comphead commented on code in PR #7359:
URL: https://github.com/apache/datafusion-comet/pull/7359#discussion_r4485785776
##########
spark/src/test/scala/org/apache/comet/CometIcebergNativeSuite.scala:
##########
@@ -2241,6 +2241,67 @@ class CometIcebergNativeSuite
}
}
+ // No schema change, so the schema evolution fallback does not apply.
+ test("data file that omits an optional nested field stays native") {
+ assume(icebergAvailable, "Iceberg not available in classpath")
+
+ withHadoopCatalog("test_cat") {
+ withSQLConf(
+ CometConf.COMET_ENABLED.key -> "true",
+ CometConf.COMET_EXEC_ENABLED.key -> "true",
+ CometConf.COMET_ICEBERG_NATIVE_ENABLED.key -> "true") {
+ val table = "test_cat.db.omitted_nested"
+ spark.sql(s"""
+ CREATE TABLE $table (id INT, s STRUCT<a: INT, b: INT>, l
ARRAY<STRUCT<x: INT, y: INT>>)
+ USING iceberg
+ """)
+ withSQLConf(CometConf.COMET_ENABLED.key -> "false") {
+ spark.sql(s"""
+ INSERT INTO $table VALUES
+ (1, named_struct('a', 1, 'b', 10), array(named_struct('x', 1,
'y', 100)))
+ """)
+ // Set after the insert: accept-any-schema matches columns by name.
+ spark.sql(
+ s"ALTER TABLE $table SET TBLPROPERTIES
('write.spark.accept-any-schema' = 'true')")
+ // This file has no s.b and no l.element.y.
+ spark
+ .sql("""
+ SELECT 2 AS id, named_struct('a', 2) AS s,
array(named_struct('x', 20)) AS l
+ UNION ALL
+ SELECT 3, CAST(NULL AS STRUCT<a: INT>), CAST(NULL AS
ARRAY<STRUCT<x: INT>>)
+ """)
+ .writeTo(table)
+ .append()
+ }
+
+ // Check the footers, so the test fails if a writer starts filling in
the missing fields.
+ val fileLeaves = spark.sql(s"SELECT file_path FROM
$table.files").collect().map { row =>
+ val reader = org.apache.parquet.hadoop.ParquetFileReader.open(
+ org.apache.parquet.hadoop.util.HadoopInputFile.fromPath(
+ new org.apache.hadoop.fs.Path(row.getString(0)),
+ spark.sessionState.newHadoopConf()))
+ try {
+ reader.getFooter.getFileMetaData.getSchema.getColumns.asScala
+ .map(_.getPath.mkString("."))
+ .toSet
+ } finally {
+ reader.close()
+ }
Review Comment:
This is the third copy in this suite of opening a footer with
`ParquetFileReader.open(HadoopInputFile.fromPath(new Path(...),
spark.sessionState.newHadoopConf()))` and closing it in a `finally`, after the
page-index test around line 1024 and the INT96 test around line 4653. Each uses
the reader differently, so a small private `withParquetReader[T](path:
String)(f: ParquetFileReader => T): T` would cover all three and keep the fully
qualified names in one place. Would you add it here?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
