We planned to remove the statistical information when reading the file, but this optimization was hindered by the need for statistical information later for whole bucket filter.
In this PR, we refactor this logical to drop stats and compute filter selected result, in this way, we can keep whole bucket filter.
Purpose
We planned to remove the statistical information when reading the file, but this optimization was hindered by the need for statistical information later for whole bucket filter.
In this PR, we refactor this logical to drop stats and compute filter selected result, in this way, we can keep whole bucket filter.
Tests
API and Format
Documentation