Drop support for v0 and v1 blocks

[joe-elliott]

What this PR does:
Drops support for v0 and v1 blocks and consolidates all code into v2.

TODO

• Investigate Performance regression indicated by benchmarks
• Figure out what to do with buffers

Checklist

• Tests updated
• Documentation added
• CHANGELOG.md updated - the order of entries should be [CHANGE], [FEATURE], [ENHANCEMENT], [BUGFIX]

[joe-elliott]

This is ready to go, but leaving draft status up b/c I want to merge #806 first. There will be some small conflicts.

[annanay25]

LGTM, just a minor non-blocking nit

[annanay25]

append() will copy over all the items from old to new array right? and the resultant pagesBuffer array might have a cap larger than len(compressedPages), worth just creating a new array with cap(pagesBuffer) = len(compressedPages) ? The reuse is primarily useful when cap(pagesBuffer) > len(compressedPages) I guess

[joe-elliott]

and the resultant pagesBuffer array might have a cap larger than len(compressedPages)

I believe after this line len(pagesBuffer) = cap(pagesBuffer) = len(compressedPages)

worth just creating a new array with cap(pagesBuffer) = len(compressedPages)

by using whatever currently exists in pagesBuffer we reduce byte slice allocations. after this line the first N entries in pagesBuffer already have allocated byte slices that can be reused in ReadAllWIthBuffer. if we recreated the slice we would lose that.