MINOR: Remove Variant type id gaps

[gene-db]

Rationale for this change

There were some unnecessary gaps in the Variant type id numbering. This removes the id number gaps.

Also, some trivial spacing/formatting changes.

What changes are included in this PR?

Updates to the Variant binary encoding spec.

Do these changes have PoC implementations?

n/a

[gene-db]

@aihuaxu Could you please take a look? I just wanted to make the type ids consecutive and remove the unnecessary gap. Thanks!

[aihuaxu]

@gene-db Initially we had interval types and that's why I leave the gap there.

Thanks for fixing that.

[rdblue]

Do we need to close this gap? I noticed it as well, but I thought that it was reasonable to reserve these for potential future interval types.

[aihuaxu]

Do we need to close this gap? I noticed it as well, but I thought that it was reasonable to reserve these for potential future interval types.

Yeah. Initially it was there because we had interval types in Spark variant spec. I saw they were removed from Spark spec, assuming they are not getting used at all and Iceberg/Spark will use new typeID if we want to introduce them back. @gene-db and @rdblue

[gene-db]

Yeah, there is no reason for the gap. Spark had the intervals ids momentarily, but they were soon removed, and there were no releases that actually had the interval type ids. Once (or if) we want to add intervals, we can add them with new ids.

[rdblue]

Spark had the intervals ids momentarily, but they were soon removed, and there were no releases that actually had the interval type ids

Sounds good to me. Thanks!