feat: Add time extension function for temporal expressions in Kotlin and Java

[joc-a]

• Added time extension functions for temporal expressions in Kotlin and Java.
• Modified Time function to make it work for all database dialects.
• Modified JodaTimeMiscTableTest and JodaTimeDefaultsTest to extend DatabaseTestsBase instead of JodaTimeBaseTest to match the way it is in the Kotlin and Java tests.
• H2 V1 is excluded from the tests because of a bug in the driver that messes up the fractional seconds.

I'm planning to add more tests to cover more types in later pull requests.

[obabichevjb]

I see that it's duplicated in other datetime packages, and I believe that it works, but is it possible that containsDatePart is false here?

I wanted to write that if we use method inside DateColumnType class, and this method takes date: String argument, it must have date part, but now I see.

I don't expect that it's the problem of this PR, but it might be a problem in the long term. We have a DateColumnType class that is expected to be used with exact values from the database (actually also different values from different dialects, but it's another problem), but after that we query something like SUBSTRING(CONVERT(NVARCHAR, TestTable."timestamptz-column", 121), 12, 15) and expect that the column type should parse it also. And the problem is that we actually cannot say which values we expect from the database, we can rely only on what's covered by tests.

But I don't know is it easy to change with the current approach. Particularly for the time function, if the final goal is to return user the date from the whole timestamp, we can query the whole timestamp and extract time on the exposed side (instead of querying time only).

Anyway, it's just some thoughts about the potential problems with maintaining our datetime classes.

[bog-walk]

Left small comments in the test files.