Configuration

The following environment variables are available to configure the application:

Variable	Description	Required	Default
DISCORD_TOKEN	The Discord bot token.	Yes	-
BOT_CONFIG_PATH	The path to the bot's configuration file.	No	application.yaml
DISCORD_API_CONCURRENCY	The maximum number of concurrent Discord API requests.	No	3
ENVIRONMENT	The environment in which the application is running.	No	production
HF_DOWNLOAD_CONCURRENCY	The maximum number of concurrent downloads when installing transformers.	No	3
HF_HOME	The directory containing Huggingface Transformers data files.	No	hf_data
LOG_LEVEL	The minimum log level.	No	INFO
MPLCONFIGDIR	The directory containing Matplotlib configuration files.	No	/app/matplotlib
NLTK_DATA	The directory containing NLTK data files.	No	nltk_data
NLTK_DOWNLOAD_CONCURRENCY	The maximum number of concurrent downloads when installing NLTK data.	No	3
PREPROCESSING_MAX_WORD_LENGTH	The maximum word length. Longer words are dropped.	No	35
PREPROCESSING_MESSAGE_TRUNCATE_LENGTH	The maximum message length. Longer messages are truncated.	No	256
REDIS_HOST	The Redis host.	No	localhost
REDIS_PORT	The Redis port.	No	6379
REDIS_PASSWORD	The Redis password.	No	-

Required Settings

The following settings are required to start the application:

DISCORD_TOKEN

You can obtain a Discord bot token from the Discord Developer Portal. Your token should have the following scopes:

• bot

Security Warning

Do not share your token with anyone!

Optional Settings

The following settings are optional or have default values that can be overridden:

BOT_CONFIG_PATH

This specifies the location of the bot's configuration file. By default, the application searches for a file named application.yaml in the directory from which it is launched. In the Docker image, this file is located at /app/application.yaml.

Read more about the application.yaml file.

DISCORD_API_CONCURRENCY

The maximum number of concurrent Discord API requests. By default, this is set to 3.

ENVIRONMENT

The environment in which the application is running. Set this to development to enable development features such as loading development cogs. By default, this is set to production.

HF_DOWNLOAD_CONCURRENCY

The maximum number of concurrent downloads when installing Huggingface Transformers models. By default, this is set to 3.

HF_HOME

The directory containing Huggingface Transformers data files. By default, this is set to hf_data in the directory from which the application is launched. In the Docker image, this directory is located at /app/hf_data.

LOG_LEVEL

The minimum log level to display. The following levels are available:

• DEBUG
• INFO
• WARNING
• ERROR
• CRITICAL

The default log level is INFO.

MPLCONFIGDIR

The directory containing matplotlib configuration files. Uses the default matplotlib configuration directory unless otherwise configured. In the Docker image, the matplotlib configuration files are located at /app/matplotlib.

NLTK_DATA

The directory containing NLTK data files. By default, this is set to nltk_data in the directory from which the application is launched. In the Docker image, this directory is located at /app/nltk_data.

NLTK_DOWNLOAD_CONCURRENCY

The application automatically downloads NLTK data files on startup. This setting controls the number of concurrent downloads. By default, this is set to 3.

PREPROCESSING_MAX_WORD_LENGTH

The maximum word length. Words longer than this value are dropped. By default, this is set to 35.

PREPROCESSING_MESSAGE_TRUNCATE_LENGTH

The maximum message length. Messages longer than this value are truncated. By default, this is set to 256.

REDIS_HOST

The hostname of the Redis server. Defaults to localhost.

REDIS_PORT

The port of the Redis server. Defaults to 6379.

REDIS_PASSWORD

The password of the Redis server. Set this variable if your Redis server requires authentication. No password is set by default.

Security Warning

Do not share your Redis password with anyone!

application.yaml

The application.yaml file is a configuration file that specifies the cogs to load, the NLTK datasets to download, and the Huggingface Transformers models to install on startup. It has the following structure:

# List of cogs to load when the bot starts, identified by their package name.
cogs:
  - <PACKAGE_NAME>
  - <PACKAGE_NAME>
# List of cogs to load in development mode only, identified by their package name.
dev-cogs:
  - <PACKAGE_NAME>
  - <PACKAGE_NAME>
# List of NLTK datasets to download on startup.
nltk:
  - <DATASET_NAME>
  - <DATASET_NAME>
# List of Huggingface Transformers models to download on startup.
transformers:
  - <MODEL_NAME>
  - <MODEL_NAME>

Advanced Users Only

For standard use, the application.yaml file does not need to be changed. We recommend against modifying this file unless you know what you're doing.

One possible use case is to add additional feature or disable existing features by modifying the list of cogs. To add a new cog, add the package name to the cogs list. To disable a feature, remove the package name from the list.

Below is a list of all cogs that are loaded by default:

Package Name	Description
courageous_comets.cogs.about	Provides information about the bot.
courageous_comets.cogs.keywords.search_command	Searches for keywords using a slash command.
courageous_comets.cogs.keywords.search_context_menu	Searches for keywords using a context menu item.
courageous_comets.cogs.keywords.topics_command	Lists the most popular keywords for a given context using a slash command.
courageous_comets.cogs.keywords.user_context_menu	Lists the most popular keywords for a particular user using a context menu item.
courageous_comets.cogs.messages	Listens for messages and passes them on for internal processing.
courageous_comets.cogs.ping	Responds to ping requests.
courageous_comets.cogs.sentiment.message_context_menu	Analyzes the sentiment of a message using a context menu item.
courageous_comets.cogs.sentiment.search_command	Searches for messages with similar sentiment using a slash command.
courageous_comets.cogs.sentiment.search_context_menu	Searches for messages with similar sentiment using a context menu item.
courageous_comets.cogs.sentiment.user_context_menu	Analyzes the sentiment of a user's messages using a context menu item.
courageous_comets.cogs.frequency	Provides an overview of the amount of recent messages for a given timespan.