With UCA-14.0.0 support working across multiple character sets, it becomes programaticly onerious to pull the character set out of of a column and explicitly list all columns.

A more convenient user experience is to, with an explict collation specified, to convert all the textual columns that can be converted to the collation, while keeping the same character set of the original columns.

This syntax drops the need for an explict charset to be specified as that would change existing columns which may not be desired.