Github user yhuai commented on a diff in the pull request:
https://github.com/apache/spark/pull/4563#discussion_r24617792
--- Diff:
sql/core/src/main/scala/org/apache/spark/sql/parquet/newParquet.scala ---
@@ -187,31 +190,49 @@ case class ParquetRelation2
@transient private val fs =
FileSystem.get(sparkContext.hadoopConfiguration)
private class MetadataCache {
+ // `FileStatus` objects of all "_metadata" files.
private var metadataStatuses: Array[FileStatus] = _
+
+ // `FileStatus` objects of all "_common_metadata" files.
private var commonMetadataStatuses: Array[FileStatus] = _
+
+ // Parquet footer cache.
private var footers: Map[FileStatus, Footer] = _
- private var parquetSchema: StructType = _
+ // `FileStatus` objects of all data files (Parquet part-files).
var dataStatuses: Array[FileStatus] = _
+
+ // Partition spec of this table, including names, data types, and
values of each partition
+ // column, and paths of each partition.
var partitionSpec: PartitionSpec = _
+
+ // Schema of the actual Parquet files, without partition columns
discovered from partition
+ // directory paths.
+ var parquetSchema: StructType = _
+
+ // Schema of the whole table, including partition columns.
var schema: StructType = _
- var dataSchemaIncludesPartitionKeys: Boolean = _
+ // Indicates whether partition columns are also included in Parquet
data file schema. If not,
+ // we need to fill in partition column values into read rows when
scanning the table.
+ var partitionKeysIncludedInParquetSchema: Boolean = _
+
+ /**
+ * Refreshes `FileStatus`es, footers, partition spec, and table schema.
+ */
def refresh(): Unit = {
- val baseStatuses = {
- val statuses = paths.distinct.map(p =>
fs.getFileStatus(fs.makeQualified(new Path(p))))
- // Support either reading a collection of raw Parquet part-files,
or a collection of folders
- // containing Parquet files (e.g. partitioned Parquet table).
- assert(statuses.forall(!_.isDir) || statuses.forall(_.isDir))
- statuses.toArray
- }
+ // Support either reading a collection of raw Parquet part-files, or
a collection of folders
+ // containing Parquet files (e.g. partitioned Parquet table).
+ val baseStatuses = paths.distinct.map { p =>
+ fs.getFileStatus(fs.makeQualified(new Path(p)))
+ }.toArray
+ assert(baseStatuses.forall(!_.isDir) || baseStatuses.forall(_.isDir))
val leaves = baseStatuses.flatMap { f =>
--- End diff --
Maybe `leafSummaryFiles?`
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
