python2.7で追加されたCounterを試しに使ってみた。従来の辞書との使い方の比較。

new_counter.py

#!/usr/bin/env python
# -*- coding:utf-8 -*-
import MeCab
from collections import Counter

def func():
    m = MeCab.Tagger()
    c = Counter()

    for i in xrange(100000):
        src = "本日は晴天なり"

        res = m.parseToNode(src)

        while res:
            c[res.surface] += 1
            res = res.next
    
    for i in c.most_common()[::-1]:
        print i

def profile(fname='test.Prof'):
    import hotshot
    prof = hotshot.Profile(fname)
    prof.run('func()')
    prof.close

    import hotshot.stats
    stats = hotshot.stats.load(fname)
    stats.strip_dirs()
    stats.sort_stats('cumulative')
    stats.print_stats(80)

if __name__ == '__main__':
    profile()

old_dict.py

#!/usr/bin/env python
# -*- coding:utf-8 -*-
import MeCab

def func():
    m = MeCab.Tagger()
    d = {}

    for i in xrange(100000):
        src = "本日は晴天なり"

        res = m.parseToNode(src)

        while res:
            d[res.surface] = d.get(res.surface,0) + 1
            res = res.next

def profile(fname='test.Prof'):
    import hotshot
    prof = hotshot.Profile(fname)
    prof.run('func()')
    prof.close

    import hotshot.stats
    stats = hotshot.stats.load(fname)
    stats.strip_dirs()
    stats.sort_stats('cumulative')
    stats.print_stats(80)

if __name__ == '__main__':
    profile()