def text_to_word(text, conv):
word = Word()
word.kaki = text
word.yomi = conv.do(text)
# 同じ文字の三回以上の繰り返しを消し去る
# ex. おはよ！！！ → おはよ！
word.yomi = re.sub(r'(.)\1{2,}', r'\1', word.yomi)
# 括弧以降を無視
# ex. ちょん↑ぱぁ！(しょうり) → ちょん↑ぱぁ！
word.yomi = re.sub(r'^([^()（）「」]+)[(（「].*$', r'\1', word.yomi)
# ひらがなと一部の記号のみにする
# ex. ちょん↑ぱぁ！ → ちょんぱぁ
word.yomi = "".join(re.findall(r'[ぁ-ん、。ー]+', word.yomi))
print(f"kaki: {word.kaki}, yomi: {word.yomi}")
return word
def main():
auth = tweepy.OAuthHandler(consumer_key, consumer_secret)
auth.set_access_token(access_token, access_secret)
api = tweepy.API(auth)
kakasi = kks()
kakasi.setMode('K', 'H')
kakasi.setMode('J', 'H')
conv = kakasi.getConverter()
results = api.user_timeline(screen_name=ACCOUNT, count=COUNT)
for r in results:
hashtag_tags = r.entities['hashtags']
hashtag = ""
if hashtag_tags:
hashtag = hashtag_tags[0]['text']
text = r.text
text = filter_text(text, hashtag)
if not text:
pass
lines = text.splitlines()
for l in lines:
if find_word(l):
word = Word()
word.kaki = l
word.yomi = conv.do(l)
print(f"kaki: {word.kaki}, yomi: {word.yomi}")
find_or_add_word(session, word)
tweet = Tweet()
tweet.twitterId = r.id
tweet.text = text
find_or_add_tweet(session, tweet)
import unittest
from analysis import save_word_from_tweet
from models import find_or_add_tweet, Word, Tweet, session
""" ImportErrorで、直接は実行できないのでインタープリタでドーンするしかNASA """
sample = [("チョン↑パァ！(勝利)", [], "ちょんぱぁ", "チョン↑パァ！(勝利)", 10),
("ばにしぇだよ〜wwwww", [], "ばにしぇだよ", "ばにしぇだよ〜wwwww", 20),
("任せてほ↑しい", [], "まかせてほしい", "任せてほ↑しい", 30)]
for text, hashtags, yomi, kaki, n in sample:
for i in range(1, 6):
tweet = Tweet()
tweet.text = text
tweet.twitterId = 200 + n + i
find_or_add_tweet(session, tweet)
word = Word()
word.yomi = yomi
word.kaki = kaki
tmp = save_word_from_tweet(text, hashtags)
print("generated: ", word.yomi, " => ", word.kaki)
print("expected: ", tmp.yomi, " => ", tmp.kaki)
print(word.yomi == tmp.yomi and word.kaki == tmp.kaki)