Given a dictionary, find words for which all rotations occur in the dictionary.

draaideur.py

import sys, collections
def rotations(a):
    return {a[x:] + a [:x] for x in range(1, len(a))}

lexicon = collections.defaultdict(set)
for a in open(sys.argv[1]):
    lexicon[len(a) - 1].add(a.strip())
for length in sorted(lexicon):
    if length == 1:
        continue
    while lexicon[length]:
        word = lexicon[length].pop()
        if rotations(word).issubset(lexicon[length]):
            lexicon[length].difference_update(rotations(word))
            lexicon[length].difference_update({word})
            print word, rotations(word)

output.txt

$ python draaideur.py /usr/share/dict/american-english
ha set(['ah'])
ti set(['it'])
ma set(['am'])
em set(['me'])
eh set(['he'])
um set(['mu'])
no set(['on'])
oh set(['ho'])
spa set(['pas', 'asp'])
eat set(['tea', 'ate'])

$ python draaideur.py /usr/share/dict/dutch
BG set(['GB'])
oh set(['ho'])
la set(['al'])
ra set(['ar'])
dc set(['cd'])
af set(['fa'])
ge set(['eg'])
SP set(['PS'])
ei set(['ie'])
IX set(['XI'])
es set(['se'])
er set(['re'])
ms set(['sm'])
MA set(['AM'])
op set(['po'])
is set(['si'])
ha set(['ah'])
as set(['sa'])
ai set(['ia'])
ik set(['ki'])
nee set(['een', 'ene'])
ere set(['eer', 'ree'])
ast set(['sta', 'tas'])