Hyve

http://github.com/Tawlk/hyve

About

Hyve (pronounced: "hive") is a Javascript library aiming to be a general solution for streaming the latest data from multiple social media networks, for any given topic, and abstracting all of them into one simple API.

Current Features

• Plays nice with all major JS frameworks (jQuery, Mootools etc.)
• Can run any Javascript function of your choice against each posting retrieved.
• Outputs all data in a normalized format - USMF
• Supports searching Facebook, Twitter, Identica, Foursquare, Flickr, Picasa, Wordpress, Youtube, Reddit, Digg, and Delicious
• Indirectly supports Imgur and Vimeo by catching links shared via other services.
• Expands urls from t.co and bitly
• Streams from all services (default), or optionally only selected services
• Streams personal feed of subscriptions using Oauth tokens (for supported networks)
• Runs in the browser or on the server (Node.js)

Requirements

For running Hyve in Node.js, you'll need the request library.

You can grab it from npm:

$ npm install request

For running Hyve in the browser in production you'll need to compile it with the grunt library.

You can grab it from npm:

$ npm -g install grunt

Usage

Configuration Options

Data Queueing

When this is enabled all incoming data will be placed into a FIFO queue by so your application can dequeue items as appropriate.

Queued items will be stored in 'hyve.queue' as an array:

console.log(hyve.queue)

>>> {'text':[],'link':[],'video':[],'image':[],'checkin':[]}

Enable Queuing:

hyve.queue_enable = true

Dequeue Queued Item:

var item = hyve.queue.text[0]

hyve.dequeue(item)

Data Recall (Browser Only)

If enabled in conjunction with hyve.queue hyve will cache all queued items using HTML5 localStorage. Hyve will recall any cached items if the page containing hyve is closed and restarted.

Enable Recall:

hyve.recall_enable = true

API Key requirements

In order to use some functions within some services you must supply API Keys.

Foursquare will only return data if the client has HTML5 Geolocaton support. You must also define client_id and client_secret like so:

hyve.feeds.foursquare.client_id = 'your_application_client_id'
hyve.feeds.foursquare.client_secret = 'your_application_client_secret'

Flickr will work without authentication, however the 'views' statistic is not returned without auth. Without 'views' the 'weight' metric will be set to 0.

To get 'views' in Flickr you must specify an API key like so:

hyve.feeds.flickr.api_key = 'your_api_key',

To utilize bitly url un-shortening you must supply a username and API key:

hyve.feeds.bitly.login = 'your_username'
hyve.feeds.bitly.api_key = 'your_api_key'

OAuth Token definitions

In order to use functions of hyve that require OAuth authentication (like the friends/subcriptions stream) you must supply OAuth tokens for each involved service. Obtaining OAuth tokens is outside the scope of hyve.

// Facebook (Oauth2)
hyve.feeds.facebook.oauth_token = 'oauth_token'

// Flickr (Oauth2)
hyve.feeds.flickr.oauth_token = 'oauth_token'

// Youtube (Oauth2)
hyve.feeds.youtube.oauth_token = 'oauth_token'

// Twitter (Oauth1a)
hyve.feeds.twitter.oauth_token = 'oauth_token'
hyve.feeds.twitter.oauth_consumer_key = 'oauth_consumer_key'
hyve.feeds.twitter.oauth_signature = 'oauth_signature'
hyve.feeds.twitter.oauth_nonce = 'oauth_nonce'
hyve.feeds.twitter.oauth_timestamp = 'oauth_timestamp'

// Instagram (Oauth2)
hyve.feeds.instagram.access_token = 'access_token'

Commands

Public search real-time stream

hyve.search.stream(search_terms,callback_function,optional_service_list)

Public recently popular search

hyve.search.popular(search_term,callback_function,optional_service_list)

Personal friends/subscriptions stream

hyve.friends.stream(callback_function,optional_service_list)

Stop running streams

hyve.stop()

Deployment

Browser - Production

For production use in a browser you should use a single-file minified version of hyve.

Compile using grunt:

$ grunt concat min

You will then find a minified version of hyve at dist/hyve.min.js

Browser - Development

During development you can simply source the files you intend to work with:

<script type="text/javascript" src="hyve/src/hyve.core.js">
<script type="text/javascript" src="hyve/src/hyve.twitter.js">
<script type="text/javascript" src="hyve/src/hyve.facebook.js">

Node.JS

In Node.js simply require hyve.core.js and any modules you need.

var hyve = require('src/hyve.core.js')
require('src/hyve.twitter.js')
require('src/hyve.facebook.js')

Examples

Simple echo of Twitter, Facebook, and Identica with pure JS:

var myFunction = function(data){console.log(' + data.service +' : '+ data.text  +')}
var myQuery = 'android'
var myServices = ['Twitter','Facebook','Identica']

hyve.search.stream(myQuery,myFunction,myServices)

Simple Node.js example to output data from all services:

var hyve = require('src/hyve.core.js')
require('src/hyve.twitter.js')
require('src/hyve.facebook.js')

hyve.stream('android', function(data){
    console.log(data.service +' : '+ data.text);
})

Basic live search engine with jQuery:

<!DOCTYPE HTML>
<html>
    <head>
        <script src="https://ajax.googleapis.com/ajax/libs/jquery/1.4.2/jquery.min.js"></script>
        <script src="dist/hyve.min.js" type="text/javascript"></script>
        <script type="text/javascript">
            $(document).ready(function() {
                $('#search').bind("click",function(){
                    query=$('#query').val();
                    hyve.search.stream(query,function(data){
                        $('#output').prepend($('<p>' + data.service +' : '+ data.text  +'</p>'))
                    });
                });
            });
        </script>
    </head>
    <body>
        <div id="input">
            <input id="query" type="text" />
            <button id="search">search</button>
        </div>
        <div id="output"></div>
    </body>
</html>

You can find more examples in the demos directory.

Notes

Questions/Comments? Please check us out on IRC via irc://freenode.net/#tawlk