# coding: UTF-8
import time
import re
import os
import MeCab
from urllib.request import urlopen
from bs4 import BeautifulSoup

tagger = MeCab.Tagger('-Owakati')
save_folder = 'save'
narou_url = 'https://ncode.syosetu.com'


def make_bsobj(url):
    '''
    引数:(str)対象のURL
    返り値:(BeautifulSoup)BeautifulSoupオブジェクト
    サーバーへの負担軽減の為アクセスの間隔を1秒開けます
    '''
    time.sleep(1)
    return BeautifulSoup(urlopen(url).read().decode(
        'utf-8', 'ignore'), "html.parser")


def format_text(t):
    '''
    引数:(str)テキスト
    返り値:(str)フォーマット後のテキスト
    markovifyが読み込むとエラーを吐き出す文字があるので、それを除去する
    '''
    return re.sub(r'[\．_－―─！＠＃＄％＾＆\-‐|\\＊\“（）＿■×+α※÷⇒—●★☆〇◎◆▼◇△□(：〜～＋=)／*&^%$#@!~`){}［］…\[\]\"\'\”\’:;<>?＜＞〔〕〈〉？、。・,\./『』【】「」→←○《》≪≫\r\u3000\u2000]+', "", t)


def get_bodies_from_url(url):
    '''
    引数:(str)各ページのURL(リゼロならhttps://ncode.syosetu.com/n2267be/1/ など)
    返り値:(str)分かち書き後、文章単位で改行された本文
    '''
    bsobj = make_bsobj(url)
    bodies = ''
    for body in bsobj.findAll("div", {"id": "novel_honbun"})[0].findAll("p"):
        formatted = format_text(body.get_text())
        if formatted == '':
            continue
        bodies += tagger.parse(formatted)+'\n'
    return bodies


def get_num_of_pages(url):
    '''
    引数:(str)目次ページURL(リゼロならhttps://ncode.syosetu.com/n2267be/)
    返り値:(int)ページ数
    '''
    bsobj = make_bsobj(url+'/1')
    return int(bsobj.select('#novel_no')[0].get_text().replace('1/', ''))


def main(url, pages_num):
    id = url[url.strip('/').rfind('/'):len(url)]  # saveフォルダ構成用に識別番号を取得
    all_pages_num = get_num_of_pages(url)  # 全ページ数の取得
    for i in range(1, pages_num+1):
        if i > all_pages_num:
            break
        bodies = get_bodies_from_url(url+'/'+str(i))  # 本文の取得
        folder_path = save_folder+id
        os.makedirs(folder_path, exist_ok=True)  # 保存用フォルダの構成
        with open(folder_path+'/'+str(i)+'.txt', mode="w", encoding='utf-8') as f:
            f.write(bodies)


if __name__ == '__main__':
    url = input(
        "Enter the URL of the novel.\n(ex:https://ncode.syosetu.com/n2267be)\n>")  # url
    pages_num = int(
        input("Enter the number of pages you want to save.\n>"))  # 保存するページ数
    main(url, pages_num)

# coding: UTF-8
import save_novel
import os
from glob import glob
import markovify


def main(url, word_size,  out_max, out_min, out_num, first_word):
    id = url[url.strip('/').rfind('/'):len(url)]  # saveフォルダ識別番号
    bodies = ''
    if not os.path.exists(save_novel.save_folder+id):
        print('Path does not exist.')
        exit()
    for file in glob(save_novel.save_folder+id+'/*.txt'):
        with open(file, mode="r", encoding='utf-8') as f:
            bodies += f.read()  # txtファイルを読み込んで結合
    markov_model = markovify.NewlineText(
        bodies, state_size=word_size, well_formed=False)  # マルコフ連鎖モデルを構成
    for _ in range(out_num):
        if(first_word != 'none'):
            try:
                sentence = markov_model.make_sentence_with_start(
                    max_words=out_max, min_words=out_min, tries=100, beginning=first_word).replace(' ', '')  # 文章生成(最初の言葉あり)
            except:
                print(
                    'The word you entered may not exist in the novel.\nOr it may contain unidentifiable characters.')
                exit()
        else:
            sentence = markov_model.make_sentence(
                max_words=out_max, min_words=out_min, tries=100).replace(' ', '')  # 文章生成(ランダム)
        print('out', sentence)


if __name__ == '__main__':
    print('If you haven\'t generated the data for your novel yet, run save_novel.py and run.')
    url = input(
        'Enter the URL of the novel.\n(ex:https://ncode.syosetu.com/n2267be)\n>')
    word_size = int(input(
        'Enter the state history (the more, the closer to the original).\n>'))
    out_max = int(
        input('Enter the maximum length of the text to be generated.\n>'))
    out_min = int(
        input('Enter the minimum length of the text to be generated.\n>'))
    out_num = int(input('Enter the number of texts to be generated.\n>'))
    first_word = input(
        'Enter the first word of the sentence to be generated.To generate randomly, enter "none".\n>')
    main(url, word_size,  out_max, out_min, out_num, first_word)