Struct U16PartitionedTableWriter 

pub struct U16PartitionedTableWriter<PartitionWriter: TableWriter + Send> { /* private fields */ }

Wraps N TableWriter in such a way that they each write to base/0/x.parquet, …, base/N-1/x.parquet instead of base/x.parquet.

This allows Hive partitioning while writing with multiple threads (x is the thread id in the example above).

If num_partitions is None, disables partitioning.

Implementations

impl<PartitionWriter: TableWriter + Send> U16PartitionedTableWriter<PartitionWriter>

pub fn partitions(&mut self) -> &mut [PartitionWriter]

Trait Implementations

impl<PartitionWriter: TableWriter + Send> TableWriter for U16PartitionedTableWriter<PartitionWriter>

type Schema = (String, Option<NonZero<u16>>, <PartitionWriter as TableWriter>::Schema)

(partition_column, num_partitions, underlying_schema)

type CloseResult = Vec<<PartitionWriter as TableWriter>::CloseResult>

type Config = <PartitionWriter as TableWriter>::Config

fn new( path: PathBuf, (partition_column, num_partitions, schema): Self::Schema, config: Self::Config, ) -> Result<Self>

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

Calls .into() on the internal builder, and writes its result to disk.

fn close(self) -> Result<Self::CloseResult>