fix bugs caused by stitching empty and non-empty data.

[yu-shipit]

The current Transient row can only handle a schema (Long, Double). However, there are different schemas for histogram, Avg and so on. Create a NaNRowReader to handle any schemas.

Pull Request checklist

• [x] The commit(s) message(s) follows the contribution guidelines ?
• [x] Tests for the changes have been added (for bug fixes / features) ?
• [ ] Docs have been added / updated (for bug fixes / features) ?

Current behavior : (link exiting issues here : https://help.github.com/articles/basic-writing-and-formatting-syntax/#referencing-issues-and-pull-requests)

New behavior :

BREAKING CHANGES

If this PR contains a breaking change, please describe the impact and migration path for existing applications. If not please remove this section.

Breaking changes may include:

• Any schema changes to any Cassandra tables
• The serialized format for Dataset and Column (see .toString methods)
• Over the wire formats for Akka messages / case classes
• Changes to the HTTP public API
• Changes to query parsing / PromQL parsing

Other information:

[alextheimer]

@yu-shipit Could you also add a brief description to the PR header?