difin commented on code in PR #1432:
URL: https://github.com/apache/orc/pull/1432#discussion_r1131606891
##########
java/core/src/test/org/apache/orc/impl/TestInStream.java:
##########
@@ -957,4 +957,47 @@ public void testMultiRangeCompressHeader() throws
IOException {
assertEquals((byte)i, inBuffer[i], "position " + i);
}
}
+
+ private static final byte[] uncompressed = input(
+ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9);
+
+ @Test
+ public void testStreamResetWithIncreasedLength() throws IOException {
+ // Set up an initial buffer of PREVIOUS_LENGTH followed by our stream
+ // at START.
+ final long START = 1_000;
+ final int PREVIOUS_LENGTH = 30;
+ BufferChunkList list = new BufferChunkList();
+ byte[] previous = new byte[PREVIOUS_LENGTH];
+ Arrays.fill(previous, (byte) -1);
+ list.add(new BufferChunk(ByteBuffer.wrap(previous), START -
PREVIOUS_LENGTH));
+ list.add(new BufferChunk(ByteBuffer.wrap(uncompressed), START));
+ // Creating a stream of 10 bytes, but with a length of 5
+ InStream inStream = InStream.create("test", list.get(), START, 5, new
InStream.StreamOptions());
+ // Resetting the stream with the increased length
+ inStream.reset(list.get(), 10);
+ // Reading the stream and expecting to read 10 bytes
+ byte[] inBuffer = new byte[10];
+ assertEquals(10, inStream.read(inBuffer));
+ }
+
+ @Test
+ public void testStreamResetWithoutIncreasedLength() throws IOException {
+ // Set up an initial buffer of PREVIOUS_LENGTH followed by our stream
+ // at START.
+ final long START = 1_000;
+ final int PREVIOUS_LENGTH = 30;
+ BufferChunkList list = new BufferChunkList();
+ byte[] previous = new byte[PREVIOUS_LENGTH];
+ Arrays.fill(previous, (byte) -1);
+ list.add(new BufferChunk(ByteBuffer.wrap(previous), START -
PREVIOUS_LENGTH));
+ list.add(new BufferChunk(ByteBuffer.wrap(uncompressed), START));
+ // Creating a stream of 10 bytes, but with a shorter length of 5
+ InStream inStream = InStream.create("test", list.get(), START, 5, new
InStream.StreamOptions());
+ // Resetting the stream without updating its length
+ inStream.reset(list.get());
Review Comment:
Yes, and this causes "Can't finish byte read from uncompressed stream DATA
position" exception in Hive (HIVE-27128). We plan to fix it in Hive by
upgrading its ORC version once a new ORC version is released with this
fix(ORC-1393) and with ORC-1384 which is a blocker for the upgrade. Thanks a
lot for merging my PR!
--
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]
