… This requires using aliases. Updated the documentation and similar to descrbe that.

…equests. Removed range calls from for loops. I asume that's faster.

…at reduced duplication in getVideoRequests / getBannerRequest

…ty method for the cookie, but also, since I don't need to send the cookie for a banner request, rearanged things a bit to save some bytes on that request.

…n validation that I don't think is me. It seems like with mixed baner/video request, validation gets mixed up and sends video requests to banner endpoints and vis versa. When I run the exact same request (well, switching 'bidder' to 'beachfront') it works fine. For now I've puled the two 'mixed' validation tests out. I still have > 88% coverage. I'll dig deeper into this in a bit and if it looks like an actual bug in the validation system I'll report that.

…..reqs = make... line in MakeRequests. I was appending to a already sized slice, so winding up with a bunch of empty adapter.RequestData items.

…th test cases that didn't make sense to me - on multi-video.json. I was getting 2 errors when I ran validation, but none when I ran the same request from postman and stepped through in a debugging build. I need to look at that and at the. On unmarshal-error-video.json, I was getting a request sent with empty appId. I even added a check for it to bail right at the end, but the test system was still saying it would get through, and I don't see any way that is possible in a real request. Validation of the Ext object would stop it IRL. I'll look into this further. I do like test cases, but these don't seem to be reflecting reality.

…in keeping the indexes straight in MakeRequests. I've added back a couple of test cases, which now pass.

…em was that an empty adapters.RequestData was being added. Also added test case unmarshal-error-but-another-good-video.json, which throws and tracks the unmarshal error and dumps that bad one, then adds and requests the good video imp.