STEP3_chapter_summary.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 环境变量"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "output_dir='./output_zero/'\n",
    "faiss_path=output_dir+'faiss_index'\n",
    "\n",
    "novel_chunk_dir=output_dir+'novel_chunks/'\n",
    "distilled_log_dir=output_dir+'distilled_log/'\n",
    "chapter_summary_dir=output_dir+'chapter_summary/'\n",
    "merged_log_path=output_dir+'merged_log.json'\n",
    "rename_map_path=output_dir+'rename_map.json'\n",
    "all_chapters_path=output_dir+'all_chapters.json'\n",
    "all_summary_path=output_dir+'all_summary.txt'\n",
    "\n",
    "\n",
    "\n",
    "bg_hint_len=500\n",
    "retrospect_len=100\n",
    "summary_chunk_len=500 # shouldn't be too large. because chatgpt's summary is very short. need to fine-tune this number\n",
    "summary_chunk_overlap=0\n",
    "final_summary_chunk_len=1300\n",
    "\n",
    "import os, json\n",
    "rename_map=json.load(open(rename_map_path,'r',encoding='utf-8')) if os.path.exists(rename_map_path) else {}\n",
    "\n",
    "if not os.path.exists(chapter_summary_dir):\n",
    "    os.makedirs(chapter_summary_dir)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 加载llm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'你好！有什么我可以帮助你的吗？'"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from my_llm import llm_chatgpt as llm\n",
    "llm('你好')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 加载FAISS背景知识索引"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "【露易丝 露易丝·法兰西斯·露·布朗·杜·拉·瓦利埃尔 本作女主角，作品中一般称为露易丝，有着夹杂金色的粉红长卷发、茶褐色的眼瞳（动画为桃红色）。】\n",
      "【露易丝 起初对身为贵族的身份有着很强的自觉心，拥有常人N倍的自尊心，后来受才人影响，逐渐放下这份执着，转而为才人着想。在后期已经可以对才人敞开心扉，坦诚相待。】\n",
      "【露易丝 动画第三期中为救出塔芭莎而放弃贵族身份，但被公主赏识，成为王位继承人。喜欢平贺才人。】\n",
      "【平贺才人 已与露易丝同居中。同时被书中许多女性角色爱慕著。喜欢露易丝。】\n",
      "【露易丝 的性格易怒，容易骂人，有点糊涂（跟才人比），为了小事常常生气。】\n",
      "【露易丝 在托里斯汀东北拥有领土的名门拉·瓦利埃尔公爵家的三女儿、托里斯汀魔法学院的二年级学生。】\n",
      "【安丽埃塔 小说16-17卷有欲从露易丝那里里夺取才人，羡慕著露易丝。】\n",
      "【露易丝 为露易丝配音的钉宫理惠，亦是J.C. STAFF前一部作品《灼眼的夏娜》中夏娜的声优。同时也是傲娇的著名角色之一，钉宫四萌之一】\n",
      "【平贺才人 亚成婚，后与露易丝回到自己所属的世界，回到自己家中。】\n"
     ]
    }
   ],
   "source": [
    "# import importlib,my_faiss;importlib.reload(my_faiss)\n",
    "from my_faiss import FAISS,embeddings,get_bg_hint\n",
    "get_bg_hint.vector_store=FAISS.load_local(faiss_path, embeddings)\n",
    "print(get_bg_hint('露易丝'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 处理gpt返回的json"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "load chunks: 2333\n",
      "num of distilled chunks: 2318\n",
      "Total items: 33564\n",
      "save merged log to ./output_zero/merged_log.json\n"
     ]
    }
   ],
   "source": [
    "import json,random\n",
    "from glob import glob\n",
    "import os\n",
    "from tqdm.auto import tqdm\n",
    "\n",
    "\n",
    "# # separate info, act into individual items\n",
    "def purge_content(all_content):\n",
    "    for item in all_content:\n",
    "        if 'info' in item:\n",
    "            info_str=item['info']\n",
    "            if 'expression' in item:\n",
    "                info_str=item['expression']+' '+info_str\n",
    "            if 'char' in item:\n",
    "                info_str=item['char']+' '+info_str\n",
    "            yield {'info':info_str,'src_id':item['src_id']}\n",
    "        if 'char' in item:\n",
    "            base_info={'char':item['char'],'src_id':item['src_id']}\n",
    "            if 'expression' in item:\n",
    "                base_info['expression']=item['expression']\n",
    "            if 'act' in item:\n",
    "                yield {**base_info,'act':item['act']}\n",
    "            else:\n",
    "                if 'think' in item:\n",
    "                    base_info['think']=item['think']\n",
    "                if 'say' in item:\n",
    "                    base_info['say']=item['say']\n",
    "                if 'think' in item or 'say' in item:\n",
    "                    yield base_info\n",
    "\n",
    "def reload_chunks(slient=False):\n",
    "    global chunks,all_content\n",
    "    chunks={}\n",
    "    for path in glob(os.path.join(novel_chunk_dir,'CHUNK*.json')):\n",
    "        with open(path) as f:\n",
    "            chunk=json.load(f)\n",
    "            chunks[chunk['id']]=chunk\n",
    "    for path in glob(os.path.join(distilled_log_dir,'RESPONSE*.json')):\n",
    "        with open(path) as f:\n",
    "            chunk=json.load(f)\n",
    "            if 'content' in chunk and chunk['content']:\n",
    "                for item in chunk['content']:\n",
    "                    item['src_id']=chunk['id']\n",
    "            chunks[chunk['id']]=chunk\n",
    "    if not slient:print('load chunks:',len(chunks))\n",
    "    if not slient:print('num of distilled chunks:',len([chunk for chunk in chunks.values() if 'content' in chunk]))\n",
    "\n",
    "    chunks=dict(sorted(chunks.items(), key=lambda item: item[0]))\n",
    "\n",
    "    all_content=[item for chunk in chunks.values() if 'content' in chunk and chunk['content'] for item in chunk['content']]\n",
    "    all_content=list(purge_content(all_content))\n",
    "    all_content[:10]\n",
    "    if not slient:print('Total items:', len(all_content))\n",
    "    json.dump(all_content, open(merged_log_path, 'w', encoding='utf-8'), ensure_ascii=False, indent=2)\n",
    "    print('save merged log to',merged_log_path)\n",
    "\n",
    "reload_chunks()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 统计角色出现的频率，统一名字中的同音字"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# run this block to avoid crash when loading chinese font in matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "import numpy as np\n",
    "plt.figure()\n",
    "plt.plot([1,2,3])\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "load, update and save rename_map at path: ./output_zero/rename_map.json\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x200 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "list of characters and their frequencies: {'才人': 5817, '露易丝': 4821, '塔芭莎': 1688, '基修': 738, '希儿菲朵': 725, '丘鲁克': 616, '安丽埃塔': 603, '谢斯塔': 524, '安莉艾塔': 352, '约瑟夫': 298, '亚尼艾斯': 267, '西艾斯塔': 241, '安莉埃塔': 236, '库尔凯': 224, '德尔弗伦格': 216, '蒂法妮娅': 211, '瓦尔德': 202, '伊莎贝拉': 195, '蒂法尼亚': 194, '夏洛特': 180, '朱利奥': 155, '蒙莫朗西': 138, '朱里奥': 137, '蒙莫兰希': 134, '芙凯': 130, '克伦威尔': 123, '柯尔贝尔': 123, '奥利凡': 107, '格鲁贝鲁': 103, '伊尔库库': 100, '苛尔贝尔': 96, '卡特莉娅': 94, '瓦鲁多': 92, '维尔斯': 87, '敏努维尔': 87, '马利寇尔奴': 86, '吉露': 82, '马里科尔努': 81, '贝亚托里斯': 78, '谢菲尔德': 71, '里修蒙': 67, '特尔弗林格': 67, '马里科尔奴': 67, '琉琉': 66, '爱尔莎': 64, '雷纳尔': 63, '卡特蕾亚': 62, '洁西卡': 61, '威尔士': 60, '奥斯曼': 59, '男人': 58, '马萨林': 57, '斯卡隆': 55, '卡斯特莫尔': 54, '德福林哥': 51, '迪法尼亚': 51, '维雷尔少尉': 50, '士兵': 47, '维特里奥': 47, '莎夏': 47, '莫里哀夫人': 46, '奥斯曼校长': 45, '拉尔卡斯': 44, '塔巴撒': 43, '阿尼亚斯': 43, '尤西亚': 42, '奥斯曼氏': 42, '艾蕾欧诺尔': 42, '朗格贝尔': 40, '尤尔班': 39, '少女': 38, '学生们': 37, '店主': 36, '琉西': 36, '贝塔夏尔': 35, '普里米尔': 35, '托马斯': 34, '村民们': 33, '吉尔摩阿': 33, '霍金斯将军': 33, '多德': 33, '德?罗雷努': 32, '约赛特': 32, '公爵': 31, '密斯尼特伦': 31, '克拉威鲁卿': 31, '雅妮特': 30, '尼可拉': 30, '阿妮艾斯': 30, '约瑟夫国王': 29, '侍女': 28, '商人': 27, '卢涅': 27, '平贺才人': 26, '水之精灵': 26, '斯蒂克斯': 26, '雪斯塔': 26, '艾伯特': 25, '谢弗勒兹': 25, '阿尼艾斯': 25, '塞雷斯坦': 24, '博伍德': 24, '客人': 24, '裘露珂': 24, '尤尔姆冈特': 24, '萨姆': 23, '马里科鲁': 23, '风凯': 23, '村长': 21, '男子': 20, '罗德巴尔德男爵夫人': 20, '女孩子们': 20, '德鲁福林哥': 20, '维尔斯王子': 20, '梦莫兰希': 20, '圣埃吉斯三十二世': 20, '亚尔托瓦伯': 19, '尼娜': 19, '骑士': 19, '加里奴': 19, '少年们': 19, '贵族们': 18, '船长': 18, '埃蕾欧诺尔': 18, '卡哥伊鲁': 17, '老婆婆': 17, '爱夏': 17, '队长': 17, '卫兵': 17, '贝鲁斯兰': 17, '士官': 17, '霍金斯': 17, '鲁尼': 17, '米诺塔洛斯': 16, '母亲': 16, '女生们': 16, '拉·瓦利埃尔公爵': 16, '圣堂骑士': 16, '教皇': 16, '年轻骑士': 16, '蒂法妮婭': 16, '女孩': 15, '龙骑士': 15, '温普芬': 15, '强盗': 15, '贝尔斯兰': 15, '老骑士': 15, '拜多里奥': 15, '芭莱莉': 15, '灰色卿': 15, '嘉内特': 15, '托马': 14, '安丽艾塔': 14, '朱力欧': 14, '杰西卡': 14, '伊露库库': 14, '基姆利': 14, '卡路罗': 14, '作者': 14, '贝达夏尔': 14, '村民': 13, '水兵': 13, '奥斯曼先生': 13, '学生': 13, '马扎利意尼': 13, '威尔士·都铎': 13, '马尔托大叔': 13, '拉.瓦利艾尔公爵': 13, '罗拉': 13, '监督官': 13, '朗格贝尔小姐': 12, '老板': 12, '安莉艾塔公主': 12, '马萨里尼': 12, '艾丽欧诺尔·瓦利埃尔': 12, '贵族士官': 12, '女骑士': 12, '德·坡瓦奇': 12, '骑士们': 12, '修普露兹夫人': 12, '女王安莉艾塔': 12, '莫莫兰希': 12, '加尔罗': 12, '摩莉艾尔夫人': 12, '奥尔里昂公夫人': 12, '贵族': 11, '弗雷姆': 11, '雌火龙': 11, '琉吉尼恩子爵': 11, '女头目': 11, '哥布灵萨满': 11, '奥尔良公': 11, '魔杖': 11, '瓦鲁多子爵': 11, '使魔露易丝': 11, '德·坡瓦奇将军': 11, '副官': 11, '贝塔夏尔卿': 11, '維托里奧': 11, '布里杰塔': 11, '石匠们': 11, '军官': 10, '马鲁托': 10, '盖托': 10, '士兵们': 10, '艾丽欧诺尔': 10, '德?坡瓦奇': 10, '财务卿': 10, '蒙莫兰溪': 10, '伊瓦尔汀': 10, '吉姆利': 10, '官吏': 10, '约路蒙钢特': 10, '库拉威尔卿': 9, '假面男': 9, '楚连努': 9, '队员们': 9, '柯尔贝尔老师': 9, '费尔南': 9, '风龙': 9, '指挥官': 9, '孩子们': 9, '佩利逊': 9, '秘书官': 9, '米斯蔻尔男爵': 9, '凯蒂': 9, '科尔贝尔': 9, '威廉少尉': 9, '索瓦松男爵': 9, '夏露尔': 9, '幻想先生': 9, '女学生们': 8, '同学们': 8, '德鲁弗林加': 8, '伯伍德': 8, '军官们': 8, '众人': 8, '队员': 8, '大队长': 8, '将军': 8, '敌人': 8, '吉努梅尔': 8, '哈登贝格侯爵': 8, '马利寇里奴': 8, '市民们': 8, '水精灵骑士队': 8, '莉捷特': 8, '修布尔兹夫人': 8, '露易絲': 8, '海伦婆婆': 8, '贝尔丹蒂': 7, '山口': 7, '奥尔良公爵夫人': 7, '女孩子': 7, '德鲁': 7, '贝露丹蒂': 7, '船员': 7, '舰长': 7, '水兵们': 7, '老农夫': 7, '龙骑士们': 7, '甲板士官': 7, '亚修': 7, '将军们': 7, '佣兵头目': 7, '可鲁贝尔': 7, '火枪队队长阿尼亚斯': 7, '教皇圣埃吉斯三十二世': 7, '卡特莉雅': 7, '教皇拜多里奥': 7, '阿利克山大': 6, '智慧短剑“地下水”': 6, '总司令库拉威尔卿': 6, '希尔菲': 6, '女人': 6, '白色假面男': 6, '空贼头儿': 6, '里维斯拉谢尔': 6, '玛丽安努': 6, '约翰斯顿先生': 6, '水精灵': 6, '塔芭萨': 6, '棋尔可': 6, '魔法师': 6, '其中一个部下': 6, '卢涅等人': 6, '大臣': 6, '阿尔布雷希特三世': 6, '民众': 6, '马力科鲁': 6, '马利蔻尔奴': 6, '店主人': 6, '米斯寇尔男爵': 6, '圣堂骑士们': 6, '市长': 6, '罗马尼亚人': 6, '玛丽安': 6, '修道院长': 6, '卫士': 5, '地下水': 5, '中年贵族': 5, '叽叽的母亲': 5, '叽叽的父亲': 5, '粉发少女': 5, '夫人': 5, '熊': 5, '山口升': 5, '佛雷姆': 5, '比塔夏尔': 5, '子爵': 5, '马扎利意尼军机臣边': 5, '首领': 5, '费比斯': 5, '德尔弗伦': 5, '猫头鹰': 5, '贵族少年们': 5, '其他女学生': 5, '精灵': 5, '丘尔可': 5, '艾雷奥诺尔': 5, '少女们': 5, '羅倫': 5, '差役': 5, '琉加利昂子爵': 5, '副队长': 5, '玛丽安太后': 5, '粗糙男': 4, '亚历山大': 4, '教师': 4, '兔冢先生': 4, '叽叽': 4, '男人1': 4, '奥尔良公爵': 4, '男人们': 4, '舰队首脑们': 4, '老板娘': 4, '使者': 4, '女子': 4, '格鲁贝鲁先生': 4, '马里克卢姆': 4, '布鲁': 4, '教师们': 4, '披风男': 4, '佣兵': 4, '头儿': 4, '蓝发女孩': 4, '女官': 4, '拉?拉梅': 4, '父亲': 4, '亨利?伯伍德': 4, '琪尔克': 4, '曼提柯尔队的队长': 4, '鹰马队': 4, '女王': 4, '妖精们': 4, '？': 4, '空尉': 4, '朱里奥.塞萨雷': 4, '亚兹洛': 4, '其他将军': 4, '巡逻兵': 4, '凯迪': 4, '青发女性': 4, '特尔弗': 4, '龙': 4, '士兵2': 4, '缪兹尼多尼伦': 4, '蒂法妮亚': 4, '卡托丽雅': 4, '室友们': 4, '普里米爾': 4, '连队长': 4, '印刷长': 4, '琉基尼子爵': 4, '声音': 4, '一贵族': 4, '德尔弗': 4, '贵族男子': 4, '侍女们': 3, '草药师雷昂': 3, '杰伊克': 3, '杰克': 3, '男人2': 3, '小女孩': 3, '马尔特老爹': 3, '黑发少年': 3, '极乐鸟': 3, '阿尔维': 3, '黑发少女': 3, '奇美拉龙': 3, '莫特索格尼尔': 3, '基修的朋友们': 3, '奥斯曼长老': 3, '西布鲁兹夫人': 3, '全员': 3, '公主安莉艾塔': 3, '戴头巾的少女': 3, '黑船': 3, '胖子': 3, '空贼': 3, '基修正': 3, '兽人鬼': 3, '舰长费比斯': 3, '会议室众人': 3, '观众们': 3, '霍雷肖': 3, '马里科努尔': 3, '塔芭莎的母亲': 3, '醉鬼们': 3, '年轻的贵族': 3, '魔法师们': 3, '未知人物': 3, '公爵夫人拉·瓦利埃尔': 3, '女仆': 3, '少年兵': 3, '朱利安': 3, '敏努维尔的部下': 3, '贵族少年3': 3, '两名枪士队的女性成员': 3, '吉尔伯特': 3, '德罗雷努': 3, '朱利奥·塞萨雷': 3, '金发少女': 3, '马萨利尼': 3, '苛尔贝尔老师': 3, '年轻的女王': 3, '女性': 3, '三人': 3, '拉.瓦利艾尔公爵夫人': 3, '士兵1': 3, '卫兵1': 3, '哥雷姆': 3, '夏尔洛': 3, '空中装甲骑士团': 3, '骑士队长': 3, '学院长奥斯曼氏': 3, '少年': 3, '未知人物1': 3, '圣弗尔萨迪': 3, '水精灵骑士队员们': 3, '唯一的客人': 3, '水精灵骑士队队员们': 3, '加尔罗·克里斯蒂亚诺·特伦博迪诺': 3, '缪斯尼多尼伦': 3, '管那边的提督': 3, '教皇維托里奧': 3, '普里米尔教徒': 3, '水精灵骑士团少年': 3, '行人': 3, '武官': 3, '群众': 3, '水精灵骑士队的少年们': 3, '罗马尼亚士兵': 3, '对方': 3, '加里亚军': 3, '登': 3, '水精灵骑士团': 3, '扮演才人的男人': 3, '马萨里尼卿': 3, '艾蕾欧诺尔姐姐': 3, '修女少女': 3, '公': 3, '翼人': 2, '守卫': 2, '艾尔莎': 2, '卫士们': 2, '短剑': 2, '人偶': 2, '德·罗纳尔伯爵夫人': 2, '侍从': 2, '英俊男子': 2, '客人们': 2, '多米尼克婆婆': 2, '男人3': 2, '男人4': 2, '火蜥蜴弗雷姆': 2, '神官': 2, '火龙': 2, '雄火龙': 2, '花坛骑士塔芭莎': 2, '修女琉西': 2, '韵龙之女': 2, '管家男子': 2, '罗德巴尔德男爵夫人样子的卡哥伊鲁': 2, '追兵': 2, '奥尔良公夫人': 2, '奥尔良夫人': 2, '某人': 2, '其他学生': 2, '凯特': 2, '基修的朋友': 2, '格雷姆': 2, '修特意克斯': 2, '巨像兵': 2, '戴着头巾的少女': 2, '看守': 2, '老贵族': 2, '巴里': 2, '詹姆斯一世': 2, '王党派的贵族': 2, '皇太子': 2, '金发少年': 2, '火蜥蜴': 2, '太后玛丽安努': 2, '基修的七个瓦里丘尔': 2, '龙之羽衣': 2, '老师': 2, '拉?拉梅伯爵': 2, '莱克星顿号': 2, '某大臣': 2, '传令官': 2, '魔法卫士队众队员': 2, '周围的人': 2, '鼹鼠': 2, '琪尔可得': 2, '魔法卫队队长': 2, '亚尔比昂的贵族们': 2, '基托先生': 2, '贵族男人': 2, '亚尼艾斯修瓦里埃德米兰': 2, '年轻将军': 2, '拉·瓦利埃尔公爵夫人': 2, '罗杰姆': 2, '德?维努依佑': 2, '士官候补生': 2, '年轻女性': 2, '托罗尔鬼们': 2, '女王安丽埃塔': 2, '参谋长': 2, '德．坡瓦奇将军': 2, '谢雷斯坦': 2, '敌方佣兵': 2, '德?坡瓦奇将军': 2, '第二龙骑士中队队长': 2, '赛弗林': 2, '少年骑士们': 2, '侍童': 2, '兵站参谋': 2, '传令士兵': 2, '骑士队队长': 2, '联络士兵': 2, '小孩子们': 2, '同学': 2, '吉姆': 2, '露可夏娜': 2, '特里斯塔尼亚市民': 2, '同班学生': 2, '缪兹尼多尼隆': 2, '精灵男子': 2, '韵龙': 2, '塔芭莎的义妹': 2, '老爹': 2, '卡莉娜': 2, '蓝发女子': 2, '精灵贝塔夏尔': 2, '威托利奥': 2, '卫兵2': 2, '雀斑少年': 2, '同班女学生': 2, '褐发少女': 2, '一年级女孩子们': 2, '战少女': 2, '泽尔布斯特小姐': 2, '另一个少年': 2, '追捕者': 2, '同僚们': 2, '教皇维特里奥·圣·李巴利': 2, '维特里奥·塞雷瓦雷': 2, '布里吉塔': 2, '鼹鼠贝尔丹蒂': 2, '骑乘者': 2, '金发女骑士': 2, '阿鲁奴鲁夫': 2, '东蔷薇骑士团': 2, '教皇维托里奥': 2, '野狼': 2, '炮术士官': 2, '炮龟兵': 2, '加里亚领事': 2, '朱力欧·切萨雷': 2, '加里亚士兵': 2, '秃头男人': 2, '女性代表': 2, '巴索·卡斯特莫尔': 2, '女孩们': 2, '警察': 2, '威尤氏': 2, '中年女性': 2, '观众': 2, '老人': 2, '猎人': 1, '龙形卡哥伊鲁': 1, '阿婆': 1, '人群中一个人': 1, '人群中另一个人': 1, '被害者的两亲': 1, '老父亲': 1, '老妇人': 1, '玛萨塔婆婆': 1, '雷昂': 1, '化妆师': 1, '居民': 1, '娜塔莉': 1, '曼提柯尔像': 1, '学生1': 1, '学生2': 1, '风龙希儿菲朵': 1, '侍女少女': 1, '少女侍从': 1, '店员': 1, '黑衣总管': 1, '接待女子': 1, '周围客人': 1, '老板吉尔摩阿满': 1, '平民客人': 1, '庄家': 1, '多米尼克': 1, '装成米诺塔洛斯的男人': 1, '装成米诺塔洛斯的男子': 1, '尼娜的母亲': 1, '侍从们': 1, '巨大火龙': 1, '使魔': 1, '希尔菲朵': 1, '绅士': 1, '兜帽男子': 1, '官员': 1, '头': 1, '小女孩儿': 1, '男村民': 1, '哥布灵神官': 1, '罗德巴尔德': 1, '男爵夫人': 1, '尤尔班的卡哥伊鲁': 1, '罗宾': 1, '吉露妹妹': 1, '露易丝的魔杖': 1, '中年魔法装扮者': 1, '格拉贝鲁': 1, '中年的魔法使': 1, '未知': 1, '奥鲁德·奥斯曼': 1, '美丽卷发的少女': 1, '教师、路易丝': 1, '马鲁科卢姆': 1, '教室里的某人': 1, '丘鲁克的火蜥': 1, '马尼坎': 1, '艾依基亚克斯': 1, '格姆力': 1, '塔芭莎特': 1, 'MEIJI': 1, '盖托老师': 1, '公主': 1, '学院长奥斯曼': 1, '巨大鼹鼠': 1, '家伙们': 1, '服务员': 1, '男人5': 1, '男人6': 1, '男人7': 1, '男人8': 1, '男人9': 1, '男人10': 1, '男人11': 1, '男人12': 1, '男人13': 1, '男人14': 1, '男人15': 1, '男人16': 1, '男人17': 1, '戴着白色假面的男人': 1, '胖胖的店主': 1, '白假面男子': 1, '空贼首领': 1, '年迈的王': 1, '老王': 1, '王党派的贵族们': 1, '威尔士皇太子': 1, '希尔芙德': 1, '魔法卫士队队员': 1, '曼提柯尔队魔法师': 1, '金发美少女': 1, '眼镜小女孩': 1, '身材高挑女性': 1, '香水之蒙莫兰希': 1, '巨大的鼹鼠': 1, '奥利弗·克伦威尔': 1, '亨利·博伍德': 1, '长发女性': 1, '瓦尔德德': 1, '青铜女武神': 1, '塔巴撒的风龙': 1, '基修父亲': 1, '塔芭撒': 1, '阿比昂舰队': 1, '舰队军官': 1, '拉罗舍尔舰长': 1, '约翰斯顿': 1, '另一大臣': 1, '某贵族': 1, '军机马萨里尼': 1, '龙骑士队员': 1, '敌方人员': 1, '未知人': 1, '亚尔比昂军的贵族们': 1, '威尔斯': 1, '男人的同伙': 1, '青蛙': 1, '袭击者(高个子)': 1, '袭击者(小个子)': 1, '近卫魔法卫士': 1, '鹰马队成员': 1, '受伤的人': 1, '马匹': 1, '棋尔': 1, '才人和琪尔可': 1, '他们': 1, '店员们': 1, '店里的女孩们': 1, '店里的女孩子们': 1, '第三个男生': 1, '高年级生': 1, '男生们': 1, '托妮': 1, '周围女生': 1, '二年级男生': 1, '托妮?夏兰特': 1, '火焰的女王': 1, '年长的贵族': 1, '商人男人': 1, '德?拉?瓦利埃尔': 1, '佣人': 1, '枪士队观众': 1, '瓦尔德子爵': 1, '土块之芙卡': 1, '议长兼初代皇帝': 1, '另一位将军': 1, '胖胖的将军': 1, '姐姐们': 1, '公爵夫人': 1, '母亲和姐姐': 1, '画眉鸟': 1, '托罗尔鬼': 1, '枢机卿马萨林': 1, '小个子男人': 1, '穿铁盔甲的将军': 1, '贵族少年1': 1, '贵族少年2': 1, '贵族少年': 1, '另一个贵族少年': 1, '第三个贵族少年': 1, '胖少年': 1, '一骑龙骑士': 1, '蛇形玩偶': 1, '年长的枪士队队员': 1, '另一个枪士队队员': 1, '某个队员': 1, '某女生': 1, '好几根魔法箭矢': 1, '小胖子龙骑士': 1, '长官': 1, '亚修.彭德尔顿': 1, '吉尔伯特和赛弗林': 1, '第二中队的龙骑士们': 1, '龙骑士中队成员': 1, '参谋总部将军': 1, '亲卫队士兵': 1, '阁僚': 1, '老枪兵们': 1, '泥人偶们': 1, '兽人鬼们': 1, '佣兵副官': 1, '卢涅若': 1, '基修的哥哥': 1, '传说之剑': 1, '瓦利埃尔小姐': 1, '卢涅和基修': 1, '马尔可': 1, '下属参谋': 1, '魔法骑士们': 1, '指挥缨枪队的魔法师指挥官': 1, '指挥弓兵队的年轻指挥官': 1, '敌方剑士': 1, '剑士': 1, '马里科尔努和基修': 1, '爱玛': 1, '妈妈': 1, '风龙亚兹洛': 1, '童女生们': 1, '得尔弗伦格': 1, '约瑟夫的仆人': 1, '安丽埃塔等人': 1, '黑色长袍的女性': 1, '魔法人偶': 1, '贝儿朵莉丝': 1, '阿里': 1, '希瓦尔兹老师': 1, '女学生': 1, '厨师': 1, '马车中的安莉艾塔': 1, '女性民众': 1, '提姆利': 1, '戴眼镜男学生': 1, '醉汉': 1, '夜之女郎': 1, '西瓦尔兹老师': 1, '才人和塔芭莎': 1, '旁人': 1, '青发女子': 1, '曼提苛尔队长': 1, '曼提苛尔队队长德-塞萨尔': 1, '父王': 1, '理查': 1, '西奥门德': 1, '马力科鲁姆': 1, '莫莫兰西': 1, '其他人': 1, '红发女郎': 1, '粉发和金发的少女': 1, '基修/马利蔻尔奴': 1, '才人等人': 1, '高个精灵': 1, '拉.瓦利艾尔家人': 1, '六人': 1, '贵族女学生': 1, '女佣和女子应援团的女孩子们': 1, '沙罗曼蛇': 1, '女图书管理员': 1, '图书管理员': 1, '才人和西艾斯塔': 1, '国王': 1, '约瑟夫的心腹们': 1, '缪斯': 1, '维塔夏尔': 1, '土之哥雷姆': 1, '问题人物': 1, '水精灵骑士队的队员们': 1, '另一人': 1, '褐发女孩': 1, '贝亚托里斯的手下': 1, '骑士团团长': 1, '基修的青铜之战少女': 1, '一个骑士': 1, '空中骑士团成员': 1, '马利寇尔奴大人': 1, '所有人': 1, '未知人物2': 1, '未知人物3': 1, '未知人物4': 1, '未知人物5': 1, '其他少女们': 1, '孩子': 1, '年长的孩子': 1, '男学生们': 1, '骑士队的学生们': 1, '加尔罗·克里斯狄亚诺·特伦博迪诺': 1, '炎之龙': 1, '水精灵骑士队员': 1, '切扎雷阁下': 1, '圣乐队': 1, '圣堂骑士团': 1, '亞尼艾斯': 1, '马达夫': 1, '基修一群人': 1, '旅行者': 1, '守望员': 1, '士官们': 1, '传令': 1, '罗马尼亚舰队': 1, '阿尔比昂': 1, '华都': 1, '参拜者': 1, '才人的母亲': 1, '骑士傀儡们': 1, '军官和骑士们': 1, '虎式坦克': 1, '水精灵骑士的队员们': 1, '丘鲁克和水精灵骑士队的少年们': 1, '队友': 1, '城下的人': 1, '水精灵骑士团成员': 1, '骑士团成员': 1, '马萨里尼枢機卿': 1, '副长': 1, '年长的圣堂骑士': 1, '圣骑士': 1, '艾多利昂': 1, '基德老师': 1, '琪可尔': 1, '平贺氏': 1, '威尤': 1, '拉·瓦里埃尔': 1, '斯卡隆店长': 1, '男性': 1, '骑兵队': 1, '前面的女骑士': 1, '工人': 1, '其他石匠': 1, '巴利贝里尼': 1, 'DM': 1, '领地上的人民': 1, '其他贵族们': 1, '大贵族': 1, '贵族女子': 1, '银发少女': 1, '红发少女': 1, '达米安': 1, '拉瓦利艾尔公爵': 1}\n",
      "rename_map: {'塔巴萨': '塔芭莎', '塔巴莎': '塔芭莎', '雪丝塔': '谢斯塔', '摩莫朗希': '蒙莫朗西', '夏尔洛特': '夏洛特', '希尔菲德': '希儿菲朵', '西布鲁兹': '谢弗勒兹', '谢丝塔': '谢斯塔', '路易丝': '露易丝', '露易斯': '露易丝', '路易斯': '露易丝', '琪尔可': '丘鲁克', '露易兹': '露易丝', '芙卡': '芙凯'}\n",
      "selected_char: ['才人', '露易丝', '塔芭莎', '基修', '希儿菲朵', '丘鲁克', '安丽埃塔', '谢斯塔', '安莉艾塔', '约瑟夫', '亚尼艾斯', '西艾斯塔', '安莉埃塔', '库尔凯', '德尔弗伦格', '蒂法妮娅', '瓦尔德', '伊莎贝拉', '蒂法尼亚', '夏洛特', '朱利奥', '蒙莫朗西', '朱里奥', '蒙莫兰希', '芙凯', '克伦威尔', '柯尔贝尔', '奥利凡', '格鲁贝鲁', '伊尔库库', '苛尔贝尔', '卡特莉娅', '瓦鲁多', '维尔斯', '敏努维尔', '马利寇尔奴', '吉露', '马里科尔努', '贝亚托里斯', '谢菲尔德', '里修蒙', '特尔弗林格', '马里科尔奴', '琉琉', '爱尔莎', '雷纳尔', '卡特蕾亚', '洁西卡', '威尔士', '奥斯曼', '男人', '马萨林', '斯卡隆', '卡斯特莫尔', '德福林哥', '迪法尼亚', '维雷尔少尉', '士兵', '维特里奥', '莎夏', '莫里哀夫人', '奥斯曼校长', '拉尔卡斯', '塔巴撒', '阿尼亚斯', '尤西亚', '奥斯曼氏', '艾蕾欧诺尔', '朗格贝尔', '尤尔班', '少女', '学生们', '店主', '琉西', '贝塔夏尔', '普里米尔', '托马斯', '村民们', '吉尔摩阿', '霍金斯将军', '多德', '德?罗雷努', '约赛特', '公爵', '密斯尼特伦', '克拉威鲁卿', '雅妮特', '尼可拉', '阿妮艾斯', '约瑟夫国王', '侍女', '商人', '卢涅', '平贺才人', '水之精灵', '斯蒂克斯', '雪斯塔', '艾伯特', '谢弗勒兹', '阿尼艾斯', '塞雷斯坦', '博伍德', '客人', '裘露珂', '尤尔姆冈特', '萨姆', '马里科鲁', '风凯', '村长', '男子', '罗德巴尔德男爵夫人', '女孩子们', '德鲁福林哥', '维尔斯王子', '梦莫兰希', '圣埃吉斯三十二世', '亚尔托瓦伯', '尼娜', '骑士', '加里奴', '少年们', '贵族们', '船长', '埃蕾欧诺尔', '卡哥伊鲁', '老婆婆', '爱夏', '队长', '卫兵', '贝鲁斯兰', '士官', '霍金斯', '鲁尼', '米诺塔洛斯', '母亲', '女生们', '拉·瓦利埃尔公爵', '圣堂骑士', '教皇', '年轻骑士', '蒂法妮婭', '女孩', '龙骑士', '温普芬', '强盗', '贝尔斯兰', '老骑士', '拜多里奥', '芭莱莉', '灰色卿', '嘉内特', '托马', '安丽艾塔', '朱力欧', '杰西卡', '伊露库库', '基姆利', '卡路罗', '作者', '贝达夏尔', '村民', '水兵', '奥斯曼先生', '学生', '马扎利意尼', '威尔士·都铎', '马尔托大叔', '拉.瓦利艾尔公爵', '罗拉', '监督官', '朗格贝尔小姐', '老板', '安莉艾塔公主', '马萨里尼', '艾丽欧诺尔·瓦利埃尔', '贵族士官', '女骑士', '德·坡瓦奇', '骑士们', '修普露兹夫人', '女王安莉艾塔', '莫莫兰希', '加尔罗', '摩莉艾尔夫人', '奥尔里昂公夫人', '贵族', '弗雷姆', '雌火龙', '琉吉尼恩子爵', '女头目', '哥布灵萨满', '奥尔良公', '魔杖', '瓦鲁多子爵', '使魔露易丝', '德·坡瓦奇将军', '副官', '贝塔夏尔卿', '維托里奧', '布里杰塔', '石匠们', '军官', '马鲁托', '盖托', '士兵们', '艾丽欧诺尔', '德?坡瓦奇', '财务卿', '蒙莫兰溪', '伊瓦尔汀', '吉姆利', '官吏', '约路蒙钢特', '库拉威尔卿', '假面男', '楚连努', '队员们', '柯尔贝尔老师', '费尔南', '风龙', '指挥官', '孩子们', '佩利逊', '秘书官', '米斯蔻尔男爵', '凯蒂', '科尔贝尔', '威廉少尉', '索瓦松男爵', '夏露尔', '幻想先生', '女学生们', '同学们', '德鲁弗林加', '伯伍德', '军官们', '众人', '队员', '大队长', '将军', '敌人', '吉努梅尔', '哈登贝格侯爵', '马利寇里奴', '市民们', '水精灵骑士队', '莉捷特', '修布尔兹夫人', '露易絲', '海伦婆婆', '贝尔丹蒂', '山口', '奥尔良公爵夫人', '女孩子', '德鲁', '贝露丹蒂', '船员', '舰长', '水兵们', '老农夫', '龙骑士们', '甲板士官', '亚修', '将军们', '佣兵头目', '可鲁贝尔', '火枪队队长阿尼亚斯', '教皇圣埃吉斯三十二世', '卡特莉雅', '教皇拜多里奥', '阿利克山大', '智慧短剑“地下水”', '总司令库拉威尔卿', '希尔菲', '女人', '白色假面男', '空贼头儿', '里维斯拉谢尔', '玛丽安努', '约翰斯顿先生', '水精灵', '塔芭萨', '棋尔可', '魔法师', '其中一个部下', '卢涅等人', '大臣', '阿尔布雷希特三世', '民众', '马力科鲁', '马利蔻尔奴', '店主人', '米斯寇尔男爵', '圣堂骑士们', '市长', '罗马尼亚人', '玛丽安', '修道院长', '卫士', '地下水', '中年贵族', '叽叽的母亲', '叽叽的父亲', '粉发少女', '夫人', '熊', '山口升', '佛雷姆', '比塔夏尔', '子爵', '马扎利意尼军机臣边', '首领', '费比斯', '德尔弗伦', '猫头鹰', '贵族少年们', '其他女学生', '精灵', '丘尔可', '艾雷奥诺尔', '少女们', '羅倫', '差役', '琉加利昂子爵', '副队长', '玛丽安太后', '粗糙男', '亚历山大', '教师', '兔冢先生', '叽叽', '男人1', '奥尔良公爵', '男人们', '舰队首脑们', '老板娘', '使者', '女子', '格鲁贝鲁先生', '马里克卢姆', '布鲁', '教师们', '披风男', '佣兵', '头儿', '蓝发女孩', '女官', '拉?拉梅', '父亲', '亨利?伯伍德', '琪尔克', '曼提柯尔队的队长', '鹰马队', '女王', '妖精们', '？', '空尉', '朱里奥.塞萨雷', '亚兹洛', '其他将军', '巡逻兵', '凯迪', '青发女性', '特尔弗', '龙', '士兵2', '缪兹尼多尼伦', '蒂法妮亚', '卡托丽雅', '室友们', '普里米爾', '连队长', '印刷长', '琉基尼子爵', '声音', '一贵族', '德尔弗', '贵族男子', '侍女们', '草药师雷昂', '杰伊克', '杰克', '男人2', '小女孩', '马尔特老爹', '黑发少年', '极乐鸟', '阿尔维', '黑发少女', '奇美拉龙', '莫特索格尼尔', '基修的朋友们', '奥斯曼长老', '西布鲁兹夫人', '全员', '公主安莉艾塔', '戴头巾的少女', '黑船', '胖子', '空贼', '基修正', '兽人鬼', '舰长费比斯', '会议室众人', '观众们', '霍雷肖', '马里科努尔', '塔芭莎的母亲', '醉鬼们', '年轻的贵族', '魔法师们', '未知人物', '公爵夫人拉·瓦利埃尔', '女仆', '少年兵', '朱利安', '敏努维尔的部下', '贵族少年3', '两名枪士队的女性成员', '吉尔伯特', '德罗雷努', '朱利奥·塞萨雷', '金发少女', '马萨利尼', '苛尔贝尔老师', '年轻的女王', '女性', '三人', '拉.瓦利艾尔公爵夫人', '士兵1', '卫兵1', '哥雷姆', '夏尔洛', '空中装甲骑士团', '骑士队长', '学院长奥斯曼氏', '少年', '未知人物1', '圣弗尔萨迪', '水精灵骑士队员们', '唯一的客人', '水精灵骑士队队员们', '加尔罗·克里斯蒂亚诺·特伦博迪诺', '缪斯尼多尼伦', '管那边的提督', '教皇維托里奧', '普里米尔教徒', '水精灵骑士团少年', '行人', '武官', '群众', '水精灵骑士队的少年们', '罗马尼亚士兵', '对方', '加里亚军', '登', '水精灵骑士团', '扮演才人的男人', '马萨里尼卿', '艾蕾欧诺尔姐姐', '修女少女', '公', '翼人', '守卫', '艾尔莎', '卫士们', '短剑', '人偶', '德·罗纳尔伯爵夫人', '侍从', '英俊男子', '客人们', '多米尼克婆婆', '男人3', '男人4', '火蜥蜴弗雷姆', '神官', '火龙', '雄火龙', '花坛骑士塔芭莎', '修女琉西', '韵龙之女', '管家男子', '罗德巴尔德男爵夫人样子的卡哥伊鲁', '追兵', '奥尔良公夫人', '奥尔良夫人', '某人', '其他学生', '凯特', '基修的朋友', '格雷姆', '修特意克斯', '巨像兵', '戴着头巾的少女', '看守', '老贵族', '巴里', '詹姆斯一世', '王党派的贵族', '皇太子', '金发少年', '火蜥蜴', '太后玛丽安努', '基修的七个瓦里丘尔', '龙之羽衣', '老师', '拉?拉梅伯爵', '莱克星顿号', '某大臣', '传令官', '魔法卫士队众队员', '周围的人', '鼹鼠', '琪尔可得', '魔法卫队队长', '亚尔比昂的贵族们', '基托先生', '贵族男人', '亚尼艾斯修瓦里埃德米兰', '年轻将军', '拉·瓦利埃尔公爵夫人', '罗杰姆', '德?维努依佑', '士官候补生', '年轻女性', '托罗尔鬼们', '女王安丽埃塔', '参谋长', '德．坡瓦奇将军', '谢雷斯坦', '敌方佣兵', '德?坡瓦奇将军', '第二龙骑士中队队长', '赛弗林', '少年骑士们', '侍童', '兵站参谋', '传令士兵', '骑士队队长', '联络士兵', '小孩子们', '同学', '吉姆', '露可夏娜', '特里斯塔尼亚市民', '同班学生', '缪兹尼多尼隆', '精灵男子', '韵龙', '塔芭莎的义妹', '老爹', '卡莉娜', '蓝发女子', '精灵贝塔夏尔', '威托利奥', '卫兵2', '雀斑少年', '同班女学生', '褐发少女', '一年级女孩子们', '战少女', '泽尔布斯特小姐', '另一个少年', '追捕者', '同僚们', '教皇维特里奥·圣·李巴利', '维特里奥·塞雷瓦雷', '布里吉塔', '鼹鼠贝尔丹蒂', '骑乘者', '金发女骑士', '阿鲁奴鲁夫', '东蔷薇骑士团', '教皇维托里奥', '野狼', '炮术士官', '炮龟兵', '加里亚领事', '朱力欧·切萨雷', '加里亚士兵', '秃头男人', '女性代表', '巴索·卡斯特莫尔', '女孩们', '警察', '威尤氏', '中年女性', '观众', '老人', '猎人', '龙形卡哥伊鲁', '阿婆', '人群中一个人', '人群中另一个人', '被害者的两亲', '老父亲', '老妇人', '玛萨塔婆婆', '雷昂', '化妆师', '居民', '娜塔莉', '曼提柯尔像', '学生1', '学生2', '风龙希儿菲朵', '侍女少女', '少女侍从', '店员', '黑衣总管', '接待女子', '周围客人', '老板吉尔摩阿满', '平民客人', '庄家', '多米尼克', '装成米诺塔洛斯的男人', '装成米诺塔洛斯的男子', '尼娜的母亲', '侍从们', '巨大火龙', '使魔', '希尔菲朵', '绅士', '兜帽男子', '官员', '头', '小女孩儿', '男村民', '哥布灵神官', '罗德巴尔德', '男爵夫人', '尤尔班的卡哥伊鲁', '罗宾', '吉露妹妹', '露易丝的魔杖', '中年魔法装扮者', '格拉贝鲁', '中年的魔法使', '未知', '奥鲁德·奥斯曼', '美丽卷发的少女', '教师、路易丝', '马鲁科卢姆', '教室里的某人', '丘鲁克的火蜥', '马尼坎', '艾依基亚克斯', '格姆力', '塔芭莎特', 'MEIJI', '盖托老师', '公主', '学院长奥斯曼', '巨大鼹鼠', '家伙们', '服务员', '男人5', '男人6', '男人7', '男人8', '男人9', '男人10', '男人11', '男人12', '男人13', '男人14', '男人15', '男人16', '男人17', '戴着白色假面的男人', '胖胖的店主', '白假面男子', '空贼首领', '年迈的王', '老王', '王党派的贵族们', '威尔士皇太子', '希尔芙德', '魔法卫士队队员', '曼提柯尔队魔法师', '金发美少女', '眼镜小女孩', '身材高挑女性', '香水之蒙莫兰希', '巨大的鼹鼠', '奥利弗·克伦威尔', '亨利·博伍德', '长发女性', '瓦尔德德', '青铜女武神', '塔巴撒的风龙', '基修父亲', '塔芭撒', '阿比昂舰队', '舰队军官', '拉罗舍尔舰长', '约翰斯顿', '另一大臣', '某贵族', '军机马萨里尼', '龙骑士队员', '敌方人员', '未知人', '亚尔比昂军的贵族们', '威尔斯', '男人的同伙', '青蛙', '袭击者(高个子)', '袭击者(小个子)', '近卫魔法卫士', '鹰马队成员', '受伤的人', '马匹', '棋尔', '才人和琪尔可', '他们', '店员们', '店里的女孩们', '店里的女孩子们', '第三个男生', '高年级生', '男生们', '托妮', '周围女生', '二年级男生', '托妮?夏兰特', '火焰的女王', '年长的贵族', '商人男人', '德?拉?瓦利埃尔', '佣人', '枪士队观众', '瓦尔德子爵', '土块之芙卡', '议长兼初代皇帝', '另一位将军', '胖胖的将军', '姐姐们', '公爵夫人', '母亲和姐姐', '画眉鸟', '托罗尔鬼', '枢机卿马萨林', '小个子男人', '穿铁盔甲的将军', '贵族少年1', '贵族少年2', '贵族少年', '另一个贵族少年', '第三个贵族少年', '胖少年', '一骑龙骑士', '蛇形玩偶', '年长的枪士队队员', '另一个枪士队队员', '某个队员', '某女生', '好几根魔法箭矢', '小胖子龙骑士', '长官', '亚修.彭德尔顿', '吉尔伯特和赛弗林', '第二中队的龙骑士们', '龙骑士中队成员', '参谋总部将军', '亲卫队士兵', '阁僚', '老枪兵们', '泥人偶们', '兽人鬼们', '佣兵副官', '卢涅若', '基修的哥哥', '传说之剑', '瓦利埃尔小姐', '卢涅和基修', '马尔可', '下属参谋', '魔法骑士们', '指挥缨枪队的魔法师指挥官', '指挥弓兵队的年轻指挥官', '敌方剑士', '剑士', '马里科尔努和基修', '爱玛', '妈妈', '风龙亚兹洛', '童女生们', '得尔弗伦格', '约瑟夫的仆人', '安丽埃塔等人', '黑色长袍的女性', '魔法人偶', '贝儿朵莉丝', '阿里', '希瓦尔兹老师', '女学生', '厨师', '马车中的安莉艾塔', '女性民众', '提姆利', '戴眼镜男学生', '醉汉', '夜之女郎', '西瓦尔兹老师', '才人和塔芭莎', '旁人', '青发女子', '曼提苛尔队长', '曼提苛尔队队长德-塞萨尔', '父王', '理查', '西奥门德', '马力科鲁姆', '莫莫兰西', '其他人', '红发女郎', '粉发和金发的少女', '基修/马利蔻尔奴', '才人等人', '高个精灵', '拉.瓦利艾尔家人', '六人', '贵族女学生', '女佣和女子应援团的女孩子们', '沙罗曼蛇', '女图书管理员', '图书管理员', '才人和西艾斯塔', '国王', '约瑟夫的心腹们', '缪斯', '维塔夏尔', '土之哥雷姆', '问题人物', '水精灵骑士队的队员们', '另一人', '褐发女孩', '贝亚托里斯的手下', '骑士团团长', '基修的青铜之战少女', '一个骑士', '空中骑士团成员', '马利寇尔奴大人', '所有人', '未知人物2', '未知人物3', '未知人物4', '未知人物5', '其他少女们', '孩子', '年长的孩子', '男学生们', '骑士队的学生们', '加尔罗·克里斯狄亚诺·特伦博迪诺', '炎之龙', '水精灵骑士队员', '切扎雷阁下', '圣乐队', '圣堂骑士团', '亞尼艾斯', '马达夫', '基修一群人', '旅行者', '守望员', '士官们', '传令', '罗马尼亚舰队', '阿尔比昂', '华都', '参拜者', '才人的母亲', '骑士傀儡们', '军官和骑士们', '虎式坦克', '水精灵骑士的队员们', '丘鲁克和水精灵骑士队的少年们', '队友', '城下的人', '水精灵骑士团成员', '骑士团成员', '马萨里尼枢機卿', '副长', '年长的圣堂骑士', '圣骑士', '艾多利昂', '基德老师', '琪可尔', '平贺氏', '威尤', '拉·瓦里埃尔', '斯卡隆店长', '男性', '骑兵队', '前面的女骑士', '工人', '其他石匠', '巴利贝里尼', 'DM', '领地上的人民', '其他贵族们', '大贵族', '贵族女子', '银发少女', '红发少女', '达米安', '拉瓦利艾尔公爵']\n"
     ]
    }
   ],
   "source": [
    "matplotlib.rcParams['font.sans-serif'] = ['SimHei']\n",
    "\n",
    "import os\n",
    "rename_map=json.load(open(rename_map_path,'r',encoding='utf-8')) if os.path.exists(rename_map_path) else {}\n",
    "rename_map_added={'塔巴萨':'塔芭莎','塔巴莎':'塔芭莎','雪丝塔':'谢斯塔','摩莫朗希':'蒙莫朗西','夏尔洛特':'夏洛特','希尔菲德':'希儿菲朵','西布鲁兹':'谢弗勒兹','谢丝塔':'谢斯塔','路易丝':'露易丝','琪尔可':'丘鲁克','芙卡':'芙凯','露易兹':'露易丝'}\n",
    "rename_map.update(rename_map_added)\n",
    "json.dump(rename_map, open(rename_map_path, 'w', encoding='utf-8'), ensure_ascii=False, indent=2)\n",
    "print('load, update and save rename_map at path:', rename_map_path)\n",
    "\n",
    "# we use the original name in history text, but the new name for labeling\n",
    "\n",
    "char_freq={}\n",
    "for item in all_content:\n",
    "    if 'char' in item:\n",
    "        char=rename_map.get(item['char'], item['char'])\n",
    "        if char not in char_freq:\n",
    "            char_freq[char]=0\n",
    "        char_freq[char]+=1\n",
    "char_freq={k: v for k, v in sorted(char_freq.items(), key=lambda item: item[1], reverse=True)}\n",
    "selected_char=list(sorted(char_freq.keys(), key=lambda x: char_freq[x], reverse=True)[:9999])\n",
    "\n",
    "def list_char_items(content, char):\n",
    "    for item in content:\n",
    "        if item.get('char', None) == char:\n",
    "            print(item)\n",
    "\n",
    "\n",
    "plt.figure(figsize=(15, 2))\n",
    "plt.bar(list(char_freq.keys())[:60], list(char_freq.values())[:60])\n",
    "plt.xticks(rotation=75)\n",
    "plt.ylabel('count')\n",
    "plt.yscale('log')\n",
    "plt.ylim(10,np.max(list(char_freq.values()))*2)\n",
    "plt.show()\n",
    "\n",
    "print('list of characters and their frequencies:',char_freq)\n",
    "print('rename_map:', rename_map)\n",
    "print('selected_char:', selected_char)\n",
    "\n",
    "# revising character name typo\n",
    "\n",
    "def revise_name(obj):\n",
    "    if isinstance(obj,str):\n",
    "        for from_name,to_name in rename_map.items():\n",
    "            obj=obj.replace(from_name,to_name)\n",
    "        return obj,True\n",
    "    else:\n",
    "        return obj,False\n",
    "\n",
    "def revise_json_object_recursive(obj,revise_func):\n",
    "    obj,terminate=revise_func(obj)\n",
    "    if not terminate:\n",
    "        if isinstance(obj,dict):\n",
    "            for key,value in obj.items():\n",
    "                obj[key]=revise_json_object_recursive(value,revise_func)\n",
    "        if isinstance(obj,list):\n",
    "            for i in range(len(obj)):\n",
    "                obj[i]=revise_json_object_recursive(obj[i],revise_func)\n",
    "    return obj\n",
    "\n",
    "def fix_names(slient=False):\n",
    "    global chunks,all_content\n",
    "    all_content=revise_json_object_recursive(all_content,revise_func=revise_name)\n",
    "    chunks=revise_json_object_recursive(chunks,revise_func=revise_name)\n",
    "    print('revise name done (not saved to original chunks)')\n",
    "    json.dump(all_content, open(merged_log_path, 'w', encoding='utf-8'), ensure_ascii=False, indent=2)\n",
    "    print('save merged_log_path:', merged_log_path)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 将chunk合并成章节，并进行总结"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2544b45c30c14641af85bff2501f927d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/188 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from glob import glob\n",
    "\n",
    "def reload_chapters(slient=False):\n",
    "    global chapters\n",
    "    chapters={}\n",
    "    for chapter_summary_path in glob(os.path.join(chapter_summary_dir, f'CHAPTER*.json')):\n",
    "        if not slient:print('load',chapter_summary_path)\n",
    "        chapter_summary = json.load(open(chapter_summary_path, 'r', encoding='utf-8'))\n",
    "        iChapter = int(chapter_summary_path.split('.')[-2].split('_')[-1])\n",
    "        chapters[iChapter]=chapter_summary\n",
    "    chapters={k: v for k, v in sorted(chapters.items(), key=lambda item: item[0])}\n",
    "    if not slient:print('reload',len(chapters),'chapters')\n",
    "\n",
    "\n",
    "def save_chapter(iChapter,chapter,slient=False,do_save_all_chapters=True):\n",
    "    global chapters\n",
    "    chapters[iChapter]=chapter\n",
    "    json.dump(chapter, open(os.path.join(chapter_summary_dir, f'CHAPTER_{iChapter}.json'), 'w', encoding='utf-8'), ensure_ascii=False, indent=2)\n",
    "    if not slient: print('save chapter',iChapter,'to',os.path.join(chapter_summary_dir, f'CHAPTER_{iChapter}.json'))\n",
    "    if do_save_all_chapters:\n",
    "        save_all_chapters(slient=slient)\n",
    "    \n",
    "\n",
    "def save_all_chapters(slient=False):\n",
    "    global chapters\n",
    "    json.dump(chapters, open(os.path.join(all_chapters_path), 'w', encoding='utf-8'), ensure_ascii=False, indent=2)\n",
    "    if not slient:print('save all chapters to',all_chapters_path)\n",
    "    with open(os.path.join(all_summary_path), 'w', encoding='utf-8') as f:\n",
    "        for iChapter,chapter in sorted(chapters.items()):\n",
    "            f.write(chapter['title']+'\\n')\n",
    "            if 'summary' in chapter:\n",
    "                f.write(chapter['summary']+'\\n')\n",
    "            if 'characters' in chapter:\n",
    "                for char,info in chapter['characters'].items():\n",
    "                    all_info=' '.join([str(info.get(k,'')) for k in ['role','goal','obstacle','plan']])\n",
    "                    f.write(f'{char}: {all_info}\\n')\n",
    "            f.write('\\n')\n",
    "            if 'chunk_summaries' in chapter:\n",
    "                for summary in chapter['chunk_summaries']:\n",
    "                    f.write(summary+'\\n')\n",
    "            f.write('\\n\\n')\n",
    "    if not slient:print('save all summaries to',all_summary_path)\n",
    "\n",
    "\n",
    "def update_chapters_from_chunks(slient=False):\n",
    "    global chapters, chunks\n",
    "\n",
    "    for chapter in chapters.values():\n",
    "        chapter['chunk_summaries']=[]\n",
    "        chapter['chunk_ids']=[]\n",
    "    for iChunk,chunk in sorted(chunks.items()):\n",
    "        chapter=chapters.get(int(chunk['chapter_id']), {}) \n",
    "        chapter['chapter_id']=chunk['chapter_id']\n",
    "        chapter['title']=chunk['title']\n",
    "        chapter['chunk_ids']=chapter.get('chunk_ids', [])+[chunk['id']]\n",
    "        if 'summary' in chunk:\n",
    "            chapter['chunk_summaries']=chapter.get('chunk_summaries', [])+[chunk['summary']]\n",
    "        chapters[int(chunk['chapter_id'])]=chapter\n",
    "\n",
    "    for iChapter,chapter in tqdm(chapters.items()):\n",
    "        save_chapter(int(iChapter),chapter,slient=slient)\n",
    "\n",
    "\n",
    "\n",
    "reload_chapters(slient=True)\n",
    "\n",
    "update_chapters_from_chunks(slient=True)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 将chunk的summary和bg_hint合并，作为chapter的summary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain.chains.base import Chain\n",
    "import json\n",
    "import re\n",
    "from langchain.docstore.document import Document\n",
    "from langchain.text_splitter import RecursiveCharacterTextSplitter\n",
    "\n",
    "class FurtherSummaryStoryChain(Chain):\n",
    "    llm:object=None\n",
    "    get_bg_hint:object=None\n",
    "    refine_bg_hint:bool=True\n",
    "    bg_hint_num:int=10\n",
    "    bg_hint_len:int=400\n",
    "    retrospect_len:int=100\n",
    "    @property\n",
    "    def input_keys(self):\n",
    "        return ['text']\n",
    "    @property\n",
    "    def output_keys(self):\n",
    "        return ['summary']\n",
    "    prmopt='''\n",
    "可能有关的背景材料\n",
    "{background}\n",
    "【前情提要：{retrospect}】\n",
    "请去除下面情节梗概中重复的部分，请尽量不要丢掉信息。不要加入额外的推测，请忠实于原著。不要加入没有意思的套话。\n",
    "{text}\n",
    "精简后的情节梗概：'''\n",
    "    def _call(self,inputs:dict)->dict:\n",
    "        print('requesting summary')\n",
    "        bghint_input=inputs.get('summary',inputs['text'])\n",
    "        if self.get_bg_hint is not None:\n",
    "            bg_hint=self.get_bg_hint(bghint_input,k=self.bg_hint_num,max_text_len=self.bg_hint_len)\n",
    "        else:\n",
    "            bg_hint='（无）'\n",
    "        inputs['background']=bg_hint[-self.bg_hint_len:]\n",
    "        inputs['retrospect']=inputs.get('retrospect','(无)')[-self.retrospect_len:]\n",
    "        query=self.prmopt.format(**inputs)\n",
    "        response=self.llm(query)\n",
    "        response=re.sub(r'[\\s]*[\\d]+[\\.、][\\s]*','',response)\n",
    "        if self.refine_bg_hint and 'summary' not in inputs:\n",
    "            print('refine the summary by using the bg hint genereated by the current summary')\n",
    "            return self._call({**inputs,'summary':response})\n",
    "        else:\n",
    "            return {'summary':response}\n",
    "    \n",
    "\n",
    "class FinalSummaryStoryChain(Chain):\n",
    "    llm:object=None\n",
    "    get_bg_hint:object=None\n",
    "    refine_bg_hint:bool=False\n",
    "    bg_hint_num:int=10\n",
    "    bg_hint_len:int=400\n",
    "    retrospect_len:int=100\n",
    "    @property\n",
    "    def input_keys(self):\n",
    "        return ['text']\n",
    "    @property\n",
    "    def output_keys(self):\n",
    "        return ['summary']\n",
    "    prmopt='''\n",
    "可能有关的背景材料\n",
    "{background}\n",
    "【前情提要：{retrospect}】\n",
    "请非常简要地概括下面的故事，字数限制在五十字以内，抓住重点，抓住关键的情节走向和可能对后文产生影响的重大细节。不要加入额外的推测，请忠实于原著。不要加入没有意思的套话。\n",
    "{text}\n",
    "精简后的情节梗概：'''\n",
    "    def _call(self,inputs:dict)->dict:\n",
    "        print('requesting summary')\n",
    "        bghint_input=inputs.get('summary',inputs['text'])\n",
    "        if self.get_bg_hint is not None:\n",
    "            bg_hint=self.get_bg_hint(bghint_input,k=self.bg_hint_num,max_text_len=self.bg_hint_len)\n",
    "        else:\n",
    "            bg_hint='（无）'\n",
    "        inputs['background']=bg_hint[-self.bg_hint_len:]\n",
    "        inputs['retrospect']=inputs.get('retrospect','(无)')[-self.retrospect_len:]\n",
    "        query=self.prmopt.format(**inputs)\n",
    "        response=self.llm(query)\n",
    "        response=re.sub(r'[\\s]*[\\d]+[\\.、][\\s]*','',response)\n",
    "        if self.refine_bg_hint and 'summary' not in inputs:\n",
    "            print('refine the summary by using the bg hint genereated by the current summary')\n",
    "            return self._call({**inputs,'summary':response})\n",
    "        else:\n",
    "            return {'summary':response}\n",
    "\n",
    "\n",
    "further_summary_story_chain=FurtherSummaryStoryChain(llm=llm,get_bg_hint=get_bg_hint,\n",
    "                                            bg_hint_len=bg_hint_len,retrospect_len=retrospect_len)\n",
    "final_summary_story_chain=FinalSummaryStoryChain(llm=llm,get_bg_hint=get_bg_hint,\n",
    "                                            bg_hint_len=bg_hint_len,retrospect_len=retrospect_len) \n",
    "text_splitter=RecursiveCharacterTextSplitter(chunk_size=summary_chunk_len,chunk_overlap=summary_chunk_overlap)\n",
    "\n",
    "def further_summary_story(summaries,retrospect=None,final_summary_chunk_len=250):\n",
    "    all_summary_text='\\n'.join([summary.replace('\\n','') for summary in summaries])\n",
    "    retrospect=retrospect or '（无）'\n",
    "    if len(all_summary_text)<=final_summary_chunk_len:\n",
    "        print('length of all summaries is less than final_summary_chunk_len, no further summarizing')\n",
    "        final_summary=final_summary_story_chain({'text':all_summary_text,'retrospect':retrospect})['summary']\n",
    "        if len(summaries)==1:\n",
    "            return [final_summary]\n",
    "        else:\n",
    "            return [final_summary,['\\n'.join(summaries)]]\n",
    "    else:\n",
    "        print('further summarizing story, num of summaries at current level:',len(summaries), 'total length:',sum([len(summary) for summary in summaries]))\n",
    "        text_chunks=text_splitter.split_text(all_summary_text)\n",
    "        chunk_summaries=[]\n",
    "        chunk_retrospect=''\n",
    "        for text_chunk in text_chunks:\n",
    "            all_retrospect=(retrospect+'。'+chunk_retrospect).replace('\\w','')\n",
    "            chunk_summary=further_summary_story_chain({'text':text_chunk,'retrospect':all_retrospect})['summary']\n",
    "            chunk_retrospect=chunk_summary\n",
    "            chunk_summaries.append(chunk_summary)\n",
    "        cascades=further_summary_story(chunk_summaries,retrospect,final_summary_chunk_len=final_summary_chunk_len)\n",
    "        return cascades+[chunk_summaries]\n",
    "\n",
    "# llm.show_query=llm.show_response=True\n",
    "\n",
    "# summaries=chapters[22]['chunk_summaries']\n",
    "# # summaries=chapters[21]['chunk_summaries']+chapters[22]['chunk_summaries']+chapters[23]['chunk_summaries']\n",
    "# summary_cascades=further_summary_story(summaries,final_summary_chunk_len=final_summary_chunk_len)\n",
    "# print('final summary:',summary_cascades[0])\n",
    "# print('semifinal summary:',summary_cascades[1])\n",
    "# print('summary cascade:',summary_cascades)\n",
    "\n",
    "# llm.show_query=llm.show_response=False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "from my_chains import Text2JsonChain\n",
    "\n",
    "\n",
    "\n",
    "class ExtractStoryElementsChain(Text2JsonChain):\n",
    "    get_bg_hint:object=None\n",
    "    bg_hint_num:int=10\n",
    "    bg_hint_len:int=400\n",
    "    summary_len:int=100\n",
    "    retrospect_len:int=100\n",
    "    story_summary_chain:object=None\n",
    "    prompt='''可能有关的背景信息：（不要加入到回复中）\n",
    "{background}\n",
    "【前情提要：{retrospect}】\n",
    "【本段概要：{summary}】\n",
    "对下文中的主要角色，分条提取角色的目标动机goal，遇到的主要阻碍obstacle，所做出的行为与决策plan，以及角色在故事中起到的作用role。\n",
    "只需要输出主要角色，按你认为的角色的重要性排序。\n",
    "输出格式：{{\n",
    "    \"小明\":{{\n",
    "        \"goal\":\"上街买菜\",\n",
    "        \"obstacle\":\"没钱\",\n",
    "        \"plan\":\"找人借钱\",\n",
    "        \"role\":\"主角的朋友，负责买菜\"\n",
    "    }}\n",
    "}}\n",
    "确保json格式正确。\n",
    "不加自己的推测想法，请忠实于原著，不要盲目正面。不要加入没有意思的套话。请修正可能被搞混淆的角色名字。\n",
    "可能有关的背景信息：（不要加入到回复中）\n",
    "文本：\n",
    "{text}\n",
    "json：'''\n",
    "    @property\n",
    "    def output_keys(self):\n",
    "        return super().output_keys+['summary','background','retrospect']\n",
    "    def _call(self,inputs:dict)->dict:\n",
    "        summary=self.story_summary_chain({**inputs,'summary':inputs['text']})['summary'] if self.story_summary_chain is not None and 'summary' not in inputs else inputs.get('summary','（无）')\n",
    "        bg_hint_input=summary if 'summary' in inputs else inputs['text']\n",
    "        bg_hint=self.get_bg_hint(bg_hint_input,k=self.bg_hint_num,max_text_len=self.bg_hint_len) if self.get_bg_hint is not None else '（无）'\n",
    "        retrospect=inputs.get('retrospect','（无）')\n",
    "\n",
    "        inputs['summary']=summary.replace('\\n','').strip()[:self.summary_len]\n",
    "        inputs['background']=bg_hint.strip()[-self.bg_hint_len:]\n",
    "        inputs['retrospect']=retrospect.replace('\\n','').strip()[-self.retrospect_len:]\n",
    "\n",
    "        print('requesting ExtractStoryElementsChain')\n",
    "        outputs=super()._call(inputs)\n",
    "        outputs['summary']=summary.replace('\\n','')\n",
    "        outputs['background']=bg_hint\n",
    "        outputs['retrospect']=retrospect.replace('\\n','')\n",
    "        return outputs\n",
    "    \n",
    "extract_story_elements_chain=ExtractStoryElementsChain(llm=llm,get_bg_hint=get_bg_hint,story_summary_chain=further_summary_story_chain,\n",
    "                                            bg_hint_len=bg_hint_len,retrospect_len=retrospect_len)\n",
    "\n",
    "\n",
    "# llm.show_query=llm.show_response=True\n",
    "# story_elements=extract_story_elements_chain({'text':'\\n'.join(summary_cascades[1]),'summary':summary_cascades[0]})\n",
    "# llm.show_query=llm.show_response=False"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 一条一条提交给llm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "save merged log to ./output_zero/merged_log.json\n",
      "revise name done (not saved to original chunks)\n",
      "save merged_log_path: ./output_zero/merged_log.json\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8a6724303396434e891ce9ad45511905",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/188 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "nothing to do\n"
     ]
    }
   ],
   "source": [
    "reload_chunks(slient=True)\n",
    "fix_names(slient=True)\n",
    "reload_chapters(slient=True)\n",
    "update_chapters_from_chunks(slient=True)\n",
    "llm.show_query=llm.show_response=False\n",
    "\n",
    "failed_iChapters_this_run=set()\n",
    "import time\n",
    "\n",
    "while True:\n",
    "    # sleep .5 s\n",
    "    time.sleep(.5)\n",
    "    # reload_chunks(slient=True)\n",
    "    # fix_names()\n",
    "    reload_chapters(slient=True)\n",
    "    # update_chapters_from_chunks(slient=True)\n",
    "    found=False\n",
    "    for iChapter,chapter in chapters.items():\n",
    "        if 'characters' in chapter and 'characters' in chapter:\n",
    "            continue\n",
    "        if iChapter in failed_iChapters_this_run:\n",
    "            continue\n",
    "        if 'chunk_summaries' not in chapter:\n",
    "            continue\n",
    "        if len(chapter['chunk_summaries'])!=len(chapter['chunk_ids']):\n",
    "            continue\n",
    "        found=True\n",
    "        break\n",
    "    if not found:\n",
    "        break\n",
    "    try:\n",
    "        print('processing chapter',iChapter,chapter['title'])\n",
    "        retrospect=chapters[iChapter-1]['summary'] if iChapter-1 in chapters and 'summary' in chapters[iChapter-1] else '（无）'\n",
    "        if 'summary_cascades' not in chapter:\n",
    "            chapter['summary_cascades']=further_summary_story(chapter['chunk_summaries'],final_summary_chunk_len=final_summary_chunk_len)\n",
    "            chapter['summary']=chapter['summary_cascades'][0]\n",
    "            save_chapter(iChapter,chapter)\n",
    "        if 'characters' not in chapter:\n",
    "            summary=chapter['summary']\n",
    "            text='\\n'.join(chapter['summary_cascades'][1]) if len(chapter['summary_cascades'])>1 else summary\n",
    "            response=extract_story_elements_chain({'text':text,'summary':summary,'retrospect':retrospect})\n",
    "            if response['content'] is not None:\n",
    "                chapter['characters']=response['content']\n",
    "            else:\n",
    "                print('story elements extraction failed')\n",
    "                print(response['error'])\n",
    "                chapter['characters_failed']=response['response']\n",
    "                failed_iChapters_this_run.add(iChapter)\n",
    "            save_chapter(iChapter,chapter)\n",
    "    except Exception as e:\n",
    "        failed_iChapters_this_run.add(iChapter)\n",
    "        print(e)\n",
    "        print('chapter',iChapter,'failed')\n",
    "        continue\n",
    "print('nothing to do')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Trash code after here"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "assert False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "class FinalSummaryStoryChain(Chain):\n",
    "    llm:object=None\n",
    "    get_bg_hint:object=None\n",
    "    bg_hint_num:int=10\n",
    "    bg_hint_len:int=400\n",
    "    retrospect_len:int=100\n",
    "    @property\n",
    "    def input_keys(self):\n",
    "        return ['text','retrospect']\n",
    "    @property\n",
    "    def output_keys(self):\n",
    "        return ['summary']\n",
    "    prmopt='''\n",
    "可能有关的背景材料\n",
    "{background}\n",
    "【前情提要：{retrospect}】\n",
    "问：下面是一个完整的小说章节中不同段落的分段梗概，请用连贯的语言详述一下本小说章节。不要加入额外的推测，请忠实于原著，不要加入没有意思的概概而谈的套话。\n",
    "{text}\n",
    "答：'''\n",
    "    def _call(self,inputs:dict)->dict:\n",
    "        bghint_input=inputs['text']\n",
    "        if self.get_bg_hint is not None:\n",
    "            bg_hint=self.get_bg_hint(bghint_input,k=self.bg_hint_num,max_text_len=self.bg_hint_len)\n",
    "        else:\n",
    "            bg_hint='（无）'\n",
    "        inputs['background']=bg_hint[-self.bg_hint_len:]\n",
    "        inputs['retrospect']=inputs['retrospect'][-self.retrospect_len:]\n",
    "        query=self.prmopt.format(**inputs)\n",
    "        print('requesting summary')\n",
    "        response=self.llm(query)\n",
    "        response=re.sub(r'[\\s]*[\\d]+[\\.、][\\s]*','',response)\n",
    "        return {'summary':response}"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "tensorflow",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.10"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}