Column types in DataCube are kept exactly as reported by the engine, including
precise primitives (e.g. meta::pure::precisePrimitives::Varchar with type
variable values [200]). Rather than rewriting them into their base primitive,
code should ask this registry what a type can do (is it numeric? an integer?
which literal type should be used for its values?).
Types DataCube can display and group by, but has no way to filter or aggregate,
e.g. StrictTime (a time of day, which is neither a date nor a date-time) and Byte
Column types in DataCube are kept exactly as reported by the engine, including precise primitives (e.g.
meta::pure::precisePrimitives::Varcharwith type variable values[200]). Rather than rewriting them into their base primitive, code should ask this registry what a type can do (is it numeric? an integer? which literal type should be used for its values?).See https://github.com/finos/legend-pure/blob/master/legend-pure-core/legend-pure-m3-precisePrimitives/src/main/resources/platform_precise_primitives/precisePrimitives.pure