save right and wrong file in test_regex.py and improve accuracy stats

Save away the "right" and "wrong" variables returned in test_regex.py in a json file. Also, tie back the rule_id/rule_group numbers back to the original regex. Print some stats, by rule_id, rule_group and overall regex.

Start by reading and running test_regex.py to trace how the right and wrong variables, rule_id, rule_group, and original regex are produced. Done means the results are saved in JSON, rule identifiers remain tied to their source regexes, and statistics are printed per rule_id, per rule_group, and overall.

Written by the indexing model from the issue text.