I would like to configure an Apache Calcite combined jdbc connection spanning over Oracle and Snowflake, but without using json model file. I thought this is the way to do it with JVM system properties:

-Dcalcite.schema.oracle.type=JDBC
-Dcalcite.schema.oracle.url=jdbc:oracle:thin:@//myserverb:1521/whatever
-Dcalcite.schema.oracle.jdbc.driver=oracle.jdbc.OracleDriver
-Dcalcite.schema.oracle.jdbc.user=orauser
-Dcalcite.schema.oracle.jdbc.password=xxxx
-Dcalcite.schema.snowflake.type=JDBC
-Dcalcite.schema.snowflake.jdbc.url=jdbc:snowflake://account.snowflakecomputing/?database=SOMEDB
-Dcalcite.schema.snowflake.jdbc.user=snfuser
-Dcalcite.schema.snowflake.jdbc.password=yyyy

So, when I connect with a jdbc-enabled IDE (IntelliJ DataGrip), and all add all jars for oracle, snowflake and calcite in a custom JDBC driver, running this query:

select * from oracle.my_shema.my_table

gives me an error: Object 'oracle' not found

For the Calcite jdcb url in the IntelliJ data tool I use: ``` jdbc:calcite:shemaType=CUSTOM

with no use/password specified.

What am I doing wrong and how do I even begin to troubleshoot this? 
The 2 jdbc connections work when tested separately - no issue there.