Node bindings for Imageflow

Quickly scale or modify images and optimize them for the web.

If the AGPLv3 does not work for you, you can get a commercial license on a sliding scale. If you have more than 1 server doing image processing your savings should cover the cost.

API docs are here.

Installation

npm install @imazen/imageflow

Usage

const {
    MozJPEG,
    Steps,
    FromURL,
    FromFile,
    FromStream,
    FromBuffer,
} = require('@imazen/imageflow')
const fs = require('fs')

let step = new Steps(new FromURL('https://jpeg.org/images/jpeg2000-home.jpg'))
    .constrainWithin(500, 500)
    .branch((step) =>
        step
            .constrainWithin(400, 400)
            .branch((step) =>
                step
                    .constrainWithin(200, 200)
                    .rotate90()
                    .colorFilterGrayscaleFlat()
                    .encode(new FromFile('./branch_2.jpg'), new MozJPEG(80))
            )
            .copyRectangle(
                (canvas) =>
                    canvas.decode(
                        new FromStream(fs.createReadStream('./test.jpg'))
                    ),
                { x: 0, y: 0, w: 100, h: 100 },
                10,
                10
            )
            .encode(new FromFile('./branch.jpg'), new MozJPEG(80))
    )
    .constrainWithin(100, 100)
    .rotate180()
step.encode(new FromBuffer(null, 'key'), new MozJPEG(80))
    .execute()
    .then(console.log)
    .catch(console.log)

Examples

Reading a file from disk. FromFile provide an easy method for reading and writing images to disk.

const { MozJPEG, Steps, FromFile } = require('@imazen/imageflow')

const output = await new Step(new FromFile('path/to/file'))
    .rotate180()
    .encode(new FromFile('./path/to/output/file'))
    .execute()

Reading from a stream. FromStream can read and write to a stream.

const { MozJPEG, Steps, FromStream } = require('@imazen/imageflow')

const output = await new Step(new FromStream(req))
    .constrainWithin(400, 400)
    .encode(new FromStream(res))
    .execute()
res.end()

Reading from a url. FromURL can make a GET request to download and POST request to upload the image.

const { MozJPEG, Steps, FromURL } = require('@imazen/imageflow')

const output = await new Step(new FromURL('url to image'))
    .colorFilterGrayscaleFlat()
    .encode(new FromURL('url to image upload'))
    .execute()

Providing buffer. FromBuffer can read and provide the output buffer. To read the output a key should be provided, which used later to access buffer in the output.

const { MozJPEG, Steps, FromBuffer } = require('@imazen/imageflow')

const output = await new Step(new FromBuffer(getSomeBuffer()))
    .colorFilterGrayscaleFlat()
    .encode(new FromBuffer(null, 'key'))
    .execute()
console.log(output.key)

Performing Batch operations. branch, decode, and encode are used together to perform batch operation. This example shows how to create varying size images from a single image.

const { MozJPEG, Steps, FromStream, FromFile } = require('@imazen/imageflow')

const test = new Steps(new FromStream(req))
    .constrainWithin(800, 800)
    .branch((step) => step.encode(new FromFile('large.jpg'), new MozJPEG()))
    .branch((step) =>
        step
            .constrainWithin(400, 400)
            .branch((step) =>
                step
                    .constrainWithin(200, 200)
                    .branch((step) =>
                        step
                            .constrainWithin(100, 100)
                            .encode(new FromFile('tiny.jpg'), new MozJPEG())
                    )
                    .encode(new FromFile('small.jpg'), new MozJPEG())
            )
            .encode(new FromFile('medium.jpg'), new MozJPEG())
    )
    .execute()

Example with complex graph of operations

const {
    MozJPEG,
    Steps,
    FromURL,
    FromFile,
    FromStream,
    FromBuffer,
} = require('@imazen/imageflow')
const fs = require('fs')

let step = new Steps(new FromURL('https://jpeg.org/images/jpeg2000-home.jpg'))
    .constraintWithin(500, 500)
    .branch((step) =>
        step
            .constraintWithin(400, 400)
            .branch((step) =>
                step
                    .constraintWithin(200, 200)
                    .rotate90()
                    .colorFilterGrayscaleFlat()
                    .encode(new FromFile('./branch_2.jpg'), new MozJPEG(80))
            )
            .copyRectangle(
                (canvas) =>
                    canvas.decode(
                        new FromStream(fs.createReadStream('./test.jpg'))
                    ),
                { x: 0, y: 0, w: 100, h: 100 },
                10,
                10
            )
            .encode(new FromFile('./branch.jpg'), new MozJPEG(80))
    )
    .constraintWithin(100, 100)
    .rotate180()
step.encode(new FromBuffer(null, 'key'), new MozJPEG(80))
    .execute()
    .then(console.log)
    .catch(console.log)

Using query style commands

await new Steps().executeCommand(
    'width=100&height=100&mode=max',
    new FromBuffer(str),
    new FromBuffer(null, 'key')
)

Using Decode Options

const output = await new Step(
    new FromBuffer(getSomeBuffer()),
    new DecodeOptions()
        .ignoreColorProfileError()
        .setJPEGDownscaleHint(100, 100)
        .setWebpDecoderHints(100, 100)
        .discardColorProfile()
)
    .colorFilterGrayscaleFlat()
    .encode(new FromBuffer(null, 'key'))
    .execute()
console.log(output.key)

Local Setup

git clone https://github.com/imazen/imageflow-node
cd imageflow-node
yarn
yarn test