pub fn execute_columnar(
rows: &[Row],
filter: Option<&Expression>,
aggregates: &[Expression],
schema: &CombinedSchema,
) -> Option<Result<Vec<Row>, ExecutorError>>Expand description
Execute a query using columnar processing (AST-based interface)
This is the entry point for columnar execution that accepts AST expressions and converts them to the columnar execution pipeline.
§Arguments
rows- The rows to processfilter- Optional WHERE clause expressionaggregates- SELECT list aggregate expressionsschema- Schema for resolving column names to indices
§Returns
Some(Result) if the query can be optimized using columnar execution, None if the expressions are too complex for columnar optimization.
Note: This function uses LLVM auto-vectorization for vectorized execution.