A work in progress tool for converting Eijiro database into something more usable.

eijiro.py

# A tool for converting Eijiro database into something more usable.
# A work in progress.

import sys

import MeCab
from lxml import etree

from multiprocessing import Process, Pipe

KEEP_WORD_TYPES = [
    ["名詞", "一般"],
    ["名詞", "形容動詞語幹"],
    ["名詞", "固有名詞"],
    ["名詞", "サ変接続"],
    ["名詞", "代名詞"]
]

def get_words(tagger, data):
    words = []
    node = tagger.parseToNode(data.split('\xe2\x97\x86')[0])
    node = node.next
    while node:
        print node.feature.split(",")[:2]
        if node.feature.split(",")[:2] in KEEP_WORD_TYPES:
            words.append(node.surface)
        node = node.next
    return words


def process_trans(conn):
    tagger = MeCab.Tagger("-Ochasen")

    data = conn.recv()
    while data is not None:
        node = tagger.parseToNode(data.split('\xe2\x97\x86')[0])
        node = node.next
        while node:
            node = node.next
            type
        conn.send()

        data = conn.recv()

def aggregate(connections):
    # TODO
    pass


class Eijiro(object):
    in_word = False
    word = None
    in_trans = False
    trans = None

    def start(self, tag, attrib):
        if tag == "word":
            self.in_word = True
        elif tag == "trans":
            self.in_trans = True

    def end(self, tag, attrib):
        if tag == "word":
            self.in_word = False
        elif tag == "trans":
            self.in_trans = False
        elif tag == "record":
            self.in_word = False
            self.in_trans = False

            # TODO: Do stuff

    def data(self, data):
        if self.in_word is True:
            self.word = data
        elif self.in_trans is True:
            self.trans = data.split('\xe2\x97\x86')



def main():
    pass