import tweepy
import csv

import keys

consumer_key = keys.consumer_key
consumer_secret = keys.consumer_secret
access_key = keys.access_key
access_secret = keys.access_secret

auth = tweepy.OAuthHandler(consumer_key, consumer_secret)
auth.set_access_token(access_key, access_secret)
api = tweepy.API(auth)

q = r'"あいみょん" "って歌詞"  -filter:retweets'
count = 100

tweet_list = []

tweets = api.search(q=q, locale="ja", count=count, tweet_mode='extended')
for tweet in tweets:
    tweet_list.append([[tweet.user.id, tweet.user.followers_count,
                        tweet.user.friends_count, tweet.user.description],
                       [tweet.id, tweet.full_text, tweet.favorite_count,
                        tweet.retweet_count]])

with open("data/Tweetsdata.csv", "w", encoding="utf-8", newline="") as f:
    writer = csv.writer(f)
    writer.writerow(["tweet"])
    for tweet_info in tweet_list:
        writer.writerow([tweet_info[1][1].replace("\n", "")])
        print([tweet_info[1][1].replace("\n", "")])

import csv
import re
from markov_janome import generate_text


# 歌詞を抜き取る部分
def phrase_split(tweet):
    # tweetから歌詞を抜き取る
    try:
        phrase = str(re.search(r'あいみょんの(.+)って歌詞', tweet).group(1))
        phrase = re.sub(r'[「『｢"」』｣]', "", phrase)
        return phrase
    except:
        return ""


tweet_list = []
with open('data/Tweetsdata.csv', encoding='utf-8_sig') as f:
    reader = csv.reader(f)
    tweet_list = list(reader)[1:]
    tweet_list = list(map(lambda t: t[0], tweet_list))

phrase_list = list(set(list(map(phrase_split, tweet_list))))

# マルコフ連鎖
text = '。'.join(phrase_list)
generate_text(text)