Wordcount code in python

mapとかlambdaを駆使してみた。やりすぎかもしれない。ポイントはファイルの各行を別々にsplitしたものをマージするタイミングかもしれない。ここでは、一気に全部やってしまっている。

import string, sys, re

def wordcount_str(line):
    words = filter( None, #same as lambda x: x,
                    re.compile(r'\W+').split(line) )#line.split()) # ()[]{}<>=,.
    return  map(None, words, [1]*len(words))

def wordcount_file(filename):
    '''file->line->word'''
    f = open(filename, 'r')
    try:
        return filter( lambda x: x, map(wordcount_str, f.readlines()))
    finally:
        f.close()

def wordcount( files ):
    '''input file list and returns a word count list.'''
    count_lists = map( wordcount_file,files)
    merge = lambda l1, l2: l1+l2 #merging lists
    return  multi_reduce( lambda x, y: x+y, #adding ints
                          reduce(merge, 
                                 reduce(merge, count_lists)))

def multi_reduce(reduce_func, mid_data):
    mid = {}
    for t in mid_data: 
        #print t, len(t)
        if( t[0] in mid ):
            mid[t[0]] = reduce_func( mid[t[0]],  t[1] )
        else:
            mid[t[0]] = t[1]
    return mid.items()

if __name__=='__main__':
    l = wordcount( sys.argv )

    print  reduce(lambda x, y: x+y, [x[1] for x in l]), 'tokens.'

    adict = {}
    adict.update( filter( lambda x: (x[1]>1), l) )

    #http://blog.modp.com/2007/11/sorting-python-dict-by-value.html
    alist = sorted(adict.items(), key=lambda (k,v): (v,k))
    n = 10
    print 'top %s: %s' % (n, alist[-n:]) # top 5