Textual representation of single value ranges for a particular column, consisting of a list of String
s.
Textual representation of single value ranges for a particular column, consisting of a list of String
s.
First element denotes the type of range, which can currently be of two types:
A ValueRanges object is a sequence of ValueRange elements and it is associated with a certain column.
A ValueRanges object is a sequence of ValueRange elements and it is associated with a certain column. As such, all string bounds of all the ValueRanges are expected to be literals of that column's VectorDataType.
The interpretation is that the individual ranges are union-ed (or-ed) together.
Filters the given df
according to any ValueRanges that its columns
might have.
Creates Some(Column)
that represents the condition ("predicate"), in Spark's terms,
corresponding to the given columnMetadata
, if the latter has some ValueRanges specified
and is of a type for which Predicate Pushdown is safe.
Creates Some(Column)
that represents the condition ("predicate"), in Spark's terms,
corresponding to the given columnMetadata
, if the latter has some ValueRanges specified
and is of a type for which Predicate Pushdown is safe. Otherwise returns None
.
typeIsSafeForPredPD
Creates a Column
that represents the condition ("predicate"), in Spark's terms, that is expressed by the given range
Says whether or not Predicate Pushdown is supported on columns of the given vectorTypeName
.
Says whether or not Predicate Pushdown is supported on columns of the given vectorDataType
.
Converts the given vectorLiteral
to the corresponding Spark literal, according to the specified VectorDataType
.
Object that brings together a set of functions that are used for implementing "Predicate Pushdown": A Vector query such as "SELECT * from external_table where col = 10" should not result in the provider sending the whole table to Vector, but only those tuples that pass the "col = 10" predicate.