Skip to content

AlintaEnergy/cucumber-rest-bdd

 
 

Repository files navigation

Cucumber Rest BDD

Build Status Gem Version

A set of Behavioural tests that can be run against a REST API.

Giphy

This is based from: http://gregbee.ch/blog/effective-api-testing-with-cucumber

A list of Steps shows the comparison between Behavioural and Functional tests provided by this package.

Usage

You can include this as a gem in your features, or run directly through docker

Gem:

$ gem install cucumber-rest-bdd

Docker:

$ docker run --rm -v $(pwd):/opt/src -e endpoint=http://server/ graze/cucumber-rest-bdd

Configuration

The following environment variables modify how this will operate:

  • endpoint - (string) the base url to call for each request
  • data_key - (string) the root data key (if applicable) (for example: "data" if all responses have a {"data":{}} field)
  • error_key - (string) this will ignore the data_key when checking for errors
  • field_separator - (string) the separator used between words by the api
  • field_camel - (bool [true|false]) does this endpoint use camelCase for fields (default: false)
  • resource_single - (bool [true|false]) if each resource should be singularized or not (default: false)
  • set_parent_id - (bool [true|false]) when creating sub resources, automatically add parent ids

Examples

  • For a full list of steps see: STEPS
  • These examples are taken from the test features

Retrieve items

Given I am a client
When I request the post "1"
Then the request was successful
And the response has the following attributes:
    | attribute | type    | value |
    | User Id   | numeric | 1     |
    | Id        | numeric | 1     |
    | Title     | string  | sunt aut facere repellat provident occaecati excepturi optio reprehenderit |
    | Body      | string  | quia et suscipit\\nsuscipit recusandae consequuntur expedita et cum\\nreprehenderit molestiae ut ut quas totam\\nnostrum rerum est autem sunt rem eveniet architecto |
Given I am a client
When I request the photo "1" for album "1" for user "1"
Then the request was successful
And the response has the attributes:
    | attribute | type   | value                                              |
    | title     | string | accusamus beatae ad facilis cum similique qui sunt |
Given I am a client
When I request a list of posts with:
    | User Id | 2 |
Then the request is successful
And the response is a list of at least 2 posts
And one response has the following attributes:
    | attribute | type    | value |
    | User Id   | numeric | 2     |
    | Id        | numeric | 11    |
    | Title     | string  | et ea vero quia laudantium autem |
    | Body      | string  | delectus reiciendis molestiae occaecati non minima eveniet qui voluptatibus\\naccusamus in eum beatae sit\\nvel qui neque voluptates ut commodi qui incidunt\nut animi commodi |

You can inspect child objects by using : in between the names

Given I am a client
When I request the comment "1" with:
    | `_expand` | post |
Then the response has the following attributes:
    | attribute    | type   | value |
    | name         | string | id labore ex et quam laborum |
    | email        | string | Eliseo@gardner.biz |
    | body         | string | laudantium enim quasi est quidem magnam voluptate ipsam eos\\ntempora quo necessitatibus\\ndolor quam autem quasi\\nreiciendis et nam sapiente accusantium |
    | post : title | string | sunt aut facere repellat provident occaecati excepturi optio reprehenderit |
    | post : body  | string | quia et suscipit\\nsuscipit recusandae consequuntur expedita et cum\\nreprehenderit molestiae ut ut quas totam\\nnostrum rerum est autem sunt rem eveniet architecto |

Alternatively you can inspect child arrays and objects by describing the path of the object with attributes

Given I am a client
When I set JSON request body to:
    """
    {"title":"test","body":"multiple",
    "comments":[
        {"common":1,"id":1,"title":"fish","body":"cake","image":{"href":"some_url"}},
        {"common":1,"id":2,"title":"foo","body":"bar","image":{"href":"some_url"}}
    ]}
    """
And I send a POST request to "http://test-server/posts"
Then the response has the attributes:
    | attribute | type   | value    |
    | title     | string | test     |
    | body      | string | multiple |
And the response has a list of comments
And the response has a list of 2 comments
And the response has two comments with attributes:
    | attribute | type    | value |
    | common    | integer | 1     |
And the response has two comments with an image with attributes:
    | attribute | type    | value    |
    | href      | string  | some_url |
And the response has one comment with attributes:
    | attribute | type    | value |
    | Id        | integer | 1     |
    | Title     | string  | fish  |
    | Body      | string  | cake  |
And the response has one comment with attributes:
    | attribute | type    | value |
    | Id        | integer | 2     |
    | Title     | string  | foo   |
    | Body      | string  | bar   |

Each numeric request can be prefixed with a modifier to modify the number specified

Given I am a client
When I request a list of posts with:
    | `_embed` | comments |
Then the response is a list of posts
Then the response is a list of more than 5 posts
Then the response is a list of at least 10 posts
Then more than three posts have the attributes:
    | attribute | type    | value |
    | User Id   | integer | 5     |
Then less than 200 posts have more than four comments
Then more than 50 posts have less than six comments
Then more than 80 posts have a list of comments
Then at least 90 posts have a list of five comments
Then more than 10 posts have five comments
Then less than 200 posts have five comments

Errors

If the error_key environment variable is set, if that key is used as the initial step it will ignore any data_key setting.

Example: error_key=error, data_key=data

Then the response has a list of posts       | {"data":[{"id": 12}]}

Then the response has one error             | {"errors":[{"message": "Error"}]}

Then the response has an error              | {"error": {"message": "Error}}

Example: error_key=, data_key=data

Then the response has a list of posts       | {"data":[{"id": 12}]}

Then the response has one error             | {"data": {"errors":[{"message": "Error"}]}}

Then the response has an error              | {"data": {"error": {"message": "Error}}}

Creation

Given I am a client
When I request to create a post with:
    | attribute | type    | value |
    | Title     | string  | foo   |
    | Body      | string  | bar   |
    | User Id   | numeric | 1     |
Then the request is successful and a post was created
And the response has the following attributes:
    | attribute | type    | value |
    | User Id   | numeric | 1     |
    | Title     | string  | foo   |
    | Body      | string  | bar   |

If the environment variable: set_parent_id is set to true then when you create sub resources it will add the level aboves id into the json, otherwise it will rely on the api to do it for you

Given I am a client
When I request to create a photo in album "2" for user "1" with:
    | attribute | type   | value |
    | title     | string | foo   |
Then the comment was created
And the response has the attributes:
    | attribute | type   | value |
    | Album Id  | int    | 2     |
    | Title     | string | foo   |
When I request a list of photos for album "2" for user "1"
Then the request was successful
And one comment has the attributes:
    | attribute | type   | value |
    | Title     | string | foo   |

Removal

Given I am a client
When I request to remove the post "20"
Then the request is successful

Modification

Given I am a client
When I request to modify the post "21" with:
    | attribute | type    | value |
    | Title     | string  | foo   |
Then the request is successful
And the response has the following attributes:
    | attribute | type    | value |
    | User Id   | numeric | 3     |
    | Title     | string  | foo   |
    | Body      | string  | repellat aliquid praesentium dolorem quo\\nsed totam minus non itaque\\nnihil labore molestiae sunt dolor eveniet hic recusandae veniam\\ntempora et tenetur expedita sunt |

Multiple Requests

Given I am a client
When I request to create a post with:
    | attribute | type    | value |
    | Title     | string  | foo   |
    | Body      | string  | bar   |
    | User Id   | integer | 12    |
Then the request is successful
When I save "id"
And I request the post "{id}"
Then the request is successful
And the response has the following attributes:
    | attribute | type    | value |
    | Title     | string  | foo   |
    | Body      | string  | bar   |
    | User Id   | numeric | 12    |
    | Id        | numeric | {id}  |

Resources

A resource "name" is attempted to be retrieved from the given name of the item to be retrieved. This pluralises, ensures everything is lower case, removes any unparameterisable characters and uses a - separator.

Token -> tokens
User -> users
Big Life -> big-lifes
octopus -> octopi

If the environment variable: resource_single is set to true then it will not attempt to pluralise the resources.

Token -> token
User -> user

You can directly pass what you want using:

`field`

this will not modify the field.

Attributes

### Types

Attribute types: The following types are supported:

type other names example
integer numeric, number, long 12
float double, decimal 4.8
string text "text"
array array ["a"]
object object {"a":"b"}
null nil
bool boolean true

Name conversion

attributes are converted into singular parametrised versions of the provided name:

The conversion is based on the provided environment variables field_camel and field_separator

Default

field_camel=false
field_separator=_

Someid -> someid
Product Id -> product_id
Bodies -> body

CamelCase

field_camel=true
field_separator=_

Someid -> someid
Product Id -> productId
Bodies -> body

Other separator

field_camel=false
field_separator=-

Someid -> someid
Product Id -> product-id
Bodies -> body

About

Behavioural testing suite for RESTful APIs

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Ruby 57.1%
  • Gherkin 40.8%
  • Other 2.1%