Omit navigational elements of a HAL response.

Author: dblock

.rubocop_todo.yml

@@ -1,5 +1,5 @@
 # This configuration was generated by `rubocop --auto-gen-config`
-# on 2014-09-17 21:23:50 -0400 using RuboCop version 0.26.0.
+# on 2014-09-18 12:21:47 -0400 using RuboCop version 0.26.0.
 # The point is for the user to remove these configuration records
 # one by one as the offenses are removed from the code base.
 # Note that changes in the inspected code, or installation of new
@@ -8,28 +8,36 @@
 # Offense count: 1
 # Configuration parameters: CountComments.
 Metrics/ClassLength:
-  Max: 106
+  Max: 118
 
-# Offense count: 36
+# Offense count: 2
+Metrics/CyclomaticComplexity:
+  Max: 8
+
+# Offense count: 39
 # Configuration parameters: AllowURI.
 Metrics/LineLength:
-  Max: 140
+  Max: 145
 
 # Offense count: 1
 # Configuration parameters: CountComments.
 Metrics/MethodLength:
   Max: 14
 
+# Offense count: 2
+Metrics/PerceivedComplexity:
+  Max: 9
+
 # Offense count: 1
 Style/AsciiComments:
   Enabled: false
 
-# Offense count: 3
+# Offense count: 2
 # Configuration parameters: EnforcedStyle, SupportedStyles.
 Style/ClassAndModuleChildren:
   Enabled: false
 
-# Offense count: 15
+# Offense count: 14
 Style/Documentation:
   Enabled: false
 
@@ -41,7 +49,7 @@ Style/DoubleNegation:
 Style/Lambda:
   Enabled: false
 
-# Offense count: 6
+# Offense count: 5
 # Configuration parameters: MaxSlashes.
 Style/RegexpLiteral:
   Enabled: false

CHANGELOG.md

@@ -1,8 +1,21 @@
 == Next
+
+* backwards incompatible changes
+  * All navigational structure methods, including `links`, `get` or `post`, have been renamed to `_link`, `_get`, or `_post` respectively (by @dblock).
+
 * enhancements
   * [#58](https://github.com/codegram/hyperclient/issues/58): Automatically follow redirects (by [@dblock](https://github.com/dblock)).
+  * [#63](https://github.com/codegram/hyperclient/pull/63): You can omit the navigational elements, `api.links.products` is now equivalent to `api.products` (by @dblock).
+  * Implemented Rubocop, Ruby-style linter (by @dblock).
+
+* bug fixes
+  * Nothing.
+
+* deprecations
+  * Nothing.
 
 == 0.3.1
+
 * backwards incompatible changes
   * Nothing.
 

Readme.md README.mdReadme.md renamed to README.md

@@ -16,12 +16,11 @@ Example API client:
 ```ruby
 api = Hyperclient.new('http://myapp.com/api').tap do |api|
   api.digest_auth('user', 'password')
-  api.headers.update({'accept-encoding' => 'deflate, gzip'})
+  api.headers.update('accept-encoding' => 'deflate, gzip')
 end
 ```
 
-By default, Hyperclient adds `application/json` as `Content-Type` and `Accept`
-headers. It will also sent requests as JSON and parse JSON responses.
+By default, Hyperclient adds `application/json` as `Content-Type` and `Accept` headers. It will also sent requests as JSON and parse JSON responses.
 
 [More examples][examples]
 
@@ -38,15 +37,22 @@ Hyperclient will try to fetch and discover the resources from your API.
 Accessing the links for a given resource is quite straightforward:
 
 ```ruby
-api.links.posts_categories
+api._links.posts_categories
+# => #<Resource ...>
+```
+
+Or omit `_links`, which will look for a link called "posts_categories" by default:
+
+```ruby
+api.posts_categories
 # => #<Resource ...>
 ```
 
 You can also iterate between all the links:
 
 ```ruby
-api.links.each do |name, link|
-  puts name, link.url
+api._links.each do |name, link|
+  puts name, link._url
 end
 ```
 
@@ -55,15 +61,21 @@ Actually, you can call any [Enumerable][enumerable] method :D
 If a Resource doesn't have friendly name you can always access it as a Hash:
 
 ```ruby
-api.links['http://myapi.org/rels/post_categories']
+api._links['http://myapi.org/rels/post_categories']
 ```
 
 ### Embedded resources
 
 Accessing embedded resources is similar to accessing links:
 
 ```ruby
-api.embedded.posts
+api._embedded.posts
+```
+
+Or omit the `_embedded` keyword. By default Hyperclient will look for a "posts" link, then for an embedded "posts" collection:
+
+```ruby
+api.posts
 ```
 
 And you can also iterate between them:
@@ -80,10 +92,25 @@ You can even chain different calls (this also applies for links):
 api.embedded.posts.first.links.author
 ```
 
+Or omit the navigational structures:
+
+```ruby
+api.posts.first.author
+```
+
+If you have a named link that retrieves an embedded collection of the same name, you can collapse the nested reference. The following statements produce identical results:
+
+```ruby
+api.links.posts.embedded.posts.first
+```
+
+```ruby
+api.posts.first
+```
+
 ### Attributes
 
-Not only you might have links and embedded resources in a Resource, but also
-its attributes:
+Not only you might have links and embedded resources in a Resource, but also its attributes:
 
 ```ruby
 api.embedded.posts.first.attributes
@@ -95,6 +122,9 @@ api.embedded.posts.first.attributes
 You can access the attribute values via attribute methods, or as a hash:
 
 ```ruby
+api.posts.first.title
+# => 'Linting the hell out of your Ruby classes with Pelusa'
+
 api.embedded.posts.first.attributes.title
 # => 'Linting the hell out of your Ruby classes with Pelusa'
 
@@ -107,35 +137,38 @@ api.embedded.posts.first.attributes.fetch('title')
 
 ### HTTP
 
-OK, navigating an API is really cool, but you may want to actually do something
-with it, right?
+OK, navigating an API is really cool, but you may want to actually do something with it, right?
 
-Hyperclient uses [Faraday][faraday] under the hood to perform HTTP calls. You can
-call any valid HTTP method on any Resource:
+Hyperclient uses [Faraday][faraday] under the hood to perform HTTP calls. You can call any valid HTTP method on any Resource:
 
 ```ruby
-post = api.embedded.posts.first
-post.get
-post.head
-post.put({title: 'New title'})
-post.patch({title: 'New title'})
-post.delete
-post.options
-
-posts = api.links.posts
-posts.post({title: "I'm a blogger!", body: 'Wohoo!!'})
+post = api._embedded.posts.first
+post._get
+post._head
+post._put(title: 'New title')
+post._patch(title: 'New title')
+post._delete
+post._options
+
+posts = api._links.posts
+posts._post(title: "I'm a blogger!", body: 'Wohoo!!')
 ```
 
 If you have a templated link you can expand it like so:
 
 ```ruby
-api.links.post.expand(:id => 3).first
+api._links.post._expand(id: 3).first
+# => #<Resource ...>
+```
+
+You can omit the "_expand" keyword.
+
+```ruby
+api.post(id: 3).first
 # => #<Resource ...>
 ```
 
-You can access the Faraday connection (to add middlewares or do whatever
-you want) by calling `connection` on the entry point. As an example, you could use the [faraday-http-cache-middleware](https://github.com/plataformatec/faraday-http-cache)
-:
+You can access the Faraday connection (to add middlewares or do whatever you want) by calling `connection` on the entry point. As an example, you could use the [faraday-http-cache-middleware](https://github.com/plataformatec/faraday-http-cache):
 
 ```ruby
 api.connection.use :http_cache
@@ -168,7 +201,7 @@ There's also a PHP library named [HyperClient](https://github.com/FoxyCart/Hyper
 
 ## License
 
-MIT License. Copyright 2012 [Codegram Technologies][codegram]
+MIT License. Copyright 2012-2014 [Codegram Technologies][codegram]
 
 [hal]: http://stateless.co/hal_specification.html
 [contributors]: https://github.com/codegram/hyperclient/contributors

examples/cyberscore.rb

@@ -44,7 +44,7 @@ def print_links(links)
     if link.is_a?(Array)
       print_links(link)
     else
-      puts %(Found "#{name}" at "#{link.url}" )
+      puts %(Found "#{name}" at "#{link._url}" )
     end
   end
 end

examples/hal_shop.rb

@@ -4,7 +4,7 @@
 def print_resources(resources)
   resources.each do |name, resource|
     begin
-      puts %(Found #{name} at #{resource.url})
+      puts %(Found #{name} at #{resource._url})
     rescue
       puts %(Found #{name})
     end

features/steps/api_navigation.rb

@@ -2,32 +2,32 @@ class Spinach::Features::ApiNavigation < Spinach::FeatureSteps
   include API
 
   step 'I should be able to navigate to posts and authors' do
-    api.links.posts.resource
-    api.links['api:authors'].resource
+    api._links.posts._resource
+    api._links['api:authors']._resource
 
     assert_requested :get, 'http://api.example.org/posts'
     assert_requested :get, 'http://api.example.org/authors'
   end
 
   step 'I search for a post with a templated link' do
-    api.links.search.expand(q: 'something').resource
+    api._links.search._expand(q: 'something')._resource
   end
 
   step 'the API should receive the request with all the params' do
     assert_requested :get, 'http://api.example.org/search?q=something'
   end
 
   step 'I load a single post' do
-    @post = api.links.posts.links.last_post
+    @post = api._links.posts._links.last_post
   end
 
   step 'I should be able to access it\'s title and body' do
-    @post.attributes.title.wont_equal nil
-    @post.attributes.body.wont_equal nil
+    @post._attributes.title.wont_equal nil
+    @post._attributes.body.wont_equal nil
   end
 
   step 'I should also be able to access it\'s embedded comments' do
-    comment = @post.embedded.comments.first
-    comment.attributes.title.wont_equal nil
+    comment = @post._embedded.comments.first
+    comment._attributes.title.wont_equal nil
   end
 end

features/steps/default_config.rb

@@ -11,19 +11,19 @@ class Spinach::Features::DefaultConfig < Spinach::FeatureSteps
 
   step 'I send some data to the API' do
     stub_request(:post, 'http://api.example.org/posts')
-    assert_equal 200, api.links.posts.post(title: 'My first blog post').status
+    assert_equal 200, api._links.posts._post(title: 'My first blog post').status
   end
 
   step 'it should have been encoded as JSON' do
     assert_requested :post, 'api.example.org/posts', body: '{"title":"My first blog post"}'
   end
 
   step 'I get some data from the API' do
-    @posts = api.links.posts
+    @posts = api._links.posts
   end
 
   step 'it should have been parsed as JSON' do
-    @posts.attributes.total_posts.to_i.must_equal 9
-    @posts.attributes['total_posts'].to_i.must_equal 9
+    @posts._attributes.total_posts.to_i.must_equal 9
+    @posts._attributes['total_posts'].to_i.must_equal 9
   end
 end

features/support/api.rb

@@ -15,7 +15,7 @@ def api
   end
 
   step 'I connect to the API' do
-    api.links
+    api._links
   end
 
   after do

lib/hyperclient/entry_point.rb

@@ -27,7 +27,7 @@ def initialize(url)
     #
     # Returns a Faraday::Connection.
     def connection
-      @connection ||= Faraday.new(url, { headers: default_headers }, &default_faraday_block)
+      @connection ||= Faraday.new(_url, { headers: default_headers }, &default_faraday_block)
     end
 
     private