MeCab implementation for Python using subprocess

2015-01-25-005613.py

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

import subprocess
import itertools as itt


class MeCab():

    def __init__(self, opts=[]):
        self.opts = opts
        self._process = subprocess.Popen(
            list(itt.chain(['mecab'], opts)),
            stdin=subprocess.PIPE,
            stdout=subprocess.PIPE,
            universal_newlines =True,
        )

        if '-Owakati' in opts or 'wakati' in opts:
            self.parse = self.wakati_parse
        else:
            self.parse = self.default_parse

    def wakati_parse(self, iterable):
        for line in iterable:
            self._process.stdin.write(line+'\n')
            output = self._process.stdout.readline()
            yield output.strip().split()

    def default_parse(self, iterable):
        buff = []
        for line in iterable:
            self._process.stdin.write(line+'\n')
            while True:
                output = self._process.stdout.readline().strip()
                if output.startswith('EOS'):
                    break
                buff.append(Morph.from_mecab_format(output))
            yield buff
            buff = []


class Morph():

    def __init__(self):
        self.surface = ''
        self.attrs = []

    @classmethod
    def from_mecab_format(cls, line):
        surface, attrs = line.split()
        attrs = attrs.split(',')

        morph = Morph()
        morph.surface = surface
        morph.attrs = attrs
        morph.pos = attrs[0]
        return morph


def test_mecab():
    parser = MeCab(['-Owakati'])
    morphs = next(parser.parse(['すもももももももものうち']))
    assert morphs == ['すもも', 'も', 'もも', 'も', 'もも', 'の', 'うち']


if __name__ == '__main__':
    test_mecab()