KSQL doesn't prevent/warn column aliases and object names that are function names

[rmoff]

If you use a function name (built in or UDF) as an CSAS/CTAS object name, and/or column alias, KSQL doesn't stop or warn you.

Whilst this is possible, it's eminently silly and we should save users from doing stupid things by mistake:

ksql> CREATE TABLE COUNT AS SELECT EVENTID, COUNT(*) AS COUNT FROM CARSENSOR GROUP BY EVENTID;

 Message
---------------------------
 Table created and running
---------------------------
ksql> DESCRIBE COUNT;

Name                 : COUNT
 Field   | Type
-------------------------------------
 ROWTIME | BIGINT           (system)
 ROWKEY  | VARCHAR(STRING)  (system)
 EVENTID | INTEGER
 COUNT   | BIGINT
-------------------------------------
For runtime statistics and query details run: DESCRIBE EXTENDED <Stream,Table>;
ksql>
ksql> SELECT COUNT FROM COUNT;
173
^CQuery terminated
ksql>

It's also inconsistent with how other reserved words are handled, where an error (albeit an unhelpful one) is thrown:

ksql> CREATE TABLE COUNT AS SELECT EVENTID, COUNT(*) AS CREATE FROM CARSENSOR GROUP BY EVENTID;
line 1:51: mismatched input 'CREATE' expecting {'ADD', 'APPROXIMATE', 'AT', 'CONFIDENCE', 'NO', 'SUBSTRING', 'POSITION', 'TINYINT', 'SMALLINT', 'INTEGER', 'DATE', 'TIME', 'TIMESTAMP', 'INTERVAL', 'YEAR', 'MONTH', 'DAY', 'HOUR', 'MINUTE', 'SECOND', 'ZONE', 'OVER', 'PARTITION', 'RANGE', 'ROWS', 'PRECEDING', 'FOLLOWING', 'CURRENT', 'ROW', 'STRUCT', 'VIEW', 'REPLACE', 'GRANT', 'REVOKE', 'PRIVILEGES', 'PUBLIC', 'OPTION', 'EXPLAIN', 'ANALYZE', 'FORMAT', 'TYPE', 'TEXT', 'GRAPHVIZ', 'LOGICAL', 'DISTRIBUTED', 'TRY', 'SHOW', 'TABLES', 'SCHEMAS', 'CATALOGS', 'COLUMNS', 'COLUMN', 'USE', 'PARTITIONS', 'FUNCTIONS', 'FUNCTION', 'TO', 'SYSTEM', 'BERNOULLI', 'POISSONIZED', 'TABLESAMPLE', 'RESCALED', 'ARRAY', 'MAP', 'SET', 'RESET', 'SESSION', 'DATA', 'START', 'TRANSACTION', 'COMMIT', 'ROLLBACK', 'WORK', 'ISOLATION', 'LEVEL', 'SERIALIZABLE', 'REPEATABLE', 'COMMITTED', 'UNCOMMITTED', 'READ', 'WRITE', 'ONLY', 'CALL', 'NFD', 'NFC', 'NFKD', 'NFKC', 'IF', 'NULLIF', 'COALESCE', IDENTIFIER, DIGIT_IDENTIFIER, QUOTED_IDENTIFIER, BACKQUOTED_IDENTIFIER}
Caused by: org.antlr.v4.runtime.InputMismatchException
ksql> CREATE TABLE CREATE AS SELECT EVENTID, COUNT(*) AS COUNT FROM CARSENSOR GROUP BY EVENTID;
line 1:14: no viable alternative at input 'CREATE TABLE CREATE'
Caused by: org.antlr.v4.runtime.NoViableAltException
ksql>

[agavra]

I'm not sure I understand why this is a bug @rmoff - I feel like I should be allowed to specify schemas that happen to share the name of a function (e.g. if my data has count as a field, why can't my schema without backquoting it? that seems artificially restricting our parser).

We can differentiate between function calls and identifiers because function calls have parens. This is similar to the concept of a "non-reserved" word in SQL standard (https://www.postgresql.org/docs/current/sql-keywords-appendix.html)

[rmoff]

OK, I'm convinced :)