get_noun.py

# -*- coding: utf-8 -*-

import sqlite3
import time
import MeCab


def connect_db():
    conn = sqlite3.connect('tl.db')
    return conn

def get_text():
    #被解析用データ
    conn = connect_db()
    query = u'select * from status'
    texts = conn.execute(query)
    return texts

def word_select(text):
    #形態素解析
    words = []
    m = MeCab.Tagger('-Ochasen')
    n = m.parseToNode(text)
    n = n.next
    while n.next:
        if (38 <= n.posid and n.posid <= 47) and n.posid != 40:
            words.append(n.surface)
        n = n.next
    return words
    
def main():
    noun={}
    texts = get_text()
    for text in texts:
        words = word_select(text[2].encode('utf-8'))
        for word in words: 
            if word in noun:
                num=noun[word]
                noun[word] = num + 1
            else:
                noun[word] = 1
        for k, v in sorted(noun.items(), key=lambda x:x[1],reverse=True):
                print k, v

if __name__ == '__main__':
        main()