Originally this was called fmap and was only an implementation of the fmap function for Python3. Now that I've been tinkering around some more I thought it'd be nice to collect a few things together. (For about a day it was then called PythFun but then I saw the error of my ways...)
- fmap: apply a function to all elements of a collection style object.
- pattern_match: A hybrid of Haskell's pattern matching and Clojure's destructuring.
- dispatch: single and multiple dispatch for your Python functions.
- prelude: a collection of functional programming functions.
Any suggestions for improvements are welcome and if you'd like to hack away and submit a pull request for a feature then raise an issue and let me know!
I hope you enjoy!
def fibgen():
'''Because EVERYONE needs a list of Fibonacci numbers...'''
yield 1
yield from iscanl(fibgen(), add, 2)
>>> take(20, fibgen())
[1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, 233, 377, 610, 987, 1597, 2584, 4181, 6765, 10946]
@tcall
def fact(n, acc=1):
'''And compute some factorials...'''
if n == 0:
return acc
else:
return fact, (n-1, acc*n)
>>> fact(99)
9332621544394415268169923885626670049071596826438162146859296389521759999322991560894146397
61565182862536979208272237582511852109168640000000000000000000000
>>> import sys
>>> sys.getrecursionlimit()
2000
>>> fact(9999)
284625968091705451890641321211986889014805140170279923079417999427...
# This one is 35656 digits long...
# From prelude.py
def cmap(func, col):
'''
Concat-Map: map a function that takes a value and returns a list over an
iterable and concatenate the results
'''
return foldl(map(func, col))
def flatten(col):
'''
Flatten an arbitrarily nested list of lists into a single list.
'''
if not iscollection(col):
return [col]
else:
return cmap(flatten, col)
>>> l = [1,2,[3,4,[5,6,7],[8,9]],[10,11,12],13]
>>> flatten(l)
[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]