Rename and possibly refactor the inner types of IterByColRange

pub enum IterByColRange<'a, R: RangeBounds<AlgebraicValue>> {
    /// When the column in question does not have an index.
    Scan(ScanIterByColRange<'a, R>),

    /// When the column has an index, and the table
    /// has been modified this transaction.
    Index(IndexSeekIterInner<'a>),

    /// When the column has an index, and the table
    /// has not been modified in this transaction.
    CommittedIndex(CommittedIndexIter<'a>),
}

The types here could have better names. Also does it need to be explicitly parameterized with a range or should it agnostic to the type of index?

Search the codebase for IterByColRange, ScanIterByColRange, IndexSeekIterInner, and CommittedIndexIter, then inspect how the R: RangeBounds parameter is used. Determine whether the names and range parameter should change, and consider the impact on each enum variant. Done means the naming and parameterization decision is agreed and consistently applied.