Strike-discord-framework NPM

strike-discord-framework is a simple, lightweight command and permission framework for Discord.js

Setup

const framework = new Framework(opts);
await framework.init();

This code will start the framework, the opts object has the following stucture

interface FrameworkClientOptions {
	token: string // Discord bot token,
	name: string // Application name, used for logging and database setup,
	commandsPath: string // Path to the commands folder,
	defaultPrefix: string // Prefix
	databaseOpts: {
		databaseName: string, // Name of the mongoDB database
		url: string // URL to connect to mongoDB
	}
	loggerOpts: { // (Optional)
		logChannels: Record<LogLevel, string>; // (Optional) A map between the log level, and where it should go
		logToFile: boolean; // If the framework should log out to a file
		filePath: string; // (Optional) Filepath to store log files, each log file is separated by day
	}
	ownerID: string; // (Optional) Your discord Id for permission override
	dmPrefixOnPing? boolean; // (Optional) If the bot should DM user the prefix whenever its pinged
	permErrorSilently: boolean; // (Optional) If permission errors should be quietly ignored, rather then sending the user an error
	dmErrorSilently: boolean; // (Optional) If the bot should tell users that they cannot run that command in DMs when its set as such
	clientOptions: ClientOptions; // (Optional) Discord client options, passed directly to DJS. Its highly recommend you set this as otherwise the intents are set to everything
}

When you init the framework, you may also pass in an optional object that will be attached to all CommandEvent's

await framework.init(some_object)

In order to load default commands use the loadBotCommands method with a path pointing to the frameworks defaultCommands method The following default commands are loadable:

Default Commands

misc - help - Uses the help object on commands to show user help - ping - A simple ping-pong command to show the bot is running
admin - eval - Allows the owner of the bot to execute javascript code - override - Enables override of all permissions - prefix - Allows a server admin to change the prefix value in a guild - perm - Allows for the control of command permission flags

This code should point to the default commands path in most nodejs project

framework.loadBotCommands(`${process.cwd()}/node_modules/strike-discord-framework/dist/defaultCommands/`);

Commands

Command folder structure

Within your commands folder, you should have multiple sub directories for each category of command. The above shows the commands and how they are arranged. There is a second type of command called "MultiCommand" that allows for a set of commands that fall under the same parent command, for the above perm is a multi command so the folder structure looks like this

misc - help.ts - ping.ts
admin - eval.ts - override.ts - prefix.ts - perm - grant.ts - list.ts - perm.ts - remove.ts

perm is a multicomamnd as there is a perm.ts file within it. Multi-commands are defined by having a file, with the same name as the folder in it. xxx/xxx.ts = multi command. All other files within the folder will be counted as child commands of the multi-command

Command

To create a command, place a file within the commands directory. commands/category_name/filename.ts The file should have a default export, that is a class that extends the Command The command class has the following structure

abstract class Command {
	abstract name: string; // (Required) The command name, what the user run
	allowDM: boolean = true; // If this command can be ran in DMs
	permissions: string[] = []; // The permission flags this command has
	altNames: string[] = []; // Alternative names for this command
	help: { msg?: string; usage?: string; } = {} // A help object used for the help command
	noPermError(event: CommandEvent, ...args: BotCommandArgument[]): BotCommandReturn; // An optional method that will be called whenever a user without permissions executes the command
	abstract run(event: CommandEvent, ...args: BotCommandArgument[]): BotCommandReturn; // (Required) The method that is run when a user executes a command
}

BotCommandReturn

BotCommandReturn can either be void, or a Sendable. It can also be a promise of either of those.

type Sendable = string | Discord.MessageEmbed | { embeds: Discord.MessageEmbed[] };

BotCommandArgument

id - Role or user Id
value - The actual object, either a Discord.User or Discord.Role
type - Either "user" or "role"

Multi-Command

The MultiCommand class extends the Command class. All sub commands are refrenced via the primary command, such as when the user runs a command it has the following syntax:

[prefix][multi-command-name] [subcommand] [arguments] A multi-command has one additional property to regular commands, check. Check is optional, and by default will not modify command execution.

check(event: CommandEvent): MultiCommandRet | Promise<MultiCommandRet>

All other commands of a multi-command should be just regular multi-commands. If the base command cannot be executed for whatever reason (DM when DMs are set to false, no permissions, etc) then the child command will not be ran

MultiCommandRet

An object with the following structure

{
	event: CommandEvent; // An updated CommandEvent, very useful if you would like to extend the command event class and add additional properties. This updated command event will be passed to the run method of the child command
	pass: boolean; // If this command should continue execution and run the child command
	failMessage: Sendable; // If pass is set to false, this will be sent to the user
}

CommandEvent

Command event is the primary method to interact with the framework when commands execute. The class has the following structure

class CommandEvent<T = any> {
	command: BotCommand; // The bot command that this event was triggered by, if its a part of a multi-command this will be the child command that gets executed
	app: T; // This is the value that was passed into the framework on initialization
	framework: FrameworkClient; // A reference to the framework instance
	message: Discord.Message; // The discord message that triggered this command
	args: string[]; // A list of arguments the user gave, split by spaces, but where anything in quotations will be grouped. hello world "hello world"
	constructor(event: CommandEvent); // If you extend the command event class then pass in the original command event to assign the values 
}

Argument Parsing

To use automatic argument parsing there are some pre-requisites that must be completed first

reflect-metadata is included, and loaded before the framework
You are using typescript
tsconfig.json has the following: - "experimentalDecorators": true - "emitDecoratorMetadata": true

The argument types that can be parsed are stated above, to enable argument parsing simply decorate a command with the @CommandRun tag. This would look something like

class MyCommand{
	name = "command"
	@CommandRun
	async run(event: CommandEvent, role: Discord.Role, name: string, user: Discord.string) {}
}

All the arguments will be automatically parsed and given to the run command. If any argument fails to parse it will not run the command

String and number arguments can have specific restrictions applied to them using the Arg decorator

Numbers can restrict using min and max properties
Strings can restrict using the regex property
Both can have the following properties: - optional - If this value is strictly required, or can be omitted (this can be applied to the other types as well) - options - Strictly sets what values are valid, is an array of possibilities

Example

Get a number greater than 5, and optionally get a string thats either "hello" or "world"

class MyCommand{
	name = "command"
	@CommandRun
	async run(event: CommandEvent, @Arg({min:5}) num: number, @Arg({optional: true, options: ["hello", "world"]}) str: string) {}
}

Utilities

All utilities are accessed via framework.utils, which will contain several utility methods

Display Id

framework.utils.displayId(id: string, guild?: Discord.Guild, opts?: DisplayIdOpts)

Parse Quotes

framework.utils.parseQuotes(str: string)

React Confirm

framework.utils.reactConfirm(prompt: string, message: Discord.Message, opts?: ConfirmOptions)

ConfirmOptions has the following structure, every property is optional

interface ConfirmOptions {
	visual: boolean; // If the replys should be ephemeral or not
	onCancelMessage: Sendable; // The message sent when the user presses cancel
	onConfirmMessage: Sendable; // Message sent when user presses confirm
	onConfirm: () => Sendable | Promise<Sendable>; // Callback on pressing confirm, return value is sent to user
	onCancel: () => Sendable | Promise<Sendable>; // Callback on pressing cancel, return value is sent to user 
};

Resolve User

framework.utils.resolveUser(resolvable: string, guild?: Discord.Guild)

User input helpers

There are four primary methods to get user input, those are

getString - Gets a single message from the user and returns its content
getSelect - Creates a drop down select and gets N number of options from it
getButton - Creates a message with buttons on it, and returns the first the user clicks
getButtonSelect - Creates a message with buttons, and allows the user to press multiple buttons

getString

Argument	type	Description
message	Discord.Message	The message from the user to target user and channel
prompt	Discord.MessageEmbed	The embed to send as the prompt
returns	Promise\	the user entered value

getSelect

Argument	type	Description
message	Discord.Message	The message from the user to target user and channel
prompt	Discord.MessageEmbed	The embed to send as the prompt
options	SelectOption[]	The selection options
values	number = 1	How many values for the user to enter, defaults to one
returns	Promise\<string[]>	the user selected value(s)

SelectOption has the structure

interface SelectOption {
	name: string; 
	description?: string;
	emoji?: string;
	value?: string; // If value is defined, the returned value is this, otherwise its the name
}

getButton

Argument	type	Description
message	Discord.Message	The message from the user to target user and channel
prompt	Discord.MessageEmbed	The embed to send as the prompt
options	ButtonOption[]	The button options
returns	Promise\	the value of the button the user pressed

ButtonOption has the structure

interface ButtonOption {
	name: string;
	emoji?: string;
	value?: string; // If value is defined, the returned value is this, otherwise its the name
	disabled?: boolean;
	style: "PRIMARY" | "SECONDARY" | "SUCCESS" | "DANGER";
}

getButtonSelect

Argument	type	Description
message	Discord.Message	The message from the user to target user and channel
prompt	Discord.MessageEmbed	The embed to send as the prompt
options	ButtonSelectOption[]	The button options
returns	Promise\

ButtonSelectOption has the structure

interface ButtonSelectOption {
	button: ButtonOption;
	onSelect: (itr: Discord.ButtonInteraction, updateButtons: (options: ButtonSelectOption[]) => Promise<void>) => void | Promise<void>;
}

onSelect gets two values, the interaction, and an updateButtons callback that can be used to update the selection options after the user clicks a button, it is not required that you update anything

Paged Embeds

There are two types of paged embeds, NamedPageEmbed or NumberedPageEmbed, where named uses string indexs selected by the user, and numbered uses sequential numeric indexs.

For a numbered page embed simply replace NamedPage with NumberedPage and use framework.utils.numberPageEmbed

EmbedCallback is a simple callback type that resolves an embed

type EmbedCallback = () => Discord.MessageEmbed | Promise<Discord.MessageEmbed>;

NamedPage is an object to describe a single page of the embed

interface NamedPage {
	name: string; // Name shown in the select
	description?: string; // Description in the select
	emoji?: string; // Emoji in the select
	get(existing: Discord.MessageEmbed, name: string): Discord.MessageEmbed | Promise<Discord.MessageEmbed> // Callback that will be executed when this page is selected, should return the updated embed
}

NumberedPage is simular to NamedPage but for numrically indexed pages

type NumberedPage = (existing: Discord.MessageEmbed, index: number) => Discord.MessageEmbed | Promise<Discord.MessageEmbed>;

Object Builder

ObjectBuilder class

Question

There are three types of questions, each extend from the base question interface

interface BaseQuestion {
	type: QuestionType; // QuestionType is an enum with str, button, and select values
	handle?: (value: string) => { passes: boolean; value: any }| Promise<{ passes: boolean; value: any }>; // Called when the user inputs a value, allows for modification of the value type and for rejecting the value outright
	prop: string; // The property on the object to set
	name: string; // The name of this question, used to select to show the user options in a drop down
	prompt: string; // The prompt to ask the user when asking for input
}

QuestionType.str (StringQuestion) - Gets a string from the user, no additional properties over BaseQuestion
QuestionType.select (SelectQuestion) - Gets an option from the user via a select, has an options property that should be of type SelectOption[]. Uses the getSelect util
QuestionType.button (ButtonQuestion) - Promopts the user to press a button, with the options being defined in an options property that should be of type ButtonOption[]. Uses the getButton util

Ask One Question

ObjectBuilder.askOneQuestion(obj: Obj)

Asks a single question from the user, where they can select the question they would like to answer. If the user presses exit then the method will return null