Skip to content

Javascript helpers to manipulate .clap files, the data format dedicated to AI cinema and AI gaming

License

Notifications You must be signed in to change notification settings

jbilcke-hf/aitube-clap

Repository files navigation

@aitube/clap

Types and helpers to manipulate .clap files

Introduction

This library is the reference implementation of OpenClap for NodeJS and browser environments.

OpenClap is a prompt container file format that was initially created for my AiTube.at project. I am also using it in my other AI demos, and I think you should use it, too!

Installation

To install the package, run the following command (yaml is a peer dependency):

npm install @aitube/clap yaml

Getting Started

import {
  // types
  ClapSegmentCategory,
  ClapOutputType,
  ClapSegmentStatus,
  ClapAuthor,
  ClapAssetSource,
  ClapEntityGender,
  ClapEntityAppearance,
  ClapEntityRegion,
  ClapEntityTimbre,
  ClapEntityAudioEngine,
  ClapSegmentFilteringMode,
  ClapVoice,
  ClapHeader,
  ClapMeta,
  ClapSceneEvent,
  ClapScene,
  ClapSegment,
  ClapEntity,
  ClapProject, 
  ClapMediaOrientation,

// defaults
  defaultMediaOrientation,

  // factories
  newClap,
  newEntity,
  newSegment,

  // main input/output handlers
  parseClap,
  serializeClap,
  fetchClap,
  updateClap,

  // utilities
  filterSegments,
  filterSegmentsWithinRange,
  generateSeed,
  getClapAssetSourceType,
  getValidNumber,
  isValidNumber,
  parseMediaOrientation,
  parseOutputType,
  parseSegmentCategory,
  parseSegmentStatus,
  UUID,

  // converters
  blobToDataUri,
  dataUriToBlob,
  clapToDataUri,

  // helpers
  buildEntityIndex,
  filterAssets,
  filterSegmentsByCategory,
  generateClapFromSimpleStory,
  getEmptyClap,
  removeGeneratedAssetUrls,
} from "@aitube/clap"


// fetch a Clap file
const res = await fetch("https://..../file.clap")
const file = await res.blob()

// open the Clap file
const clap: ClapProject = await parseClap(file)

// perform arbitrary changes in the project

clap.segments.at(64).assetUrl = await generateVideoWithAI(....)

const segment: ClapSegment = clap.segments.at(42)
segment.prompt = "a camel in the desert, medium-shot, award-winning, 4k, Canon EOS"

const storyboards = clap.segment.filter(s => s.category === ClapSegmentCategory.STORYBOARD)

// save the Clap file
const newFile: ClapProject = await serializeClap(clap)

Build Instructions

Install Bun

Run the following commands:

bun install

bun run build

To publish:

bun run build

bun run build:declaration

bun run publish

Contributing

We welcome contributions! Please feel free to submit a pull request.

License

This package is under the MIT License. See LICENSE file for more details.