dongjoon-hyun commented on code in PR #1080:
URL: https://github.com/apache/orc/pull/1080#discussion_r842931210
##########
java/core/src/test/org/apache/orc/TestVectorOrcFile.java:
##########
@@ -4384,6 +4384,91 @@ public void testPredicatePushdownWithNan(Version
fileFormat) throws Exception {
assertEquals(0, batch.size);
}
+ @ParameterizedTest
+ @MethodSource("data")
+ public void testPredicatePushdownWithSumOverflow(Version fileFormat) throws
Exception {
+ TypeDescription schema = TypeDescription.createStruct()
+ .addField("double1", TypeDescription.createDouble())
+ .addField("float1", TypeDescription.createFloat());
+
+ Writer writer = OrcFile.createWriter(testFilePath,
+ OrcFile.writerOptions(conf)
+ .setSchema(schema)
+ .stripeSize(400000L)
+ .compress(CompressionKind.NONE)
+ .bufferSize(500)
+ .rowIndexStride(1000)
+ .version(fileFormat));
+ VectorizedRowBatch batch = schema.createRowBatch();
+ batch.ensureSize(3500);
+ batch.size = 3500;
+ batch.cols[0].noNulls = true;
+ batch.cols[1].noNulls = true;
+
+ DoubleColumnVector dbcol = ((DoubleColumnVector) batch.cols[0]);
+ DoubleColumnVector fcol = ((DoubleColumnVector) batch.cols[1]);
+
+ // First two rows of data cause sum overflow, sum is not a finite value,
+ // but this does not prevent pushing down (range comparisons work fine)
+ // The same applies to the middle stripe
Review Comment:
Could you add some more illustration about how many stripes are used 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]
