Prevent UnionArray with Repeated Type IDs

[tustvold]

Is your feature request related to a problem or challenge? Please describe what you are trying to do.

Currently DataType::Union has no protection against the same type ID appearing multiple times in its DataType.

Describe the solution you'd like

We should probably validate that the type_ids are unique, and prevent constructing an invalid DataType

Describe alternatives you've considered

We could not do this, much like currently permit invalid DataType::Decimal #2362

Additional context

[Weijun-H]

Could I pick this ticket up?

[tustvold]

Sure, as the maximum id is 127, you can probably just use an u128 bitset to check for duplicates within UnionFields

[Weijun-H]

Should we guarantee that typeid is unsigned? The current implementation does not guarantee it.

[tustvold]

Should we guarantee that typeid is unsigned

The arrow format specifies the following - https://arrow.apache.org/docs/format/Columnar.html#union-layout

A buffer of 8-bit signed integers. Each type in the union has a corresponding type id whose values are found in this buffer. A union with more than 127 possible types can be modeled as a union of unions.

So they should be signed integers, but never negative 😅. Most likely this was originally a workaround for Java

[tustvold]

label_issue.py automatically added labels {'arrow'} from #4070

[tustvold]

label_issue.py automatically added labels {'parquet'} from #3981