Added PlayByPlay Regex for Ejections. Restructured PlayByPlay Tests.

[rsforbes]

Grouped tests for PlayByPlay into folder. Separated out data from tests to be more inclusive of disparate data that is identified throughout the PlayByPlay feed.

TestData is structured into a collections.defaultdict. This allows the data to be looked up by a key (e.g. FieldGoalMade). The values are stored as a k:v pair that work against the description key/value which contains the actual play as presented by the NBA. The other key\values correspond to the regex capture groups.

Overall, this will make testing much easier as new data is found and regex is modified to accommodate.

EXAMPLES OF DIFFERENT DATA:

#Field Goal Made (With Assist)
data["FieldGoalMade"].append({"description" : "S. Hill 24' 3PT Jump Shot (3 PTS) (Mahinmi 1 AST)", "player" : "S. Hill", "distance" : "24", "field_goal_type" : "3PT Jump Shot", "points" : "3", "player_ast" : "Mahinmi", "assists" : "1"})

#Field Goal Made (Without Assist)
data["FieldGoalMade"].append({"description" : "Aldridge 6' Turnaround Hook Shot (8 PTS)", "player" : "Aldridge", "distance" : "6", "field_goal_type" : "Turnaround Hook Shot", "points" : "8", "player_ast" : None, "assists" : None})

#Field Goal Made (Without Distance)
data["FieldGoalMade"].append({"description" : "Broekhoff 3PT Jump Shot (3 PTS) (Lee 2 AST)", "player" : "Broekhoff", "distance" : None, "field_goal_type" : "3PT Jump Shot", "points" : "3", "player_ast" : "Lee", "assists" : "2"})