We'll introduce a new table INFORMATION_SCHENA.COLLATION_PROPERTIES with
the following structure:

CREATE TABLE INFORMATION_SCHEMA.COLLATION_PROPERTIES

{

  ID INT NOT NULL,

  TAILORING TEXT CHARACTER SET utf8,

  USE_COUNT BIGINT NOT NULL

};

Column description

• The column TAILORING will return collation definition rules for UCA based
  collations (as defined in strings/ctype-uca.c or in Index.xml) and NULL for non-UCA collations. For example:

  MariaDB [test]> SELECT TAILORING FROM INFORMATION_SCHEMA.COLLATIONS WHERE ID=202;

  +-----------------------------------------------------+

  | TAILORING |

  +-----------------------------------------------------+

  | & C < \u010D <<< \u010C ... & Z < \u017E <<< \u017D |

  +-----------------------------------------------------+

• The column USE_COUNT will contain collation usage statistics,
  similar to what was previously added into INFORMATION_SCHEMA.PLUGINS
  under terms of MDEV-6274. Unlike I_S.PLUGINS (which exists only if the
  feedback plugins is loaded into the server), the new table
  INFORMATION_SCHEMA.COLLATION_PROPERTIES will always
  exist, thus the column USE_COUNT will give an alternative always available
  access to the collation usage statistics.

• In the future there will be more fields for other properties,
  like case sensitivity, accent sensitivity, Kana sensitivity,
  Unicode version, number of levels for comparison and sort, etc.