#!/usr/bin/python3
import csv
import pprint
import MeCab
import collections
import sys
from sys import argv

m = MeCab.Tagger('-Ochasen  -d /usr/lib/x86_64-linux-gnu/mecab/dic/mecab-ipadic-neologd')    #形態素解析

mode = 'none'
wcs = {}
ec = ['Twitter', 'Imgur'] #特定文字列
target = ['名詞']    #統計対象品詞

if len(sys.argv) < 2:
        sys.exit()
input_file = sys.argv[1]

with open(input_file, encoding='utf8') as file:
        lines = file.readlines()
lines_strip = [line.strip() for line in lines]

for line in lines_strip:
        if line == '':
                continue
        if '--------------------------------------------' in line:
                mode = 'thread'
                continue
        if '---' in line:
                mode = 'title'
                continue

        if mode == 'thread':
                mode = 'none'
                continue
        if mode == 'title':
                wc = line.split()[3][:9]
                mode = 'content'
                continue
        if mode == 'content':
                if wc not in wcs.keys():
                        wcs[wc] = dict()
                        wcs[wc]['text'] = ''
                        wcs[wc]['count'] = 0
                        wcs[wc]['ec'] = dict()
                wcs[wc]['text'] += line
                wcs[wc]['count'] += 1
                mode = 'title'
                continue

for key in wcs:
        text = wcs[key]['text']
        node = m.parseToNode(text)
        words=[]
        while node:
                hinshi = node.feature.split(",")[0]
                if hinshi in target:
                        origin = node.feature.split(",")[6]
                        words.append(origin)
                node = node.next

        wcs[key]['result'] = collections.Counter(words)

        if wcs[key]['count'] > 0:
                for k in wcs[key]['result']:
                        if k in ec:
                                wcs[key]['ec'][k] = '{:.2f}'.format(wcs[key]['result'][k]/wcs[key]['count'])
#        print(key, wcs[key]['count'],  c.most_common(20))
        print(key, wcs[key]['count'], wcs[key]['ec'])