Check out this toy example, first you create a Lazy object:
var Lazy = ;var lazy = ;lazy;
This code says that 'lazy' is going to be a lazy list that filters even numbers, takes first five of them, then multiplies all of them by 2, and then calls the join function (think of join as in threads) on the final list.
And now you can emit 'data' events with data in them at some point later,
The output will be produced by the 'join' function, which will output the expected [0, 4, 8, 12, 16].
And here is a real-world example. Some time ago I wrote a hash database for node.js called node-supermarket (think of key-value store except greater). Now it had a similar interface as a list, you could .forEach on the stored elements, .filter them, etc. But being asynchronous in nature it lead to the following code, littered with callbacks and temporary lists:
var Store = ;var db = filename : 'users.db' json : true ;var users_over_20 = ;db
This code selects first five users who are over 20 years old and stores them in users_over_20.
But now we changed the node-supermarket interface to return lazy lists, and the code became:
var Store = ;var db = filename : 'users.db' json : true ;db;
This is so much nicer!
Here is the latest feature: .lines. Given a stream of data that has \n's in it, .lines converts that into a list of lines.
Here is an example from node-iptables that I wrote the other week,
var Lazy = ;var spawn = spawn;var iptables = ;lines// skips the two lines that are iptables header;
This example takes the
iptables -L -n -v command and uses .lines on its output.
Then it .skip's two lines from input and maps a function on all other lines that
creates a data structure from the output.
Supports the following operations:
The Lazy object itself has a .range property for generating all the possible ranges.
Here are several examples:
Then you can use other lazy functions on these ranges.