Make responses from html type queries more succinct (and consistent)

[AaronAcerboni]

Responses from HTML type queries differ from json, xml and feed type queries because html queries take into account an aspect to extract. eg. text, html or some attribute.

Example:

Response from a query to json document

{
  results: [
    {
      foo: 'hello'
    }
  ],
  created: ...
}

Response from a query to a html document

{
  results: [
    {
      foo: {
        text: 'hello'
      }
    }
  ],
  created: ...
}

Responses to html queries should mirror the responses to the other query types in that foo should be a direct value and not an object. That object can still exist but only if either:

• The extract is specifically specified in the query (the default extract can change to text from html to match with most use cases)

OR

• only if multiple extract properties are specified.