asynchronous copy-on-write btree
npm install btree
|2||downloads in the last week|
|5||downloads in the last month|
|Version||1.0.1 last updated a year ago|
|License||Apache License, Version 2.0|
|Keywords||btree, async, copy-on-write, cow, asynchronous, database, big data, b-tree, data structure|
see http://en.wikipedia.org/wiki/B-tree for details on b-tree's. they're a fundamental building-block of databases or "big data," and one of the best and most powerful ways known to mankind for enabling random access to large structures.
to install the b-tree code and then run a simple demo using node/npm.js @ https://npmjs.org/package/btree ---
npm install btree npm start btree
"asynchronous" means that all methods require a callback and return nothing directly. in other words, this implementation is completely non-blocking, event-driven, and thus naturally high-performance.
"copy-on-write" (cow) means that the entire b-tree is a pure immutable value, represented by its root key. all mutations create a new root object; i.e., a new version three levels down in the b-tree creates new versions of nodes in the upper levels as well, including creating the new root which references that change down below.
the two main methods are
really just special cases. the b-tree relies on a
key/value store, the simplest example of which is included as
simple.js for an ultra-simple demo, and
demo.js for a more
bug reports and pull requests will be welcomed! thanks in advance.
initially inspired by Justin Wetherell's java code at http://code.google.com/p/java-algorithms-implementation/