[question] reuse subtree for follow-up query of same user

[hpvd]

Reuse common expressions in a query #14196 introduces reuse of subtrees within same query.

When using Pinot for user facing analytics where

• users slice and dice their data and
• often a user starts relatively broad (e.g. dashboard standard views) and narrow the results further with multiple steps:

Example:
Simple analyse of website usage
(e.g. number of visitors
or time on page
or number of subpages visited...)
1. start by week,
2. narrow to days,
3. narrow origin country
4. narrow to mobile

• with freshness requirements of minutes (for this part of pinot usage..)

=> Questions:
A) what are the current possibilities to reuse the result subtree of 1. in 2., of 2. in 3. ... instead of always do an possibly expensive full query?
B) or would this kind of load reduction not look like to be worth it (opinion)?

inspired by
@gortiz in Reuse common expressions in a query #14196

and
Splitting Queries Across Data Warehouses
https://cotera.co/blog/how-era-brings-last-mile-analytics-to-any-data-warehouse-via-duckdb

[hpvd]

C) Is this /can this be maybe some how related to pagination support as described here by @siddharthteotia
#5246 (comment)

where the first query may fetch all results and "deliver" only the first part, but the follow-up queries p2, p3.. reuse the full results and deliver the requested part?