Skip to content

demian85/gnip

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

NodeJS Gnip module

Connect to Gnip streaming API and manage rules. You must have a Gnip account with any data source available, like Twitter Power Track.

Currenly, this module only supports JSON activity stream format, so you must enable data normalization in your admin panel.

Gnip.Stream

This class is an EventEmitter and allows you to connect to the stream and start receiving data.

Constructor options

  • timeout As requested in the Gnip docs (http://support.gnip.com/apis/powertrack/api_reference.html), this option in the constructor allows us to set a read timeout in the client. The recommended value is >=30 seconds, so the constructor will throw an error if a smaller timeout is provided. The default value for this option is 35 seconds.
  • backfillMinutes Number of minutes to backfill after connecting to the stream. Optional. Value should be 0 - 5.
  • partition Partition of the Firehose stream you want to connect to. Only required for Firehose streams.
  • parser Parser library for incoming JSON data. Optional, but defaults to the excellent json-bigint library.
    Matching tag IDs are sent to us as big integers which can't be reliably parsed by the native JSON library in Node.js. More info on this issue can be found at StackOverflow

API methods

  • start() Connect to the stream and start receiving data. At this point you should have registered at least one event listener for any of these events: 'data', 'object' or 'tweet'.

  • end() Terminates the connection.

Events

  • ready Emitted when the connection has been successfully established
  • data Emitted for each data chunk (decompressed)
  • error Emitted when any type of error occurs. An error is raised if the response status code is not 20x. {error: String} objects are also checked here.
  • object Emitted for each JSON object.
  • tweet Emitted for each tweet.
  • delete Emitted for each deleted tweet.
  • end Emitted when the connection is terminated. This event is always emitted when an error occurs and the connection is closed.

Gnip.Rules

This class allows you to manage an unlimited number of tracking rules.

Constructor options

  • user GNIP account username.
  • password GNIP account password.
  • url GNIP Rules endpoint url e.g. https://gnip-api.twitter.com/rules/${streamType}/accounts/${account}/publishers/twitter/${label}.json
  • batchSize The batch size used when adding/deleting rules in bulk. (Defaults to 5000)
  • parser Much like the parser option allowed in the Gnip Stream constructor, you can pass a custom parser handler/library for incoming JSON data. This is optional, and defaults to the json-bigint library. More details.
  • cacheFile Internally Gnip.Rules uses a file for caching the current state of the rules configuration, the default path is in the directory of the package. This optional configuration allows you to change the path as the default one may cause problems in applications where node_modules is in a read-only filesystems (e.g. AWS Lambda).

API methods

  • getAll(callback) Get cached rules.

  • update(rules: Array, callback) Creates or replaces the live tracking rules.
    Rules are sent in batches of options.batchSize, so you can pass an unlimited number of rules.
    The current tracking rules are stored in a local JSON file so you can update the existing rules efficiently without having to remove them all. The callback receives an object as the 2nd argument and contains the number of added and deleted rules.

  • clearCache(callback) Clears cached rules.

The following methods uses Gnip API directly and ignores the local cache. Avoid usage if you are working with too many rules!

  • live.update(rules: Array, callback)
  • live.add(rules: Array, callback)
  • live.remove(rules: Array, callback)
  • live.getAll(callback)
  • live.getByIds(ids: Array, callback)
  • live.removeAll(callback)

Gnip.Search

This class is an EventEmitter and allows you to connect to either the 30 day or full archive search API and start receiving data.

Constructor options

  • user GNIP account username.
  • password GNIP account password.
  • url GNIP Search endpoint url e.g. https://gnip-api.twitter.com/search/30day/accounts/{ACCOUNT_NAME}/{LABEL}.json
  • query Rule to match tweets.
  • fromDate The oldest date from which tweets will be gathered. Date given in the format 'YYYYMMDDHHMM'. Optional.
  • toDate The most recent date to which tweets will be gathered. Date given in the format 'YYYYMMDDHHMM'. Optional.
  • maxResults The maximum number of search results to be returned by a request. A number between 10 and 500. Optional.
  • tag Used to segregate rules and their matching data into different logical groups. Optional.
  • bucket The unit of time for which count data will be provided. Options: "day", "hour", "minute". Optional, for /counts calls.
  • rateLimiter A limiter object, used to control the rate of collection. Optional. If unspecified, a rate limit of 30 requests a minute will be shared between Search streams. If you have a non-standard rate limit, you should pass this parameter.
const RateLimiter = require('limiter').RateLimiter;
// Allow 60 requests per minute
const limiter = new RateLimiter(60, 'minute');
const stream = new Gnip.Search({
	rateLimiter : limiter,
  ...
});

API methods

  • start() Start receiving data. At this point you should have registered at least one event listener for 'object' or 'tweet'.

  • end() Terminates the connection.

Events

  • ready Emitted when tweets have started to be collected.
  • error Emitted when a recoverable (non fatal) error occurs.
  • object Emitted for each JSON object.
  • tweet Emitted for each tweet.
  • end Emitted when the connection is terminated. If the stream has ended due to a fatal error, the error object will be passed.

Gnip.Usage

This class allows you to track activity consumption across Gnip products.

Constructor options

const usage = new Gnip.Usage({
	url : 'https://gnip-api.twitter.com/metrics/usage/accounts/{ACCOUNT_NAME}.json',
	user : 'xxx',
	password : 'xxx'
});

API Methods

usage.get({ bucket:'day', fromDate:'201612010000', toDate:'201612100000' },function( err, body )
{
	...
});

Installation

npm install gnip

Example Usage

const Gnip = require('gnip');

const stream = new Gnip.Stream({
  url : 'https://gnip-stream.twitter.com/stream/powertrack/accounts/xxx/publishers/twitter/prod.json',
  user : 'xxx',
  password : 'xxx',
  backfillMinutes: 5 // optional
});
stream.on('ready', function() {
  console.log('Stream ready!');
});
stream.on('tweet', function(tweet) {
  console.log(tweet);
});
stream.on('error', function(err) {
  console.error(err);
});

const rules = new Gnip.Rules({
  url : 'https://gnip-api.twitter.com/rules/powertrack/accounts/xxx/publishers/twitter/prod.json',
  user : 'xxx',
  password : 'xxx',
  batchSize: 1234 // not required, defaults to 5000
});

const newRules = [
  '#hashtag', 
  'keyword', 
  '@user',
  {value: 'keyword as object'},
  {value: '@demianr85', tag: 'rule tag'}
];

rules.update(newRules, function(err) {
  if (err) throw err;
  stream.start();
});

const search = new Gnip.Search({
  url : 'https://gnip-stream.twitter.com/stream/powertrack/accounts/xxx/publishers/twitter/prod.json',
  user : 'xxx',
  password : 'xxx',
  query : '@user'
});

search.on('tweet', function(tweet) {
  console.log(tweet);
});

search.on('error', function(err) {
  console.error(err);
});

search.on('end', function(err) {
  if( err ) 
    console.error(err);
});

// search counts usage
const counts = new Gnip.Search({
  url : 'https://gnip-stream.twitter.com/stream/powertrack/accounts/xxx/publishers/twitter/prod/counts.json',
  user : 'xxx',
  password : 'xxx',
  query : '@user',
  bucket: 'day'
});

counts.on('object', function(object) {
  console.log(object.results);
  counts.end();
});

More details and tests soon...