Improve the logic of extract hive job url in JdbcInterpreter

Currently only mr job url is extracted, tez job url is not extracted. And it looks like the latest hive version (3.1.2) don't provide yarn app id in its logs