ready now so we can test each function individually and for this we will pass mock objects to the function under test by invoking the function from the main_module into the test module & recording its output.
  1. import main_module
  2. output = main_module.func_example1 (1,2)
  3. print("the output of func_example1 is" + output)
I invoked only the func_example1 from the main_module and passed the objects ‘1’ & ‘2’ (which may be different from the actual data) & I could now record & test the behavior of this function in isolation which will help me to zero in on any possible bug or identify any improvement needed in that particular block of code.
Brute force method to Test the program end to end
  1. def func_example1 (a,b):
  2.   Print("1st function is now invoked")
  3.   return a+b
  4. def func_example2 (a,b):
  5.   Print("2nd function is now invoked")
  6.   return a*b
  7. def func_main(func_example1(a,b), func_example2(a,b)):
  8.   Print("main function is now invoked")
  9.   result = func_example1 + func_example2
  10.   return result
  11. func_main(func_example1(1,2), func_example2(3,4))
In the above pseudo code I have inserted simple print statements inside every function so that each time that function is invoked anywhere the statement inside the print () will be displayed on the screen along with the regular return value or action of that function which can help us test if the desired function/code block is invoked and if the control is traversing the code in the expected manner.
Using print statements like above may be a brute force way or may be a novice way to test the control flow but it’s simple & is very useful way to quickly devise a system so that you can track your program flow. I am not saying that this is the only way or it is the best way to do things but it’s one of the practical way to test your code in a short period of time and it’s more visual. If you hate the “print” statements then you can create another function which will write the output message into a log file which can be later analysed to check if your program flow is as expected or you can refer the log in case you have a break in your code. So, folks this was all for this week but more stuffs are coming your way next week, till then happy coding.
If you found this post helpful and you have any suggestion to improve my blog then do post your comments below. Additionally you can also bug me on Quora. You can also find me on Twitter or Facebook. So, why wait go ahead and start exploring.

If you wanna get started in learning Python then why not get started with the python basics chapter. Check out this link to access the Python Basic Chapter for free.

If you want to download a copy of the chapter then purchase the chapter by clicking on the following link.

No comments:

Post a Comment