abstract class PartitioningAwareFileIndex extends FileIndex with Logging
An abstract class that represents FileIndexs that are aware of partitioned tables. It provides the necessary methods to parse partition data based on a set of files.
- Alphabetic
- By Inheritance
- PartitioningAwareFileIndex
- Logging
- FileIndex
- AnyRef
- Any
- Hide All
- Show All
- Public
- Protected
Instance Constructors
- new PartitioningAwareFileIndex(sparkSession: SparkSession, parameters: Map[String, String], userSpecifiedSchema: Option[StructType], fileStatusCache: FileStatusCache = NoopCache)
- parameters
as set of options to control partition discovery
- userSpecifiedSchema
an optional user specified schema that will be use to provide types for the discovered partitions
Abstract Value Members
- abstract def partitionSpec(): PartitionSpec
Returns the specification of the partitions inferred from the data.
- abstract def refresh(): Unit
Refresh any cached file listings
Refresh any cached file listings
- Definition Classes
- FileIndex
- abstract def rootPaths: Seq[Path]
Returns the list of root input paths from which the catalog will get files.
Returns the list of root input paths from which the catalog will get files. There may be a single root path from which partitions are discovered, or individual partitions may be specified by each path.
- Definition Classes
- FileIndex
Concrete Value Members
- def allFiles(): Seq[FileStatus]
- def inputFiles: Array[String]
Returns the list of files that will be read when scanning this relation.
Returns the list of files that will be read when scanning this relation.
- Definition Classes
- PartitioningAwareFileIndex → FileIndex
- def listFiles(partitionFilters: Seq[Expression], dataFilters: Seq[Expression]): Seq[PartitionDirectory]
Returns all valid files grouped into partitions when the data is partitioned.
Returns all valid files grouped into partitions when the data is partitioned. If the data is unpartitioned, this will return a single partition with no partition values.
- partitionFilters
The filters used to prune which partitions are returned. These filters must only refer to partition columns and this method will only return files where these predicates are guaranteed to evaluate to
true. Thus, these filters will not need to be evaluated again on the returned data.- dataFilters
Filters that can be applied on non-partitioned columns. The implementation does not need to guarantee these filters are applied, i.e. the execution engine will ensure these filters are still applied on the returned files.
- Definition Classes
- PartitioningAwareFileIndex → FileIndex
- def metadataOpsTimeNs: Option[Long]
Returns an optional metadata operation time, in nanoseconds, for listing files.
Returns an optional metadata operation time, in nanoseconds, for listing files.
We do file listing in query optimization (in order to get the proper statistics) and we want to account for file listing time in physical execution (as metrics). To do that, we save the file listing time in some implementations and physical execution calls it in this method to update the metrics.
- Definition Classes
- FileIndex
- def partitionSchema: StructType
Schema of the partitioning columns, or the empty schema if the table is not partitioned.
Schema of the partitioning columns, or the empty schema if the table is not partitioned.
- Definition Classes
- PartitioningAwareFileIndex → FileIndex
- def sizeInBytes: Long
Sum of table file sizes, in bytes
Sum of table file sizes, in bytes
- Definition Classes
- PartitioningAwareFileIndex → FileIndex
- def toString(): String
- Definition Classes
- FileIndex → AnyRef → Any