A simple library to do extensive things with JSON.

JSON Toolkit

This is a simple toolkit to make life with JSON easier in Node. On top of traditional parse and stringify methods, this toolkit helps retrieve JSON and traverse keys/values.

npm install json-toolkit

To use, simply require the package. JSON-Toolkit also has a Helper class (which is really the bulk of this thing) which is attached to the main object.

var json = require("json-toolkit"),
    JSONHelper = json.Helper;

Parses str into Javascript. Internally uses JSON.parse.

Parses a obj into JSON. Internally uses JSON.stringify.

Parses a obj into formatted JSON. indent is a string to use for indents or a number for the amount of spaces in an indent. Default for indent is \t.

Gets file contents and parses into Javascript. callback gives two arguments, error and data.

Parses data as JSON and saves to file. Using indent forces formatted printing (see above for usage). callback is called with one argument, error.

Creates a new Helper object. The constructor's first argument will take anything. If it's a string, it assumes that it is raw JSON or a filename (see options). If it's falsy, the internal data is set to {}. If it's anything else, it uses that.

Helper extends the Node class EventEmitter and has several events to capture. If using a filename in the constructor, you must capture the "ready" event before the object can be used.

One major thing to point out is that this uses path keys to traverse the data object. Anytime a method below has an argument named key, you can retrieve "deep" information by creating a "path" to it. This path is seperated by the option key_sep. Example: get("dependencies:underscore") -> data["dependencies"]["underscore"]

  • from_file, Default: false, Tells Helper that first argument into constructor is a filename and not raw JSON.
  • key_sep, Default: :, The string seperator used to divide paths.
  • pretty_output, Default: false, When saving, print formatted JSON.
  • indent, Default: \t, When turning this object into formatted JSON, this string is used as the indent. (See above)
  • data : Holds all of the parsed JSON data.
  • file : Name of the file to save to. This is set on construction and each save.
  • ready : Emitted when the object is ready for use. Not necesary unless using a file for data instead of a string or an object. Actually, any "ready" event you declare while using synchronous operations (ie, no FS calls) won't be fast enough and likely won't ever be called.
  • error : Emitted when object encouters an error. It should be noted that errors have the possibility of being thrown and not captured by this event. If you come across one, please let me know.
  • set : Emitted when a key is set. Gives two arguments, keys, which is an array of keys representing the path and value.
  • save : Emitted when a file is successfully saved. Gives a single argument, file.

Retrieves the value at key in data. To transcend objects, seperate each key with key_sep.

Sets value at key in data.

Returns an array of all keys (deeply) that match key. On top of using the "path" for key, match introduces two special characters: * and **. A single * will match anything up to the next key_sep. A **, however, will match anything including key_sep.

For example, deps:* would match deps:underscore and deps:foo, but not deps or deps:foo:bar. deps** would match deps and deps:foo and deps:foo:bar but not foo:deps.

This method turns string key into a Regular Expression so any valid regex in key is compatible. This method will also accept a Regular Expression for key and will be significantly faster than using a string.

Deeply loops through data at key start and calls iterative for each pair. iterative is a function with two arguments, key and value. If value is an object, iterative is called and then it will be called on all children values (hence the deep part).

Test the value at key against t. If t is a string, then it is used in a typeof to test the value as a specifc type. If t is a function, then it is called with value as the only argument. If t is a regular expression, the value is tested for matches. Anything else is matched for equality.

For example, test("dependencies:underscore", "string") would return true on this module's package.json.

Deeply search data for value and return the first key that matches.

Deeply search data for value and return all keys that match.

Saves data as JSON to file. If file is not provided, the internal file property is used instead. On completion, the internal file property is set to file. callback is called with one argument, error.

Returns a formatted JSON string. indent is optional and by default will use options.indent.

Returns an unformatted JSON string.

Returns internal data object. This is useful for embedding into objects that will be converted into JSON.

var job = new JSONHelper("package.json", {
    from_file: true,
    pretty_output: true

job.on("ready", function() {
    job.set("dependencies:underscore", "1.4.4");

job.on("save", function(file) {
    console.log("Saved JSON to "+file+".");

job.on("error", function(err) {

I really want to extend this library to include many different JSON tools. These are only the ones that I have personally needed in the past. If you have a suggested addition to this library, I'd love to hear it (and probably add it). Of course, I (Tyler Johnson) am always reachable at tyler@vintyge.com.