npm install btreejs
var btree = require"btreejs";var Tree = btreecreate2 btreenumcmp;var tree = ;treeput0 "null";treeput1 "one";treeput2 "two";treedel1;treeget2; // == "two"
Creates and returns a Tree class of the specified order with the specified comparator.
|order||Order of the Tree-class to build as a number. Defaults to |
|compare||function(a: *, b: *):number returning -1 if a<b, 1 if a>b and 0 otherwise. Defaults to |
Strict string comparator that compares strings character by character.
Constructs a Tree instance of the order specified to
Puts a non-undefined, non-null key with the given non-undefined value into the tree. You have to type check keys on your own.
Gets the value of the specified key. Returns
undefined if the key does not exist. There is no Tree#exists method or
similar because it just would encourage multiple lookups if one is already sufficient. So the correct usage is:
var value = treegetmyKey;if typeof value == 'undefined'// Key does not existelse// Key exists, value may be null
Deletes a key from the tree. Returns
true on success or
false if there is no such key.
Walks the range [minKey, ..., maxKey] in ascending order by calling the callback for every key/value pair found.
|minKey||Minimum key or |
|maxKey||Maximum key or |
To break the loop, let callback explicitly return
Walks the range [minKey, ..., maxKey] in descending order by calling the callback for every key/value pair found.
Counts the number of keys in the range [minKey, ..., maxKey]. See
Tree#del throw an
Error only if the key is
null or the value
undefined. Other methods do not throw.
The test suite contains a 100k benchmark:
Ran in node v0.10.5 on one core of an 3.40Ghz Intel Core i7-2600K working with premium DDR3 ram (I'm too lazy to look
up the exact model). To test on your own hardware, simply run
npm [-g] install testjs &&
testjs itself is an optimized wrapper around node's native assert module.
Also includes a cluster-based (runs in a dedicated worker) optimizer in
tests/optimize.js that takes ranges of orders
to calculate the optimal order for a specific tree size:
times parameter specifies how many benchmarks for each order shall be run to eliminate random peeks and is
responsible for how long a test takes. Results may vary depending on the actual hardware configuration used.
Beware: The default settings that have been used to find btree's default order of 52 (orders 0 to 200, 100k items, 20
times) will take a while to process.
Apache License, Version 2.0