Multiple clustering keys

TokuDB allows a secondary key to be defined as a clustering key. This means that all of the columns in the table are clustered with the secondary key. This blueprint is for the parser support.

Clustering key grammar:
https://github.com/Tokutek/mysql56/commit/c054b455374c709eb3106fb3ae2a6db5ea03f6a6

Query planner changes have been moved to https://blueprints.launchpad.net/percona-server/+spec/multiple-clustering-keys-query-opt.

The parser has been extended to support:
CREATE TABLE ... ( ..., CLUSTERING KEY identifier (column list), ...
CREATE TABLE ... ( ..., UNIQUE CLUSTERING KEY identifier (column list), ...
CREATE TABLE ... ( ..., CLUSTERING UNIQUE KEY identifier (column list), ...
CREATE TABLE ... ( ..., CONSTRAINT identifier UNIQUE CLUSTERING KEY identifier (column list), ...
CREATE TABLE ... ( ..., CONSTRAINT identifier CLUSTERING UNIQUE KEY identifier (column list), ...

CREATE TABLE ... (... column type CLUSTERING [UNIQUE] [KEY], ...)
CREATE TABLE ... (... column type [UNIQUE] CLUSTERING [KEY], ...)

ALTER TABLE ..., ADD CLUSTERING INDEX identifier (column list), ...
ALTER TABLE ..., ADD UNIQUE CLUSTERING INDEX identifier (column list), ...
ALTER TABLE ..., ADD CLUSTERING UNIQUE INDEX identifier (column list), ...
ALTER TABLE ..., ADD CONSTRAINT identifier UNIQUE CLUSTERING INDEX identifier (column list), ...
ALTER TABLE ..., ADD CONSTRAINT identifier CLUSTERING UNIQUE INDEX identifier (column list), ...

CREATE CLUSTERING INDEX identifier ON ...