Functional based library to support monads and other functional programming concepts


Keywords
functional-programming, inmutable, lambdas, monads, python
License
MIT
Install
pip install pynction==0.4.1

Documentation

Pynction 🐍

continuous_integration codecov

Functional based library to support haskell monads like Either, Maybe in a scala fashion style. The library also contains Try monad inspired from vavr and a Stream class which is pretty similar to scala and java stream API

Inspired in: VΛVR

Why should you use it ?

Probably if you have reached this library you already know something about functional programming and Monads. Well this library is another one that empowers your imperative code to start using functional programming concepts. This type of programming makes your code declarative as long as give you support to the most famous monads like Maybe and Either. These monads make your interfaces explicit for error handling so paraphrasing If it compiles, it works this time it is If mypy is happy, your code works

Basic examples

Stream examples

from pynction import stream_of, stream


foo = (
    stream_of([1, 2, 3, 4])
    .map(lambda a: a + 1)
    .filter(lambda n: n % 2 == 0)
    .flat_map(lambda n: [n, n * 2])
    .to_list
)

# foo => [2, 4, 4, 8]

bar = (
    stream("example", "e", "something")
    .take_while(lambda s: s.startswith("e"))
    .to_list
)

# bar => ["example", "e"]

Maybe examples

from pynction import maybe, nothing

def divide_10_by(n: int) -> Maybe[int]:
    if n == 0:
        return nothing
    return maybe(10 / n)

result = divide_10_by(2).get_or_else_get(-1)
# result => 5
result = divide_10_by(0).get_or_else_get(-1)
# result => -1

Try examples

from pynction import try_of


def add_10(n: int) -> int:
    if n > 10:
        raise Exception("n must be less than 10")
    return n + 10

try_example_1 = try_of(lambda: add_10(11)).map(lambda a: a + 1)
try_example_1.on(
    on_success=lambda a: print(f"Result: {a}"),
    on_failure=lambda e: print(f"Error: {e}"),
)
# ==> Will print "Error: n must be less than 10"

try_example_2 = try_of(lambda: add_10(9)).map(lambda a: a + 1)
try_example_2.on(
    on_success=lambda a: print(f"Result: {a}"),
    on_failure=lambda e: print(f"Error: {e}"),
)
# ==> Will print "Result: 20"

Either examples

from pynction import left, right, Either


LESS_THAN_10_LETTERS = Literal["LESS_THAN_10_LETTERS"]
GREATER_THAN_100 = Literal["GREATER_THAN_100"]
WordTransformationError = Literal[LESS_THAN_10_LETTERS, GREATER_THAN_100]

def make_upper_case_first_n_letters(word: str, number: int) -> Either[WordTransformationError, str]:
    if len(word) < 10:
        return left("LESS_THAN_10_LETTERS")
    elif number > 100:
        return left("GREATER_THAN_100")
    else:
        return right(word.upper()[0:number])

result = make_upper_case_first_n_letters("example", 10)
print(result) # ==> Will be Left("LESS_THAN_10_LETTERS")

API

Stream

Factory methods

  1. stream(*args: T) -> Stream[T]
  2. stream_of(elems: Iterable[T]) -> Stream[T]

Stream api

  1. map(f: Callable[[T], S]) -> Stream[S]
  2. filter(condition: Callable[[T], bool]) -> Stream[T]
  3. flat_map(f: Callable[[T], Iterable[S]]) -> Stream[S]
  4. take_while(condition: Callable[[T], bool]) -> Stream[T]
  5. to_list() -> List[T]
  6. to_set() -> Set[T]

Maybe

Factory methods

  1. maybe(elem: Optional[T]) -> Maybe[T]
  2. just(elem: T) -> Just[T]
  3. nothing: It is a global instance of Nothing

Maybe api

  1. is_empty -> bool
  2. map(f: Callable[[T], V]) -> Maybe[V]
  3. flat_map(f: Callable[[T], Maybe[V]]) -> Maybe[V]
  4. get_or_else(default: T) -> T
  5. to_either(error: L) -> Either[L, T]

Either

Factory methods

  1. right(elem: T) -> Either[Any, T]
  2. left(elem: T) -> Either[T, Any]

Either api

  1. is_left -> bool
  2. is_right -> bool
  3. map(f: Callable[[R], R1]) -> Either[L, R1]
  4. filter_or_else(predicate: Callable[[R], bool], leftValue: L) -> Either[L, R]
  5. get_or_else_get(f: Callable[[L], R]) -> R

Try

Factory methods

  1. try_of(f: Callable[[], T]) -> Try[T]

Try api

  1. map(f: Callable[[T], S]) -> Try[S]
  2. flat_map(f: Callable[[T], Try[S]]) -> Try[S]
  3. get_or_else_get(self, default: Callable[[Exception], T]) -> T
  4. on( on_success: Callable[[T], None] = None, on_failure: Callable [[Exception], None] = None ) -> Try[T]
  5. catch(f: Callable[[Exception], T]) -> Try[T]
  6. and_finally(f: Callable[[], None]) -> Try[T]
  7. to_either() -> Either[Exception, T]

Do notation

The following functions are python decorators that enables your decoratee function to use the do notation way of programming like Haskell or Scala. To better understand this decorators let's take a look at 2 quick examples

  1. do_either
@do_either
def example_with_union() -> DoEither[str, Union[int, str], str]:
    name = yield get_eihter_name()
    age = yield right(25)
    lastname = yield right("wick")
    return f"{name} {lastname} with age {age}"

result: Either[str, str] = example_with_union()
  1. do_maybe
@do_maybe
def do_maybe_example() -> DoMaybe[Union[str, int], str]:
    name = yield get_maybe_name()
    age = yield get_maybe_age()
    return f"{name} {age}"

result: Maybe[str] = do_maybe_example()