ag.funutils

Fun functional utilities for python. This library provides a chain method and several operators that you can chain together to create pipelines of functional transformations, like with Lodash's chain or Clojure's threading macros.

The provided functions are "sugary". They do NOT provide performance optimizations, error handling, or guarantees of statelessness.

Each operator will take a transformation and return a function that applies that transformation to an iterable. Because most operators are wrappers around built-in functions like of the same name, they return iterables and the results of most chains will need to be converted to a list to be immediately useful. The documentation for associated standard library functions should largely be applicable to the provided operators.

Examples

Also see the tests.

from ag.funutils import fun

add_one = fun.map(lambda x: x + "1")
upper = fun.map(str.upper)

fun.chain(
    ["a", "b", "c", "d"],
    add_one,
    upper,
    list
) # => ["A1", "B1", "C1", "D1"]

big_transform = [add_one, upper]

fun.chain(
    ["a", "b", "c", "d"],
    *big_transform,
    list
) # => ["A1", "B1", "C1", "D1"])

fun.chain(
    ["a", "b", "c", "d"],
    fun.tap(print), # => "[a, b, c, d]"
    *big_transform,
    fun.tap(print), # => "[A1, B1, C1, D1]"
    fun.sort(reverse=True),
    list
) # => ["D1", "C1", "B1", "A1"]

fun.chain(
    ["a", "b", "c", "d"],
    *big_transform,
    fun.reduce(lambda acc, x: acc + x)
) # => "A1B1C1D1"

# Values that are tuples will be spread into the transformations,
# which lets you work with dicts.

data = {
    'beep': 1,
    'boop': 2,
    'buup': 3,
}

add_one = fun.map(lambda k, v: (k, v + 1))
evens = fun.filter(lambda k, v: v % 2 == 0)
beep_buup = fun.reduce(lambda acc, k, v: f'{acc}{k}{v}', '')

result = fun.chain(
    data.items(),
    add_one,
    evens,
    add_one,
    beep_buup
) # => 'beep3buup5'

Reference

• ag.funutils.chain(data, *transforms):
  • Provides data as an argument to the first transform, then the result of each transform to the following one. Each transform should be a function that takes a single, iterable argument and returns an iterable. The exception is the reduce operator, which can return a single value if it is the last operator in the chain.
• ag.funutils.map(transform):
  • returns a function which takes an iterable and applies transform to each item of the iterable
• ag.funutils.filter(condition):
  • returns a function which takes an iterable and returns an iterable with only items matching the condition
• ag.funutils.sort(key=None, reverse=False):
  • returns a function which takes an iterable and return an iterable sorted according to the value returned by the key function. Compares items directly by default.
• ag.funutils.reduce(transform, initial=None):
  • returns a function which takes an iterable and reduces it to a single value. If no initial value is provided, the first item is used as the initial value.
• ag.funutils.tap(fn):
  • returns a function which takes a single argument, passes it to fn, and returns it. Intended for debugging purposes, in particular: ag.funutils.tap(print).

Development

Requires pipenv and python 3.7.

$ ./scripts/setup.sh
$ ./scripts/test.sh