A python slack bot
Simply executing main.py will allow Bottomly to run, connect to slack, and listen for events.
The dockerfile allows the easy creation of a simple docker image based on Alpine which will allow the app to be run from within a container and hosted in your container framework of choice.
Modules are broadly divided by responsibility. Files in the root are related to global config and application spin up, everyhing else is in a module.
Classes in here implement the actions sent to the bot via the delivery channel(s). In most cases they will peform some action and return a result directly, without calling out to other modules, however, if access to the persistence layer is required, it should be done through the model module.
This defines the persistence layer of the bot. All behaviours which relate to persisting information beyond a simple request/response model should be defined here, e.g. describing user histories, their karma, different user lists etc.
Implements the slack delivery channel, which is currently the only delivery channel. In future, the potential exists to define other ones, such as IRC, without needing to re-implement the command or model modules.
Uses the SlackSocket implementation of Slack's RTM protocl.
slack_event_handler.py handles the broad event loop, and loops through the custom handlers to find one that can handle the given message event.
*_event_handler.py files implement custom handlers and implement AbstractEventHandler. This mandates that they implement the following functions:
can_handle(slack_event): registers that this handler for a particular slack event. Returns a boolean.handle(slack_event): extracts the relevant information from the givenslack_eventand passes that to a correspondingcommandobject for execution._get_command_symbol: Returns the trigger character(s) which specify what messages this handler will call on, e.g..gfor a google search command (assuming the bot is configured to use.as a common prefix).
Has an internal structure matching the rest of the app, and each app file should have a corresponding test file.
The following environment variables must be configured for the app to run & all unit tests pass:
bottomly_env: A string describing the active environment. If it is not set toliveall output messages will display marked as "DEBUG".bottomly_prefix: The prefix for bottomly commands. While events can overload this, for example karma changes can trigger on++or--regardless of the common prefix, most commands should require a single character prefix such as.or!. This sets that prefix globally, allowing for easy switching between test and production, for example.bottomly_google_api_key: A valid google API keybottomly_google_cse_id: A valid google custom search engine IDbottomly_mongo_conn_str: A mongo DB connection string. Since we're usingpymodmfor data access, this connection string must include a DB name.bottomly_slack_bot_token: The "Bot User OAuth Access Token" from slack to allow access to their RTM endpoint.bottomly_giphy_api_key: a valid giphy API key.
Most actions will be carried out by adding a handler/command pair in the relevant modules, with the command handing off to the model layer for any work that requires persistence. PRs will not be merged without proper test coverage.