Can't create guild command about pycord HOT 12 CLOSED

Cannot reproduce. Can you confirm that this still occurs?

from pycord.

Can still reproduce. Maybe a caching issue? Just tried again, I can't reproduce it with my 3-guilds bot, but with a 800-guilds bot.

from pycord.

It runs fetch_guilds and then adds all the guild ids to the dict containing an empty list for each

async for guild in self.fetch_guilds(limit=None):
    update_guild_commands[guild.id] = []

After this, it processes the guild_ids passed into all of the commands

    for command in [cmd for cmd in self.pending_application_commands if cmd.guild_ids is not None]:
        as_dict = command.to_dict()
        for guild_id in command.guild_ids:
            to_update = update_guild_commands[guild_id]

It appears this error could occur if a guild_id passed into one of the commands doesn't actually belong to a guild that the bot is in, could you check if that is true? A simple fix to this would be changing

to_update = update_guild_commands[guild_id]

to

to_update = update_guild_commands.get(guild_id, [])

but that would just be avoiding a bigger issue so I don't think it should be used.

from pycord.

I've tried it again in logging debug mode double checking the guild id, the guild was sent in the guild_create events on startup.

from pycord.

Hello, I am having this issue as well as I am trying to test slash commands in my bot which is in 6.1k servers:

# testing slash commands
@client.slash_command(guild_ids=[667412571728379940])
async def new(
	ctx,
	name: Option(str, "Enter your name"),
	gender: Option(str, "Choose your gender", choices=["Male", "Female", "Other"]),
	age: Option(int, "Enter your age", required=False, default=18),
):
	await ctx.send(f'Hello {name}, it looks like you are {age} years old and {gender}.')
	return

Traceback (most recent call last):
  File "/usr/local/lib/python3.8/dist-packages/discord/client.py", line 351, in _run_event
    await coro(*args, **kwargs)
  File "/usr/local/lib/python3.8/dist-packages/discord/bot.py", line 370, in on_connect
    await self.register_commands()
  File "/usr/local/lib/python3.8/dist-packages/discord/bot.py", line 191, in register_commands
    to_update = update_guild_commands[guild_id]
KeyError: 667412571728379940

from pycord.

@BobDotCom should take a look at this.

from pycord.

I believe this may be an issue with the optimization of guild commands at the moment for larger bots in our library. At the time being I would recommend not using guild commands for large bots until this issue is fixed.

It's also possible it could be happening only on sharded bots, but i haven't explored this possibility yet. Could @lukenamop and @Miolus confirm that they were using sharded bots?

from pycord.

Can reproduce it sharded and non sharded.

from pycord.

from pycord.

Try adding

@bot.event
async def on_ready():
    print(bot.guilds)
    await bot.register_commands()

See if the server you're looking for is there. (You can use a different way to check, but do it in on_ready and before register_commands). It might be an issue with guilds not being cached.

from pycord.

from pycord.

@BobDotCom first I added the check (your print(bot.guilds), but I just made it iterate through the list of guilds and print a confirmation if my guild was found. on_ready() waited until all 8 shards were loaded and then ran, confirming that my target guild was found. Then I added the await bot.register_commands() after my check and un-commented the slash command and tried again. I received the error as soon as I started to execute the file, which to me means the bot tried to register the command before I explicitly called await bot.register_commands() which means it happened before on_ready() which is why the guild is not found.

TL;DR: @client.slash_command(guild_ids=[]) seems to be attempting to register commands as soon as the code is executed, rather than adding a flag to be registered after the bot is ready. As such, the guild doesn't exist in the bot's cache, and it errors out and breaks.

from pycord.