funcy-chain

Simple Python data processing using method-chaining style and the funcy library.

>>> users = [
...   { 'first_name': 'barney',  'age': 36 },
...   { 'first_name': 'fred',    'age': 40 },
...   { 'first_name': 'pebbles', 'age': 1 }
... ]
>>> (Chain(users)
...     .sort(key=itemgetter("age"))
...     .map(itemgetter("first_name"))
...     .thru(lambda names: names + ["wilma"])
...     .map(str.capitalize)
... ).value
['Pebbles', 'Barney', 'Fred', 'Wilma']

Why?

• Method chaining can make some multi-stage data processing easier to read and write.
• Funcy is great, but doesn't support method chaining.
• Other method-chaining implementations have too much "magic", are overly complex, and support too many ways of doing one thing. (see: pydash, fluentpy)
• Being simple and straightforward is a great boon, making it easier to reason about code, debug, etc.

Key Features

• Useful out of the box, supporting built-in, stdlib and funcy functions (see below).
• Easy to compose with additional utilities (see Chain.thru()).
• Both "immediate" (list-based) and "lazy" (iterator-based) computations supported, via parallel Chain and IterChain classes.

>>> from funcy_chain import Chain
>>> (Chain([1, 2, 3, 7, 6, 5, 4])
...     .without(3)
...     .filter(lambda x: x > 2)
...     .remove(lambda x: x > 6)
...     .sort(reverse=True)
... ).value
[6, 5, 4]

Supported Methods

built-in

• enumerate
• max
• min
• reduce
• reverse
• slice
• sort
• sum
• zip
• dict.items
• dict.keys
• dict.values
• dict.update

Note: funcy's map and filter are used rather than the built-ins.

stdlib

• itertools.starmap
• itertools.zip_longest
• heapq.nlargest
• heapq.nsmallest
• random.choice
• random.choices
• random.sample
• random.shuffle

funcy

The following is a sub-set of the funcy cheat sheet including only functions supported as methods on Chain and IterChain objects.

Sequences

Collections