level-indexer

Generic indexer for leveldb. Only stores document keys for space efficiency.

npm install level-indexer

Usage

var indexer = require('level-indexer')
 
// create a index (by country)
var country = indexer(db, ['country']) // index by country
 
country.add({
  key: 'mafintosh',
  name: 'mathias',
  country: 'denmark'
})
 
country.add({
  key: 'maxogden',
  name: 'max',
  country: 'united states'
})
 
var stream = country.find({
  gte:{country:'denmark'},
  lte:{country:'denmark'}
})
 
// or using the shorthand syntax
 
var stream = country.find('denmark')
 
stream.on('data', function(key) {
  console.log(key) // will print mafintosh
})

The stored index is prefix with the index key names which means you can use the same levelup instance to store multiple indexes.

API

index = indexer(db, [prop1, prop2, ...], [options])

Creates a new index using the given properties. Options include

{
  map: function(key, cb) {
    // map find results to another value
    db.get(key, cb)
  }
}

index.add(doc, [key], [cb])

Add a document to the index. The document needs to have a key or provide one. Only the key will be stored in the index.

index.remove(doc, [key], [cb])

Remove a document from the index.

index.key(doc, [key])

Returns the used leveldb key. Useful if you want to batch multiple index updates together yourself

var batch = [{type:'put', key:index.key(doc), value:doc.key}, ...]

stream = index.find(options, [cb])

Search the index. Use options.{gt,gte,lt,lte} to scope your search.

// find everyone in the age range 20-50 in denmark
 
var index = indexer(db, ['country', 'age'])
 
...
var stream = index.find({
  gt: {
    country: 'denmark',
    age: 20
  },
  lt: {
    country: 'denmark',
    age: 50
  }
})

Optionally you can specify the ranges using arrays

var stream = index.find({
  gt: ['denmark', 20],
  lt: ['denmark', 50]
})

Or if you do not care about ranges

var stream = index.find(['denmark', 20])
 
// equivalent to
 
var stream = index.find({
  gte: ['denmark', 20],
  lte: ['denmark', 20]
})

The stream will contain the keys of the documents that where found in the index. Use options.map to map the to the document values.

Options also include the regular levelup db.createReadStream options.

If you set cb the stream will be buffered and passed as an array.

index.findOne(options, cb)

Only find the first match in the index and pass that to the callbck

License

MIT