Autocompletion

This page will covers information about autocompletion support for slash command.

Autocompletion works by responding to a users typing a query for something and we (bot) responded to it by a list of choices.

For quick example, here’s how you can answer to an autocompletion:

@app.slash_command()
@app.option("word", str, autocomplete=True)
async def guess(ctx, word: str):
    # .autocompleting should return the focused parameter name
    # so if user is autocompleting "word", it should be "word"
    # It can be none if nothing is being focused.
    if ctx.autocompleting == "word":
        await ctx.autocomplete(["a", "b", "c"])
        # OR
        await ctx.autocomplete([
            OptionChoice("A", "a"),
            OptionChoice("B", "b"),
        ])
        return

    await ctx.send(f"Word is: {word}")

That is just a simple way to answer to autocompletion interaction. It will just keep sending a and b and c to user.

You can obviously create a autocompletion system on your bot, but we will just provide a simple way to answer to autocompletion.

You can use either a string list of response, or using OptionChoice for more granular response. You can also mix it up, the library will automatically convert it to dict for responding to Discord.

Attributes

With this changes, we introduced an attribute called ApplicationContext.autocompleting.

That attribute will return the current parameter that need responding. If there’s nothing need autocompletion that attribute will be None.

Limitations

• Autocompletion can only be done on option that use str type.

• The choices response is currently limited to 25 choices, this is Discord side and would be subject to changes.

• If you enabled autocompletion, you cannot set choices on the Option decorator. It will raise ValueError.