Struct DistinctResult 

pub struct DistinctResult { /* private fields */ }

Streaming distinct result that removes duplicate rows on-the-fly

This streams rows and only stores seen row values for deduplication. This enables:

• Early termination with LIMIT (no need to scan all rows)
• Lower latency to first row
• Streaming output

Implementations

impl DistinctResult

pub fn new(inner: Box<dyn QueryResult>) -> Self

Create a new streaming distinct result

pub fn with_column_count( inner: Box<dyn QueryResult>, distinct_columns: Option<usize>, ) -> Self

Create a distinct result that only considers the first distinct_columns columns for uniqueness comparison. This is used when ORDER BY references columns not in SELECT.

For example: SELECT DISTINCT a FROM t ORDER BY b

• The result has columns [a, b] for sorting
• But distinctness should only compare column a

Trait Implementations

impl QueryResult for DistinctResult

fn columns(&self) -> &[String]

Returns the column names in the result

fn next(&mut self) -> bool

Moves the cursor to the next row

fn scan(&self, dest: &mut [Value]) -> Result<()>

Scans the current row into the provided values

fn row(&self) -> &Row

Returns the current row directly without copying

fn take_row(&mut self) -> Row

Takes ownership of the current row (avoids clone)

fn close(&mut self) -> Result<()>

Closes the result set and releases resources

fn rows_affected(&self) -> i64

Returns the number of rows affected by an INSERT, UPDATE, or DELETE

fn last_insert_id(&self) -> i64

Returns the last inserted ID for an INSERT operation

fn last_error(&mut self) -> Option<Error>

Returns a pending error from the last next() call, if any.

fn with_aliases( self: Box<Self>, aliases: FxHashMap<String, String>, ) -> Box<dyn QueryResult>

Sets column aliases for this result

fn columns_arc(&self) -> Option<CompactArc<Vec<String>>>

Returns column names as Arc for zero-copy sharing

fn try_into_arc_rows(&mut self) -> Option<CompactArc<Vec<Row>>>

Try to extract all rows as CompactArc<Vec<Row>> for zero-copy joins

fn estimated_count(&self) -> Option<usize>

Returns an estimate of the total number of rows in the result.

fn supports_typed_batches(&self) -> bool

Whether the result can yield decoded typed column batches without constructing a row/value object for every result record.

fn typed_batch_fallback_reason(&self) -> Option<TypedBatchFallbackReason>

If supports_typed_batches() is false, return the best semantic reason.

fn next_typed_batch(&mut self) -> Result<Option<TypedColumnBatch>, Error>

Advance to the next typed column batch.