The aim of HungryMoose is to make documenting tested APIs almost effortless. By using a unified data format to generate both tests and Web-based documentation, developers can rest assured that a test and its documentation will form an atomic unit. As a user, you are guaranteed that any examples you see in the documentation are functional examples.
HungryMoose is built around structured YAML files. These files are used by both the JUnit runner and the documentation generator to complete their respective tasks. You can learn more about what both of these projects do in the documentation and testing projects.
The YAML file can contain multiple examples known as scenarios. They are broken up in standard YAML fashion by triple
dashes --- Each scenario is made up of four fields as described below.
| Field | Description |
|---|---|
| name | The name of the scenario. Should be something easily understood like the name of a test |
| detail | Additional context on what the scenario is for or trying to accomplish |
| request | The text representation of the request including the url, verb, headers, and body |
| response | The text representation of the response including the status code, headers, and body |
The request object would look like something similar to this:
POST /endpoint/to/hit
Content-Type: application/json
{"data": "foo" }
and a response would look somthing like this:
200 OK
Content-Type: application/json
{"data":"bar"}
A complete file with two examples would look like this:
name: Can send arbitrary HTTP headers
request: |
GET /echo
Something: somewhere
Everything: everywhere
response: |
200 OK
Content-Type: application/json
{
"method": "GET",
"headers": {
"something": ["somewhere"],
"everything": ["everywhere"]
},
"body": "...",
"uri": "..."
}
---
name: Can submit a PUT request
request: |
PUT /echo
Content-Type: application/xml
{"data":"here ya go"}
response: |
200 OK
Content-Type: application/json
{
"uri": "/echo",
"method": "PUT",
"headers": "...",
"body": {"data":"here ya go"}
}