hyperindex

A keyword based search index built on Hypertrie

npm install hyperindex

Usage

const Hyperindex = require('hyperindex')
const ram = require('random-access-memory')
const idx = new Hyperindex(ram, { valueEncoding: 'json' })

idx.add({
  some: 'data',
  you: 'want to index'
}, {
  keywords: ['good', 'data']
}, function () {
  console.log('object indexed to the trie ...')
})

To do lookups in the index use the lookup API:

const stream = idx.lookup('good')

stream.on('data', function (document) {
  console.log('document with keyword "good":', document)
})

API

index = new Hyperindex(storage, [key], [options])

Create a new index. All arguments are forwarded to the Hypertrie instance. To create an index from an existing Hypercore feed you can use the feed option

new Hyperindex(null, { feed: existingFeed })

If you're storing JSON documents, set valueEncoding: 'json' in the options as well.

index.add(document, options, [callback])

Add a document to the index. Options should include

{
  keywords: ['...'], // array of keywords to index this item at
  key: '...' // optionally set a unique key for this document
             // if not set one will be generated internally.
}

Callback is called with callback(err, key) where the key is the identifier you need if you want to remove this document.

index.remove(key, options, [callback])

Removes a document from the index. options should contain the same keywords as above.

stream = index.lookup(keyword)

Do a streaming lookup to get all the results for a given keyword. The stream will contain the documents for that keyword and is hash-ordered.

stream = index.and(...wordsOrStreams)

Do a streaming intersect of multiple keywords.

stream = index.or(...wordsOrStreams)

Do a streaming union of multiple keywords.

replicationStream = index.replicate(...args)

Make a replication stream. Options forwarded to Hypertrie.

License

MIT