Extracting character conversations in Arknights
First, git clone ArknightsData
(Thanks to Dimbreath's project!)
Simply run parse.py after specifying ArknightsData directory path and lang:
line 8: repo = "D:/code/github/ArknightsData"
line 9: lang = "en-US"
This results in story.jsonl in data folder. Each line is a JSON dict, containing the following fields:
context: description of story context
conversation: ChatGPT-message-like list of dict, with role, content fields
meta: meta info
{
"context": "混入龙门的整合运动数量远超想象,并且他们的行动越来越不掩饰。\n在碎骨的带领下,他们竟然主动向罗德岛发起进攻。",
"conversation": [
{
"role": "雷蛇",
"content": "——"
},
{
"role": "雷蛇",
"content": "怎么有这么多整合运动!"
},
{
"role": "整合运动成员",
"content": "诶呀!"
},
{
"role": "芙兰卡",
"content": "......"
},
...
{
"role": "阿米娅",
"content": "(趁他们不敢继续进攻的时候,向低层转移!)"
},
{
"role": "阿米娅",
"content": "(现在必须要扩大战场!)"
},
{
"role": "芙兰卡",
"content": "(知道了,那就边跑边打吧!)"
},
{
"role": "阿米娅",
"content": "与你有什么关系?"
},
{
"role": "碎骨",
"content": "......你!"
},
{
"role": "碎骨",
"content": "......"
},
{
"role": "碎骨",
"content": "我的同胞,动手吧!"
}
],
"meta": {
"source": "obt/main\\level_main_02-09_beg.txt",
"characters": [
"碎骨",
"雷蛇",
"芙兰卡",
"整合运动成员",
"阿米娅"
]
}
}