Some JDBC datasources will fail to query or load parts of tables, like GaussDB, responding to KE

Currently KE keep all the table metadata as upper case, so it will fail to contacting with some JDBC datasources, like GaussDB, which are sensitive to the letters.

So to solve this issue, a new boolean property kylin.source.jdbc.convert-to-lowercase will be introduced, false by default, to transform KE metadata to lowercase.