Struct light_snowflake_connector::QueryResponse

pub struct QueryResponse { /* private fields */ }

The result of SQL that returns rows

The first partition is included immediately, but additional partitions are streamed lazily and incur additional IO.

You might consider using QueryResponse::only_partition if you only need one partition.

Implementations

impl QueryResponse

pub fn num_rows(&self) -> usize

Get the number of rows across all partitions

pub fn num_columns(&self) -> usize

Get the number of columns

pub fn num_partitions(&self) -> usize

Get the number of columns in the response

pub fn column_types(&self) -> &[ColumnType]

Column types in the result set

In most cases Cell should already expose the data you need, but if you use the raw strings or want information about nullability, etc, this can be useful.

pub fn only_partition(self) -> SnowflakeResult<Partition>

A convenience method to assert that there is only one partition and return it

This never causes IO, is not async, and can only error with SnowflakeError::MultiplePartitions

pub async fn partition( &self, index: usize ) -> SnowflakeResult<Option<Partition>>

Get a single partition from the response

If this is the first partition, you get it immediately, otherwise it will incur an additional request to get the partition

Returns an error if the requested partition does not exist.

pub fn partitions( &self ) -> impl TryStream<Ok = Partition, Error = SnowflakeError> + '_

Stream over all partitions in the response

This incurs IO, so try to only use this once.

In order to improve concurrency, this will buffer one partition, so you can have one partition in flight while processing another.

pub async fn concat_partitions(&self) -> SnowflakeResult<Partition>

Concatenate all partitions into a single partition

This incurs IO, so try to only use this once.

This could use an unbounded amount of memory, but it could save time for uses cases requiring multiple passes.

pub fn rows( &self ) -> impl TryStream<Ok = Vec<Cell>, Error = SnowflakeError> + '_

Stream over all rows in the response

This incurs IO, so try to only use this once.

In order to improve concurrency, this will buffer one partition, so you can have one partition in flight while processing another.

If you only need one partition, it may be simpler to use partition and then stream over the rows in that partition.

pub fn json_tables( &self ) -> impl TryStream<Ok = Vec<Value>, Error = SnowflakeError> + '_

Stream over all rows in the response as JSON tables

This incurs IO, so try to only use this once.

In order to improve concurrency, this will buffer one partition, so you can have one partition in flight while processing another.

pub fn json_objects( &self ) -> impl TryStream<Ok = Value, Error = SnowflakeError> + '_

Stream over all rows in the response as JSON objects

This incurs IO, so try to only use this once.

In order to improve concurrency, this will buffer one partition, so you can have one partition in flight while processing another.

Trait Implementations

impl Debug for QueryResponse

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.