Creates a single node Context

By default any TIMESTAMP literal expression is handled as TIMESTAMP_NANOSECONDS Using the `Reinterpret` clause we can get the right TIMESTAMP unit usi

* Set the MetadataDictionary */

Shutdown memory manager.

* Returns a function to use in pseudos for positionals * @param {Function} fn

Use get_projections and if there are no projections or expression is empty then returns a filled array with the sequence of all columns (0, 1, ..., n)

input: min_keys=[MIN($0) OVER (PARTITION BY $2 ORDER BY $3, $1 DESC)], n_nationkey=[$0] output: < [3, 1], [cudf::ASCENDING, cudf::DESCENDING], [cudf::

input: min_keys=[MIN($0) OVER (PARTITION BY $1, $2 ORDER BY $3)], n_nationkey=[$0] output: < [1, 2], [cudf::ASCENDING, cudf::ASCENDING], [cudf::null_o

For this function `indices` vector starts empty inputs: expression: "$0, $1, $3" indices: [] output: "+(+(CAST($0):INTEGER, CAST($1):INTEGER), CAST($3

Calcite translates the `col_1 IS NOT DISTINCT FROM col_3` subquery to the OR(AND(IS NULL($1), IS NULL($3)), IS TRUE(=($1 , $3))) expression. So let's

Feeds an input cache

input: -($3) output: -(0, $3)

input: min_keys=[MIN($0) OVER (PARTITION BY $1, $2 ORDER BY $0)] output: PARTITION BY $1, $2 ORDER BY $0