Validate User Input in Python Function

Writing type hints for function arguments is a good start, but as the name suggests, it only hints at the data type of argument. When the functionality of the code depends on the data type of argument, then we must put validations in place.

Introduction

Type hints are great, they help you to write better code and reduce documentation required for external users or downstream software developers. However, type hints do not validate the input from the user. In my opinion, without any validation, you are playing the odds, but if you are up for a gamble then good luck to you! For the non-gamblers, we will explore how to do this with an example.

To validate type-hint or not

Code

Let us write a simple function (add) which has a single argument vals. The logic for the function is:

• Validate that the input for valsargument is a List of either int or string.
• Return a list of integer after adding 1 or list of string with period appended to the argument vals based on the user input.

You can validate the user input with a simple isinstance function. The above code works, but it's not scalable for the below reasons:

1. Have to write validations for every argument of the function.
2. Probably requires a lot of effort to rewrite, when the function argument is modified.
3. Don’t forget about the test cases required for type validations.

Tests

Positive:

Positive Test Cases

Negative:

# Test Function add for int
add(1)
# Test Function add for list of float
add([1.0])
# Test Function add for list of mixed data types
add([1, 1.0])

Error message for negative test cases

Better Way

Python’s type hints and type_extensions is changing rapidly over the last few iterations, so my recommendation is to use an external package like Typeguard for the following reasons:

1. Helps to abstract away the code for type validation with help of a decorator.
2. Provides additional functionality (way more than what we wrote or need).
3. Support for multiple Python versions have been tested.
4. Error message is detailed and easy to understand.

To replicate our example from above, we just have to add the typechecked decorator before the function definition and remove the validations for type checks.

Negative Tests:

# Test Function add_ext for int
add_ext(1)

Error message for negative test case — Int

# Test Function add_ext for list of float
add_ext([1.])

Error message for negative test case — Float

# Test Function add_ext for list of mixed data types
add_ext([1, 1.0])

Error message for negative test case — Mixed list data

Takeaway

If you are working on something serious, then don’t forget to validate user inputs.

Further Reading

20 Type Hinting Techniques and Tools for Better Python Code