TheForcer/vision-nio

A general purpose matrix chatbot based on matrix-nio

Vision

A matrix bot with some cool functionality, which is based on matrix-nio and the corresponding template by anoadragon453/nio-template.

Installation

Clone the repo to your preferred location:

git clone https://github.com/theforcer/vision-nio
cd vision-nio

Copy and rename the config file. For basic functionality the matrix credentials are required.

cp config/sample.config.yaml config/config.yaml
nano/vim/pico config/config.yaml

Start up the bot/container with docker-compose (add -d if you want the process to run in the background):

docker-compose up

Usage

PiHole

The PiHole module is used to get a quick overlook of the total amount of (blocked) DNS queries. Configure the respective section in the config file with your PiHole's IP/FQDN, and with a quick !v ads you'll get the information you need.

UptimeRobot

After supplying a (read-only) API Key from UptimeRobot in the settings file, the bot will scrape all your available monitors and output the latest Response Time (in ms) and the duration since the latest status change for each of the monitors (symbolized by ✅ and ❌ also :D). The command for that is !v uptime

Project structure

main.py

Initialises the config file, the bot store, and nio's AsyncClient (which is used to retrieve and send events to a matrix homeserver). It also registering some callbacks on the AsyncClient to tell it to call some functions when certain events are received (such as an invite to a room, or a new message in a room the bot is in).

It also starts the sync loop. Matrix clients "sync" with a homeserver, by asking constantly asking for new events. Each time they do, the client gets a sync token (stored in the next_batch field of the sync response). If the client provides this token the next time it syncs (using the since parameter on the AsyncClient.sync method), the homeserver will only return new event since those specified by the given token.

This token is saved in the database created by storage.py so even if the bot quits unexpectantly, it will continue syncing where it left off next time it is started.

config.py

This file reads a config file at a given path (hardcoded as config.yaml in main.py), processes everything in it and makes the values available to the rest of the bot's code so it knows what to do. Most of the options in the given config file have default values, so things will continue to work even if an option is left out of the config file. Obviously there are some config values that are required though, like the homeserver URL, username, access token etc. Otherwise the bot can't function.

storage.py

Creates (if necessary) and connects to a SQLite3 database and provides commands to put or retrieve data from it. Table definitions should be specified in _initial_setup, and any necessary migrations should be put in _run_migrations. There's currently no defined method for how migrations should work though.

The sync_token table should be left in tact so that the bot can save its progress when syncing events from the homeserver.

sync_token.py

A simple class that can load and save a sync token to/from the database.

A SyncToken is an instance of a sync token, which is simply a string retrieved from a matrix homeserver when querying the /sync endpoint (which clients use to retrieve new events). It is given to the next call of the /sync endpoint in order to specify the starting point in the event timeline you would like to receive messages from.

callbacks.py

Holds callback methods which get run when the bot get a certain type of event from the homserver during sync. The type and name of the method to be called are specified in main.py. Currently there are two defined methods, one that gets called when a message is sent in a room the bot is in, and another that runs when the bot receives an invite to the room.

The message callback function, message, checks if the message was for the bot, and whether it was a command. If both of those are true, the bot will process that command.

The invite callback function, invite, processes the invite event and attempts to join the room. This way, the bot will auto-join any room it is invited to.

bot_commands.py

Where all the bot's commands are defined. New commands should be defined in process with an associated private method. echo and help commands are provided by default.

A Command object is created when a message comes in that's recognised as a command from a user directed at the bot (either through the specified command prefix (defined by the bot's config file), or through a private message directly to the bot. The process command is then called for the bot to act on that command.

message_responses.py

Where responses to messages that are posted in a room (but not necessarily directed at the bot) are specified. callbacks.py will listen for messages in rooms the bot is in, and upon receiving one will create a new Message object (which contains the message text, amongst other things) and calls process() on it, which can send a message to the room as it sees fit.

A good example of this would be a Github bot that listens for people mentioning issue numbers in chat (e.g. "We should fix #123"), and the bot sending messages to the room immediately afterwards with the issue name and link.

chat_functions.py

A separate file to hold helper methods related to messaging. Mostly just for organisational purposes. Currently just holds send_text_to_room, a helper method for sending formatted messages to a room.

errors.py

Custom error types for the bot. Currently there's only one special type that's defined for when a error is found while the config file is being processed.

sample.config.yaml

The sample configuration file. People running your bot should be advised to copy this file to config.yaml, then edit it according to their needs. Be sure never to check the edited config.yaml into source control since it'll likely contain sensitive details like an access token!