npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2024 – Pkg Stats / Ryan Hefner

@fullstackcraftllc/codevideo-backend-engine

v1.4.0

Published

Create shockingly realistic automated software videos! The backend / CLI tool from CodeVideo to create videos.

Downloads

20

Readme

CodeVideo Backend Engine

NPM Version Code Quality Code Coverage Tests Moral Tool Abilities

Create shockingly realistic automated software videos!

Any sufficiently advanced technology is indistinguishable from magic. 😉

Quick Start - Programmatic Usage

Install this package:

npm install @fullstackcraft/codevideo-backend-engine

Use it in your project!

import fs from 'fs'
import { generateVideoFromActions } from '@fullstackcraft/codevideo-backend-engine'
import { IGenerateVideoFromActionsOptions } from "./interfaces/IGenerateVideoFromActionsOptions.js";

const videoOptions: IGenerateVideoFromActionsOptions = {
  actions: [
    {
      name: "speak-before",
      value: "I'm gonna type some code!"
    },
    {
      name: "type-editor",
      value: "console.log('Hello, world!');"
    },
  ],
  language: "en-US",
  textToSpeechOption: "coqui-ia" // free & decent quality voice, but requires coqui-ai to be installed: 'pip install TTS'
}


// 'video' is a Buffer representing an mp4 video
const video = generateVideoFromActions(videoOptions)

// for example, save the video to a file
fs.writeFileSync('hello-world.mp4', video)

Quick Start - CLI Usage

Clone this repository:

git clone https://github.com/codevideo/codevideo-ai.git

Move into the repo:

cd codevideo-ai

Install dependencies:

npm install

Run the simple hello world example (creates a video on how to use the console.log function in JavaScript):

npm run start ./examples/hello-world.json

Let it run & have faith! A headless chrome browser is running in the background, creating your video. As long as you don't see errors on your console it will eventually finish and the video will be pop out in the ./video directory.

There will be a variety of files created:

  • ./audio/*.mp3 - the various speaking action audio files (generated before the video is made)
  • ./audio/hello-world/combined.mp3 - the properly spaced speaking action audios combined and used to interleave into the editor video
  • ./video/hello-world.mp4 - the final video with typing animation and audio

That's it! Create any automation you want by creating a new actions (either .json or .ts) and passing it to the start command. See the defining actions section below for more information.

Other Examples

Looks like GitHub isn't very friendly for video embeds, you can see a few examples and a Web MVP on our website at codevideo.ai.

Implementation of a Fibonacci function in TypeScript in a Monaco editor at localhost, driven by puppeteer:

Using the actual Visual Studio Code application, demonstrating how to use JavaScript's console.log function in Node.js, including opening the file and starting the script from the terminal. Driven by nut-js

Other Text To Speech Engines

Free

Windows and Mac

You can use sayjs, but this will be a very robotic sounding voice.

Linux

We have a festival voice available for Linux users. Keep in mind with this option you must have festival and lame installed on your system, i.e.

sudo apt-get install festival lame

### Paid

Currently, we support the following AI voices:

- Eleven Labs (including custom cloned voices*)
- OpenAI TTS (text-to-speech) 

*For inspiration on what to record to train a voice for Eleven Labs, we recommend recording yourself reading blog posts or other text that you find interesting - we had great results with just 45 minutes of recorded audio, when the 'optimum' amount recommended by Eleven Labs is supposed to be 3 hours or more.

To use Eleven Labs or OpenAI as the voice, you will need to set the `ELEVEN_LABS_API_KEY` + `ELEVEN_LABS_VOICE_ID` or `OPENAI_API_KEY` environment variables respectively.

Then pass as a second argument specifying if you want to use `elevenlabs` or `openai` respectively:

```bash
npm run start ./examples/hello-world.json elevenlabs
npm run start ./examples/hello-world.json openai

sayjs is also acceptable as an engine:

npm run start ./examples/hello-world.json sayjs

this is anyway the default and it doesn't need to be explicitely included.

Defining Actions

The API works via an array of actions (type IAction), each represented by an object with a name and value property. These actions can be defined in a actions file. The actions file can be either a json file or a TypeScript file that contains an array of actions.

Here is an example of an action file, the very same one used to drive the hello-world example above:

[
  {
    "name": "speak-before",
    "value": "To represent that this file is 'index.js', I'll just put a comment here"
  },
  {
    "name": "type-editor",
    "value": "// index.js"
  },
  {
    "name": "speak-before",
    "value": "Then, we'll add a simple console.log statement to the file."
  },
  {
    "name": "enter",
    "value": "1"
  },
  {
    "name": "type-editor",
    "value": "console.log('Hello, world!');"
  },
  {
    "name": "speak-before",
    "value": "For example, if I wanted to write the value of some variable, I could do that with console.log."
  },
  {
    "name": "backspace",
    "value": "29"
  },
  {
    "name": "type-editor",
    "value": "const myVariable = 'Important variable I want to keep track of';"
  },
  {
    "name": "enter",
    "value": "1"
  },
  {
    "name": "type-editor",
    "value": "console.log(myVariable);"
  },
  {
    "name": "speak-before",
    "value": "And that's it! We now know how to use the console.log function in JavaScript."
  }
]

Alternatively, you can define the actions directly in TypeScript. This is useful if you want to use the editor's Copilot / ChatGPT / intellisense to help you write the actions. The only important thing to note is that you must use export default syntax to export the actions. The equivalent TypeScript version of the above actions file would look like this:

const helloWorldActions: Array<IAction> = [
  {
    name: "speak-before",
    value: "To represent that this file is 'index.js', I'll just put a comment here"
  },
  {
    name: "type-editor",
    value: "// index.js"
  },
  {
    name: "speak-before",
    value: "Then, we'll add a simple console.log statement to the file."
  },
  {
    name: "enter",
    value: "1"
  },
  {
    name: "type-editor",
    value: "console.log('Hello, world!');"
  },
  {
    name: "speak-before",
    value: "For example, if I wanted to write the value of some variable, I could do that with console.log."
  },
  {
    name: "backspace",
    value: "29"
  },
  {
    name: "type-editor",
    value: "const myVariable = 'Important variable I want to keep track of';"
  },
  {
    name: "enter",
    value: "1"
  },
  {
    name: "type-editor",
    value: "console.log(myVariable);"
  },
  {
    name: "speak-before",
    value: "And that's it! We now know how to use the console.log function in JavaScript."
  }
]

// IMPORTANT!!! You must use the export default syntax here!!!
export default helloWorldActions

To see a list of all available actions, see the action names as string directly in the codevideo-types repo.

Other Commands

clean

npm run clean

This command cleans out the ./dist, ./audio, and ./video directories.

***Be careful! After running this, you'll have to regenerate everything again for each video made. This could get expensive if you are doing a lot of text to speech using a paid plan.

make-video

npm run make-video ./examples/hello-world.json

Just an alias for start.

scripts-health-check

It sometimes happens that the text to speech model produces artifacts in the audio. This command will run a health check on all audio files produced by an actions file by transcripting them and comparing their transcript to the original transcript using Levenshtein distance. If the transcripts are different, it will print out the file name and the transcript, and attempt to create a new audio file. This is most useful when using an AI generated voice, as the say.js library is a deterministic library and will always produce the same audio file for the same transcript.

Ex. check the accuracy of your Eleven Labs audio files:

npm run scripts-health-check ./examples/hello-world.json elevenlabs

Ex. check the accuracy of your OpenAI audio files:

npm run scripts-health-check ./examples/hello-world.json openai

scripts-character-count

As is always the case with character or word based cost models, the amount of text that will be converted is always a question. This script prints out the character, word, and page count of all speaking-based actions in a given actions file.

npm run scripts-character-count ./examples/hello-world.json

code-health-check

This command runs a health check of code after each type event, checking for any strange formatting or syntax errors. If it finds any, it will print out the file name and the error.

npm run code-health-check ./examples/hello-world.json

Experimental: Visual Studio Code on Desktop Automation

Experimental: use a running instance of Visual Studio Code and robotjs to record in the desktop environment:

npm run visual-studio-code-driver ./examples/hello-world.json sayjs desktop

A few caveats are required to properly run desktop automation actions:

  • Need exactly one 1920x1080 monitor
  • The desktop to the right of where you issue the above script must be an empty instance of Visual Studio Code

Patches!

Due to an ugly bug with fluent-ffmpeg, which is used by puppeteer-screen-recorder, the fluent-ffmpeg library has been patched with patch-package. For those interested, the patch is in ./patches/fluent-ffmpeg+2.1.2.patch.