README.md
# Airborne
[](https://travis-ci.org/brooklynDev/airborne)
[](https://coveralls.io/r/brooklynDev/airborne?branch=master)
[](https://codeclimate.com/github/brooklynDev/airborne/maintainability)
[](http://rubygems.org/gems/airborne)
[](http://rubygems.org/gems/airborne)
[](http://rubygems.org/gems/airborne)
RSpec driven API testing framework
## Looking for Project Maintainers
I am looking for project maintainers to help keep airborne up to date and bug-free while avoiding feature creep and maintaining backwards compatibility.
Comment [here](https://github.com/brooklynDev/airborne/issues/134) if you would like to help out.
## Installation
Install Airborne:
```shell
$ gem install airborne
```
Or add it to your Gemfile:
```ruby
gem 'airborne'
```
## Creating Tests
```ruby
require 'airborne'
describe 'sample spec' do
it 'should validate types' do
get 'http://example.com/api/v1/simple_get' #json api that returns { "name" : "John Doe" }
expect_json_types(name: :string)
end
it 'should validate values' do
get 'http://example.com/api/v1/simple_get' #json api that returns { "name" : "John Doe" }
expect_json(name: 'John Doe')
end
end
```
When calling expect_json_types, these are the valid types that can be tested against:
* `:int` or `:integer`
* `:float`
* `:bool` or `:boolean`
* `:string`
* `:date`
* `:object`
* `:null`
* `:array`
* `:array_of_integers` or `:array_of_ints`
* `:array_of_floats`
* `:array_of_strings`
* `:array_of_booleans` or `:array_of_bools`
* `:array_of_objects`
* `:array_of_arrays`
If the properties are optional and may not appear in the response, you can append `_or_null` to the types above.
```ruby
describe 'sample spec' do
it 'should validate types' do
get 'http://example.com/api/v1/simple_get' #json api that returns { "name" : "John Doe" } or { "name" : "John Doe", "age" : 45 }
expect_json_types(name: :string, age: :int_or_null)
end
end
```
Additionally, if an entire object could be null, but you'd still want to test the types if it does exist, you can wrap the expectations in a call to `optional`:
```ruby
it 'should allow optional nested hash' do
get '/simple_path_get' #may or may not return coordinates
expect_json_types('address.coordinates', optional(latitude: :float, longitude: :float))
end
```
Additionally, when calling `expect_json`, you can provide a regex pattern in a call to `regex`:
```ruby
describe 'sample spec' do
it 'should validate types' do
get 'http://example.com/api/v1/simple_get' #json api that returns { "name" : "John Doe" }
expect_json(name: regex("^John"))
end
end
```
When calling `expect_json` or `expect_json_types`, you can optionally provide a block and run your own `rspec` expectations:
```ruby
describe 'sample spec' do
it 'should validate types' do
get 'http://example.com/api/v1/simple_get' #json api that returns { "name" : "John Doe" }
expect_json(name: -> (name){ expect(name.length).to eq(8) })
end
end
```
Calling `expect_json_sizes` actually make use of the above feature and call `expect_json` under the hood:
```ruby
describe 'sample spec' do
it 'should validate types' do
get 'http://example.com/api/v1/simple_get_collection' #json api that returns { "ids" : [1, 2, 3, 4] }
expect_json_sizes(ids: 4)
end
end
```
## Making requests
Airborne uses `rest_client` to make the HTTP request, and supports all HTTP verbs. When creating a test, you can call any of the following methods: `get`, `post`, `put`, `patch`, `delete`, `head`, `options`. This will then give you access the following properties:
* `response` - The HTTP response returned from the request
* `headers` - A symbolized hash of the response headers returned by the request
* `body` - The raw HTTP body returned from the request
* `json_body` - A symbolized hash representation of the JSON returned by the request
For example:
```ruby
it 'should validate types' do
get 'http://example.com/api/v1/simple_get' #json api that returns { "name" : "John Doe" }
name = json_body[:name] #name will equal "John Doe"
body_as_string = body
end
```
When calling any of the methods above, you can pass request headers to be used.
```ruby
get 'http://example.com/api/v1/my_api', { 'x-auth-token' => 'my_token' }
```
For requests that require a body (`post`, `put`, `patch`) you can pass the body as well:
```ruby
post 'http://example.com/api/v1/my_api', { :name => 'John Doe' }, { 'x-auth-token' => 'my_token' }
```
The body may be any JSON-serializable type, as long as you want to post `application/json` content type.
You may set a different content type and post a string body this way:
```ruby
post 'http://example.com/api/v1/my_api', "Hello there!", { content_type: 'text/plain' }
```
For requests that require Query params you can pass a params hash into headers.
```ruby
post 'http://example.com/api/v1/my_api', { }, { 'params' => {'param_key' => 'param_value' } }
```
### (Not) Verifying SSL Certificates
SSL certificate verification is enabled by default (specifically, `OpenSSL::SSL::VERIFY_PEER`).
Carefully consider how you use this. It's not a solution for getting around a failed SSL cert verification; rather, it's intended for testing systems that don't have a legitimate SSL cert, such as a development or test environment.
You can override this behavior per request:
```ruby
verify_ssl = false
post 'http://example.com/api/v1/my_api', "Hello there!", { content_type: 'text/plain' }, verify_ssl
```
or with a global Airborne configuration:
```ruby
Airborne.configure do |config|
config.verify_ssl = false # equivalent to OpenSSL::SSL::VERIFY_NONE
end
```
Note the per-request option always overrides the Airborne configuration:
```ruby
before do
Airborne.configuration.verify_ssl = false
end
it 'will still verify the SSL certificate' do
verify_ssl = true
post 'http://example.com/api/v1/my_api', "Hello there!", { content_type: 'text/plain' }, verify_ssl
end
```
You can use the `verify_ssl` setting to override your global defaults in test blocks like this:
```ruby
describe 'test something', verify_ssl: false do
end
```
OR
```ruby
describe 'test something' do
Airborne.configuration.verify_ssl = false
end
```
This feature currently isn't supported when testing loaded Rack applications (see "Testing Rack Applications" below). If you need to set `verify_ssl: false`, then we recommend starting your Rack app server and sending the `airborne` HTTP requests as you would when testing any other service.
## Testing Rack Applications
If you have an existing Rack application like `sinatra` or `grape` you can run Airborne against your application and test without actually having a server running. To do that, just specify your rack application in your Airborne configuration:
```ruby
Airborne.configure do |config|
config.rack_app = MySinatraApp
end
```
Under the covers, Airborne uses [rack-test](https://github.com/brynary/rack-test) to make the requests.
## Rails Applications
If you're testing an API you've written in Rails, Airborne plays along with `rspec-rails`:
```ruby
require 'rails_helper'
RSpec.describe HomeController, :type => :controller do
describe 'GET index' do
it 'returns correct types' do
get :index, :format => 'json' #if your route responds to both html and json
expect_json_types(foo: :string)
end
end
end
```
## API
* `expect_json_types` - Tests the types of the JSON property values returned
* `expect_json` - Tests the values of the JSON property values returned
* `expect_json_keys` - Tests the existence of the specified keys in the JSON object
* `expect_json_sizes` - Tests the sizes of the JSON property values returned, also test if the values are arrays
* `expect_status` - Tests the HTTP status code returned
* `expect_header` - Tests for a specified header in the response
* `expect_header_contains` - Partial match test on a specified header
## Path Matching
When calling `expect_json_types`, `expect_json`, `expect_json_keys` or `expect_json_sizes` you can optionally specify a path as a first parameter.
For example, if our API returns the following JSON:
```json
{
"name": "Alex",
"address": {
"street": "Area 51",
"city": "Roswell",
"state": "NM",
"coordinates": {
"latitude": 33.3872,
"longitude": 104.5281
}
}
}
```
This test would only test the address object:
```ruby
describe 'path spec' do
it 'should allow simple path and verify only that path' do
get 'http://example.com/api/v1/simple_path_get'
expect_json_types('address', street: :string, city: :string, state: :string, coordinates: :object)
#or this
expect_json_types('address', street: :string, city: :string, state: :string, coordinates: { latitude: :float, longitude: :float })
end
end
```
Or, to test the existence of specific keys:
```ruby
it 'should allow nested paths' do
get 'http://example.com/api/v1/simple_path_get'
expect_json_keys('address', [:street, :city, :state, :coordinates])
end
```
Alternatively, if we only want to test `coordinates` we can dot into just the `coordinates`:
```ruby
it 'should allow nested paths' do
get 'http://example.com/api/v1/simple_path_get'
expect_json('address.coordinates', latitude: 33.3872, longitude: 104.5281)
end
```
When dealing with `arrays`, we can optionally test all (`*`) or a single (`?` - any, `0` - index) element of the array:
Given the following JSON:
```json
{
"cars": [
{
"make": "Tesla",
"model": "Model S"
},
{
"make": "Lamborghini",
"model": "Aventador"
}
]
}
```
We can test against just the first car like this:
```ruby
it 'should index into array and test against specific element' do
get '/array_api'
expect_json('cars.0', make: 'Tesla', model: 'Model S')
end
```
To test the types of all elements in the array:
```ruby
it 'should test all elements of the array' do
get 'http://example.com/api/v1/array_api'
expect_json('cars.?', make: 'Tesla', model: 'Model S') # tests that one car in array matches the tesla
expect_json_types('cars.*', make: :string, model: :string) # tests all cars in array for make and model of type string
end
```
`*` and `?` work for nested arrays as well. Given the following JSON:
```json
{
"cars": [
{
"make": "Tesla",
"model": "Model S",
"owners": [
{
"name": "Bart Simpson"
}
]
},
{
"make": "Lamborghini",
"model": "Aventador",
"owners": [
{
"name": "Peter Griffin"
}
]
}
]
}
```
===
```ruby
it 'should check all nested arrays for specified elements' do
get 'http://example.com/api/v1/array_with_nested'
expect_json_types('cars.*.owners.*', name: :string)
end
```
## Dates
JSON has no support for dates, however airborne gives you the ability to check for dates using the following. For `expect_json_types` you would use it as you would for any of the other types:
```ruby
it 'should verify date type' do
get '/get_date' #api that returns {createdAt: "Mon Oct 20 2014 16:10:42 GMT-0400 (EDT)"}
expect_json_types(createdAt: :date)
end
```
However if you want to check the actual date data with `expect_json`, you need to call the `date` function:
```ruby
it 'should verify correct date value' do
get '/get_date' #api that returns {createdAt: "Mon Oct 20 2014 16:10:42 GMT-0400 (EDT)"}
prev_day = DateTime.new(2014,10,19)
next_day = DateTime.new(2014,10,21)
#within the date callback, you can use regular RSpec expectations that work with dates
expect_json(createdAt: date { |value| expect(value).to be_between(prev_day, next_day) })
end
```
## Configuration
When setting up Airborne, you can call `configure` just like you would with `rspec`:
```ruby
#config is the RSpec configuration and can be used just like it
Airborne.configure do |config|
config.include MyModule
end
```
Additionally, you can specify a `base_url` and default `headers` to be used on every request (unless overridden in the actual request):
```ruby
Airborne.configure do |config|
config.base_url = 'http://example.com/api/v1'
config.headers = { 'x-auth-token' => 'my_token' }
end
describe 'spec' do
it 'now we no longer need the full url' do
get '/simple_get'
expect_json_types(name: :string)
end
end
```
You can also control the strictness of `expect_json` and `expect_json_types` with the global settings `match_expected_default` and `match_actual_default` like this.
```ruby
Airborne.configure do |config|
config.match_expected_default = true
config.match_actual_default = false
end
```
`match_expected_default` requires all the keys in the expected JSON are present in the response.
`match_actual_default` requires that the keys in the response are tested in the expected Hash.
So you can do the following combinations:
`match_expected_default=false`, `match_actual_default=false` - check only intersection
`match_expected_default=false`, `match_actual_default=true` - raise on extra key in response
`match_expected_default=true`, `match_actual_default=false` - raise on missing key in response
`match_expected_default=true`, `match_actual_default=true` - expect exact match
Airborne sets `match_expected_default` to `true` and `match_actual_default` to `false` by default.
You can use the `match_expected` and `match_actual` settings to override your global defaults in test blocks like this.
```ruby
describe 'test something', match_expected: true, match_actual: false do
end
```
OR
```ruby
describe 'test something' do
Airborne.configuration.match_expected = true
Airborne.configuration.match_actual = false
end
```
## Run it from the CLI
```shell
$ cd your/project
$ rspec spec
```
## Authors
* [Seth Pollack](https://github.com/sethpollack)
* [Alex Friedman](https://github.com/brooklynDev)
## Contributors
https://github.com/brooklynDev/airborne/graphs/contributors
Inspired by [frisby.js](https://github.com/vlucas/frisby)
## License
The MIT License
Copyright (c) 2014 [brooklyndev](https://github.com/brooklynDev), [sethpollack](https://github.com/sethpollack)
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.