A Ruby interface to the Twitter API.
gem install twitter
Looking for the Twitter command-line interface? It was removed from this gem in version 0.5.0 and now exists as a separate project.
https://github.com/sferik/twitter/tree/master/examples
You should follow @gem on Twitter for announcements and updates about this library.
Please direct questions about this library to the mailing list.
Does your project or organization use this gem? Add it to the apps wiki!
Global configuration has been removed, as it was not threadsafe. Instead, you
can configure a Twitter::REST::Client
by passing it a block when it's
initialized.
client = Twitter::REST::Client.new do |config|
config.consumer_key = "YOUR_CONSUMER_KEY"
config.consumer_secret = "YOUR_CONSUMER_SECRET"
config.access_token = "YOUR_ACCESS_TOKEN"
config.access_token_secret = "YOUR_ACCESS_SECRET"
end
Note: oauth_token
has been renamed to access_token
and oauth_token_secret
is now access_token_secret
to conform to the terminology used in Twitter's
developer documentation.
This library now offers support for the Twitter Streaming API.
Site Streams are restricted to whitelisted accounts. To apply for access, follow the steps in the Site Streams documentation. User Streams do not require prior approval.
Unlike the rest of this library, this feature is not well tested and not recommended for production applications. I've decided to ship it as an experimental feature and make it more robust over time. Patches in this area are particularly welcome.
Hopefully, by the time version 6 is released, this gem can fully replace TweetStream, em-twitter, twitterstream, and twitter-stream. Special thanks to Steve Agalloco, Tim Carey-Smith, and Tony Arcieri for helping to develop this feature.
Configuration works just like Twitter::REST::Client
client = Twitter::Streaming::Client.new do |config|
config.consumer_key = "YOUR_CONSUMER_KEY"
config.consumer_secret = "YOUR_CONSUMER_SECRET"
config.access_token = "YOUR_ACCESS_TOKEN"
config.access_token_secret = "YOUR_ACCESS_SECRET"
end
Stream mentions of coffee or tea
topics = ["coffee", "tea"]
client.filter(:track => topics.join(",")) do |object|
puts object.text if object.is_a?(Twitter::Tweet)
end
Stream a random sample of all tweets
client.sample do |object|
puts object.text if object.is_a?(Twitter::Tweet)
end
Stream tweets, events, and direct messages for the authenticated user
client.user do |object|
case object
when Twitter::Tweet
puts "It's a tweet!"
when Twitter::DirectMessage
puts "It's a direct message!"
when Twitter::Streaming::StallWarning
warn "Falling behind!"
end
end
An object
may be one of the following:
Twitter::DirectMessage
Twitter::Streaming::DeletedTweet
Twitter::Streaming::Event
Twitter::Streaming::FriendList
Twitter::Streaming::StallWarning
Twitter::Tweet
The Twitter::Cursor
class has been completely redesigned with a focus on
simplicity and performance.
Notes | Version 4 | Version 5 | ||
---|---|---|---|---|
Code | HTTP GETs | Code | HTTP GETs | |
Are you at the start of the cursor? |
|
Θ(1) |
|
Θ(1) |
Return your most recent follower. |
|
Θ(1) |
|
Θ(1) |
Return an array of all your friends. |
|
Θ(n+1) |
|
Θ(n) |
Collect your 20 most recent friends. |
|
Θ(n+1) |
|
Θ(1) |
Collect your 20 most recent friends twice. |
|
Θ(2n+2) |
|
Θ(1) |
In the examples above, n varies with the number of people the authenticated
user follows on Twitter. This resource returns up to 20 friends per HTTP GET,
so if the authenticated user follows 200 people, calling
client.friends.take(20)
would make 11 HTTP requests in version 4. In version
5, it makes just 1 HTTP request. Keep in mind, eliminating a single HTTP
request to the Twitter API will reduce the latency of your application by
about 500 ms.
The last example might seem contrived ("Why would I call
client.friends.take(20)
twice?") but it applies to any
Enumerable
method you might call on a cursor, including:
#all?
, #collect
, #count
, #each
, #inject
, #max
, #min
, #reject
,
#reverse_each
, #select
, #sort
, #sort_by
, and #to_a
. In version 4,
each time you called one of those methods, it would perform n+1 HTTP
requests. In version 5, it only performs those HTTP requests the first time any
one of those methods is called. Each subsequent call fetches data from a
cache.
The performance improvements are actually even better than the table above
indicates. In version 5, calling Twitter::Cursor#each
(or any
Enumerable
method) starts yielding results immediately and
continues yielding as each response comes back from the server. In version 4,
#each
made a series of requests and waited for the last one to complete
before yielding any data.
Here is a list of the interface changes to Twitter::Cursor
:
#all
has been replaced by#to_a
.#last
has been replaced by#last?
.#first
has been replaced by#first?
.#first
now returns the first element in the collection, as prescribed byEnumerable
.#collection
and its aliases have been removed.
The Twitter::SearchResults
class has also been redesigned to have an
Enumerable
interface. The #statuses
method and its aliases
(#collection
and #results
) have been replaced by #to_a
. Additionally,
this class no longer inherits from Twitter::Base
. As a result, the #[]
method has been removed.
The #trends
method now returns an Enumerable
Twitter::TrendResults
object instead of an array. This object provides
methods to determine the recency of the trend (#as_of
), when the trend
started (#created_at
), and the location of the trend (#location
). This data
was previously unavailable.
Similarly, the #reverse_geocode
, #geo_search
, and #similar_places
methods
now return an Enumerable
Twitter::GeoResults
object instead
of an array. This object provides access to the token to create a new place
(#token
), which was previously unavailable.
The Twitter::Tweet
object has been cleaned up. The following methods have been
removed:
#from_user
#from_user_id
#from_user_name
#to_user
#to_user_id
#to_user_name
#profile_image_url
#profile_image_url_https
These attributes can be accessed via the Twitter::User
object, returned
through the #user
method.
The Twitter::User
object has also been cleaned up. The following aliases have
been removed:
#favorite_count
(use#favorites_count
)#favoriters_count
(use#favorites_count
)#favourite_count
(use#favorites_count
)#favouriters_count
(use#favorites_count
)#follower_count
(use#followers_count
)#friend_count
(use#friends_count
)#status_count
(use#statuses_count
)#tweet_count
(use#tweets_count
)#update_count
(use#tweets_count
)#updates_count
(use#tweets_count
)#translator
(use#translator?
)
Earlier versions of this library aliased favourites
to favorites
. These
aliases have been removed. Ruby is implemented in American English. The
initialize
method is spelled with a "z", not an "s", and Ruby provides no
alias. Likewise, this library does not provide aliases for Commonwealthers.
Merica. 🇺🇸
All create, destroy, add, and remove methods have been renamed to put the verb at the beginning:
#direct_message_create
is now#create_direct_message
#direct_message_destroy
is now#destroy_direct_message
#list_create
is now#create_list
#list_destroy
is now#destroy_list
#list_remove_member
is now#remove_list_member
#list_remove_members
is now#remove_list_members
#list_add_member
is now#add_list_member
#list_add_members
is now#add_list_members
#lists_owned
is now#owned_lists
#saved_search_create
is now#create_saved_search
#saved_search_destroy
is now#destroy_saved_search
#status_destroy
is now#destroy_status
In version 4, methods you would expect to return a Twitter
object would
return nil
if that object was missing. This may have resulted in a
NoMethodError
. To prevent such errors, you may have introduced checks for the
truthiness of the response, for example:
status = client.status(55709764298092545)
if status.place
# Do something with the Twitter::Place object
elsif status.geo
# Do something with the Twitter::Geo object
end
In version 5, all such methods will return a Twitter::NullObject
instead of
nil
. This should prevent NoMethodError
but may result in unexpected
behavior if you have truthiness checks in place, since everything is truthy in
Ruby except false
and nil
. For these cases, there are now predicate
methods:
status = client.status(55709764298092545)
if status.place?
# Do something with the Twitter::Place object
elsif status.geo?
# Do something with the Twitter::Geo object
end
The Twitter::List
, Twitter::Tweet
, and Twitter::User
objects all have a
#uri
method, which returns an HTTPS URI to twitter.com. This clobbers the
Twitter::List#uri
method, which previously returned the list URI's path (not
a URI).
These methods are aliased to #url
for users who prefer that nomenclature.
Twitter::User
previously had a #url
method, which returned the user's
website. This URI is now available via the #website
method.
All #uri
methods now return Addressable::URI
objects instead of strings. To convert an
Addressable::URI
object to a string, call #to_s
on it.
Twitter API v1.1 requires you to authenticate via OAuth, so you'll need to register your application with Twitter. Once you've registered an application, make sure to set the correct access level, otherwise you may see the error:
Read-only application cannot POST
Your new application will be assigned a consumer key/secret pair and you will be assigned an OAuth access token/secret pair for that application. You'll need to configure these values before you make a request or else you'll get the error:
Bad Authentication data
You can pass configuration options as a block to Twitter::REST::Client.new
.
client = Twitter::REST::Client.new do |config|
config.consumer_key = "YOUR_CONSUMER_KEY"
config.consumer_secret = "YOUR_CONSUMER_SECRET"
config.access_token = "YOUR_ACCESS_TOKEN"
config.access_token_secret = "YOUR_ACCESS_SECRET"
end
After configuration, requests can be made like so:
client.update("I'm tweeting with @gem!")
All examples require an authenticated Twitter client. See the section on configuration.
Tweet (as the authenticated user)
client.update("I'm tweeting with @gem!")
Follow a user (by screen name or user ID)
client.follow("gem")
client.follow(213747670)
Fetch a user (by screen name or user ID)
client.user("gem")
client.user(213747670)
Fetch a cursored list of followers with profile details (by screen name or user ID, or by implicit authenticated user)
client.followers("gem")
client.followers(213747670)
client.followers
Fetch a cursored list of friends with profile details (by screen name or user ID, or by implicit authenticated user)
client.friends("gem")
client.friends(213747670)
client.friends
Fetch a collection of user_ids that the currently authenticated user does not want to receive retweets from
client.no_retweet_ids
Fetch the timeline of Tweets by a user
client.user_timeline("gem")
client.user_timeline(213747670)
Fetch the timeline of Tweets from the authenticated user's home page
client.home_timeline
Fetch the timeline of Tweets mentioning the authenticated user
client.mentions_timeline
Fetch a particular Tweet by ID
client.status(27558893223)
Collect the three most recent marriage proposals to @justinbieber
client.search("to:justinbieber marry me", :result_type => "recent").take(3).collect do |tweet|
"#{tweet.user.screen_name}: #{tweet.text}"
end
Find a Japanese-language Tweet tagged #ruby (excluding retweets)
client.search("#ruby -rt", :lang => "ja").first.text
For more usage examples, please see the full documentation.
This entity-relationship diagram is generated programatically. If you add or remove any Twitter objects, please regenerate the ERD with the following command:
bundle exec rake erd
This library aims to support and is tested against the following Ruby versions:
- Ruby 1.8.7
- Ruby 1.9.2
- Ruby 1.9.3
- Ruby 2.0.0
- Ruby 2.1.0
If something doesn't work on one of these versions, it's a bug.
This library may inadvertently work (or seem to work) on other Ruby versions, however support will only be provided for the versions listed above.
If you would like this library to support another Ruby version or implementation, you may volunteer to be a maintainer. Being a maintainer entails making sure all tests run and pass on that implementation. When something breaks on your implementation, you will be responsible for providing patches in a timely fashion. If critical issues for a particular implementation exist at the time of a major release, support for that Ruby version may be dropped.
This library aims to adhere to Semantic Versioning 2.0.0. Violations of this scheme should be reported as bugs. Specifically, if a minor or patch version is released that breaks backward compatibility, that version should be immediately yanked and/or a new version should be immediately released that restores compatibility. Breaking changes to the public API will only be introduced with new major versions. As a result of this policy, you can (and should) specify a dependency on this gem using the Pessimistic Version Constraint with two digits of precision. For example:
spec.add_dependency 'twitter', '~> 5.0'
Copyright (c) 2006-2013 Erik Michaels-Ober, John Nunemaker, Wynn Netherland, Steve Richert, Steve Agalloco. See LICENSE for details.