ZEPPELIN-3379. Refactoring of JDBC autocomplete.

[mebelousov]

What is this PR for?

Current implementation of autocomplete allows only completion in schema.table.column scenario. Such scenario exclude tables for default schema.
The PR fixes the scenario with tables in default schema and alias processing.
Also I want to add feature to complete columns for used tables in statement from blank cursor.

The autocomplete was tested for Greenplum 4.3 (PostgreSQL), Oracle 11g, MySQL (Hive Metastore) and Hive.
This PR is reopening of #2970

What type of PR is it?

Improvement

What is the Jira issue?

ZEPPELIN-3379

How should this be tested?

• Manually
• Tests pass

Screenshots (if appropriate)

Questions:

• Does the licenses files need update? no
• Is there breaking changes for older versions? no
• Does this needs documentation? no

[mebelousov]

Will revert when anyone could test this PR.

[anirbanch]

@mebelousov would need this functionality, how can I help?

[mebelousov]

Hi @anirbanch !

I have adopted PR to branch 0.8 and have created patch (can be founded in linked JIRA issue).
You can apply patch and build jars.

On your risks you can

1. Replace next files (from linked JIRA issue)
   $ZEPPELIN_HOME/interpreter/jdbc/zeppelin-jdbc-0.8.1-SNAPSHOT.jar
   $ZEPPELIN_HOME/interpreter/jdbc/postgresql-native-driver-sql.keywords
   $ZEPPELIN_HOME/lib/interpreter/zeppelin-interpreter-0.8.1-SNAPSHOT.jar
2. Set "completionSupport" to false in $ZEPPELIN_HOME/interpreter/jdbc/interpreter-setting.json
   also you can add "completionKey": "TAB" to editor key
3. Run Zeppelin.