RussellSpitzer commented on code in PR #9803:
URL: https://github.com/apache/iceberg/pull/9803#discussion_r1508073261
##########
spark/v3.5/spark/src/test/java/org/apache/iceberg/spark/actions/TestRewriteDataFilesAction.java:
##########
@@ -1463,6 +1465,148 @@ public void testSnapshotProperty() {
assertThat(table.currentSnapshot().summary()).containsKeys(commitMetricsKeys);
}
+ @Test
+ public void testBinPackRewriterWithSpecificUnparitionedOutputSpec() {
+ Table table = createTable(10);
+ shouldHaveFiles(table, 10);
+ // to be used for rewrite
+ int outputSpecId = table.spec().specId();
+ // create multiple partition specs with different commit
+ table.updateSpec().addField(Expressions.truncate("c2", 2)).commit();
+
+ long dataSizeBefore = testDataSize(table);
+ long count = currentData().size();
+
+ RewriteDataFiles.Result result =
+ basicRewrite(table)
+ .option(RewriteDataFiles.OUTPUT_SPEC_ID,
String.valueOf(outputSpecId))
+ .option(SizeBasedFileRewriter.REWRITE_ALL, "true")
+ .binPack()
+ .execute();
+
+ assertThat(result.rewrittenBytesCount()).isEqualTo(dataSizeBefore);
+ assertThat(currentData().size()).isEqualTo(count);
+ shouldRewriteDataFilesWithPartitionSpec(table, outputSpecId);
+ }
+
+ @Test
+ public void testBinPackRewriterWithSpecificOutputSpec() {
+ Table table = createTable(10);
+ shouldHaveFiles(table, 10);
+ // to be used for rewrite back to original un-partitioned spec
+ int outputSpecId = table.spec().specId();
+ // create multiple partition specs with different commit
+ table.updateSpec().addField(Expressions.truncate("c2", 2)).commit();
+ table.updateSpec().addField(Expressions.bucket("c3", 2)).commit();
+
+ long dataSizeBefore = testDataSize(table);
+ long count = currentData().size();
+
+ RewriteDataFiles.Result result =
+ basicRewrite(table)
+ .option(RewriteDataFiles.OUTPUT_SPEC_ID,
String.valueOf(outputSpecId))
+ .option(SizeBasedFileRewriter.REWRITE_ALL, "true")
+ .binPack()
+ .execute();
+
+ assertThat(result.rewrittenBytesCount()).isEqualTo(dataSizeBefore);
+ assertThat(currentData().size()).isEqualTo(count);
+ shouldRewriteDataFilesWithPartitionSpec(table, outputSpecId);
+ }
+
+ @Test
+ public void testBinpackRewriteWithInvalidOutputSpecId() {
+ Table table = createTable(10);
+ shouldHaveFiles(table, 10);
+ int previousSpecId = table.spec().specId();
+ // simulate multiple partition specs with different commit
+ table.updateSpec().addField(Expressions.truncate("c2", 2)).commit();
+ table.updateSpec().addField(Expressions.bucket("c3", 2)).commit();
+ Assertions.assertThatThrownBy(
+ () ->
+ actions()
+ .rewriteDataFiles(table)
+ .option(RewriteDataFiles.OUTPUT_SPEC_ID,
String.valueOf(1234))
+ .binPack()
+ .execute())
+ .isInstanceOf(IllegalArgumentException.class)
+ .hasMessage(
+ "Cannot use output spec id 1234 because the table does not contain
a reference to this spec-id.");
+ }
+
+ @Test
+ public void testSortRewriterWithSpecificOutputSpecId() {
+ Table table = createTable(10);
+ shouldHaveFiles(table, 10);
+ Integer previousSpecId = table.spec().specId();
+ // simulate multiple partition specs with different commit
+ table.updateSpec().addField(Expressions.truncate("c2", 2)).commit();
+ // to be used for rewrite
+ int outputSpecId = table.spec().specId();
+ table.updateSpec().addField(Expressions.bucket("c3", 2)).commit();
+
+ long dataSizeBefore = testDataSize(table);
+ long count = currentData().size();
+
+ RewriteDataFiles.Result result =
+ basicRewrite(table)
+ .option(RewriteDataFiles.OUTPUT_SPEC_ID,
String.valueOf(outputSpecId))
+ .option(SizeBasedFileRewriter.REWRITE_ALL, "true")
+
.sort(SortOrder.builderFor(table.schema()).asc("c2").asc("c3").build())
+ .execute();
+
+ assertThat(result.rewrittenBytesCount()).isEqualTo(dataSizeBefore);
+ assertThat(currentData().size()).isEqualTo(count);
+ shouldRewriteDataFilesWithPartitionSpec(table, outputSpecId);
+ }
+
+ @Test
+ public void testZOrderRewriteWithSpecificOutputSpecId() {
+ Table table = createTable(10);
+ shouldHaveFiles(table, 10);
+ Integer previousSpecId = table.spec().specId();
+ // simulate multiple partition specs with different commit
+ table.updateSpec().addField(Expressions.truncate("c2", 2)).commit();
+ // to be used for rewrite
+ int outputSpecId = table.spec().specId();
+ table.updateSpec().addField(Expressions.bucket("c3", 2)).commit();
+
+ long dataSizeBefore = testDataSize(table);
+ long count = currentData().size();
+
+ RewriteDataFiles.Result result =
+ basicRewrite(table)
+ .option(RewriteDataFiles.OUTPUT_SPEC_ID,
String.valueOf(outputSpecId))
+ .option(SizeBasedFileRewriter.REWRITE_ALL, "true")
+ .zOrder("c2", "c3")
+ .execute();
+
+ assertThat(result.rewrittenBytesCount()).isEqualTo(dataSizeBefore);
+ assertThat(currentData().size()).isEqualTo(count);
+ shouldRewriteDataFilesWithPartitionSpec(table, outputSpecId);
+ }
+
+ protected void shouldRewriteDataFilesWithPartitionSpec(Table table, int
outputSpecId) {
+ List<DataFile> rewrittenFiles = currentDataFiles(table);
+ // check specId matches outputSpecId
+ rewrittenFiles.stream()
+ .map(DataFile::specId)
+ .forEach(element ->
assertThat(element.intValue()).isEqualTo(outputSpecId));
+ // check partitionType matches partitionType for the output-spec-id
+ rewrittenFiles.stream()
Review Comment:
Rather than having this comment, you could have the code be
```java
foreach( fileSpecID -> fileSpecId.intValue()).isequaltTo....)
```
Especially since the comment is a little confusing
You also could change this to a "AllMatch" checker
```java
assertThat(rewrittenFiles).allMatch(file ->
file.specId.intValue().isEqualTo(outputSpecId)
```
Which I think is pretty clear as to what we are checking without a comment
--
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]
