Frequently Asked Questions

This is a list of Frequently Asked Questions regarding using discord.py and its extension modules. Feel free to suggest a new question or submit one via pull requests.

Coroutines

Questions regarding coroutines and asyncio belong here.

I get a SyntaxError around the word async! What should I do?

This SyntaxError happens because you’re using a Python version lower than 3.5. Python 3.4 uses @asyncio.coroutine and yield from instead of async def and await.

Thus you must do the following instead:

async def foo():
    await bar()

# into

@asyncio.coroutine
def foo():
    yield from bar()

Don’t forget to import asyncio on the top of your files.

It is heavily recommended that you update to Python 3.5 or higher as it simplifies asyncio massively.

What is a coroutine?

A coroutine is a function that must be invoked with await or yield from. When Python encounters an await it stops the function’s execution at that point and works on other things until it comes back to that point and finishes off its work. This allows for your program to be doing multiple things at the same time without using threads or complicated multiprocessing.

If you forget to await a coroutine then the coroutine will not run. Never forget to await a coroutine.

Where can I use await?

You can only use await inside async def functions and nowhere else.

What does “blocking” mean?

In asynchronous programming a blocking call is essentially all the parts of the function that are not await. Do not despair however, because not all forms of blocking are bad! Using blocking calls is inevitable, but you must work to make sure that you don’t excessively block functions. Remember, if you block for too long then your bot will freeze since it has not stopped the function’s execution at that point to do other things.

A common source of blocking for too long is something like time.sleep(n). Don’t do that. Use asyncio.sleep(n) instead. Similar to this example:

# bad
time.sleep(10)

# good
await asyncio.sleep(10)

Another common source of blocking for too long is using HTTP requests with the famous module requests. While requests is an amazing module for non-asynchronous programming, it is not a good choice for asyncio because certain requests can block the event loop too long. Instead, use the aiohttp library which is installed on the side with this library.

Consider the following example:

# bad
r = requests.get('http://random.cat/meow')
if r.status_code == 200:
    js = r.json()
    await client.send_message(channel, js['file'])

# good
async with aiohttp.get('http://random.cat/meow') as r:
    if r.status == 200:
        js = await r.json()
        await client.send_message(channel, js['file'])

General

General questions regarding library usage belong here.

How do I set the “Playing” status?

There is a method for this under Client called Client.change_presence(). The relevant aspect of this is its game keyword argument which takes in a Game object. Putting both of these pieces of info together, you get the following:

await client.change_presence(game=discord.Game(name='my game'))

How do I send a message to a specific channel?

If you have its ID then you can do this in two ways, first is by using Object:

await client.send_message(discord.Object(id='12324234183172'), 'hello')

The second way is by calling Client.get_channel() directly:

await client.send_message(client.get_channel('12324234183172'), 'hello')

I’m passing IDs as integers and things are not working!

In the library IDs must be of type str not of type int. Wrap it in quotes.

How do I upload an image?

There are two ways of doing it. Both of which involve using Client.send_file().

The first is by opening the file and passing it directly:

with open('my_image.png', 'rb') as f:
    await client.send_file(channel, f)

The second is by passing the file name directly:

await client.send_file(channel, 'my_image.png')

How can I add a reaction to a message?

You use the Client.add_reaction() method.

If you want to use unicode emoji, you must pass a valid unicode code point in a string. In your code, you can write this in a few different ways:

  • '👍'
  • '\U0001F44D'
  • '\N{THUMBS UP SIGN}'

In case you want to use emoji that come from a message, you already get their code points in the content without needing to do anything special. You cannot send ':thumbsup:' style shorthands.

For custom emoji, you should pass an instance of discord.Emoji. You can also pass a 'name:id' string, but if you can use said emoji, you should be able to use Client.get_all_emojis()/Server.emojis to find the one you’re looking for.

How do I pass a coroutine to the player’s “after” function?

A StreamPlayer is just a threading.Thread object that plays music. As a result it does not execute inside a coroutine. This does not mean that it is not possible to call a coroutine in the after parameter. To do so you must pass a callable that wraps up a couple of aspects.

The first gotcha that you must be aware of is that calling a coroutine is not a thread-safe operation. Since we are technically in another thread, we must take caution in calling thread-safe operations so things do not bug out. Luckily for us, asyncio comes with a asyncio.run_coroutine_threadsafe function that allows us to call a coroutine from another thread.

Warning

This function is only part of 3.5.1+ and 3.4.4+. If you are not using these Python versions then use discord.compat.run_coroutine_threadsafe.

However, this function returns a concurrent.Future and to actually call it we have to fetch its result. Putting all of this together we can do the following:

def my_after():
    coro = client.send_message(some_channel, 'Song is done!')
    fut = asyncio.run_coroutine_threadsafe(coro, client.loop)
    try:
        fut.result()
    except:
        # an error happened sending the message
        pass

player = await voice.create_ytdl_player(url, after=my_after)
player.start()

Why is my “after” function being called right away?

The after keyword argument expects a function object to be passed in. Similar to how threading.Thread expects a callable in its target keyword argument. This means that the following are invalid:

player = await voice.create_ytdl_player(url, after=self.foo())
other  = await voice.create_ytdl_player(url, after=self.bar(10))

However the following are correct:

player = await voice.create_ytdl_player(url, after=self.foo)
other  = await voice.create_ytdl_player(url, after=lambda: self.bar(10))

Basically, these functions should not be called.

How do I get a specific User/Role/Channel/Server?

There are multiple ways of doing this. If you have a specific entity’s ID then you can use one of the following functions:

If the functions above do not help you, then use of utils.find() or utils.get() would serve some use in finding specific entities. The documentation for those functions provides specific examples.

Commands Extension

Questions regarding discord.ext.commands belong here.

Is there any documentation for this?

Not at the moment. Writing documentation for stuff takes time. A lot of people get by reading the docstrings in the source code. Others get by via asking questions in the Discord server. Others look at the source code of other existing bots.

There is a basic example showcasing some functionality.

Documentation is being worked on, it will just take some time to polish it.

Why does on_message make my commands stop working?

Overriding the default provided on_message forbids any extra commands from running. To fix this, add a bot.process_commands(message) line at the end of your on_message. For example:

@bot.event
async def on_message(message):
    # do some extra stuff here

    await bot.process_commands(message)

Can I use bot.say in other places aside from commands?

No. They only work inside commands due to the way the magic involved works.

Why do my arguments require quotes?

In a simple command defined as:

@bot.command()
async def echo(message: str):
    await bot.say(message)

Calling it via ?echo a b c will only fetch the first argument and disregard the rest. To fix this you should either call it via ?echo "a b c" or change the signature to have “consume rest” behaviour. Example:

@bot.command()
async def echo(*, message: str):
    await bot.say(message)

This will allow you to use ?echo a b c without needing the quotes.

How do I get the original message?

Ask the command to pass you the invocation context via pass_context. This context will be passed as the first parameter.

Example:

@bot.command(pass_context=True)
async def joined_at(ctx, member: discord.Member = None):
    if member is None:
        member = ctx.message.author

    await bot.say('{0} joined at {0.joined_at}'.format(member))

How do I make a subcommand?

Use the group decorator. This will transform the callback into a Group which will allow you to add commands into the group operating as “subcommands”. These groups can be arbitrarily nested as well.

Example:

@bot.group(pass_context=True)
async def git(ctx):
    if ctx.invoked_subcommand is None:
        await bot.say('Invalid git command passed...')

@git.command()
async def push(remote: str, branch: str):
    await bot.say('Pushing to {} {}'.format(remote, branch))

This could then be used as ?git push origin master.