Instead, the storage engine scans the intermediate index page and builds a list of the leaf pages that must be read.

• 相反,存储引擎将扫描中间索引页并生成必须要读取的叶级页的列表。