[
https://issues.apache.org/jira/browse/FLINK-8230?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16285249#comment-16285249
]
ASF GitHub Bot commented on FLINK-8230:
---------------------------------------
GitHub user packet23 opened a pull request:
https://github.com/apache/flink/pull/5145
[FLINK-8230] [ORC] Improved mapping of Orc records to Flink Table rows.
## What is the purpose of the change
Improved mapping of Orc records to Flink table rows and ensured nested
structs are correctly mapped.
## Brief change log
- Introduced OrcRowReader to be used on a single input split
- OrcRowReader was adapted from
org.apache.orc.mapred.OrcMapredRecordReader
- Added a bunch of test datasets, mostly from orc distribution
## Verifying this change
This change added tests and can be verified as follows:
- OrcRowReaderTest maps test datasets to rows comparing count
- OrcRowReader is also covered by existing flink-orc unit tests.
- Manually verified the change by running a 4 node cluster on YARN,
reading an existing full dataset
## Does this pull request potentially affect one of the following parts:
- Dependencies (does it add or upgrade a dependency): no
- The public API, i.e., is any changed class annotated with
`@Public(Evolving)`: no
- The serializers: don't know
- The runtime per-record code paths (performance sensitive): yes
- Anything that affects deployment or recovery: JobManager (and its
components), Checkpointing, Yarn/Mesos, ZooKeeper: don't know
- The S3 file system connector: no
## Documentation
- Does this pull request introduce a new feature? no
- If yes, how is the feature documented? not applicable
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/packet23/flink FLINK-8230
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/flink/pull/5145.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #5145
----
commit 93ef978bb5a55700fa25551eda5464860d6d1ae2
Author: Sebastian Klemke <[email protected]>
Date: 2017-12-09T19:35:01Z
[FLINK-8230] [ORC] Improved mapping of Orc records to Flink Table rows.
----
> NPE in OrcRowInputFormat on nested structs
> ------------------------------------------
>
> Key: FLINK-8230
> URL: https://issues.apache.org/jira/browse/FLINK-8230
> Project: Flink
> Issue Type: Bug
> Components: Batch Connectors and Input/Output Formats
> Affects Versions: 1.4.0
> Reporter: Sebastian Klemke
> Priority: Minor
>
> OrcRowInputFormat ignores isNull and isRepeating on nested struct columns. If
> a struct column contains nulls, it tries to read struct fields, leading to
> NPE in case of string fields:
> {code}
> java.lang.NullPointerException
> at java.lang.String.checkBounds(String.java:384)
> at java.lang.String.<init>(String.java:462)
> at
> org.apache.flink.orc.OrcUtils.readNonNullBytesColumnAsString(OrcUtils.java:392)
> at org.apache.flink.orc.OrcUtils.readField(OrcUtils.java:215)
> at org.apache.flink.orc.OrcUtils.readStructColumn(OrcUtils.java:1203)
> at org.apache.flink.orc.OrcUtils.readField(OrcUtils.java:252)
> at
> org.apache.flink.orc.OrcUtils.readNonNullStructColumn(OrcUtils.java:677)
> at org.apache.flink.orc.OrcUtils.readField(OrcUtils.java:250)
> at org.apache.flink.orc.OrcUtils.fillRows(OrcUtils.java:142)
> at
> org.apache.flink.orc.OrcRowInputFormat.ensureBatch(OrcRowInputFormat.java:334)
> at
> org.apache.flink.orc.OrcRowInputFormat.reachedEnd(OrcRowInputFormat.java:314)
> at
> org.apache.flink.runtime.operators.DataSourceTask.invoke(DataSourceTask.java:165)
> at org.apache.flink.runtime.taskmanager.Task.run(Task.java:718)
> at java.lang.Thread.run(Thread.java:748)
> {code}
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
