TINYclient, a tiny HTTP/JSON crud client toolkit

Gem Version Build Status

TINYclient is inspired by Active Record and based on Curb.


  • Be sure that Curb/Curl works properly on your machine.
  • install the gem
gem install tiny_client

Getting Started


You can initialize your API by extending the TinyClient::Configuration

class MyConf < TinyClient::Configuration

  def initialize
    @url = 'http://localhost:3000/api/1.0'
    @headers = { 'Authorization' => 'token [email protected]' }
    @limit = 100
    @connect_timeout = 30 # seconds

You can use that configuration in your resource.

class Author < TinyClient::Resource
  conf MyConf.instance

  path 'authors' # query will be made on http://localhost:3000/api/1.0/authors

  fields :id, :name # your resource attributes

  nested Books # your resource nested resource

class Book < TinyClient::Resource
  conf MyConf.instance
  path 'books'
  fields :id, :title


Now you will be able to this:

author = Author.show(1) # Get /authors/1.json
author.name = 'P. K. D.'
author.save!   # PUT /authors/1.json { "author" : { "name" : "Bob" } }

book = Book.new
book.title = 'Confessions of a crap artist'
book = author.add_book(book) # POST /authors/1/books.json { "book" : { "title" : ".." }

book.id.present? # true

books = Book.index(limit: 10) # GET /books.json?limit=10

ed = Author.new
ed.name = 'Poe'
ed.save! # POST /authors.json { "author" : { "name" : "Poe" } }

ed_books = ed.books(limit: 10) # GET /authors/{ed.id}/books.json
first = ed_books.first
first.load! # GET /books/{first.id}.json

# You can also navigate through all resources

Author.index_all do |author| # It will retrieve all the authors, using limit, and offset query params to paginate
 # Do something for each author

Author.index_in_batches(limit: 1000) do |authors|
  # retrieve authors by batch of 1000

Instance methods behavior


It will perform a get request on the resource id and set the resource fields value to the value retrived by the response.

author.load! # GET /authors/{author.id}.json ->  { id: 1, name: 'Toto' ... }
author.name # 'Toto'


It will create the resource if id is not set, otherwise it will update it. The resource fields value will be updated by the response body content.

toto = MyModule::Author.new
toto.save! # POST /authors.json { author: {} }

toto.id # should have been set by through the reponse

toto.name = 'Toto'
toto.save! # PUT {author: {name: 'Toto'}} -> /authors/{toto.id}.json

Only changed values will be passed through the body content.

You can clear changes with #clear_changes! You can now which fields has been marked has changed with #changes

Changes is automatically clear when you perform a request ( i.e call, #show #index #get #put #post save! and so on)

Nested resource

You can add a nested resource thanks to the nested class methods.

class Author < TinyClient::Resource
  nested Books, Magazines

It will allows you to call your nested resource directly from an instance of your parent resource.

author = Author.show(1)
author.books(limit: 100)  # index GET /authors/1/books.json?limit=100
book = author.book(1)     # show  GET/authors/1/books/1.json
book.title = 'New title'
author.update_book(book)  # update  PUT /authors/1/books/1.json -- { 'book': { 'title': 'New title' } }
author.remove_book(book)  # destroy DELETE /authors/1/books/1.json
author.add_book(book)     # create  POST /author/1/books.json -- { 'book': { 'title': 'New title' } }
author.books_all.each     # x GET /authors/1/books.json?limit=.. -- Enumerator -- Retrieve ALL books using limit and offset to handle pagination

This is equivalent to the following:

author = Author.show(1)
author.nested_index(Book, limit: 100)
book = author.nested_show(Book, 1)
book.title = 'New title'
author.nested_all(Book, limit: 10) # retrieve all books, quering the server by batch of 10;

Constraint & Support

JSON only

TinyClient supports only JSON data. Accept: application/json header is set.

POST/PUT create/update

The content passed to the server will always be prefixed by the class name in lower case.

toto = MyModule::Author.new
toto.save! # POST { author: {} }

Pagination / Buffer

Pagination, buffer is achieve through limit and offset params.

Author.index_all({limit: 100}) # Will queries the server by batch of 100, until all authors has been retrieved through the enumerator.

Content-Encoding support

TinyClient support gzip Content-Encoding. Response with gzip Content-Encoding will be automatically decompressed. You can set the Accept-Encoding: gzip through the configuration headers.


You can run the test using: