Incorrect 'accessing url' output

[hamiltont]

Here is the bug - note that the Accessing URL field never changes, but somehow the response appears to be different!

Accessing URL http://127.0.0.1:8080/index.php/bench/db/:
  Response (trimmed to 40 bytes): "[{"id":778,"randomNumber":5795},{"id":30"
Accessing URL http://127.0.0.1:8080/index.php/bench/db/:
  Response (trimmed to 40 bytes): "{"id":5754,"randomNumber":1388}"
Accessing URL http://127.0.0.1:8080/index.php/bench/db/:
  Response (trimmed to 40 bytes): "{"id":8660,"randomNumber":4689}"
Accessing URL http://127.0.0.1:8080/index.php/bench/db/:
  Response (trimmed to 40 bytes): "[{"id":1995,"randomNumber":2759},{"id":3"
   WARN for http://127.0.0.1:8080/index.php/bench/db/0
     Top-level JSON is an object, not an array
     See http://frameworkbenchmarks.readthedocs.org/en/latest/Project-Information/Framework-Tests/#specific-test-requirements
   WARN for http://127.0.0.1:8080/index.php/bench/db/foo
     Top-level JSON is an object, not an array
     See http://frameworkbenchmarks.readthedocs.org/en/latest/Project-Information/Framework-Tests/#specific-test-requirements

The problem is that we print "Accessing URL" in _curl, but the calls to _curl do not include the full URL, only the calls to _curl_body. This is a mistake - _curl needs the full URL passed to it, just in case the returned headers are different

[hamiltont]

Ping @msmith-techempower @knewmanTE - this one is low hanging fruit if someone has a minute - basically just look thru all the files in test_types and ensure all the calls to _curl use the proper "URL + suffix" pattern