nichoth/multiblob

A content-addressable-store that supports multiple hashing algorithms, and pull-streams

multiblob

A content-addressable-store that supports multiple hashing algorithms, and pull-streams.

example

var Blobs = require('multiblob')

var blobs = Blobs({ dir: '/media/blobCollection' })

pull(
  source, // a buffer stream e.g. from pull-file
  blobs.add(function (err, hash) {
    console.log('added source to blobs:', hash)
  })
)

API

Blobs(config) => blobs

where config is an Object with properties:

• dir String - the directory to store blobs in
• alg String (optional) - the algorithm for hashing. Valid options: 'blake2s', 'sha256' (default: 'blake2s')
• encode Function (optional) - converts a buffer to a string (default: see util.js#encode)
• decode Function (optional) - recovers a string into an object { hash: Buffer, alg } (default: see util.js#decode)
• isHash Function (optional) - tests a string to check if it's a valid hash (default: see util.js#isHash)

blobs.add (hash?, cb?) => Sink

Create a sink stream for writing a blob. Expects to receive a buffer stream.

If hash was given, then it will error if the file turned out to be different. If a cb is not given and there was an error, this function will throw.

blobs.get (hash || opts) => Source

Takes the hash of blob already in the store and return a source buffer stream. If the file does not exist this stream will error.

If the argument is a hash string, then return the stream. If the argument is an opts object, with the key: hash property, retrive that blob, but error if the size does not exactly match the size property, or is over max property (in bytes)

blobs.getSlice(opts) => Source

create a source stream that reads a slice of a given blob, from the start property to the end property, in bytes. Error if the file does not exist or if the size of the whole blob does not exactly match the size property, or is over max property (in bytes).

blobs.has(hash, cb)

check if the given hash is in the store. If hash is an array of hashes, size will callback with an array of booleans.

blobs.size(hash, cb)

get the size of this blob. If hash is an array of hashes, size will callback with an array of sizes. If the hash does not exist in the store, size will callback null.

blobs.ls() => Source

source stream that reads the list of hashes available in the store.

blobs.rm(hash, cb)

remove a hash from the store.

blobs.isEmptyHash(hash)

Check if a given hash is actually the empty hash. If something has the empty hash, that is probably a bug. The above methods will act like the empty file is already in the store.

blobs.meta

???

blobs.resolve

???

todo

maybe emit events when blobs are stored?

License

MIT