rdblue commented on code in PR #16936:
URL: https://github.com/apache/iceberg/pull/16936#discussion_r4223566423
##########
core/src/test/java/org/apache/iceberg/TestTrackedFileAdapters.java:
##########
@@ -671,6 +733,312 @@ void unknownSpecIdThrows() {
.hasMessageContaining("Cannot find partition spec for spec ID");
}
+ @Test
+ void dataTrackedFileAdapterFromDataFile() {
+ TrackedFile result =
TrackedFileAdapters.forDataFile(TABLE_SCHEMA).wrap(DATA_FILE);
+
+ assertThat(result.tracking()).isNull();
+ assertWrappedDataFileMatchesFileFields(result, DATA_FILE);
+ assertThatThrownBy(result::formatVersion)
+ .isInstanceOf(IllegalStateException.class)
+ .hasMessage("Format version is assigned at write time");
+ }
+
+ @Test
+ void dataTrackedFileAdapterFromExistingManifestEntry() {
+ TrackedFile result =
+ TrackedFileAdapters.forDataFile(TABLE_SCHEMA)
+ .wrap(
+ newEntry()
+ .wrapExisting(
+ SNAPSHOT_ID, DATA_SEQUENCE_NUMBER,
FILE_SEQUENCE_NUMBER, DATA_FILE));
+
+ assertThat(result.tracking().status()).isEqualTo(EntryStatus.EXISTING);
+ assertThat(result.tracking().snapshotId()).isEqualTo(SNAPSHOT_ID);
+
assertThat(result.tracking().dataSequenceNumber()).isEqualTo(DATA_SEQUENCE_NUMBER);
+
assertThat(result.tracking().fileSequenceNumber()).isEqualTo(FILE_SEQUENCE_NUMBER);
+ assertThat(result.tracking().firstRowId()).isEqualTo(FIRST_ROW_ID);
+ assertManifestPosition(result.tracking(), DATA_FILE);
+ assertWrappedDataFileMatchesFileFields(result, DATA_FILE);
+ }
+
+ @Test
+ void dataTrackedFileAdapterReuse() {
+ TrackedFileAdapters.DataTrackedFile adapter =
TrackedFileAdapters.forDataFile(TABLE_SCHEMA);
+
+ adapter.wrap(DATA_FILE);
+ assertWrappedDataFileMatchesFileFields(adapter, DATA_FILE);
+ assertThat(adapter.tracking()).isNull();
+
+ DataFile file2 =
+ new GenericDataFile(
+ UNPARTITIONED_SPEC.specId(),
+ "s3://bucket/data/file2.parquet",
+ FileFormat.PARQUET,
+ PartitionData.EMPTY,
+ 2048L,
+ new Metrics(200L, null, null, null, null),
+ null,
+ ImmutableList.of(0L),
+ null,
+ null);
+ assignManifestPosition(file2, "s3://bucket/table/manifest-2.parquet", 8L);
+ adapter.wrap(
+ newEntry().wrapExisting(SNAPSHOT_ID, DATA_SEQUENCE_NUMBER,
FILE_SEQUENCE_NUMBER, file2));
+ assertWrappedDataFileMatchesFileFields(adapter, file2);
+ assertThat(adapter.tracking().status()).isEqualTo(EntryStatus.EXISTING);
+ assertManifestPosition(adapter.tracking(), file2);
+ }
+
+ @Test
+ void dataTrackedFileAdapterRejectsNullFile() {
+ TrackedFileAdapters.DataTrackedFile adapter =
TrackedFileAdapters.forDataFile(TABLE_SCHEMA);
+ assertThatThrownBy(() -> adapter.wrap((DataFile) null))
+ .isInstanceOf(IllegalArgumentException.class)
+ .hasMessageContaining("Invalid file: null");
+ }
+
+ @Test
+ void dataTrackedFileAdapterContentStats() {
+ TrackedFile result =
TrackedFileAdapters.forDataFile(TABLE_SCHEMA).wrap(DATA_FILE_WITH_METRICS);
+
+ ContentStats stats = result.contentStats();
+ assertThat(stats).isNotNull();
+
assertThat(stats.fieldStats()).extracting(FieldStats::fieldId).containsExactlyInAnyOrder(1,
2);
+
+ FieldStats<?> idStats = stats.statsFor(1);
+ assertThat(idStats.valueCount()).isEqualTo(100L);
+ assertThat(idStats.lowerBound()).isEqualTo(1);
+ assertThat(idStats.upperBound()).isEqualTo(1000);
+ }
+
+ @Test
+ void dataTrackedFileAdapterWithoutMetricsHasNoContentStats() {
+ TrackedFile result =
TrackedFileAdapters.forDataFile(TABLE_SCHEMA).wrap(DATA_FILE);
+
+ assertThat(result.contentStats()).isNull();
+ }
+
+ @Test
+ void dataTrackedFileAdapterKeepsPartitionTuple() {
+ DataFile partitioned =
+ new GenericDataFile(
+ PARTITIONED_SPEC.specId(),
+ DATA_FILE_LOCATION,
+ FileFormat.PARQUET,
+ PARTITION,
+ 1024L,
+ new Metrics(100L, null, null, null, null),
+ null,
+ ImmutableList.of(0L),
+ null,
+ null);
+ TrackedFile result =
TrackedFileAdapters.forDataFile(TABLE_SCHEMA).wrap(partitioned);
+
+ assertThat(result.partition())
+ .usingComparator(Comparators.forType(PARTITIONED_SPEC.partitionType()))
+ .isEqualTo(PARTITION);
+ assertThat(TrackedFileAdapters.asDataFile(result,
specsById(PARTITIONED_SPEC)).partition())
Review Comment:
What is this assertion checking? `asDataFile` should return the original
`DataFile`, which I would expect to be validated by a test like the ones below
for the tracked file adapter unwrapping.
I think this assertion should be dropped since it is either testing
double-wrapping (which we don't want) or it's testing that the original is
returned by a different method (`asDataFile`) and neither of those are directly
related to `DataTrackedFile` behavior for `partition()`.
--
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]