Maybe rename DataType::Utf8 to a more reasonable name?

[leiysky]

Seems this name came from LargeUtf8 in arrow.

It looks a little bit weird since in databases like Postgres or MySQL collation and charset are seperated from types. In this case, CHAR(string) is a data type and utf8 is a charset.

We can only support utf8 charset for now, but IMO maybe it's more reasonable to rename DataType::Utf8 to String or StringUtf8?

[leiysky]

cc @sundy-li

[sundy-li]

Maybe remove DataType::Utf8 and let's keep DataType::Binary as DataType::String, this could support more charset datatypes.

[leiysky]

@sundy-li It seems there has been many macros and utility of both DataType::Utf8 and DataType::Binary, do you have any idea about how to deal with them?

[sundy-li]

Let's do it in #1603

[leiysky]

Duplicated with #1603