[
https://issues.apache.org/jira/browse/ORC-168?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15968160#comment-15968160
]
ASF GitHub Bot commented on ORC-168:
------------------------------------
Github user Citrullin commented on a diff in the pull request:
https://github.com/apache/orc/pull/104#discussion_r111478051
--- Diff: site/_docs/core-java.md ---
@@ -233,14 +233,15 @@ VectorizedRowBatch batch = schema.createRowBatch();
LongColumnVector x = (LongColumnVector) batch.cols[0];
LongColumnVector y = (LongColumnVector) batch.cols[1];
for(int r=0; r < 10000; ++r) {
- int row = batch.size++;
+ int row = batch.size;
x.vector[row] = r;
--- End diff --
Hi, I'm not that good in Java. I'm more familiar with Scala. But you take a
look into the Library I wrote. I created two branches where I changed only the
position of the up counting.
In example-1 I count both batch.size and rowBatchSize up before I add a row
to VectorizedBatch.
[See more
here](https://github.com/Citrullin/scalaOrcWriter/blob/ORC-168-wrong-example-1/src/main/scala/citrullin/orcwriter/OrcWriter.scala#L77)
In example 2 I count only batch.size up before I add the row to the batch.
rowBatchSize will up counted after a row is written.
[See more
here](https://github.com/Citrullin/scalaOrcWriter/blob/ORC-168-wrong-example-2/src/main/scala/citrullin/orcwriter/OrcWriter.scala#L77)
The working example is in the dev branch.
[More
here](https://github.com/Citrullin/scalaOrcWriter/blob/dev/src/main/scala/citrullin/orcwriter/OrcWriter.scala#L77)
You can run the Implicit ComplexMap example to see the differences.
[Here is the
source](https://github.com/Citrullin/scalaOrcWriter/blob/dev/src/main/scala/citrullin/orcwriter/examples/orcwriterimplicitapi/WriteComplexMap.scala)
> Documentation Writer Example: batch.size++ has a wrong possition
> ----------------------------------------------------------------
>
> Key: ORC-168
> URL: https://issues.apache.org/jira/browse/ORC-168
> Project: ORC
> Issue Type: Improvement
> Components: documentation, Java
> Reporter: Philipp Blum
> Priority: Critical
>
> There's one little mistake in the Java Core Example. The for loops starts
> with a batch.size++:
> for(int r=0; r < 10000; ++r) {
> int row = batch.size++;
> x.vector[row] = r;
> y.vector[row] = r * 3;
> // If the batch is full, write it out and start over.
> if (batch.size == batch.getMaxSize()) {
> writer.addRowBatch(batch);
> batch.reset();
> }
> }
> If you start with a batch.size++ the first index will be 1, so the first
> entry in the orc file will be empty.
> Correct is:
> for(int r=0; r < 10000; ++r) {
> int row = batch.size;
> x.vector[row] = r;
> y.vector[row] = r * 3;
> // If the batch is full, write it out and start over.
> if (batch.size == batch.getMaxSize()) {
> writer.addRowBatch(batch);
> batch.reset();
> }
> batch.size++;
> }
> Already tested it in scala.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
