feathers-blob

feathers-blob

Feathers abstract blob store service

Installation

npm install feathers-blob --save

Also install a abstract-blob-store compatible module.

API

const BlobService = require('feathers-blob')

blobService = BlobService(options)

• options.Model is an instantiated interface that implements the abstract-blob-store API
• options.id is a string 'key' for the blob identifier.
• returnUri defaults to true and indicates if the service output a data URI on create/get operations
• returnBuffer set it to true if you only want a buffer output instead of a data URI on create/get operations

Tip: returnUri/returnBuffer are mutually exclusive. If you need both extract the buffer from the data URI on the client-side to avoid transferring the data twice over the wire.

blobService.create(body, params)

where input body is an object with either:

• a key uri pointing to data URI of the blob,
• a key buffer pointing to raw data buffer of the blob along with its contentType (i.e. MIME type).

Optionally, you can specify in the body the blob id which can be the file path where you want to store the file, otherwise it would default to ${hash(content)}.${extension(contentType)}.

Tip: You can use feathers hooks to customize the id. You might not want the client-side to write whereever they want.

returns output 'data' of the form:

{
  [this.id]: `${hash(content)}.${extension(contentType)}`,
  uri: body.uri, // When returnUri options is set
  buffer: body.buffer, // When returnBuffer options is set
  size: length(content)
}

blobService.get(id, params)

returns output data of the same form as create.

blobService.remove(id, params)

Example

const { getBase64DataURI } = require('dauria');
const AWS = require('aws-sdk');
const S3BlobStore = require('s3-blob-store');
const feathers = require('@feathersjs/feathers');
const BlobService = require('feathers-blob');

const s3 = new AWS.S3({
  endpoint: 'https://{service}.{region}.{provider}.com',
  accessKeyId: process.env.AWS_ACCESS_KEY_ID,
  secretAccessKey: process.env.AWS_SECRET_ACCESS_KEY,
});

const blobStore = S3BlobStore({
  client: s3,
  bucket: 'feathers-blob'
});

const blob = {
  uri: getBase64DataURI(new Buffer('hello world'), 'text/plain')
}

const app = feathers();

app.use('/upload', BlobService({
  Model: blobStore
}));

const blobService = app.service('upload');

blobService.create(blob).then(function (result) {
  console.log('Stored blob with id', result.id);
}).catch(err => {
  console.error(err);
});

Should you need to change your bucket's options, such as permissions, pass a params.s3 object using a before hook.

app.service('upload').before({
  create(hook) {
    hook.params.s3 = { ACL: 'public-read' }; // makes uploaded files public
  }
});

For a more complete example, see examples/app which can be run with npm run example.

License

Copyright (c) 2018

Licensed under the MIT license.

Changelog

v2.3.0 (2020-10-04)

Full Changelog

Implemented enhancements:

• Add TypeScript definitions #78

Closed issues:

• Metadata does not been used on cb of createWriteStream #80
• docs: how to use feathers-blob with typescript #77
• passing option returnUri: false and returns uri... #75
• How to render uploaded photos? #74

Merged pull requests:

• feat: add typescript typings with dtslint tests #79 (Mairu)
• Small optimization to create\(\). #76 (davidbludlow)
• Update mocha to the latest version 🚀 #73 (greenkeeper[bot])