optional.py 0.3.0

An implementation of the Optional object in Python

Optional.py

An Implementation of the Optional Object for Python

Why

There is a difference between None as Empty and None as the result for an Error. A common bad practice is to return None to indicate the absence of something. Doing this introduces ambiguity into you code.

For example:

thing = stuff.getSomeThing().getAnotherThing()

What will happen if the result from getSomeThing returns None? We will get an AttributeError: 'NoneType' object has no attribute 'getAnotherThing'.

What can you do to prevent these kinds of exceptions? You can write defensively:

something = stuff.getSomeThing()
if something is not None:
    thing = something.getAnotherThing()

However, if we add to our chain, you can imagine how the nesting of defensive checks adds up quickly. These defensive checks obfuscate our actual business logic, decreasing readability. Furthermore, defensive checking is an error prone process, because it is easy to forget to check a required condition.

So we present you with an Optional object as an alternative.

Install

Compatible with both python 2 and 3!

$ pip install optional.py

Usage

1. You can import it using:

   from optional import Optional
   

2. You can set it to empty:

   instead of: :scream_cat:

   return None
   

   you can do: :smile_cat:

   return Optional.empty()
   

   or

   return Optional.of()
   

3. You can set it to have content:

   instead of: :scream_cat:

   return "thing"
   

   you can do: :smile_cat:

   return Optional.of("thing")
   

4. You can check if its present:

   instead of: :scream_cat:

   if thing is not None:
   

   you can do: :smile_cat:

   thing = some_func_returning_an_optional()
   if thing.is_present():
   

5. You can check if its empty:

   instead of: :scream_cat:

   if thing is None:
   

   you can do: :smile_cat:

   thing = some_func_returning_an_optional()
   if thing.is_empty():
   

6. You can get the value:

   instead of: :scream_cat:

   print(thing)
   

   you can do: :smirk_cat:

   thing = some_func_returning_an_optional()
   ...
   print(thing.get())
   

   but this is not the recommended way to use this library.

7. You can't get the value if its empty:

   instead of: :crying_cat_face:

   if thing is None:
       print(None) # very odd
   

   you can do: :smirk_cat:

   thing = some_func_returning_an_optional()
   if thing.is_empty():
       print(thing.get()) # **will raise an exception**
   

   but this will raise an exception!

8. You can get_or_default which takes a default value:

   instead of: :crying_cat_face:

   thing = some_func_may_return_none()
   if thing is None:
        thing = '23'
   

   or: :scream_cat:

   thing = Optional.of(some_func_may_return_none())
   if thing.is_empty():
       thing = '23'
   else:
       thing = thing.get()
   

   you can do: :smirk_cat:

   thing = Optional.of(some_func_may_return_none()).get_or_default('23')
   

9. You can get_or_raise to raise any exception on abscence:

   instead of: :scream_cat:

   try:
       thing = some_func_returning_an_optional()
       return thing.get()
   except OptionalAccessOfEmptyException:
       raise MyCustomException()
   

   you can do: :heart_eyes_cat:

   return some_func_returning_an_optional().get_or_raise(
       MyCustomException()
   )
   

10. Best Usage: You can chain on presence:

    instead of: :scream_cat:

    if thing is not None:
        print(thing)
    

    you can do: :heart_eyes_cat:

    thing = some_func_returning_an_optional()
    thing.if_present(lambda thing: print(thing))
    

11. Best Usage: You can chain on non presence:

    instead of: :scream_cat:

    if thing is not None:
        print(thing)
    else:
        print("PANTS!")
    

    you can do: :heart_eyes_cat:

    thing = some_func_returning_an_optional()
    thing.if_present(lambda thing: print(thing)).or_else(lambda _: print("PANTS!"))
    

    Note that the lambdas here can be swapped out for actual function names.

12. Best Usage: You can raise on non presence:

    instead of: :scream_cat:

    if thing is None:
        raise SomeException("Boom!")
    

    you can do: :heart_eyes_cat:

    thing = some_func_returning_an_optional()
    thing.if_present(lambda thing: print(thing)).or_else_raise(SomeException("Boom!"))
    

13. Best Usage: You can map a function: :heart_eyes_cat:

    def mapping_func(thing):
        return thing + "PANTS"
    
    thing_to_map = Optional.of("thing")
    mapped_thing = thing_to_map.map(mapping_func) # returns Optional.of("thingPANTS")
    

    Note that if the mapping function returns None then the map call will return Optional.empty(). Also if you call map on an empty optional it will return Optional.empty().

14. Best Usage: You can flat map a function which already returns an Optional: :heart_eyes_cat:

    def flat_mapping_func(thing):
        return Optional.of(thing + "PANTS")
    
    thing_to_map = Optional.of("thing")
    mapped_thing = thing_to_map.map(mapping_func) # returns Optional.of("thingPANTS")
    

    Note that this does not return an Optional of an Optional. Use this for mapping functions which return optionals. If the mapping function you use with this does not return an Optional, calling flat_map will raise a FlatMapFunctionDoesNotReturnOptionalException.

15. You can compare two optionals: :smile_cat:

    Optional.empty() == Optional.empty() # True
    Optional.of("thing") == Optional.of("thing") # True
    Optional.of("thing") == Optional.empty() # False
    Optional.of("thing") == Optional.of("PANTS") # False
    

Tests

There is complete test coverage and they pass in both python 2 and 3.

Running Unit Tests

First, install the requirements in test_requirements.txt using:

$ pip install -r test_requirements.txt

You can run the tests using:

$ pytest

Test Coverage

You can check the code coverage using:

$ pytest --cov=optional test/
$ coverage report