import wordcloud

import matplotlib.pyplot as plt

from datetime import datetime

from twitter import Twitter, OAuth

import neologdn

import re

import MeCab

# apiキーを記入してください

api_key =

api_secret =

access_token_key =

access_token_secret =

def normalize_string(text):

"""

文字列から余計な記号などを取り除く

"""

normalized_text = neologdn.normalize(text).lower()

replaced_text = re.sub("[!?@「」()、。・（）…/_:;\s]", "", normalized_text)

return replaced_text

def create_wordcloud(user, text):

"""

ワードクラウドを作成し、同一ディレクトリ内に保存する

"""

fpath = "/Library/Fonts/ヒラギノ丸ゴ ProN W4.ttc"

wc = wordcloud.WordCloud(background_color="white",font_path=fpath, width=900, height=500).generate(text)

plt.figure(figsize=(15,8))

plt.imshow(wc)

plt.axis("off")

plt.savefig(user + str(datetime.now()) + ".png")

def main():

target_user = input()

count = 100 #一度のアクセスで何件取ってくるか

t = Twitter(auth=OAuth(

access_token_key,

access_token_secret,

api_key,

api_secret,

))

user_tweets = []

timeline = t.statuses.user_timeline(screen_name=target_user, count=count)

for tweet in timeline:

user_tweets.append(tweet['text'])

tweet_data = list(map(normalize_string, user_tweets))

result_text = []

mec = MeCab.Tagger("-Ochasen")

for tw in tweet_data:

temp = mec.parse(tw).split('\n')

for t in temp:

word_data = t.split()

word = word_data[0]

stopwords = ['http', 'EOS', 'rt', 'http', 'tweet', 'peing', '人', 'あと',\

'感じ','httpst', 'rt', 'com', 'the', 'http', '今日']

check = [sw in word for sw in stopwords]

if any(check):

break

else:

tow = word_data[-1]

stop = ["非自立", "代名詞" , "数", "接尾"]

validation = all(map(lambda x: x not in tow, stop))

if tow[0:2] == "名詞" and validation:

result_text.append(word)

create_wordcloud(target_user, " ".join(result_text))

if __name__ == '__main__':

main()