File size: 4,734 Bytes
78efe79
440418c
f3985af
dc80b35
 
22dee1c
fe505c6
407a575
9c071a8
 
 
 
c4be42e
9c071a8
c4be42e
 
 
 
 
9c071a8
 
c4be42e
 
9c071a8
 
c4be42e
9c071a8
 
32c38ef
f3985af
440418c
1831164
440418c
22dee1c
440418c
22dee1c
 
08baccf
dc80b35
 
 
 
40d0e92
74ccf1c
12bb502
 
 
78efe79
08baccf
 
dc80b35
08baccf
78efe79
40d0e92
dc80b35
 
78efe79
5f46973
 
 
 
 
 
aab7a25
5f46973
13feae4
5f46973
 
9c071a8
13feae4
 
5f46973
 
 
aab7a25
5f46973
 
aab7a25
 
 
5f46973
ce5155e
 
9c071a8
 
aab7a25
9c071a8
aab7a25
 
9c071a8
aab7a25
 
9c071a8
 
 
 
 
ce5155e
 
9c071a8
 
ce5155e
 
9c071a8
 
 
 
a8a69ff
 
9c071a8
 
01dd380
aab7a25
 
 
 
 
 
a8a69ff
34428f1
dc80b35
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
import discord
import logging
import os
from huggingface_hub import InferenceClient
import asyncio
import subprocess
from datasets import load_dataset

# ํ˜„์žฌ ์ž‘์—… ๋””๋ ‰ํ† ๋ฆฌ ์ถœ๋ ฅ
print("Current Working Directory:", os.getcwd())

# ๋ฐ์ดํ„ฐ์…‹ ํŒŒ์ผ ์ด๋ฆ„
data_files = ['train_0.csv', 'train_1.csv', 'train_2.csv', 'train_3.csv', 'train_4.csv', 'train_5.csv']

# ํ˜„์žฌ ์ž‘์—… ๋””๋ ‰ํ† ๋ฆฌ์— ๋ชจ๋“  ํŒŒ์ผ์ด ์žˆ๋Š”์ง€ ํ™•์ธ
missing_files = [file for file in data_files if not os.path.exists(file)]
if missing_files:
    print(f"Missing files: {missing_files}")
    # ํ•„์š”ํ•œ ๊ฒฝ์šฐ ์ž‘์—… ๋””๋ ‰ํ† ๋ฆฌ ๋ณ€๊ฒฝ
    os.chdir('/home/user/app')
    print("Changed directory to:", os.getcwd())
else:
    print("All files are present in the current directory.")

# ๋ฐ์ดํ„ฐ์…‹ ๋กœ๋“œ
law_dataset = load_dataset('csv', data_files=data_files)
print("Dataset loaded successfully.")

# ๋กœ๊น… ์„ค์ •
logging.basicConfig(level=logging.DEBUG, format='%(asctime)s:%(levelname)s:%(name)s: %(message)s', handlers=[logging.StreamHandler()])

# ์ธํ…ํŠธ ์„ค์ •
intents = discord.Intents.default()
intents.message_content = True
intents.messages = True
intents.guilds = True
intents.guild_messages = True

# ์ถ”๋ก  API ํด๋ผ์ด์–ธํŠธ ์„ค์ •
hf_client = InferenceClient("CohereForAI/c4ai-command-r-plus", token=os.getenv("HF_TOKEN"))

# ํŠน์ • ์ฑ„๋„ ID
SPECIFIC_CHANNEL_ID = int(os.getenv("DISCORD_CHANNEL_ID"))

# ๋Œ€ํ™” ํžˆ์Šคํ† ๋ฆฌ๋ฅผ ์ €์žฅํ•  ์ „์—ญ ๋ณ€์ˆ˜
conversation_history = []

class MyClient(discord.Client):
    def __init__(self, *args, **kwargs):
        super().__init__(*args, **kwargs)
        self.is_processing = False

    async def on_ready(self):
        logging.info(f'{self.user}๋กœ ๋กœ๊ทธ์ธ๋˜์—ˆ์Šต๋‹ˆ๋‹ค!')
        subprocess.Popen(["python", "web.py"])
        logging.info("Web.py server has been started.")

    async def on_message(self, message):
        if message.author == self.user:
            return
        if not self.is_message_in_specific_channel(message):
            return
        if self.is_processing:
            logging.debug("Currently processing another message, skipping this one.")
            return

        self.is_processing = True
        try:
            response = await generate_response(message)
            if response.strip() == "":
                response = "์ฃ„์†กํ•ฉ๋‹ˆ๋‹ค, ์ œ๊ณตํ•  ์ˆ˜ ์žˆ๋Š” ์ •๋ณด๊ฐ€ ์—†์Šต๋‹ˆ๋‹ค."
            await message.channel.send(response)
        finally:
            self.is_processing = False
            logging.debug("Message processing completed, ready for the next one.")

    def is_message_in_specific_channel(self, message):
        channel_condition = message.channel.id == SPECIFIC_CHANNEL_ID
        thread_condition = isinstance(message.channel, discord.Thread) and message.channel.parent_id == SPECIFIC_CHANNEL_ID
        return channel_condition or thread_condition



async def generate_response(message):
    global conversation_history
    case_number = message.content.strip()
    user_mention = message.author.mention
    system_message = f"{user_mention}, ๋‹ค์Œ์€ ์š”์ฒญํ•˜์‹  ์‚ฌ๊ฑด์˜ ์ „๋ฌธ์ž…๋‹ˆ๋‹ค:"

    # ๋ฐ์ดํ„ฐ ๊ฒ€์ƒ‰ ๋ฐ ์‘๋‹ต ์ค€๋น„
    full_text = search_in_dataset_by_case_number(case_number, law_dataset)
    full_response_text = f"{system_message}\n\n{full_text}"
    max_length = 2000
    if len(full_response_text) > max_length:
        for i in range(0, len(full_response_text), max_length):
            part_response = full_response_text[i:i+max_length]
            await message.channel.send(part_response)
        logging.debug(f'Full model response sent: {full_response_text}')
        conversation_history.append({"role": "assistant", "content": full_response_text})
    else:
        await message.channel.send(full_response_text)
        logging.debug(f'Full model response sent: {full_response_text}')
        conversation_history.append({"role": "assistant", "content": full_response_text})


def search_in_dataset(query, dataset):
    for record in dataset['train']:
        # '์‚ฌ๊ฑด๋ช…'์ด None์ด ์•„๋‹Œ์ง€ ํ™•์ธํ•˜๊ณ , query๊ฐ€ ํ•ด๋‹น ๊ฐ’์— ํฌํ•จ๋˜์–ด ์žˆ๋Š”์ง€ ๊ฒ€์‚ฌ
        if record['์‚ฌ๊ฑด๋ช…'] and query in record['์‚ฌ๊ฑด๋ช…']:
            return record['์‚ฌ๊ฑด๋ฒˆํ˜ธ']
    return "๊ด€๋ จ ๋ฒ•๋ฅ  ์ •๋ณด๋ฅผ ์ฐพ์„ ์ˆ˜ ์—†์Šต๋‹ˆ๋‹ค."

def search_in_dataset_by_case_number(case_number, dataset):
    for record in dataset['train']:
        if record['์‚ฌ๊ฑด๋ฒˆํ˜ธ'] == case_number:
            return record.get('์ „๋ฌธ', "ํ•ด๋‹น ์‚ฌ๊ฑด์— ๋Œ€ํ•œ ์ „๋ฌธ ์ •๋ณด๊ฐ€ ์—†์Šต๋‹ˆ๋‹ค.")
    return "๊ด€๋ จ ๋ฒ•๋ฅ  ์ •๋ณด๋ฅผ ์ฐพ์„ ์ˆ˜ ์—†์Šต๋‹ˆ๋‹ค."


if __name__ == "__main__":
    discord_client = MyClient(intents=intents)
    discord_client.run(os.getenv('DISCORD_TOKEN'))