knife Documentation¶

knife is a powerful Python multitool loosely inspired by Underscore.js but remixed for maximum pythonicity.

knife concentrates power that is normally dispersed across the entire Python universe in one convenient shrink-wrapped package.

Vitals¶

knife works with CPython 2.6, 2.7, 3.1. and 3.2 and PyPy 1.8.

knife documentation is at http://readthedocs.org/docs/knife/en/latest/ or http://packages.python.org/knife/

Installation¶

Install knife with pip...:

$ pip install knife
[... possibly exciting stuff happening ...]
Successfully installed knife

...or easy_install...:

$ easy_install knife
[... possibly exciting stuff happening ...]
Finished processing dependencies for knife

...or old school by downloading knife from http://pypi.python.org/pypi/knife/:

$ python setup.py install
[... possibly exciting stuff happening ...]
Finished processing dependencies for knife

Development¶

Public repository: https://bitbucket.org/lcrees/knife.

Mirror: https://github.com/lcrees/knife/

Issue tracker: https://bitbucket.org/lcrees/knife/issues

License: BSD

3 second knife¶

Things go in:

>>> from knife import __
>>> gauntlet = __(5, 4, 3, 2, 1)

Things get knifed:

>>> gauntlet.initial().rest().slice(1, 2).last()
knife.lazy.lazyknife ([IN: ([3]) => WORK: ([]) => HOLD: ([]) => OUT: ([3])])

Things come out:

>>> gauntlet.get()
3

Slightly more knife¶

knife has 40 plus methods that can be chained into pipelines...

contrived example:¶

>>> __(5, 4, 3, 2, 1).initial().rest().slice(1, 2).last().get()
3

...or used object-oriented style.

contrived example:¶

>>> from knife import knife
>>> oo = knife(5, 4, 3, 2, 1)
>>> oo.initial()
knife.active.activeknife ([IN: ([5, 4, 3, 2, 1]) => WORK: ([]) => HOLD: ([]) => OUT: ([5, 4, 3, 2])])
>>> oo.rest()
knife.active.activeknife ([IN: ([5, 4, 3, 2]) => WORK: ([]) => HOLD: ([]) => OUT: ([4, 3, 2])])
>>> oo.slice(1, 2)
knife.active.activeknife ([IN: ([4, 3, 2]) => WORK: ([]) => HOLD: ([]) => OUT: ([3])])
>>> oo.last()
knife.active.activeknife ([IN: ([3]) => WORK: ([]) => HOLD: ([]) => OUT: ([3])])
>>> oo.get()
3

A knife object can roll its current state back to previous states like snapshots of immediately preceding operations, a baseline snapshot, or even a snapshot of the original arguments.

contrived example:¶

>>> undone = __(1, 2, 3).prepend(1, 2, 3, 4, 5, 6)
>>> undone.peek()
[1, 2, 3, 4, 5, 6, 1, 2, 3]
>>> undone.append(1).undo().peek()
[1, 2, 3, 4, 5, 6, 1, 2, 3]
>>> undone.append(1).append(2).undo(2).peek()
[1, 2, 3, 4, 5, 6, 1, 2, 3]
>>> undone.snapshot().append(1).append(2).baseline().peek()
[1, 2, 3, 4, 5, 6, 1, 2, 3]
>>> undone.original().peek()
[1, 2, 3]

knife knives come in two flavors: active and lazy. active.knife objects evaluate the result of calling a method immediately after the call. Calling the same method with a lazy.knife object only yields results when it is iterated over or knife.lazy.lazyknife.get is called to get results.

knife.lazy.lazyknife combines all knife methods in one knife class:

>>> from knife import lazyknife

It can be imported under its dunderscore (knife.__) alias.

>>> from knife import __

knife.active.activeknife also combines every knife method in one combo knife class:

>>> from knife import activeknife

It can be imported under its knife.knife alias:

>>> from knife import knife

knife methods are available in more focused classes that group related methods together. These classes can also be chained together into pipelines.

contrived example:¶

>>> from knife.active import mathknife, reduceknife
>>> one = mathknife(10, 5, 100, 2, 1000)
>>> two = reduceknife()
>>> one.minmax().pipe(two).merge().back().min().get()
2
>>> one.original().minmax().pipe(two).merge().back().max().get()
1000
>>> one.original().minmax().pipe(two).merge().back().sum().get()
1002

knife documentation