forbiddenfp 0.5.3

Forbidden Functional Programming in Python.

forbiddenfp

Summary

Functional-Programming (FP) in a forbidden way:

Turn arbitrary function into postfix notation in favor of function chaining.

And the library provides many useful patches for builtin/itertools functions.

Install

pip install forbiddenfp

Examples

# objects are already patched at import time
import forbiddenfp

"abc".print().len()  # print out "abc", then return 3
"abc".then(lambda s: s * 2).filter(lambda s: s == "b")  # "bb"

# A more complex one (examples/word_count.py)
print("./lorem_ipsum.txt"
      .with_context(open, lambda path, f: f.read().also(print(f"Reading {path}")))
      .then(lambda s: s.split(" "))
      .counter())

See more ./examples.

Why Functional Programming

Separate out control structs (which are functions provided by library) from business logic (which are lambda functions supplied to fill the control structs).

So we can have a clearer scope, and separate concerns, when we want to change on either side of things.

• if ... and if ... else ... are modeled by Maybe/Either monads.
• while ... and for ... in ... are generator/iterators. Additionally, stop-early behavior is takewhile of the sequence.
• I/O operations and side-effects are I/O monads.

Philosophically, think more in def transform(old_state) -> new_state, rather than state = modify(state).

Warning

This library patches builtin object (and hence ALL classes), with hacks around CPython API (provided by forbiddenfruit),

so consider this NSFW (Not Safe For Work).

Known Issues

None doesn't work well with chained keyword arguments.

import forbiddenfp

None.apply(print)  # works
None.apply(func=print)  # doesn't work