[

This library allows for managing pipelines that processes digital assets into renditions.

Components:

• Pipeline: overall system
• Orchestration: controls the flow
• Transformer: individual software components that can do a certain transformation or operation on an asset
• Plan: actual execution plan that might be passed around different components/services

Transformers are individual software components that can do a certain transformation or operation on an asset.

To create a new transformer, extend the transformer api:

class TransformerExample extends Transformer {

    constructor() {
        super("transformerExample", new Manifest({
            inputs: {
                type: ["image/png"],
                width: { "min": 1, "max": 319},
                height: { "min": 1, "max": 319 },
                colorProfile: "rbg"
            },
            outputs: {
                type:  ["image/png", "image/png"]
            }
        }));
    }
    compute(input, output, options) {
        // ... convert input.path to output.path based on output.instructions
    }
}

The compute method is where you put your custom worker code. The basic expectation of this function is to look at parameters from output.instructions and convert it into a rendition, then write this rendition to output.path.

The parameters for the compute function are: input, output, and options.

Note: This function is backwards compatible with renditionCallback from the old SDK: https://github.com/adobe/asset-compute-sdk#rendition-callback-for-worker-required The compute function does not need to return anything.

Object containing the following attributes:

Object containing following attributes:

Optional parameters to pass into workers

• disableSourceDownload: Boolean used to disable the source download (defaults to false). (keeping for backwards compatibility with the old SDK)
• disableRenditionUpload: Boolean used to disable the rendition upload (defaults to false). (keeping for backwards compatibility with the old SDK)
• any other options needed for the transformer (i.e. authorization. Pass params.auth into this options object)

At the bare minimum, the compute function must write something to the output.path.

Simplest example (copying the source file):

async function compute(input, output) => {
    // Check for unsupported file
    const stats = await fs.stat(input.path);
    if (stats.size === 0) {
        throw new SourceUnsupportedError('source file is unsupported');
    }
    // process infile and write to outfile
    await fs.copyFile(input.path, output.path);
}

The manifest is a JSON object of attributes determining what the transformer supports as inputs and outputs. Every transformer must have a valid manifest containing at least the following:

• inputs: object containing at least the following attributes
  • type: string or array of strings containing the mimetype/s the transformer supports as input
  • sourceType: string containing value URL or LOCAL depending on what the transformer supports as input
• outputs
  • type: string or array of strings containing the mimetype/s the transformer supports producing an output

The manifest can have any number of other attributes in inputs and outputs that must follow these rules:

• attribute value can be defined as range (object), preference list (array), or singular item
• if attribute value is a range (object), it must be a valid number
• if attribute value is a preference list (array), it can be a list of numbers, strings or booleans
• if attribute value is a singular item, it can be a number, string or boolean
• empty array means you don't support it
• if an attribute is not listed, your transformer should accept it as an optional parameter and also have the option to ignore as well

inputs: {
    type: ['image/tiff'],
    width: { min: 2000, max: 200000},
    height: { min: 2000, max: 200000 },
    alphaChannel: [],
    sourceType: "LOCAL"
},
outputs: {
    type: ['image/png', 'image/jpeg'],
    width: { min: 0, max: 2000},
    height: { min: 0, max: 2000 }
}

Contributions are welcomed! Read the Contributing Guide for more information.

This project is licensed under the Apache V2 License. See LICENSE for more information.