level-tier

Minimalistic LevelUP utility for namespacing keys and facilitating blazing fast range queries.

Caveat

You might want to zero pad numerical values (such as arbitrary timestamps) to have a uniform length if using them as parts of a namespaced key. Failing to do so might result in unexpected ordering, as emphasised by the example below.

'10' < '2' // -> true

'10' < '02' // -> false

level-tier will warn if using numbers as namespace keys.

Uniform length timestamp padding

level-tier includes a convenience method leveltier.now() that produces uniform length timestamps.

  Date.now() // -> 1575146509497
  leveltier.now() // -> '00000001575146509497'

This function also accepts a timestamp to pad when creating a range Key.

  leveltier.now({ timestamp: 1575146509497 }) // -> '00000001575146509497'

Putting this together, writing and retrieval would look like this

    // writing data
      var key = leveltier(['data', leveltier.now()])
      mydb.put(key, data)
    
    // retrieving data after some lower bound
      var startkey = leveltier.gte(['data', leveltier.now({ timestamp: 1575156550378 })])
      mydb.createReadStream({ gt:startkey })

Important note

This module is a minimalistic and naive implementation namespacing lexicographical keys. It simply uses \x00 and \uffff as lower and upper bounds. Allowing user input to determine the keys and not stripping out the delimiter characters could result in the NoSQL equivalent of an SQL injection attack. For this reason any characters corresponding to the upper or lower bounds will be stripped from the namespaced keys.

Other solutions

For a robust solutions when it comes to "binary serialization of arbitrarily complex structures that sort element-wise", see deanlandolt/bytewise.

Examples

Import the module

const leveltier = require('level-tier')

Creating namespaced key

leveltier(['Alice', 'Cooper']); // -> alice\x00cooper

Creating range start key

The following constructs a start key that can be used as the gte option when creating a read stream in LevelUP.

leveltier.gte(['Alice']); // -> Alice\x00

Creating range end key

The following constructs an end key that can be used as the lte option when creating a read stream in LevelUP.

leveltier.lte(['Alice']); // -> Alice\x00\uffff

Combining start and end keys

Combining {gte: leveltier.gte(['Alice']), lte: leveltier.lte(['Alice'])} when creating a read stream in LevelUP would yield all documents that have "Alice" as the first part in the name spaced key.

Parsing a key

leveltier.parse('Alice\x00Cooper'); // -> ['Alice', 'Cooper']

Test

Run unit tests;

$ npm test

License

MIT