Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP
chaos is a node.js database
JavaScript
tree: 8b48a5f520

Fetching latest commit…

Cannot retrieve the latest commit at this time

Failed to load latest commit information.
.gitignore
.npmignore
LICENCE
README.md
bench.js
chaos.js
package.json
test.js

README.md

chaos - a node.js database

Why chaos? Because we exploit the sha1 chaotic randomness to store the keys evenly in the filesystem.

The idea

The first 2 characters of the sha1 hash of a key become the top dir in the tree, the next 1 becomes the child dir in the tree and the rest of the hash becomes the filename. This means that the 1st time you create a database, it will create 4,096 directories so it can then store keys/files in them. It also means that while the sha1 has trillions of combinations, the filesystem has limitations way lower than that. It depends on the filesystem but in general, but my guess is that less than 30,000 files in a folder should perform really well. This translates to a max 122,880,000 keys. Cool, eh? Plus you can store any utf8 string or number you want, it's just a proxy to fs.readFile() and fs.writeFile().

Installation

npm install chaos

Usage

var db = require('chaos')('your database name')

Commands

db.set(key, val, function(err) {})

Sets a key value pair.

db.get(key, function(err, val) {})

Gets the value of a key.

db.del(key, function(err) {})

Deletes a key.

db.incr(key, function(err, new_number) {})

db.decr(key, function(err, new_number) {})

Increment or decrement a key value by 1 and return the new number. If a key doesn't exist or its value isn't a number it will be created starting from 0. Therefore will return 1 or -1 respectively.

db.getset(key, val, function(err, old_val) {})

Get a key value and set another afterwards.

db.getdel(key, function(err, val) {})

Get a key value and delete it afterwards.

db.getorsetget(key, default_value, function(err, val) {})

Get a key's value or if it doesn't exist, set the value and get it afterwards.

db.hset(hkey, field, val, function(err) {})

Set a hkey field value.

db.hget(hkey, field, function(err, val) {})

Get the value of a hkey field.

db.hdel(hkey, field, function(err) {})

Delete a hkey field.

Warning: This deletes only a field, to delete the hkey itself, use db.del.

db.hgetall(hkey, function(err, field_value_object) {})

Get all field value pairs from a hkey. Returns an object with fields as keys and their values.

db.hkeys(hkey, function(err, fields_array) {})

Get all field names from a hkey. Returns an unsorted array with the field names.

db.hvals(hkey, function(err, values_array) {})

Get all field values from a hkey. Returns an unsorted array with the field values.

db.watch(key, [options,] function(err, val) {})

Watch a key for changes and attempt a db.get. If options is provided, it is passed on to fs.watchFile and it's an object. The default is {persistent: true, interval: 0}.

Notes: This maps to fs.watchFile and thus is not very reliable as to when it's going to fire the callback. Don't trust it will fire on every key change. Also on hkeys, it fires only when a field is added or removed from the hkey, not when fields change.

db.unwatch(key)

Stop watching a key.

Future

  • More commands
  • Better tests
  • Optimizations
  • Who knows?

Contributions are welcome! :)

Disclaimer

It's still just a proof of concept, no real life tests are done.

Questions?

Find me on Twitter @stagas and IRC freenode.net #node.js as stagas

Something went wrong with that request. Please try again.