Compare commits

...

2 Commits

Author SHA1 Message Date
89300e74e5 Added proof of concept for discord bot with OpenAI image recognition. 2025-06-28 11:25:55 +05:00
8e9fc5db8b Initial Commit 2025-06-28 08:57:19 +05:00
2 changed files with 117 additions and 0 deletions

50
discord_bot.py Normal file
View File

@ -0,0 +1,50 @@
import os
import discord
from discord.ext import commands
from openai_interaction.image_recognition_test import image_recognition
# Загружаем токен из переменной среды
TOKEN = os.getenv('SCREAMING_OPOSSUM')
# Создаем клиент с префиксом (не обязателен для слэш-команд)
intents = discord.Intents.default()
intents.message_content = True
bot = commands.Bot(command_prefix='!', intents=intents)
# region Bot Stuff
@bot.event
async def on_ready():
print(f'Logged in as {bot.user} (ID: {bot.user.id})')
print('------')
try:
synced = await bot.tree.sync()
print(f'Synced {len(synced)} command(s).')
except Exception as e:
print(f'Failed to sync commands: {e}')
@bot.tree.command(name='hello', description='Replies with Hello World!')
async def hello_command(interaction: discord.Interaction):
await interaction.response.send_message('AAAAAAAAAAAAAAAAAAAAA!')
@bot.event
async def on_message(message: discord.Message):
if message.author == bot.user:
return
if message.attachments and bot.user in message.mentions:
message_attachments = [
{'filename': i_attachment.filename, 'url': i_attachment.url} for i_attachment in message.attachments
]
await message.channel.send(await image_recognition(message.content, message_attachments))
await bot.process_commands(message)
# endregion
if __name__ == '__main__':
if TOKEN is None:
raise RuntimeError('Environment variable SCREAMING_OPOSSUM is not set')
bot.run(TOKEN)

View File

@ -0,0 +1,67 @@
from openai import OpenAI
client = OpenAI()
SUPPORTED_IMAGE_EXTENSIONS = ('.png', '.jpg', '.jpeg', '.webp', '.gif')
# response = client.responses.create(
# model="gpt-4.1-mini",
# input=[{
# "role": "user",
# "content": [
# {"type": "input_text", "text": "Что на этих картинках и в этом PDF файле?"},
# {
# "type": "input_image",
# "image_url": "https://cdn.discordapp.com/attachments/1381924995994882150/1388385230108360855/2a264d9a62e7420586ee7873f53552d7.webp?ex=6860ca01&is=685f7881&hm=e8e21d0829cc9be3a21f63cb8f0c5065bf76c1de3952dbb524640694dd5037be&'>, <Attachment id=1388385230582321253 filename='",
# },
# {
# "type": "input_image",
# "image_url": "https://cdn.discordapp.com/attachments/1381924995994882150/1388385230582321253/u_312c5f5306d0913c58dd0639fe493bdd_800.png?ex=6860ca01&is=685f7881&hm=d0ebcfaecc2e69551da0cdefdf88c3833f8cbd59d04b035b787ea3150382bdd2&'>",
# },
# ],
# }],
# )
#
# print(response.output_text)
async def image_recognition(content: str, files: list[dict[str: str]]) -> str:
input_data = [{'type': 'input_text', 'text': content}]
input_data.extend(
[
{
'type': 'input_image',
'image_url': i_url.get('url'),
}
for i_url in files if i_url.get('filename').lower().endswith(SUPPORTED_IMAGE_EXTENSIONS)
]
)
print(input_data)
response = client.responses.create(
model='gpt-4.1-mini',
input=[{
'role': 'user',
'content': input_data,
}],
)
return response.output_text
if __name__ == '__main__':
data = {
'content': 'What\'s on these images?',
'files': [
{
'filename': '2a264d9a62e7420586ee7873f53552d7.webp',
'url': 'https://cdn.discordapp.com/attachments/1381924995994882150/1388385230108360855/2a264d9a62e7420586ee7873f53552d7.webp?ex=6860ca01&is=685f7881&hm=e8e21d0829cc9be3a21f63cb8f0c5065bf76c1de3952dbb524640694dd5037be&',
},
{
'filename': 'u_312c5f5306d0913c58dd0639fe493bdd_800.png',
'url': 'https://cdn.discordapp.com/attachments/1381924995994882150/1388385230582321253/u_312c5f5306d0913c58dd0639fe493bdd_800.png?ex=6860ca01&is=685f7881&hm=d0ebcfaecc2e69551da0cdefdf88c3833f8cbd59d04b035b787ea3150382bdd2&',
},
{
'filename': 'shitty_pdf_file.pdf',
'url': 'https://example.com/shitty_pdf_file.pdf'
}
],
}
print(image_recognition(**data))