Hacker News Reader

Python lists are array-backed, Lisp lists are singly-linked lists. Appending to the front of an array list is O(n)--you don't want to do it. Appending to the back is constant time, but is an in-place operation and doesn't return anything, which is less than ideal for FP purposes.

That's the problem. I can fake a functional paradigm within Python in some settings, but it falls apart at the seams.

You can always use 2-tuples to get lists in the Lisp sense. You'd add an item to the head of a list by just pairing it:

    lst    = (2, (3, ()))
    newlst = (1, lst)

Toss in some small helper functions and they become less unwieldy. You could pretty easily write an iterator for processing these lists:

    class List:
      def __init__(self, lst):
        self.cursor = lst
      def __iter__(self):
        return self
      def next(self):
        if self.cursor == ():
          raise StopIteration
        else:
          head = self.cursor[0]
          self.cursor = self.cursor[1]
          return head

Then you can do stuff like:

    >>> for i in List(newlst):
    ...   print i
    ...
    1
    2
    3
    >>> map(lambda x: x + 1, List(newlst))
    [2, 3, 4]

    One thing I love about Lisp is that it's
    ridiculously easy to write (cons item some-list),
    or even (cons item1 (cons item2 some-list)).

Iterators can help (and in O(1)):

    >>> from itertools import chain
    >>> some_list = range(5)
    >>> item = 99
    >>> a = chain([item], some_list)
    >>> for x in a:
    ...     print x
    ...
    99
    0
    1
    2
    3
    4

Yes. That's the hole point of the article, and matches very well with my understanding of SICP: functional programming is possible only if you model changing state in streams, which are just called iterators in python. All the fuss about python missing some parts too be really functional seem to have missed the point to me.

namedtuple is awesome, until you realise it works internally by using eval and generating the source code of a class.

And Lisps work internally by generating lists and evaluating those, but you don't see me complaining about macros!

In all seriousness, though, if I really wanted a functional implementation, I wouldn't be using Python. It's more just nice to know that I can create an object that's immutable and which functions more or less like a CLOS object, so that I can think in that mindset instead. Also, I believe it doesn't come with all of the hidden functions that it would otherwise.

In the end, if I really cared, I'd actually write a wrapper around the type() function.

namedtuple is still awesome. The fact that it's eval'ing a string to initialize it doesn't make that any less the case.

However, I'll agree that it would be disingenuous to say it's not a little peculiar, and surprising to see the source for it :)

Why does it do that, instead of some other more pythonic way?

See also http://bugs.python.org/issue3974, where someone submits a patch for namedtuple that doesn't use exec. The maintainers reject it as less clear and probably slower. I agree there is "nothing unholy in using exec", though I found this weird at first too.

if you want to append to fronts and backs of lists, you should use deque: http://docs.python.org/library/collections.html#collections....

pronounced "deck"

So it is a doubly linked list, its name is deque, and its name is pronounced "deck". The levels of indirection here sound kind of like something Lewis Carroll might write.

The name means double-ended queue. That's the reason behind the strange pronunciation.

'deque' is an irregular abbreviation for double-ended queue, and "deck" is a convenient metaphor for visualising the container - you can put things at either end of a deck of cards. c.f. std::deque<>

Maybe there's no way in the standard library, and it may be a valid point that it should be there, but you could trivially define your own 'cons' function and write cons(item, someList)

How about [item] + [some, list]?

That works but it's kind of a pain that the natural pythonic syntax: ['a'].extend(some-list) doesn't return a list.

that is not the natural pythonic syntax.

for concatenating lists, "+" is standard

Thanks - looks like plus operator concat was in there in the beginning as well:

http://docs.python.org/release/1.4/lib/node10.html#SECTION00...

Hmm, I forgot that '+' was overloaded for lists in Python - nevermind that last statement, then.

The rest still holds, though - not all functions have a mutable and immutable counterpart (like sort) and map() returning lists makes them rather unwieldy. I think they fixed this in Python 3, though I'm not sure if it was backported to 2.7.

There's in-place .sort() and the functional sorted(), which has been there since at least 2.6. There's no in-place map, but it would be a one-line function (and you could use map as an in-place operator if you use a function for its side-effects). In the case of + there's the in-place counterpart += (same as extend for lists).

So I don't really get your concern, plus mutability is a property of data-structures, not of functions.

> There's no in-place map

It's not a matter of in-place mapping, but if I remember correctly, map() returns a generator in Python3, which has the effect of simulating single-traversal and lazy evaluation like Haskell does.

> mutability is a property of data-structures, not of functions

True, but in a (purely) functional language, if all data structures are immutable and functions are simply mappings of input values to output values with no side-effects, that's inconsequential. We're talking about simulating a functional paradigm within a non-functional language with mutable data structures, so writing functions without side-effects would be more idiomatic - hence the problems with trying to write Python in a functional style.

If Python were a non-functional language, as you claim, it wouldn't have map reduce and functools. Python is not a purely functional language, it is a multi paradigm language. This difference is the main point here.

No, that would go against the philosophy of Python: only one [obvious] way to do it. Lambda, map, and reduce were all components that Guido very much wanted to remove, as he felt they had no place in Python.

Python is an object-oriented language, and the fact that it is a more purely object oriented language than (for example) Java allows one to use functions as first-class objects in ways that appear to be consistent with a functional paradigm, but that doesn't mean that the language supports a functional approach. That would require support by the interpreter for some key

Put another way, what's the difference between the following?

> [f(x) for x in xrange(10)]

> map(f, range(10))

Without knowing the difference in the underlying implementation, you can't really be sure - it's possible that map() is just a syntactic alternative to a list comprehension, and you could write an implementation of Python that did just that. If we're talking about CPython, though (which we are), then we have to know that the latter is using a function object that is retrieved once and called multiple times, and the former is using a function object that is retrieved ten times. In the latter case, we have a function object that contains a bound method which receives the implicit parameter 'self' and an integer within the range specified, and this method call is what is, in idiomatic Python, producing the values that are then used to create the list. Also, xrange is using a generator, which requires several function calls in itself, whereas range creates a list (which is handled differently).

That isn't necessarily so inconsistent with a functional paradigm, but my point is that the implementation is more important than the function names - simply saying that map() and reduce() exist doesn't mean that Python is a multiparadigm language. Python is incredibly eager-evaluating and it does not optimize tail recursion, two things which make it hard to argue that Python supports a truly functional style.

As noted by another commenter, this is not an accident; while you can fake functional paradigms syntactically with Python's extensive implementation of first-class objects (including functions), the explicit goal is to discourage use of recursion, as well as other functional ways of thinking, in favor of iteration, and other imperative/OO ways of thinking.