nft-pixels

nft-pixels

nft-pixels allows one, to programatically, generate NFT images and their corresponding metadata from a configuration file.

There are various NFT generator projects out there, this project tries to solve some of the configuration issues that exist after the nfts have been generated.

• Generates unique images, via configured traits
• Configure rarity, via weights
• Control size and type of image output
• Aims to be opensea compatible
• Print report of traits used
• Seperately creates images and metadata
• Metadata can be correctly configured with IPFS hashes, that are returned from a pinning service e.g. (pinata, nft storage)

Install

npm install nft-pixels --save-dev

Setup

nft-pixels has been configured to be used programatically, for example, in a hardhat or truffle task.

In order to run nft-pixels you must configure a folder structure, as below, see the repository example folder for a concrete example.

{folder}
- traits
  -- trait1
    --- trait1.png
    --- trait2.png
    --- trait3.png
  -- trait2
    --- trait1.png
    --- trait2.png
    --- trait3.png
  -- trait3
    --- trait1.png
    --- trait2.png
    --- trait3.png

nftpixels.generateConfig(config)

You will also need a nftconfig file; this command generates a skeleton nftConfig file, it reads the file system traits, above and generates a config file.

await nftpixels.generateConfig({
    metadata: {
      name: "NFT",
      description: "NFT collection",
      image: "ipfs://*********/",
      externalUrl: "url of dapp providing an nft minting function",
      background_color: "",
      animation_url: "",
      youtube_url: "",
      includeCreatedDate: "Birthday",
    },
    traits: '/traits/',
    outputImagePath: '/images',
    outputMetadataPath: '/metadata',
    configOutputPath: '/nftConfig.json'
  })

The file generated looks like so, this can be reused to re-generate your images.

The order of the traits structure is important, the images will be laid out based on the trait order

const nftConfig = {
  "metadata": {
    "name": "NFT",
    "description": "NFT collection",
    "image": "ipfs://***/",
    "externalUrl": "",
    "background_color": "",
    "animation_url": "",
    "youtube_url": "",
    "includeCreatedDate": "Birthday"
  },
  "files": {
    "imageSize": { "width": 350, "height": 350 },
    "outputFileType": "png",
    "deleteFolders": false,
    "printReport": true,
    "attributesPath": "example/traits/",
    "outputImagePath": "example/images",
    "outputMetadataPath": "example/metadata"
  },
  "traits": {
    "backgrounds": [
      { "file": "brown.png", "numberOfItems": 1000 },
      { "file": "lemon.png", "numberOfItems": 1000 }
    ],
    "eyes": [
      { "file": ".DS_Store", "numberOfItems": 1000 },
      { "file": "black-big-square.png", "numberOfItems": 1000 },
      { "file": "black-small-circle.png", "numberOfItems": 1000 }
    ],
    "hair": [
      { "file": "black-short.png", "numberOfItems": 1000 },
      { "file": "lemon-straight.png", "numberOfItems": 1000 }
    ],
    "mouth": [
      { "file": ".DS_Store", "numberOfItems": 1000 },
      { "file": "black-smile.png", "numberOfItems": 1000 },
      { "file": "black-straight.png", "numberOfItems": 1000 }
    ]
  }
}

Weights and rarity

The weight algorithim, is simple, it randomly selects traits; traits with numberOfItems set to larger values are more likely to be selected; numberOfItems is the maximum number of images we could generate, not the minimum number.

e.g. if we set numberOfItems to 1 we would generate a single image with this trait. a numberOfItems set to 2000, would mean that you could get 2000 images with this trait, if the number of unique images generated is > 2000.

So if you wanted a trait with a rarity of 10 images; set numberOfItems to 10. We generate a report after creating images, so you can play around with the numbers and see whats been generated via the report.

API

nftpixels.generateImages(nftConfig)`

This command takes the above nftconfig as input, and creates images {0}.png files for all images, written to a folder config.files.outputImagePath. combinations contains all the images generated (with traits), and can be used to generateMetadata

const combinations = await nftpixels.generateImages(nftConfig)

Report

If you have printReport set to true this function will print to console, a report outlining the number of images created for each trait e.g.

***********************************
Nft Generator created 1000 nfts

***********************************
Trait report

***********************************

backgrounds
- 406 images include brown
- 18 images include lemon

hair
- 99 images include black-long
- 300 images include black-shaved

mouth
- 302 images include black-grin
- 310 images include black-smile

eyes
- 62 images include black-big-square
- 33 images include black-wink

***********************************

nftpixels.generateMetadata(config, combinations)

This command takes the above nftconfig as input, and the combinations returned from generateImages and creates metadata {0}.json files for all images, which are written to a folder location config.files.outputMetadataPath.

await nftpixels.generateMetadata(nftConfig, combinations)

nftpixels.createGif(inputFolder, outputFolder)

This command allows you to create an animated gif of the images you have generated, simply takes the first 20 images and makes a gif.

await nftpixels.createGif(inputFolder, outputFolder)

Hardhat Example tasks

This is an example of this modules usecase. Here we have a hardhat task which:

• generates nft images, returns all combinations created,
• pins images to IPFS via Pinata
• generates metadata and injects the retuned IpfsHash, passes combinations created
• pins metadata to IPFS via Pinata
• update config file used by the smart contract with a image IpfsHash

So, you are now ready to deploy your smart contract, no further configuration is required, and everything is setup correctly.

task("pin", "generate and pin images and metadata to pinata").setAction(async (_, hre) => {
  const pinToPinata = true
  let imageResponse;
  let metaResponse;

  // create a stream pointing at your image collection
  const pinata = pinataSdk(PINATA_KEY, PINATA_SECRET)

  // generate images
  console.log('generating images')
  const combinations = await nftpixels.generateImages(nftConfig)

  if (pinToPinata) {
    // pin images via pinata
    console.log('pin images via pinata')
    imageResponse = await pinata.pinFromFS(nftConfig.files.outputImagePath, {
      pinataMetadata: {
        name: `${config.name}-images`
      }
    })
  }

  // generate metadata, pass image via IpfsHash
  console.log('generating metadata')
  await nftpixels.generateMetadata({
    ...nftConfig,
    metadata: {
      ...nftConfig.metadata,
      image: `ipfs://${imageResponse?.IpfsHash}`
    }
  }, combinations)

  if (pinToPinata) {
    // pin metadata via pinata
    console.log('pin metadata via pinata')
    metaResponse = await pinata.pinFromFS(nftConfig.files.outputMetadataPath, {
      pinataMetadata: {
        name: `${config.name}-metadata`
      }
    })
  }

  // update config file with image IpfsHash
  console.log('update config file with IpfsHash')
  await fs.writeFile(path.join(__dirname, '../config/config.json'), JSON.stringify({
    ...config,
    baseTokenURI: `ipfs://${metaResponse?.IpfsHash}/`
  }))
})