[Python-Dev] Candidate Itertools

[Raymond Hettinger]

Feedback is requested for two prospective itertools:

def count_elements(iterable):
    b = {}
    for elem in iterable:
        b[elem] = b.get(elem, 0) + 1
    return ((cnt, elem) for elem, cnt in b.iteritems())


def pairswap(iterable):
    return ((b,a) for a,b in iterable)



The first is designed to provide item counts for min(), max(), sorted(),
nlargest(), and nsmallest().  It accepts any iterable including
generator expressions:

>>> from heapq import nlargest
>>> words = (word  for line in open('mystory.txt')  for word in
line.split())
>>> nlargest(count_elements(words), 3)
[(200, 'super'), (129, 'hero'), (103, 'villain')]



The second handles reordering of fields in tuples of length two.  It
works well with dict(), enumerate(),  dict.iteritems(),
itertools.starmap() and anything else that accepts or returns a series
of pairs:

>>> list(pairswap(_))
[('super', 200), ('hero', 129), ('villain', 103)]

>>> inversemap = dict(pairswap(forwardmap.iteritems()))

>>> maptoposition = dict(pairswap(enumerate(mysequence)))



Raymond


P.S.  If you have naming suggestions other than pairswap() and
count_elements(), I'm all ears.