contributing.md

BetterYTM - Contributing Guide

Thank you for your interest in contributing to BetterYTM!
This guide will help you get started with contributing to the project.
If you have any questions or need help, feel free to contact me, see my homepage for contact info.
Or you can also join my Discord server to get in touch or get help.

Table of contents:

• Submitting translations
  • Adding translations for a new language
  • Editing an existing translation
• Setting up the project for local development
  • Requirements
  • CLI commands
  • Extras
  • Getting started
  • Procedure for specific tasks
• Developing a plugin that interfaces with BetterYTM
  • Shimming for TypeScript without errors & with autocomplete
  • Global functions and classes on the plugin interface

Submitting translations:

Thank you so much for your interest in translating BetterYTM!
Before submitting a translation, please check on this document if the language you want to translate to has already been translated and how many strings are still missing.

Adding translations for a new language:

Important

Please make sure you always select the develop branch when translating, as the main branch is only used for releases.

To submit a translation, please follow these steps:

1. Copy the contents of the default translation file assets/translations/en_US.json
2. Replace the en_US part of the file name with the language code and locale code of the language you want to translate to
   You can find lists of ISO 639-1 alpha-2 language codes here and ISO 3166-1 alpha-2 country codes here.
   The final locale code should be in the format languageCode_localeCode (e.g. en_US, en_UK, ...)
3. Translate the strings inside the file, while making sure not to change the keys on the left side of the colon and to preserve the placeholders with the format %n (where n is any number starting at 1).
   If you don't want to finish it in one go, please remove the extra keys before submitting the file. They can always be added back by running the command pnpm run tr-format -p -o=language_LOCALE (see this section for more info).
4. If you like, you may also create a translation for the README-summary.md file for display on the userscript distribution sites
   Please duplicate the file README-summary.md and call it README-summary-languageCode_localeCode.md and place it in the assets/translations/ folder.
5. If you want to submit a pull request with the translated file:
   1. Duplicate the en_US.json file in the folder assets/translations/ by keeping the format languageCode_localeCode.json
   2. Edit it to your translated version and keep the left side of the colon unchanged
   3. Create the mapping in assets/locales.json by copying the English one and editing it (please make sure it's alphabetically ordered)
   4. Add your name to the respective authors property in assets/locales.json
   5. Test your changes by following this section, then submit your pull request
6. Alternatively send it to me directly, see my homepage for contact info
   Make sure you also add your language to the contents of assets/locales.json

Editing an existing translation:

Important

Please make sure you always select the develop branch when translating, as the main branch is only used for releases. To edit an existing translation, please follow these steps:

1. Set up the project for local development by following this section
   Make sure you have forked the repository and cloned your fork instead of cloning the original repository.
2. Find the file for the language you want to edit in the folder assets/translations/
3. Run the command pnpm run tr-format -p -o=language_LOCALE, where language_LOCALE is the part of the file name before the .json extension
   This will prepare the file for translation by providing the missing keys once in English and once without any value and also formatting the file to have the same structure as the base file en_US.json
4. Edit the strings inside the file, while making sure not to change the keys on the left side of the colon and to preserve the placeholders with the format %n (where n is any number starting at 1).
5. Make sure there are no duplicate keys in the file
6. Run the command pnpm run tr-format -o=language_LOCALE to make sure the file is formatted correctly
7. Test for syntax errors and update translation progress with the command pnpm run tr-progress
8. Open the file assets/translations/README.md to see if you're still missing any untranslated keys (you don't have to translate them all, but it would of course be nice)
9. I highly encourage you to test your changes to see if the wording fits into the respective context by following this section
10. Submit your pull request by clicking here and setting the compare: dropdown to your fork
11. Check that the CI checks just above the comment box pass and then wait for the pull request to be reviewed and merged

Setting up the project for local development:

Requirements:

1. Have current versions of Node.js, npm and Git installed
2. Install pnpm by running npm i -g pnpm
3. Clone this repository
   If you plan on contributing to the project, please click here to fork it and clone your fork instead.
   Make sure to clone or fork from the develop branch since the main branch is only used for releases!
4. Switch to the develop branch by running git checkout -b develop in the project root.
   Skip this step if you are using your own forked repository.
5. Open a terminal in the project root and run pnpm i
6. Copy the file .env.template to .env and modify the variables inside to your needs.
7. Now you can run pnpm run dev to build the userscript and host it on a development server or check out the other commands below

These are the CLI commands available after setting up the project:

• pnpm i
  Run once to install dependencies

• pnpm run dev
  This is the command you want to use to locally develop and test BetterYTM.
  It watches for any changes, then rebuilds and serves the userscript on port 8710, so it can be updated live if set up correctly in the userscript manager (see extras).
  Once it has finished building, a link will be printed to the console. Open it to install the userscript.
  You can also configure request logging and more in .env and src/tools/serve.ts, just make sure to restart the dev server after changing anything.

• pnpm run build-prod
  Builds the userscript for production for all hosts with their respective options already set.
  Outputs the files using a suffix predefined in the package.json file.
  Use this to build the userscript for distribution on all host/CDN platforms.

• pnpm run build <arguments>
  Builds the userscript with custom options
  Arguments:

  • --config-mode=<value> - The mode to build in. Can be either production or development (default)
  • --config-branch=<value> - The GitHub branch to target. Can be any branch name, but should be main for production and develop for development (default)
  • --config-host=<value> - The host to build for. Can be either github (default), greasyfork or openuserjs
  • --config-assetSource=<value> - Where to get the resource files from. Can be either local or github (default)
  • --config-suffix=<value> - Suffix to add just before the .user.js extension. Defaults to an empty string

  Shorthand commands:

  • pnpm run build-prod-base - Used for building for production, targets the main branch and the public asset source.
    Sets --config-mode=production and --config-branch=main and --config-assetSource=github
  • pnpm run build-dev - Builds a preview version, targeting the develop branch and the public asset source so no local dev environment is needed.
    Sets --config-mode=development, --config-branch=develop and --config-assetSource=github
  • pnpm run preview - Same as pnpm run build-prod, but sets --config-host=github and --config-assetSource=local, then starts the dev server for a few seconds so the extension that's waiting for file changes can update the script and assets

• pnpm run lint
  Builds the userscript with the TypeScript compiler and lints it with ESLint. Doesn't verify the functionality of the script, only checks for syntax and TypeScript errors!

• pnpm run storybook
  Starts Storybook for developing and testing components. After launching, it will automatically open in your default browser.

• pnpm run gen-readme
  Updates the README files by inserting different parts of generated sections into them.

• pnpm run tr-progress
  Checks all translation files for missing strings and updates the progress table in assets/translations/README.md
  Will also be run automatically after every script build.

• pnpm run tr-format <arguments>
  Reformats all translation files so they match that of the base file en_US.json
  This includes sorting keys and adding the same empty lines and indentation. Arguments:

  • --prep or -p - Prepares the files for translation via GitHub Copilot by providing the missing key once in English and once without any value
  • --only="<value>" or -o="<value>" - Only applies formatting to the files of the specified locales. Has to be a quoted, case-sensitive, comma separated list! (e.g. -o="fr_FR,de_DE" or -o="pt_BR")
  • --include-based or -b - Also includes files which have a base locale specified

• pnpm run tr-prep
  Shorthand for pnpm run tr-format --prep (see above)

• pnpm run --silent invisible "<command>"
  Runs the passed command as a child process without giving any console output. (-- and double quotes are required!)
  Remove --silent to see pnpm's info and error messages.

• pnpm run node-ts <path>
  Runs the TypeScript file at the given path using the regular node binary and the ts-node ESM loader.
  Also enables source map support and disables experimental warnings.

• pnpm run dep-cruise
  Runs dependency-cruiser to show problems with the dependency tree like circular, missing or orphaned dependencies.

• pnpm run dep-graph
  Generates a dependency graph of the project, visually showing the dependencies between files and problems with the dependency tree.
  Requires the Graphviz binary dot to be in the system PATH.

Note

When you are using npm (as opposed to pnpm), read the following carefully:
You will need to use a lone -- between the command name and the arguments, for example: npm run --silent invisible -- "<command>"
This is so npm can tell the difference between arguments passed to it versus arguments passed to the script it is running.

Extras:

• When using ViolentMonkey, after letting the command pnpm run dev run in the background, open http://localhost:8710/BetterYTM.user.js and select the Track local file option.
  This makes it so the userscript automatically updates when the code changes.
  Note: the tab needs to stay open on Firefox or the script will not update itself.
• To link any local JS file (like a work-in-progress library) in the userscript, add a "link": "/path/to/script.umd.js" property to the respective library in assets/require.json (relative or absolute path)
  As this file will just be injected as-is at build time, make sure you are targeting a UMD or IIFE bundle that exports a variable with the name set by "global".
  In order to make TypeScript shut up, you will still need to link the library manually with pnpm link -g /path/to/library_root

Getting started working on the project:

After setting the project up for local development (see above), you can start working on the project.
The main files you will be working with are:

• src/index.ts - The main entry point for the userscript and where all features are initialized
• src/interface.ts - The file that contains all events and functions that are exposed to plugins
• src/types.ts - The file that contains all types and interfaces that are used throughout the project
• src/observers.ts - The file that contains all mutation observers that are used to detect changes on the page
• src/siteEvents.ts - The file that contains all site- and script-specific events that are dispatched by BetterYTM
• src/tools - The folder that contains all CLI tools and utilities that are used throughout the project
• src/components - The folder that contains all HTML component functions for reusable UI elements

Procedure for specific tasks:

• Adding a new feature:
  1. Add your feature to the FeatureConfig type in src/types.ts (after choosing a fitting category for it)
  2. Add your feature and its properties to the featInfo object in src/features/index.ts, under the correct category
  3. Create an async initialization function for your feature in the respective category's file inside the src/features folder
  4. Add the init function to the onDomLoad function in src/index.ts, under the correct "domain guard condition" and category by following the format of the other features
• Adding an asset (image, icon, stylesheet, translation file and misc. other files):
  1. Add the asset to the assets folder in the root of the project, under the correct subfolder
  2. Add the asset to the assets/resources.json file by following the format of the other entries.
     If the path begins with a slash, it will start at the project root (where package.json is), otherwise it will start at the assets folder.
  3. The asset will be immediately available in the userscript after the next build and the @resource directive will automatically point at the locally served asset or the GitHub CDN, depending on the build mode.
  4. When committing, make sure to commit the assets first, then rebuild the userscript and make a second commit.
     This needs to be done because the build script at src/tools/post-build.ts will use the previous commit hash to create version-independent URLs for the assets. These will continue to work in the future, instead of pointing to an ever-changing branch where files could be moved, renamed or deleted at any time.
• Adding a new site event:
  1. Add your event to the SiteEventsMap type in src/siteEvents.ts
  2. Dispatch the event inside initSiteEvents in src/siteEvents.ts or at another point where it is run independent of the feature configuration (the only exception being domain-specific events).
     Always use the function emitSiteEvent to dispatch the event, so it will automatically be logged and emitted to the plugin interface as well.
• Adding something to the plugin interface:
  • If you want to make a function globally available, simply add it to the globalFuncs variable in src/interface.ts under the correct category.
    If the function should require a token, create a proxy function at the bottom of the file that checks for the token and then calls the actual function (also see the bottom of the file for examples).
  • If you want to add something else like a class, constant or entire library (as long as its license allows it), add it to the props variable inside the function initInterface in src/interface.ts
• Creating a new reusable UI component:
  1. Create a new file in the src/components folder with a descriptive name
  2. Add a function that takes a single object of properties as an argument (kind of like a React component), and returns an element that extends the HTMLElement interface (like what the return value of document.createElement() is)
  3. Add a re-export inside the file src/components/index.ts
  4. If you want to expose the component to plugins, add it to the globalFuncs variable in src/interface.ts under the category Components
  5. Write some documentation for the component inside this file (contributing.md), under the global functions and classes section by following the format of the other documented components.
• Adding a locale (language & regional dialect):
  1. Add the locale code and info about the locale to the file assets/locales.json by following the format of the other entries.
     Please make sure the alphabetical order is kept.
     You can find lists of ISO 639-1 alpha-2 language codes here and ISO 3166-1 alpha-2 country codes here.
     The final locale code should be in the format languageCode_localeCode (e.g. en_US, en_UK, ...)
  2. Add a translation file for the locale by following the instructions in the translations section
  3. Your locale will be immediately available in the userscript after the next build.
• Creating a release:
  1. Make sure the version in package.json is bumped according to semantic versioning
  2. Run pnpm i so the version is updated in the lockfile
  3. Update the changelog.md with the new version and an exhaustive list of changes that were made
  4. Make sure all files are committed before the built userscript is, so the next build will have the correct build number
  5. Run pnpm run build-prod to build the userscript for all hosts
  6. Commit and push the built files
  7. Create a new release on GitHub with a tag that follows the format of the previous releases, a copy of the relevant section in the changelog and an install button that points to the built userscript on GitHub (make sure it uses the version tag in its URL to ensure the correct version is installed)
  8. Update the userscript on GreasyFork and OpenUserJS from the built files
     GreasyFork also needs the relevant section of the changelog, but the internal and plugin changes should be trimmed out
  9. Send an announcement in the Discord server linking to the install pages and the changelog, with a summary of the most important changes

Developing a plugin that interfaces with BetterYTM:

BetterYTM has a built-in interface based on events and exposed global constants and functions that allows other "plugin" userscripts to benefit from its features.
If you want your plugin to be displayed in the readme and possibly inside the userscript itself, please submit an issue using the plugin submission template

Tip

Want to quickly get started with a plugin? Check out the official plugin template.
It is based on Vite and TypeScript and has all the necessary setup for you to start developing a plugin right away.
It also contains some small examples to show you how to interact with the BYTM API.

Required knowledge:

• Intermediate JavaScript knowledge (DOM API, events, Promise API, etc.)
  • Understanding JS types and reading TypeScript type declarations
  • Semantic versioning (for versioning your plugin in the correct format)
• At least basic knowledge of userscripts (start on the GreaseSpot wiki)
• Reading the file license-for-plugins.txt to understand the licensing conditions for plugins

Recommended knowledge:

• TypeScript (for type safety and better autocomplete)
• The source code of BetterYTM
  This is especially regarding the files src/interface.ts, src/types.ts, src/siteEvents.ts and src/observers.ts
• Being on the lookout for pull requests, since they will list new features and changes to the interface that you probably want to prepare for
• This document, as it contains most of the information you need to know about the BetterYTM interface, or at least points you to the places where you can find the actual information
• The official plugin template for a quick start with a plugin
• The fact you can join my Discord server to ask questions or get help with writing your plugin

These are the ways to interact with BetterYTM; through constants, events and global functions:

• Static interaction is done through constants that are exposed through the global BYTM object, which is available on the unsafeWindow object.
  These read-only properties tell you more about how BetterYTM is currently being run.
  You can find all properties that are available and their types in the declare global block of src/types.ts

• Dynamic interaction is done through events that are dispatched on the unsafeWindow object.
  They all have the prefix bytm:eventName and are all dispatched with the CustomEvent interface, meaning their data can be read using the detail property.
  You can find all events that are available and their types in src/interface.ts

  Additionally BetterYTM has an internal system called SiteEvents. They are dispatched using the format bytm:siteEvent:eventName
  You may find all SiteEvents that are available and their types in src/siteEvents.ts
  Note that the detail property will be an array of the arguments that can be found in the event handler at the top of src/siteEvents.ts

• Another way of dynamically interacting is through global functions, which are also exposed by BetterYTM through the global BYTM object.
  You can find all functions that are available in the InterfaceFunctions type in src/types.ts
  Find a summary with examples below.

• Additionally, the following namespaces expose entire libraries for you that BetterYTM has already loaded in:

  • unsafeWindow.BYTM.UserUtils contains all exported members from the UserUtils library.
    This library can register listeners for when CSS selectors exist, intercept events, manage persistent user configurations, allow you to modify the DOM more easily and more.
  • unsafeWindow.BYTM.compareVersions has all functions from the compare-versions library.
    Use it to compare semver-compliant version strings.

All of these interactions require the use of unsafeWindow, as the regular window object is pretty sandboxed in userscript managers.

If you need specific events to be added or modified, please submit an issue.

Static interaction Example (click to expand)

Example:

const BYTM = unsafeWindow.BYTM;

console.log(`BetterYTM was built in '${BYTM.mode}' mode`);
console.log(`BetterYTM's locale is set to '${BYTM.locale}'`);
console.log(`BetterYTM's version is '${BYTM.version} #${BYTM.buildNumber}'`);

Dynamic interaction examples (click to expand)

Basic format:

unsafeWindow.addEventListener("bytm:eventName", (event) => {
  // can have any type, but usually it's an object or undefined
  const { detail } = event as CustomEvent<{ foo: string }>;

  console.log(detail.foo);
});

// for listening to SiteEvents:
unsafeWindow.addEventListener("bytm:siteEvent:eventName", (event) => {
  // always typed as array / tuple
  const { detail } = event as CustomEvent<[ foo: HTMLElement ]>;

  console.log(detail[0]);
});

Practical Example:

// listening to generic events:
unsafeWindow.addEventListener("bytm:ready", () => {
  console.log("The DOM is loaded and all BetterYTM features have been initialized");
});

unsafeWindow.addEventListener("bytm:lyricsLoaded", (event) => {
  const { detail } = event as CustomEvent<{ type: "current" | "queue", artists: string, title: string, url: string }>;

  console.log(`Lyrics URL for "${detail.artists} - ${detail.title}" has been loaded: ${detail.url}`);

  if(detail.type === "current")
    console.log("This is from the currently playing song");
  else
    console.log("This is from a song in the queue");
});

// listening to a SiteEvent:
unsafeWindow.addEventListener("bytm:siteEvent:queueChanged", (event) => {
  const { detail } = event as CustomEvent<[ queueItem: HTMLElement ]>;

  console.log(`The queue has been changed. It now contains ${detail[0].childNodes.length} items`);
});

For global function examples see below.

Shimming for TypeScript without errors & with autocomplete:

In order for TypeScript to not throw errors while creating a plugin, you need to shim the types for BYTM.
To do this, create a .d.ts file (for example bytm.d.ts) and add the following code:

declare global {
  interface Window {
    BYTM: {
      // add types here
    };
  }
}

You may specify all types that you need in this file.
To find which types BetterYTM exposes, check out the declare global block in src/types.ts
You may also just copy it entirely, as long as all the imports also exist in your project.
An easy way to do this might be to include BetterYTM as a Git submodule, as long as you stick to only using type imports

Global functions and classes:

These are the global functions and classes that are exposed by BetterYTM through the unsafeWindow.BYTM object.
The usage and example blocks on each are written in TypeScript but can be used in JavaScript as well, after removing all type annotations.

Important

Authenticated functions are marked with 🔒 and need to be passed a per-session and per-plugin authentication token. It can be acquired by calling registerPlugin()

• Meta:
  • registerPlugin() - Registers a plugin with BetterYTM with the given plugin definition object
  • getPluginInfo() 🔒 - Returns the plugin info object for the specified plugin - can be used to check if a certain plugin is registered
• BYTM-specific:
  • getDomain() - Returns the current domain of the page as a constant string (either "yt" or "ytm")
  • getResourceUrl() - Returns a blob: URL provided by the local userscript extension for the specified BYTM resource file
  • getSessionId() - Returns the unique session ID that is generated on every started session
• DOM:
  • BytmDialog - A class for creating and managing modal, fully customizable dialogs
  • ExImDialog - Subclass of BytmDialog for allowing users to export and import serializable data
  • MarkdownDialog - Subclass of BytmDialog for displaying markdown content
  • setInnerHtml() - Sets the innerHTML property of an element after sanitizing the string with DOMPurify and Trusted Types
  • addSelectorListener() - Adds a listener that checks for changes in DOM elements matching a CSS selector
  • onInteraction() - Adds accessible event listeners to the specified element for button or link-like keyboard and mouse interactions
  • getVideoTime() - Returns the current video time (on both YT and YTM)
  • getThumbnailUrl() - Returns the URL to the thumbnail of the currently playing video
  • getBestThumbnailUrl() - Returns the URL to the best quality thumbnail of the currently playing video
  • waitVideoElementReady() - Waits for the video element to be queryable in the DOM - has to be called after bytm:observersReady
  • getCurrentMediaType() - (On YTM only) returns the type of media that is currently playing (either "video" or "song")
• Components:
  • createHotkeyInput() - Creates a hotkey input element
  • createToggleInput() - Creates a toggle input element
  • createCircularBtn() - Creates a generic, circular button element with just an icon
  • createLongBtn() - Creates a generic, long and circular button element with an icon and text
  • createRipple() - Creates a click ripple effect on the given element
  • showToast() - Shows a toast notification and a message string or element
  • showIconToast() - Shows a toast notification with an icon and a message string or element
  • showPrompt() - Shows a styled prompt dialog of the type confirm, alert or prompt
• Translations:
  • setLocale() 🔒 - Sets the locale for BetterYTM
  • getLocale() - Returns the currently set locale
  • hasKey() - Checks if the specified translation key exists in the currently set locale
  • hasKeyFor() - Checks if the specified translation key exists in the specified locale
  • t() - Translates the specified translation key using the currently set locale
  • tp() - Translates the specified translation key including pluralization using the currently set locale
  • tl() - Returns the translation for the provided key and provided locale
  • tlp() - Returns the translation for the provided locale and key, including pluralization identifier
• Feature config:
  • getFeatures() 🔒 - Returns the current BYTM feature configuration object
  • saveFeatures() 🔒 - Overwrites the current BYTM feature configuration object with the provided one
• Lyrics:
  • fetchLyricsUrlTop() - Fetches the URL to the lyrics page for the specified song
  • getLyricsCacheEntry() - Tries to find a URL entry in the in-memory cache for the specified song
  • sanitizeArtists() - Sanitizes the specified artist string to be used in fetching a lyrics URL
  • sanitizeSong() - Sanitizes the specified song title string to be used in fetching a lyrics URL
• Auto-Like:
  • getAutoLikeData() 🔒 - Returns the current auto-like data object
  • saveAutoLikeData() 🔒 - Overwrites the current auto-like data object with the provided one
  • fetchVideoVotes() - Fetches the approximate like and dislike count for the video with the specified ID
• Other:
  • NanoEmitter - Abstract class for creating lightweight, type safe event emitting classes
  • formatNumber - Formats a number with the configured locale and passed or configured format

registerPlugin()

Usage:

unsafeWindow.BYTM.registerPlugin(pluginDef: PluginDef): PluginRegisterResult

Description:
Registers a plugin with BetterYTM with the given plugin definition object.

Arguments:

• pluginDef - The properties of this plugin definition object can be found by searching for type PluginDef in the file src/types.ts

The function will either throw an error if the plugin object is invalid, or return a registration result object.
The error message will contain a list of problems with the passed definition.
Search for type PluginRegisterResult in the file src/types.ts to see the properties of the returned object.

The returned properties include:

• token - A private token that is used for authenticated function calls and that should not be persistently stored beyond the current session
• events - A NanoEmitter instance that allows you to listen for plugin-specific events that are dispatched by BetterYTM.
  To find a list of all events, search for PluginEventMap in the file src/types.ts
• info - The info object that contains all data other plugins will be able to see about your plugin

Complete example (click to expand)

// Search for "type PluginDef" in "src/types.ts" to see the whole type
const pluginDef = {
  plugin: { // required
    // The name and namespace should combine to be unique across all plugins
    // Also, you should never change them after releasing the plugin, so other plugins can rely on them as an identifier
    name: "My cool plugin",                         // required
    namespace: "https://github.com/MyUsername", // required
    version: "4.2.0",                               // required
    iconUrl: "https://picsum.photos/128/128",       // required
    description: { // required
      en_US: "This plugin does cool stuff",      // required
      de_DE: "Dieses Plugin macht coole Sachen", // (all other locales are optional)
      // (see all supported locale codes in "assets/locales.json")
    },
    license: { // (optional)
      name: "MIT",                                // both required
      url: "https://opensource.org/licenses/MIT", // both required
    },
    homepage: { // required
      source: "https:/MyUsername/MyCoolBYTMPlugin",     // required
      other: "https://example.org/MyCoolBYTMPlugin",                // (optional)
      bug: "https:/MyUsername/MyCoolBYTMPlugin/issues", // (optional)
      greasyfork: "...",                                            // (optional)
      openuserjs: "...",                                            // (optional)
    },
  },
  // The intents (permissions) the plugin needs to be granted to be able to use certain functions.
  // Search for "enum PluginIntent" in "src/types.ts" to see all available values, then sum all of them together to get the final intents number.
  // If you have BYTM as a dependency/submodule, you can import the enum and add the values like so: `PluginIntent.Foo | PluginIntent.Bar`
  intents: 18,              // required
  contributors: [           // (optional)
    {                                            // (optional)
      name: "MyUsername",                        // required
      homepage: "https:/MyUsername", // (optional)
      email: "[email protected]",        // (optional)
    },
    {                                              // (optional)
      name: "SomeOtherGuy",                        // required
      homepage: "https:/SomeOtherGuy", // (optional)
      email: "[email protected]",        // (optional)
    },
  ],
};

// private token for authenticated function calls (don't store this persistently, as your plugin gets a new one every page load!)
let authToken: string | undefined;

// since some function calls require the token, this function can be called to return it once the plugin is fully registered
export function getToken() {
  return authToken;
}

unsafeWindow.addEventListener("bytm:registerPlugin", (registerPlugin) => {
  try {
    // register the plugin
    const { token, events } = registerPlugin(pluginDef);
    // listen for the pluginRegistered event
    events.on("pluginRegistered", (info) => {
      // store the private token for later use in authenticated function calls
      authToken = token;
      console.log(`${info.name} (version ${info.version}) is registered`);

      onRegistered();
    });
    // for other events search for "type PluginEventMap" in "src/types.ts"
  }
  catch(err) {
    console.error("Failed to register plugin:", err);
  }
});

// put your plugin's logic that depends on authentication in here so it only runs after registration
function onRegistered() {
  try {
    // example authenticated function call:
    const bytmFeatureConfig = unsafeWindow.BYTM.getFeatures(getToken());
    if(!bytmFeatureConfig)
      console.error("Failed to get feature config object, the token is probably somehow invalid");
    else
      console.log("Feature config object:", bytmFeatureConfig);
  }
  catch(err) {
    console.error("Failed to run authenticated function call due to an underlying error:", err);
  }
}

getPluginInfo()

Usages:

unsafeWindow.BYTM.getPluginInfo(token: string | undefined, name: string, namespace: string): PluginInfo | undefined
unsafeWindow.BYTM.getPluginInfo(token: string | undefined, pluginDef: { plugin: { name: string, namespace: string } }): PluginInfo | undefined

Description:
Returns the plugin info object for the specified plugin. It's basically a more restricted version of the plugin definition object.
This object contains all information that other plugins will be able to see about your plugin.

Arguments:

• token - The private token that was returned when the plugin was registered (if not provided, the function will always return undefined)
• either:
  • name - The "name" property of the plugin
  • namespace - The "namespace" property of the plugin
• or:
  • pluginDef - A plugin definition object containing at least the plugin.name and plugin.namespace properties

The function will return undefined if the plugin is not registered or the token is invalid.
The type of the returned object can be found by searching for type PluginInfo in the file src/types.ts

Example (click to expand)

unsafeWindow.addEventListener("bytm:pluginsRegistered", () => {
  const pluginInfo = unsafeWindow.BYTM.getPluginInfo(myToken, "My cool plugin", "https:/MyUsername");
  if(pluginInfo)
    console.log(`The plugin '${pluginInfo.name}' with version '${pluginInfo.version}' is loaded`);
  else
    console.error("The plugin 'My cool plugin' is not registered");
});

getDomain()

Usage:

unsafeWindow.BYTM.getDomain(): "yt" | "ytm"

Description:
Returns the current domain of the page as a constant string.
It will return "yt" for YouTube and "ytm" for YouTube Music.
Throws an error if the domain is not supported by BetterYTM.

Example (click to expand)

try {
  const domain = unsafeWindow.BYTM.getDomain();

  if(domain === "yt")
    console.log("Running on YouTube");
  else
    console.log("Running on YouTube Music");^
}
catch(err) {
  console.error("Running on an unsupported domain:", err);
}

getResourceUrl()

Usage:

unsafeWindow.BYTM.getResourceUrl(): Promise<string>

Description:
Returns a blob: URL for the specified BYTM resource file.
You can find a list of them by looking at the @resource directives in the userscript header or in the files assets/resources.json and src/tools/post-build.ts
The resource and its URL are provided by the userscript extension and it is locally cached for quicker fetching.

Should a resource not be defined, the function will return the equivalent URL from the GitHub repository instead.
Should that also fail, it will try to return a base64-encoded data: URI version of the resource.

Arguments:

• resourceName - The name of the resource to get the URL for.

Example (click to expand)

const deleteButtonImg = document.createElement("img");
deleteButtonImg.src = await unsafeWindow.BYTM.getResourceUrl("delete");

myElement.appendChild(deleteButtonImg);

getSessionId()

Usage:

unsafeWindow.BYTM.getSessionId(): string | null

Description:
Returns the unique session ID that is generated on every page load.
It should persist between history navigations, but not between page reloads.

⚠️ On privacy-focused browsers or if cookies are disabled, this function will return null since sessionStorage is not available.

Example (click to expand)

const sessionId = unsafeWindow.BYTM.getSessionId();

if(await GM.getValue("_myPlugin-sesId") !== sessionId) {
  console.log("New session started");
  // do something that should only be done once per session
  // or store values persistently that should be unique per session:
  await GM.setValue("_myPlugin-sesId", sessionId);
}

setInnerHtml()

Usage:

unsafeWindow.BYTM.setInnerHtml(element: HTMLElement, html: string): void

Description:
Sets the innerHTML property of the specified element to the provided string, after sanitizing it.
This is done for compatibility with the Trusted Types API and to prevent XSS attacks.
Uses the library DOMPurify on default settings to sanitize the HTML.

Arguments:

• element - The element to set the innerHTML property of
• html - The HTML string to sanitize and set as the innerHTML property

Example (click to expand)

const myElement = document.querySelector("#myElement");
const htmlString = "<img onload='alert(\"XSS attack!\")' src='https://picsum.photos/100/100'>";

unsafeWindow.BYTM.setInnerHtml(myElement, htmlString);

addSelectorListener()

Usage:

unsafeWindow.BYTM.addSelectorListener<TElem extends Element>(observerName: ObserverName, selector: string, options: SelectorListenerOptions<TElem>): void

Description:
Adds a listener to the specified SelectorObserver instance that gets called when the element/s behind the passed selector is/are found.
They are immediately checked for and then checked again whenever the part of the DOM tree changes (elements get added or removed) that is observed by that specific SelectorObserver.

The instances are chained together in a way that the least specific observer is the parent of the more specific ones.
This is done to limit the amount of checks that need to be run, especially on pages with a lot of dynamic content and if continuous listeners are used.
See the UserUtils SelectorObserver documentation for more info and example code.

⚠️ Due to this chained architecture, the selector you pass can only start with an element that is a child of the observer's base element.
If you provide a selector that starts higher up or directly on the base element, the listener will never be called.
You can check which observer has which base element in the file src/observers.ts

Arguments:

• observerName - The name of the SelectorObserver instance to add the listener to. You can find all available instances and which base element they observe in the file src/observers.ts
• selector - The CSS selector to observe for changes.
• options - The options for the listener. See the UserUtils SelectorObserver documentation

Example (click to expand)

// wait for the observers to exist
unsafeWindow.addEventListener("bytm:observersReady", () => {
  // use the "lowest" possible SelectorObserver (playerBar) to prevent unnecessary checks
  // and call the listener as soon as the passed selector is found in the DOM
  unsafeWindow.BYTM.addSelectorListener<HTMLAnchorElement>("playerBar", "#bytm-player-bar-lyrics-btn", {
    listener: (lyricsBtnElem) => {
      console.log("The player bar lyrics button was added or removed:", lyricsBtnElem);
    },
  });
});

onInteraction()

Usage:

unsafeWindow.BYTM.onInteraction(
  element: HTMLElement,
  callback: (event: MouseEvent | KeyboardEvent) => void,
  listenerOptions?: AddEventListenerOptions
): void

Description:
Adds accessible event listeners to the specified element for button or link-like keyboard and mouse interactions.
All events passed to the callback function automatically have the default behavior prevented and stop propagation, meaning no other listener of the same type will be called.
For keyboard events this only happens as long as the captured key is a valid interaction key (Space, Enter).

Arguments:

• element - The element to add the listeners to
• callback - The function to call when the element is interacted with
• listenerOptions - Optional event listener options (same as the third argument of addEventListener, shared between the keyboard and mouse event listeners)

Example (click to expand)

const myButton = document.querySelector("button#myButton");

unsafeWindow.BYTM.onInteraction(myButton, (event) => {
  if(event instanceof MouseEvent)
    console.log("The button was clicked");
  else if(event instanceof KeyboardEvent)
    console.log("The button was activated with the keyboard (Space / Enter)");
}, {
  // if `once` is set, when either the mouse or keyboard event are triggered once,
  // the other listener is automatically removed as well to prevent double triggering
  once: true,
  // you can pass `capture` to listen in the capture phase (helpful for triggering before other listeners),
  // or an AbortController's `signal` to be able to abort the listener
});

getVideoTime()

Usage:

unsafeWindow.BYTM.getVideoTime(precision?: number): Promise<number | null>

Description:
Returns the current video time in seconds, with the given precision (2 decimal digits by default).
Rounds down if the precision is set to 0. The maximum average available precision on YTM is 6.

In case the time can't be determined on YT, mouse movement is simulated to bring up the video time element and read it.
In order for that edge case not to throw an error, the function would need to be called in response to a user interaction event (e.g. click) due to the strict automated interaction policy in browsers, otherwise an error can be thrown.
Resolves with a number of seconds or null if the time couldn't be determined.

Example (click to expand)

try {
  // get the video time with 3 decimal digits
  const videoTime = await unsafeWindow.BYTM.getVideoTime(3);
  console.log(`The video time is ${videoTime}s`);
}
catch(err) {
  console.error("Couldn't get the video time, probably due to automated interaction restrictions");
}

getThumbnailUrl()

Usage:

unsafeWindow.BYTM.getThumbnailUrl(
  watchID: string,
  qualityOrIndex: "maxresdefault" | "sddefault" | "hqdefault" | "mqdefault" | "default" | 0 | 1 | 2 | 3
): string

Description:
Returns the URL to the thumbnail of the video with the specified watch/video ID and quality (resolution).
If an index number is passed, 0 will return a very low resolution thumbnail and 1-3 will return a very low resolution frame from the video (if available).

Arguments:

• watchID - The watch/video ID of the video to get the thumbnail for
• qualityOrIndex - The quality or index of the thumbnail to get. Possible quality strings sorted by highest resolution first: maxresdefault > sddefault > hqdefault > mqdefault > default. If no quality is specified, maxresdefault (highest resolution) is used.

Example (click to expand)

const thumbnailUrl = unsafeWindow.BYTM.getThumbnailUrl("dQw4w9WgXcQ", "maxresdefault");
console.log(thumbnailUrl); // "https://img.youtube.com/vi/dQw4w9WgXcQ/maxresdefault.jpg"

getBestThumbnailUrl()

Usage:

unsafeWindow.BYTM.getBestThumbnailUrl(watchID: string): Promise<string | undefined>

Description:
Returns the URL to the best resolution thumbnail of the video with the specified watch/video ID.
Will sequentially try to get the highest quality thumbnail available until one is found.
Resolution priority list: maxresdefault.jpg > sddefault.jpg > hqdefault.jpg > 0.jpg

If no thumbnail is found, the Promise will resolve with undefined
May throw if an error occurs while fetching the thumbnails.

Arguments:

• watchID - The watch/video ID of the video to get the thumbnail for

Example (click to expand)

try {
  const thumbnailUrl = await unsafeWindow.BYTM.getBestThumbnailUrl("dQw4w9WgXcQ");
  if(thumbnailUrl)
    console.log(thumbnailUrl); // "https://img.youtube.com/vi/dQw4w9WgXcQ/maxresdefault.jpg"
}
catch(err) {
  console.error("Failed to get the best thumbnail URL:", err);
}

waitVideoElementReady()

Usage:

unsafeWindow.BYTM.waitVideoElementReady(): Promise<HTMLVideoElement>

Description:
Waits for the video element to be queryable in the DOM.
The Promise could potentially take a while, since it will only resolve if the /watch page is loaded and the video element is queryable and has the media buffered and ready.
If the video element already exists, the Promise will resolve immediately.
This function has to be called after the bytm:observersReady event has been dispatched.

Example (click to expand)

unsafeWindow.addEventListener("bytm:observersReady", async () => {
  const videoElement = await unsafeWindow.BYTM.waitVideoElementReady();
  console.log("The video element:", videoElement);
});

getCurrentMediaType()

Usage:

unsafeWindow.BYTM.getCurrentMediaType(): "video" | "song"

Description:
Returns the type of media that is currently playing (works on YTM only).
It will return "video" for videos (manually uploaded to YT - plays an actual video) and "song" for songs (automatic YTM releases - only displays a static, square image).
Throws an error if waitVideoElementReady() hasn't been awaited yet or the function is called on YT.

Example (click to expand)

// only available on YTM:
if(unsafeWindow.BYTM.getDomain() === "ytm") {
  const mediaType = unsafeWindow.BYTM.getCurrentMediaType();
  console.log(`The current media type is: ${mediaType}`); // "video" or "song"
}

setLocale()

Usage:

unsafeWindow.BYTM.setLocale(token: string | undefined, locale: string): void

Description:
Sets the locale for BetterYTM's translations.
The new locale is used for all translations after this function is called.

Arguments:

• token - The private token that was returned when the plugin was registered (if not provided, the function will do nothing).
• locale - The locale to set. Refer to the file assets/locales.json for a list of available locales.

Example (click to expand)

unsafeWindow.BYTM.setLocale(myToken, "en_UK");

getLocale()

Usage:

unsafeWindow.BYTM.getLocale(): string

Description:
Returns the currently set locale.

Example (click to expand)

unsafeWindow.BYTM.getLocale(); // "en_US"

unsafeWindow.BYTM.setLocale("en_UK");

unsafeWindow.BYTM.getLocale(); // "en_UK"

hasKey()

Usage:

unsafeWindow.BYTM.hasKey(key: string): boolean

Description:
Returns true if the specified translation key exists in the currently set locale.

Arguments:

• key - The key of the translation to check for.

Example (click to expand)

unsafeWindow.BYTM.hasKey("lyrics_rate_limited"); // true
unsafeWindow.BYTM.hasKey("some_key_that_doesnt_exist"); // false

hasKeyFor()

Usage:

unsafeWindow.BYTM.hasKeyFor(locale: string, key: string): boolean

Description:
Returns true if the specified translation key exists in the specified locale.

Arguments:

• locale - The locale to check for the translation key in.
• key - The key of the translation to check for.

Example (click to expand)

unsafeWindow.BYTM.hasKeyFor("en_UK", "lyrics_rate_limited"); // true
unsafeWindow.BYTM.hasKeyFor("en_UK", "some_key_that_doesnt_exist"); // false

t()

Usage:

unsafeWindow.BYTM.t(key: TFuncKey, ...values: Stringifiable[]): string

Description:
Returns the translation for the provided translation key and currently set locale.
To see a list of translations, check the file assets/translations/en_US.json

Arguments:

• translationKey - The key of the translation to get.
• ...values - A spread parameter of values that can be converted to strings to replace the numbered placeholders in the translation with.

Example (click to expand)

const customConfigMenuTitle = document.createElement("div");
customConfigMenuTitle.textContent = unsafeWindow.BYTM.t("config_menu_title", "My cool BYTM Plugin");
// translated text: "My cool BYTM Plugin - Configuration" (if locale is en_US or en_UK)

tp()

Usage:

unsafeWindow.BYTM.tp(key: TFuncKey, num: number | unknown[] | NodeList, ...values: Stringifiable[]): string

Description:
Returns the translation for the provided translation key, including pluralization identifier and currently set locale.
To see a list of translations, check the file assets/translations/en_US.json

The pluralization identifier is determined by the number of items in the second argument.
It can be either "1" or "n" and will be appended to the translation key separated by a hyphen.

Arguments:

• translationKey - The key of the translation to get.
• num - The number of items to determine the pluralization identifier from. Can also be an array or NodeList.
• ...values - A spread parameter of values that can be converted to strings to replace the numbered placeholders in the translation with.

Example (click to expand)

try {
  const lyrics = await unsafeWindow.BYTM.fetchLyricsUrl("Michael Jackson", "Thriller");
}
catch(err) {
  if(err instanceof Error && err.status === 429) {
    // rate limited
    const retryAfter = err.response.headers["retry-after"];
    const retryAfterSeconds = retryAfter ? parseInt(retryAfter) : 60;
    const errorText = unsafeWindow.BYTM.tp("lyrics_rate_limited", retryAfterSeconds);
    // translation key: "lyrics_rate_limited-n"
    // translated text: "You are being rate limited.\nPlease wait 23 seconds before requesting more lyrics."
    alert(errorText);
  }
}

tl()

Usage:

unsafeWindow.BYTM.tl(locale: string, key: TFuncKey, ...values: Stringifiable[]): string

Description:
Returns the translation for the provided translation key and locale.
Useful to get the translation for a specific locale without changing the currently set locale.
To see a list of possible translation values, check the file assets/translations/en_US.json

Arguments:

• locale - The locale to get the translation for.
• translationKey - The key of the translation to get.
• ...values - A spread parameter of values that can be converted to strings to replace the numbered placeholders in the translation with.

For an example, see t() which behaves in the same way, but uses the currently set locale instead of a specified one.

tlp()

Usage:

unsafeWindow.BYTM.tlp(locale: string, key: TFuncKey, num: number | unknown[] | NodeList, ...values: Stringifiable[]): string

Description:
Returns the translation for the provided translation key, including pluralization identifier and locale.
Useful to get the translation for a specific locale without changing the currently set locale.
To see a list of possible translation values, check the file assets/translations/en_US.json

Arguments:

• locale - The locale to get the translation for.
• key - The key of the translation to get.
• num - The number of items to determine the pluralization identifier from. Can also be an array or NodeList.
• ...values - A spread parameter of values that can be converted to strings to replace the numbered placeholders in the translation with.

For an example, see tp() which behaves in the same way, but uses the currently set locale instead of a specified one.

getFeatures()

Usage:

unsafeWindow.BYTM.getFeatures(token: string | undefined): FeatureConfig

Description:
Returns the current feature configuration object synchronously from memory.
To see the structure of the object, check out the type FeatureConfig in the file src/types.ts
If features are set to be hidden using valueHidden: true, their value will always be undefined in the returned object.
In the future, an intent could grant access to the hidden values, but for now, they are only accessible to BetterYTM itself.

Example (click to expand)

const features = unsafeWindow.BYTM.getFeatures(myToken);
console.log(`The volume slider step is currently set to ${features.volumeSliderStep}`);

saveFeatures()

Usage:

unsafeWindow.BYTM.saveFeatures(token: string | undefined, config: FeatureConfig): Promise<void>

Description:
Overwrites the current feature configuration object with the provided one.
The object in memory is updated synchronously, while the one in GM storage is updated asynchronously once the Promise resolves.

Arguments:

• token - The private token that was returned when the plugin was registered (if not provided, the function will do nothing).
• config - The full config object to save. If properties are missing, BYTM will break!
  To see the structure of the object, check out the type FeatureConfig in the file src/types.ts

Example (click to expand)

async function updateVolSliderStep() {
  const oldConfig = unsafeWindow.BYTM.getFeatures();
  const newConfig = { ...oldConfig, volumeSliderStep: 1 };

  const promise = unsafeWindow.BYTM.saveFeatures(myToken, newConfig);
  // new config is now saved in memory, but not yet in GM storage
  // so this already returns the updated config:
  console.log(unsafeWindow.BYTM.getFeatures());

  await promise;
  // now the data is saved persistently in GM storage and the page can
  // safely be reloaded without losing the updated config data
}

updateVolSliderStep();

fetchLyricsUrlTop()

Usage:

unsafeWindow.BYTM.fetchLyricsUrlTop(artist: string, song: string): Promise<string | undefined>

Description:
Fetches the top result's URL to the lyrics page for the specified song.
If there is already an entry in the in-memory cache for the song, it will be returned without fetching anything new.
URLs that are returned by this function are added to the cache automatically.
Returns undefined if there was an error while fetching the URL.

Arguments:

• artist - The main artist of the song to fetch the lyrics URL for.
  The value needs to be sanitized with sanitizeArtists() before being passed to this function.
• song - The title of the song to fetch the lyrics URL for.
  The value needs to be sanitized with sanitizeSong() before being passed to this function.

Example (click to expand)

async function getLyricsUrl() {
  const lyricsUrl = await unsafeWindow.BYTM.fetchLyricsUrlTop("Michael Jackson", "Thriller");

  if(lyricsUrl)
    console.log(`The lyrics URL for Michael Jackson's Thriller is '${lyricsUrl}'`);
  else
    console.log("Couldn't find the lyrics URL for this song");
}

getLyricsUrl();

getLyricsCacheEntry()

Usage:

unsafeWindow.BYTM.getLyricsCacheEntry(artists: string, song: string): LyricsCacheEntry | undefined

Description:
Tries to find an entry in the in-memory cache for the specified song.
You can find the structure of the LyricsCacheEntry type in the file src/types.ts
Contrary to fetchLyricsUrlTop(), this function does not fetch anything new if there is no entry in the cache.

Arguments:

• artist - The main artist of the song to grab the lyrics URL for.
  The value needs to be sanitized with sanitizeArtists() before being passed to this function.
• song - The title of the song to grab the lyrics URL for.
  The value needs to be sanitized with sanitizeSong() before being passed to this function.

Example (click to expand)

function tryToGetLyricsUrl() {
  const lyricsEntry = unsafeWindow.BYTM.getLyricsCacheEntry("Michael Jackson", "Thriller");

  if(lyricsEntry)
    console.log(`The lyrics URL for Michael Jackson's Thriller is '${lyricsEntry.url}'`);
  else
    console.log("Couldn't find the lyrics URL for this song in cache");
}

tryToGetLyricsUrl();

sanitizeArtists()

Usage:

unsafeWindow.BYTM.sanitizeArtists(artists: string): string

Description:
Sanitizes the specified artist string to be used in fetching a lyrics URL.
This tries to strip out special characters and co-artist names, separated by a comma or ampersand.
Returns (hopefully) a single artist name with leading and trailing whitespaces trimmed.

Arguments:

• artists - The string of artist name(s) to sanitize.

Example (click to expand)

// usually artist strings will only have one of these characters but this is just an example
const sanitizedArtists = unsafeWindow.BYTM.sanitizeArtists(" Michael Jackson    • Paul McCartney & Freddy Mercury, Frank Sinatra");
console.log(sanitizedArtists); // "Michael Jackson"

sanitizeSong()

Usage:

unsafeWindow.BYTM.sanitizeSong(songName: string): string

Description:
Sanitizes the specified song title string to be used in fetching a lyrics URL.
This tries to strip out special characters and everything inside regular and square parentheses like (Foo Remix).
Returns (hopefully) a song title with leading and trailing whitespaces trimmed.

Arguments:

• songName - The string of the song title to sanitize.

Example (click to expand)

const sanitizedSong = unsafeWindow.BYTM.sanitizeSong(" Thriller (Freddy Mercury Cover) [Tommy Cash Remix]");
console.log(sanitizedSong); // "Thriller"

getAutoLikeData()

Usage:

unsafeWindow.BYTM.getAutoLikeData(token: string | undefined): AutoLikeData

Description:
Returns the current auto-like data object synchronously from memory.
To see the structure of the object, check out the type AutoLikeData in the file src/types.ts

Arguments:

• token - The private token that was returned when the plugin was registered (if not provided, the function will return an empty object).

Example (click to expand)

const autoLikeData = unsafeWindow.BYTM.getAutoLikeData(myToken);

// check if the channel is added to the auto-like list and if it's currently enabled
function isEnabledForChannel(channelId: string) {
  return autoLikeData && autoLikeData.channels.find((ch) => ch.id === channelId && ch.enabled);
}

// channelId can be in the format UC... or @username
console.log(isEnabledForChannel("UCXuqSBlHAE6Xw-yeJA0Tunw"));

saveAutoLikeData()

Usage:

unsafeWindow.BYTM.saveAutoLikeData(token: string | undefined, data: AutoLikeData): Promise<void>

Description:
Saves the provided auto-like data object synchronously to memory and asynchronously to GM storage.

Arguments:

• token - The private token that was returned when the plugin was registered (if not provided, the function will return an empty object).
• data - The full auto-like data object to save. No validation is done so if properties are missing, BYTM will break!

Example (click to expand)

async function toggleAutoLikeForChannel(channelId: string, channelName: string) {
  const autoLikeData = unsafeWindow.BYTM.getAutoLikeData(myToken);
  const channelIndex = autoLikeData.channels.findIndex((ch) => ch.id === channelId);

  if(channelIndex > -1)
    autoLikeData.channels[channelIndex].enabled = !autoLikeData.channels[channelIndex].enabled;
  else
    autoLikeData.channels.push({ id: channelId, name: channelName, enabled: true });

  await unsafeWindow.BYTM.saveAutoLikeData(myToken, autoLikeData);
}

// channelId can be in the format UC... or @username
toggleAutoLikeForChannel("UCXuqSBlHAE6Xw-yeJA0Tunw", "Linus Sex Tips").then(() => {
  const newAutoLikeData = unsafeWindow.BYTM.getAutoLikeData(myToken);
  console.log("Auto-like status for the channel was toggled. New data:", newAutoLikeData);
});

fetchVideoVotes()

Usage:

unsafeWindow.BYTM.fetchVideoVotes(videoId: string): Promise<VideoVotesObj | undefined>

Description:
Fetches the approximate like and dislike counts for the specified video ID, using the ReturnYoutubeDislike API.
RYD will approximate the votes based on historical data and users of the browser extension. The numbers will never be 100% accurate!
The object returned by this function has the structure of the VideoVotesObj type in the file src/types.ts
If the video ID is not found or the API is down, the function will return undefined

Arguments:

• videoId - The video ID to fetch the votes for (e.g. dQw4w9WgXcQ)

Example (click to expand)

async function getVotes() {
  const votes = await unsafeWindow.BYTM.fetchVideoVotes("dQw4w9WgXcQ");

  if(!votes)
    return console.error("Couldn't fetch the votes for this video");

  console.log(`The video has ${votes.likes} likes and ${votes.dislikes} dislikes`);
}

getVotes();

BytmDialog

Usage:

new unsafeWindow.BYTM.BytmDialog(options: BytmDialogOptions): BytmDialog

A class that can be used to create and manage a modal dialog with a fully customizable header, body and footer.

Features:

• Can be opened, closed, mounted, unmounted and destroyed at any time for full control.
• The dialog is fully responsive and can be used on any screen size, but will not exceed the provided maximum width and height.
• Scrollability of the body is automatically removed and the body is set to be inert (ignore keyboard input) automatically. Also works with multiple dialogs open at the same time!
• The dialog can be closed by clicking the background, pressing the escape key or clicking the close button (freely configurable).
• Features many helper methods and events to make it more flexible and easier to work with.
• Has an optional small mode for a more compact appearance.
• If needed, the relatively uniform CSS naming conventions make it easy for the appearance to be overridden by a BetterYTM plugin or userstyle.

Options properties:

Property	Description
id: string	ID that gets added to child element IDs - has to be unique and conform to HTML ID naming rules!
width: number	Maximum and target width of the dialog in pixels
height: number	Maximum and target height of the dialog in pixels
closeOnBgClick?: boolean	Whether the dialog should close when the background is clicked - defaults to true
closeOnEscPress?: boolean	Whether the dialog should close when the escape key is pressed - defaults to true
closeBtnEnabled?: boolean	Whether the close button should be enabled - defaults to true
destroyOnClose?: boolean	Whether the dialog should be destroyed when it's closed - defaults to false
small?: boolean	Whether the dialog should have a smaller overall appearance - defaults to false
verticalAlign?: string	Where the dialog should be anchored vertically ("top", "center" or "bottom") - defaults to "center"
renderBody: () => HTMLElement │ Promise<HTMLElement>	Called to render the body of the dialog
renderHeader?: () => HTMLElement │ Promise<HTMLElement>	Called to render the header of the dialog - leave undefined for a blank header
renderFooter?: () => HTMLElement │ Promise<HTMLElement>	Called to render the footer of the dialog - leave undefined for no footer

Methods:
The methods from the NanoEmitter class are also available here.
These are the additional methods that are exclusive to the BytmDialog class:

• open(e?: MouseEvent | KeyboardEvent): Promise<void>
  Opens the dialog - also mounts it if it hasn't been mounted yet.
  Prevents default action and immediate propagation if an event is passed.
  Resolves once the dialog is fully mounted and opened.
• close(e?: MouseEvent | KeyboardEvent): void
  Closes the dialog - also unmounts and destroys it if the destroyOnClose option is set to true.
  Prevents default action and immediate propagation if an event is passed.
• isOpen(): boolean
  Returns true if the dialog is currently open, false if not.
• isMounted(): boolean
  Returns true if the dialog is currently mounted, false if not.
• mount(): Promise<void>
  Mounts the dialog to the DOM without making it visible - can be called before opening the dialog for the first time to pre-load all elements.
  Resolves once the dialog is fully mounted in the DOM.
• unmount(): void
  Removes the dialog from the DOM.
• remount(): Promise<void>
  Unmounts and mounts the dialog again.
  This can be used to re-render the dialog's contents with new information.
  Resolves once the dialog is fully mounted in the DOM.
• destroy(): void
  Unmounts and removes the dialog from the DOM and removes all event listeners.
  Should be called when the dialog is no longer needed.
• static getLastDialogId(): string
  Returns the ID of the last dialog that was opened.
  This can be used to check if a dialog is currently open and to get its ID for further use.
  Static method usage: unsafeWindow.BytmDialog.getLastDialogId() (doesn't need a new instance)
• static getOpenDialogs(): string[]
  Returns the IDs of all open dialogs, topmost (last opened) first.
  Static method usage: unsafeWindow.BytmDialog.getOpenDialogs() (doesn't need a new instance)

Events:

Event	Description
on("close", () => void)	Emitted just after the dialog is closed
on("open", () => void)	Emitted just after the dialog is opened
on("render", () => void)	Emitted just after the dialog contents are rendered
on("clear", () => void)	Emitted just after the dialog contents are cleared
on("destroy", () => void)	Emitted just after the dialog is destroyed and before all listeners are removed

Example (click to expand)

const dialog = new unsafeWindow.BYTM.BytmDialog({
  id: "my-dialog",
  width: 500,
  height: 300,
  closeOnBgClick: true,
  closeOnEscPress: true,
  closeBtnEnabled: true,
  destroyOnClose: false,
  small: true,
  verticalAlign: "top", // if the content's height changes, it's better to anchor it to the top or bottom
  // add elements to the header, body and footer here, in one of these ways:
  // - setInnerHtml(foo, "..."); // (see contributing guide)
  // - foo.appendChild(document.createElement("..."));
  // - ReactDOM.render(<MyComponent />, foo);
  // - etc.
  renderHeader: () => {
    const header = document.createElement("div");
    header.textContent = "This is the header of my dialog";
    return header;
  },
  renderBody: () => {
    const body = document.createElement("div");
    body.textContent = "This is the body of my dialog";
    return body;
  },
  renderFooter: () => {
    const footer = document.createElement("div");
    footer.textContent = "This is the footer of my dialog";
    return footer;
  },
});

async function run() {
  dialog.on("close", () => {
    console.log("The dialog was closed");
  });

  await dialog.open();
  console.log("The dialog is now open");
}

run();

ExImDialog

Usage:

new unsafeWindow.BYTM.ExImDialog(options: ExImDialogOptions): ExImDialog

A subclass of BytmDialog that can be used to create and manage a generic export/import dialog.

Features:

• Has all the features of the BytmDialog class
• Can be used to export and import any kind of data that can be serialized to a string
• Built-in textareas for the user to paste or copy data to/from
• Copy to clipboard button for the export textarea
• Ability to copy a second variety of the data when shift-clicking the copy button
• Exported data is hidden by default in case it contains sensitive information
• Text can be given as a constant string or "lazy-loaded" via sync or async function

Options properties:
All properties from the BytmDialog class are available here as well, except for renderHeader, renderBody and renderFooter

Property	Description
title: string | (() => (string | Promise<string>))	Title of the dialog
descImport: string | (() => (string | Promise<string>))	Description of the dialog when importing
descExport: string | (() => (string | Promise<string>))	Description of the dialog when exporting
onImport: (data: string) => void	Callback function that gets called when the user imports data
exportData: string | (() => (string | Promise<string>))	The data to export (or a function that returns the data as string, either sync or async)
exportDataSpecial?: string | (() => (string | Promise<string>))	Optional variant of the data, used for special cases like when shift-clicking the copy button

Example (click to expand)

const exImDialog = new unsafeWindow.BYTM.ExImDialog({
  id: "my-exim-dialog",
  width: 500,
  height: 400,
  exportData,
  exportDataSpecial,
  onImport,
  title: "My ExIm Dialog",
  descImport: "Past the data you want to import below and click import!",
  descExport: "Copy the data below to save it or share it with others. Warning: may contain sensitive information!",
});

type MyDataType = { foo: string };

async function exportData() {
  // compress the data to save space
  const exportData = JSON.stringify({ foo: "bar" });
  return await unsafeWindow.BYTM.UserUtils.compress(exportData, "deflate");
}

function exportDataSpecial() {
  // return the data uncompressed when shift-clicking the copy button
  return JSON.stringify({ foo: "bar" });
}

async function onImport(data: string) {
  let decompData: string;
  try {
    // since the data could either be compressed or not, try to decompress it and see if it errors
    decompData = await unsafeWindow.BYTM.UserUtils.decompress(data, "deflate");
  }
  catch {
    // the data is not compressed, so just use it as is
    decompData = data;
  }

  let parsedData: MyDataType;
  try {
    parsedData = JSON.parse(decompData);
  }
  catch(err) {
    console.error("The user imported invalid data");
    return;
  }

  console.log("The user successfully imported data:", parsedData);
}

async function run() {
  exImDialog.on("close", () => {
    console.log("The dialog was closed");
  });

  await exImDialog.open();
  console.log("The dialog is now open");
}

run();

MarkdownDialog

Usage:

new unsafeWindow.BYTM.MarkdownDialog(options: MarkdownDialogOptions): MarkdownDialog

A subclass of BytmDialog that can be used to create and manage a dialog that renders its entire body using GitHub-flavored Markdown (and HTML mixins).
Note: the provided id will be prefixed with md- to avoid conflicts with other dialogs.

Features:

• Has all the features of the BytmDialog class
• Can be used to display any kind of information in a dialog that can be written in Markdown
• Supports GitHub-flavored Markdown and HTML mixins like <details> and <summary>

Options properties:
All properties from the BytmDialog class are available here as well, except for renderBody (which was replaced by body)

Property	Description
body: string | (() => string | Promise<string>)	Markdown content to render in the dialog. Can be a string or a sync or async function that returns a string.

Methods:
The methods from the NanoEmitter and BytmDialog classes are also available here.

• static parseMd(md: string): Promise<string>
  Parses the provided Markdown string (with GitHub flavor and HTML mixins) and returns the HTML representation as a string.
• protected renderBody(): Promise<void>
  Renders the Markdown content to the dialog's body element. You can only override this method if you create a subclass of MarkdownDialog
  If you do, you can use parseMd() to render a custom mixture of Markdown HTML and JavaScript-created elements to the body.

Example (click to expand)

const markdownDialog = new unsafeWindow.BYTM.MarkdownDialog({
  id: "my-markdown-dialog",
  width: 400,
  height: 600,
  body: `\
# Look at this table:
| Column 1 | Column 2 |
| --: | :-- |
| Hello | World |

<br /><br /><br /><br />

<details>
  <summary>Click me!</summary>
  I'm a hidden text block!
</details>`,
});

createHotkeyInput()

Usage:

unsafeWindow.BYTM.createHotkeyInput(inputProps: {
  initialValue?: HotkeyObj,
  onChange: (hotkey: HotkeyObj) => void,
}): HTMLElement

Creates a hotkey input element that can be used to let the user set a hotkey.
The HotkeyObj type has the properties code: string, shift: boolean, ctrl: boolean and alt: boolean
The function onChange is called whenever the hotkey was changed.

Example (click to expand)

const hotkeyInput = unsafeWindow.BYTM.createHotkeyInput({
  initialValue: { code: "KeyA", shift: true, ctrl: false, alt: false },
  onChange: (hotkey) => {
    console.log(`The hotkey was changed to ${hotkey.code} with shift: ${hotkey.shift}, ctrl: ${hotkey.ctrl} and alt: ${hotkey.alt}`);
  },
});

document.querySelector("#my-element").appendChild(hotkeyInput);

createToggleInput()

Usage:

unsafeWindow.BYTM.createToggleInput(toggleProps: {
  onChange: (value: boolean) => void,
  initialValue?: boolean,
  id?: string,
  labelPos?: "off" | "left" | "right",
})

Creates a toggle input element that behaves like a checkbox but looks better.

• onChange - Callback function that is called when the toggle is changed, gets passed the new value of the toggle as a boolean.
• initialValue - Initial value of the toggle - defaults to false (toggled off).
• id - Useful for getting a unique selector - if unspecified, a random ID is generated.
• labelPos - Toggle builtin label "off" or change position of the label to "right" or "left", relative to the toggle.

Example (click to expand)

const toggleInput = unsafeWindow.BYTM.createToggleInput({
  onChange: (value) => {
    console.log(`The toggle was changed to ${value}`);
  },
  initialValue: true,
  id: "my-toggle",
  labelPos: "left",
});

document.querySelector("#my-element").appendChild(toggleInput);

createCircularBtn()

Usage:

unsafeWindow.BYTM.createCircularBtn(props: CircularBtnProps): Promise<HTMLElement>

Creates a circular button element containing an icon that can be used to trigger an action or navigate to a different page.

Properties:

• title: string - The title that is displayed when hovering over the button. Also used as a description for accessibility.
• either of:
  • resourceName: string - Name of the resource starting with icon- to use as the button icon (see assets/resources.json)
  • src: string | Promise<string> - URL of the image to use as the button icon
• either of:
  • href: string - URL to navigate to when the button is clicked or interacted with.
  • onClick: (evt: MouseEvent | KeyboardEvent) => void - Function that's called when the button is clicked or interacted with

Example (click to expand)

const circularBtn = unsafeWindow.BYTM.createCircularBtn({
  title: "My cool button",
  resourceName: "icon-help",
  onClick() {
    console.log("The button was clicked");
  },
});

document.querySelector("#my-element").appendChild(circularBtn);

createLongBtn()

Usage:

unsafeWindow.BYTM.createLongBtn(props: LongBtnProps): Promise<HTMLElement>

Creates a long button element that can be used to trigger an action or navigate to a different page.
It can also be set up to act as a toggle button with rich CSS classes for customization.

Properties:

• text: string - The text to display on the button
• title: string - The title of the button that is displayed when hovering over it. Also used as a description for accessibility
• iconPosition?: "left" | "right" - The position of the icon relative to the text. Can be "left" or "right" (defaults to "left")
• either of:
  • resourceName: string - Name of the resource to use as the icon (see assets/resources.json)
  • src: string - URL of the image to use as the icon
• either of:
  • href: string - URL to navigate to when the button is clicked or interacted with.
  • onClick: (evt: MouseEvent | KeyboardEvent) => void - Function to call when the button is clicked or interacted with
  • toggle: true - Set to true to make the button act as a toggle button
    In addition, there are these props:
    • onToggle: (state: boolean, evt: MouseEvent | KeyboardEvent) => void - Function to call when the button is interacted with
    • toggleInitialState?: boolean - The initial value of the toggle button (optional, defaults to false)
    • togglePredicate?: (evt: MouseEvent | KeyboardEvent) => boolean - Gets called every toggle attempt to determine if the state should swap and onToggle should be called

Example (click to expand)

// link:
const linkBtn = unsafeWindow.BYTM.createLongBtn({
  resourceName: "icon-help",
  href: "https://example.com/help?topic=foo",
  text: "Help",
  title: "Click to open the help page for this topic",
  iconPosition: "right",
});

// button:
const btn = unsafeWindow.BYTM.createLongBtn({
  src: "https://example.com/icon.png",
  onClick(evt: MouseEvent | KeyboardEvent) {
    console.log("The button was clicked");
  },
  text: "Upload",
  title: "Click to upload a file",
});

// toggle:
const toggleBtn = unsafeWindow.BYTM.createLongBtn({
  resourceName: "icon-globe",
  toggle: true,
  toggleInitialState: true,
  togglePredicate: (evt) => {
    // don't toggle if shift is pressed and instead do something special
    evt.shiftKey && doSomething(evt);
    return !evt.shiftKey;
  },
  onToggle(state: boolean, evt: MouseEvent | KeyboardEvent) {
    console.log(`The button was toggled ${state ? "on" : "off"}`);
  },
  text: "Toggle",
  title: "Click to toggle something",
});

createRipple()

Usages:

unsafeWindow.BYTM.createRipple<TElement>(rippleElement: TElement, props?: RippleProps): TElement
unsafeWindow.BYTM.createRipple(rippleElement?: undefined, props?: RippleProps): HTMLDivElement

Creates a circular, expanding ripple effect on the specified element or creates a new one with the effect already applied if none is provided.
Returns either the new element or the initially passed one.
External CSS overrides can be used to change the color, size, speed values and opacity.
The exact speed values and variable names and locations can be found in src/components/ripple.css

Properties:

• speed?: string - The speed of the ripple effect. Can be "faster", "fast", "normal", "slow" or "slower" (defaults to "normal")

Example (click to expand)

const myBoringButton = document.querySelector("#my-boring-button");
if(myBoringButton)
  unsafeWindow.BYTM.createRipple(myBoringButton, { speed: "slowest" }); // it's as easy as this

showToast()

Usages:

unsafeWindow.BYTM.showToast(props: ToastProps): Promise<void>;
unsafeWindow.BYTM.showToast(message: string): Promise<void>;

Shows a toast notification with the specified message or element for the given duration and anchored in the specified corner of the viewport.
If a toast is already shown, it will be immediately closed and the new one will be shown shortly afterwards.
If the second overload is used, the duration will default to the value of the toastDuration option in the feature config.

Properties for first overload:

• either:
  • for showing a string:
    • message: string - The message to show in the toast
    • subtitle?: string - An optional subtitle to show below the message
  • for showing a generic element:
    • element: HTMLElement - The element to show in the toast
    • title: string - The hover and accessibility title of the toast
• and any of:
  • duration?: number - Duration in milliseconds to show the toast for (defaults to what is set in the feature config) - use Infinity for a persistent toast and 0 to not show it at all
  • position?: ToastPos - Corner position of the toast on the screen. Can be "tl", "tr", "bl" or "br" (defaults to "tr")

Example (click to expand)

unsafeWindow.BYTM.showToast({
  message: "Nothing suspicious here",
  duration: 2524140,
  position: "bl",
});

showIconToast()

Usage:

unsafeWindow.BYTM.showIconToast(props: IconToastProps): Promise<void>

Shows a toast notification with the specified icon, message and duration anchored in the specified corner of the viewport.
If a toast is already shown, it will be immediately closed and the new one will be shown shortly afterwards.

Properties:

• either:
  • for showing a string:
    • message: string - The message to show in the toast
    • subtitle?: string - An optional subtitle to show below the message
  • for showing a generic element:
    • element: HTMLElement - The element to show in the toast
    • title: string - The hover and accessibility title of the toast
• and either:
  • for using an <img> with a URL:
    • iconSrc: string | Promise<string> - URL to the image file to use as the icon
  • or when using a BYTM SVG resource:
    • icon: string - Any SVG resource name (has to start with icon-!) to use as the icon (see assets/resources.json)
    • iconFill?: string - CSS color value to set the icon's <path> elements' fill property to
• and any of:
  • duration?: number - Duration in milliseconds to show the toast for (defaults to what is set in the feature config)
  • position?: ToastPos - Corner position of the toast on the screen. Can be "tl", "tr", "bl" or "br" (defaults to "tr")

Example (click to expand)

unsafeWindow.BYTM.showIconToast({
  message: "This is an icon toast",
  icon: "icon-help",
  iconFill: "var(--bytm-warning-col)", // find all values in src/dialogs.css
  duration: 3_000,
  position: "bl",
});

showPrompt()

Usage:

unsafeWindow.BYTM.showPrompt(props: ShowPromptProps): Promise<boolean>

Shows a prompt dialog with the specified message and type.
If another prompt is already shown, it will be closed (and resolve as closed or canceled) and the new one will be shown immediately afterwards.

If the type is alert, the user can only close the prompt.
In this case the Promise always resolves with true.

For the type confirm, the user can choose between confirming or canceling the prompt.
In this case the Promise resolves with true if the user confirmed and false if the user canceled or closed.

If the type prompt is used, the user can input a text value.
In this case the Promise resolves with the entered text if the user confirmed and null if the user canceled or closed.
If the user confirms with an empty text field, the Promise resolves with an empty string.
Additionally, the property defaultValue can be used to set the preset value for the input field.

Properties:

• message: string | ((type: string) => string | Promise<string>) - The message to show in the prompt
• type: "confirm" | "alert" | "prompt" - The type of the prompt. Can be "confirm", "alert" or "prompt"
• for type "prompt" only:
  • defaultValue?: string - The default value for the input field (only has an effect when using type "prompt")
• for overriding button text and tooltips:
  • confirmBtnText?: string | ((type: string) => string | Promise<string>) - Text for the confirm button (only when using type "confirm" or "prompt")
  • confirmBtnTooltip?: string | ((type: string) => string | Promise<string>) - Tooltip for the confirm button (only when using type "confirm" or "prompt")
  • denyBtnText?: string | ((type: string) => string | Promise<string>) - Text for the deny button (shows up for all types)
  • denyBtnTooltip?: string | ((type: string) => string | Promise<string>) - Tooltip for the deny button (shows up for all types)

Example (click to expand)

const itemName = await unsafeWindow.BYTM.showPrompt({
  type: "prompt",
  message: "Enter the name of the item to delete:",
  // default value for the input field:
  defaultValue: "My Item",
});

const confirmed = itemName && await unsafeWindow.BYTM.showPrompt({
  type: "confirm",
  message: "Are you sure you want to delete this?",
  confirmBtnText: "Yes, delete",
  denyBtnText: "No, cancel",
  // can also be sync or async functions:
  confirmBtnTooltip: () => "Click to confirm the deletion",
  // and the type parameter can be used for further customization:
  denyBtnTooltip: async (type: "confirm" | "alert" | "prompt") => await getText(`prompts.${type}.cancel_deletion`),
});

if(confirmed && itemName) {
  await deleteItem(itemName);
  unsafeWindow.BYTM.showPrompt({
    type: "alert",
    message: () => `Deleted "${itemName}" successfully.`,
    // only the deny button is shown in alerts:
    denyBtnText: "Sure thing",
    denyBtnTooltip: "Click to close the dialog, bud",
  });
}
else
  console.log("The user canceled one of the prompts.");

formatNumber()

Usage:

unsafeWindow.BYTM.formatNumber(num: number, notation?: "short" | "long"): string

Formats a number according to the configured locale and notation (unless specified).
The default notation can be found in numbersFormat in src/features/index.ts

Example (click to expand)

const { formatNumber, setLocale } = unsafeWindow.BYTM;

// (underscores in numbers are ignored in JS/TS)
const num = 123_456_789;

setLocale(myToken, "de_DE");             // German's commas and dots are swapped:
console.log(formatNumber(num, "short")); // 123,5 Mio.
console.log(formatNumber(num, "long"));  // 123.456.789

setLocale(myToken, "hi_HI");             // In Hindi it's a bit different:
console.log(formatNumber(num, "long"));  // 12,34,56,789

NanoEmitter

Usage:

new unsafeWindow.BYTM.NanoEmitter<TEventMap>(settings: NanoEmitterSettings): NanoEmitter

Abstract class that can be extended to create custom event emitting classes.
The methods are fully typed through the generic TEventMap, which is an object map of event names to a callback function signature (the type is Record<string, (...args: any) => void>)

Settings properties:

Property	Description
publicEmit?: boolean	If set to true, allows emitting events through the public method emit() If false, the only way to emit events is inside your derived class using this.events.emit()

Methods:

• on(event: string, callback: Function<any>): Function
  Registers a callback for the specified event.
  Returns a function that can be called to unsubscribe from the event at any time.
• once(event: string, callback?: Function<any>): Promise<any[]>
  Registers a callback for the specified event that gets called only once.
  The callback is called and the Promise is resolved at the same time.
• emit(event: string, ...args: any[]): boolean
  Emits the specified event with the passed arguments.
  Has to be enabled through the publicEmit option in the constructor first!
  Returns true if the event was emitted successfully, false if not.
• unsubscribeAll(): void
  Unsubscribes all listeners from all events and clears the internal listener map.

Example (click to expand)

interface MyEvents {
  /** Emitted when the foo is bar */
  foo: (bar: string) => void;
}

class MyEmitter extends unsafeWindow.BYTM.NanoEmitter<MyEvents> {
  constructor() {
    // allow calling emit() from outside this class instance
    super({ publicEmit: true });
  }

  public doSomething() {
    this.emit("foo", "baz");
  }
}

function run() {
  const emitter = new MyEmitter();

  emitter.on("foo", (bar) => {
    //                ^ automatically typed as string
    console.log(`The bar is ${bar}`);
  });

  // will log "The bar is baz" to the console, see above
  emitter.doSomething();
}

run();