Skip to content

Latest commit

 

History

History
183 lines (130 loc) · 6.44 KB

README.md

File metadata and controls

183 lines (130 loc) · 6.44 KB
Raw

Oneliner-izer

Build Status

Convert any Python 2 script into a single line of code.

No newlines allowed. No semicolons, either. No silly file I/O tricks, or eval or exec. Just good, old-fashioned λ.

Play with a live demo at onelinerizer.com! Watch the presentation at PyCon 2016, or peruse the slide deck.

User Installation and Usage

Install via pip from PyPI:

$ pip install onelinerizer

Use either the command line function or the Python module:

$ echo "def f(x):\n    print x\nf(4)" > sandbox.py
$ onelinerizer sandbox.py --debug
$ onelinerizer sandbox_ol.py
from onelinerizer import onelinerize
onelinerize("def f(x):\n    print x\nf(4)")

Examples

Before:

x = 3
y = 4
print (x < y < 5)

After:

(lambda __builtin__: (lambda __print, __y, d: [[__print(d.x<d.y<5) for d.y in [(4)]][0] for d.x in [(3)]][0])(__builtin__.__dict__['print'],(lambda f: (lambda x: x(x))(lambda y: f(lambda *args: y(y)(*args)))),type('StateDict',(),__builtin__.__dict__)()))(__import__('__builtin__'))

That line looks complicated, because we need some tricks to import the print function and to support certain tricks which are needed for more complicated features such as while and if. For a program as simple as this one, though, you can think of it as working this way:

(lambda x: (lambda y: print(x<y<5))(4))(3)

Before:

def f(x):
    return x+5
print f(13)

After:

(lambda __builtin__: (lambda __print, __y, d: [__print(d.f(13)) for d.f in [(lambda x:[(d.x+5) for d.x in [(x)]][0])]][0])(__builtin__.__dict__['print'],(lambda f: (lambda x: x(x))(lambda y: f(lambda *args: y(y)(*args)))),type('StateDict',(),__builtin__.__dict__)()))(__import__('__builtin__'))

...or, if you want to think about something more simplified:

(lambda f: print(f(13)))(lambda x: x+5)

Before:

def guess_my_number(n):
    while True:
        user_input = raw_input("Enter a positive integer to guess: ")
        if len(user_input)==0 or not user_input.isdigit():
            print "Not a positive integer!"
        else:
            user_input = int(user_input)
            if user_input > n:
                print "Too big! Try again!"
            elif user_input < n:
                print "Too small! Try again!"
            else:
                print "You win!"
                return True
guess_my_number(42)

After:

(lambda __builtin__: (lambda __print, __y, d: [(lambda ___: None)(d.guess_my_number(42)) for d.guess_my_number in [(lambda n:[(__y(lambda __this: (lambda d: (lambda __after: [(lambda __after: (lambda ___: __after(d))(__print('Not a positive integer!')) if (d.len(d.user_input)==0 or (not d.user_input.isdigit())) else [(lambda __after: (lambda ___: __after(d))(__print('Too big! Try again!')) if d.user_input>d.n else (lambda __after: (lambda ___: __after(d))(__print('Too small! Try again!')) if d.user_input<d.n else (lambda ___: d.True)(__print('You win!')))(lambda d: __after(d)))(lambda d: __after(d)) for d.user_input in [(d.int(d.user_input))]][0])(lambda d: __this(d)) for d.user_input in [(d.raw_input('Enter a positive integer to guess: '))]][0] if d.True else __after(d))(lambda d: None))))(d) for d.n in [(n)]][0])]][0])(__builtin__.__dict__['print'],(lambda f: (lambda x: x(x))(lambda y: f(lambda *args: y(y)(*args)))),type('StateDict',(),__builtin__.__dict__)()))(__import__('__builtin__'))

FAQ

Oh dear God why?

Yeah. I'm sorry. But on the other hand, why not?

Can't you use semicolons?

That would be against the spirit of this exercise. Why pass up a perfectly good excuse to abuse lambda functions, ternary expressions, list comprehensions, and even the occasional Y combinator? Never pass up an opportunity to use the Y combinator.

Analysis

Output program size

O(n). No code is ever duplicated, so the one-lined code produced is linear in the size of the input code.

Runtime of one-lined code

I have no reason to believe the resulting code, however absurd, is more than a constant factor slower than the original code. Since while loops, etc. are implemented using recursive function calls, causing more overhead for setting up those function calls, the constant factor is likely to be somewhat bad.

Tips

The one-lined code tends to contain many nested lambdas; if there are too many, Python will refuse to run it.

$ python main_ol.py
s_push: parser stack overflow
MemoryError

This can be fixed using pypy.

$ pypy main_ol.py

However, since while loops and for loops are implemented with recursion, you might encounter maximum recursion depth exceeded errors during runtime if your loops go on for too long.

To get around this, you can put

import sys
sys.setrecursionlimit(new_limit)

in your original Python code. (onelinerizer will not place this command in for you.)

Not Implemented

  • from foo import *

Open Problems

  • with
  • yield

Developer Installation and Testing

$ git clone https://github.com/csvoss/onelinerizer
$ cd onelinerizer
$ python -m onelinerizer .setup.py setup.py
$ python setup.py test

To install the local module:

$ pip install .

Further Reading