[Interpreter] Introducing zeppelin via submodule

1. First, fork Apache Zeppelin's code repository in the hadoopsubmarine account, which ensures that hadoopsubmarine/zeppelin can always follow apache/zeppelin code updates.
2. Create a submarine branch in hadoopsubmarine/zeppelin and customize the development on this branch.
3. The hadoopsubmarine/zeppelin repository is contained via submodule in the apache/submarine repository.
4. In the /submarine/submarine-workbench/submarine-interpreter in the apache/submarine repository, we import the JAR package of our customized zeppelin interpreter in submodule for use.

Design doc: https://docs.google.com/document/d/1LJc5hRTWfIs6K5f7vMgGC_FUizeQf-qqiWICqO8SoR8/edit#