When there is a chain of hash joins, it's often worthwhile to create Bloom filters and push them to the earliest possible point in the chain of joins to minimize number of materialized rows.

Reporter: Sasha Krassovsky / @save-buffer
Assignee: Sasha Krassovsky / @save-buffer

Related issues:

• [C++][Compute] Introduce Bloom filters to hash join (is blocked by)
• [C++] Query engine umbrella issue (is a child of)

PRs and other links:

• GitHub Pull Request #12289

_{Note: This issue was originally created as ARROW-15498. Please see the migration documentation for further details.}