Support for HTML request

This release addresses the following issues:

• #5 Get more useful info from the response: Now carl returns some basic headers apart from structured data when it is an HTML page.
• #4 Work without having to use protocol: Now using a protocol is optional. If you don't set it, carl will use https by default.
• #3 Detect Content-Type and parse the response accordingly: Added support for text/html and application/json content types.