[
https://issues.apache.org/jira/browse/PARQUET-1719?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17483677#comment-17483677
]
ASF GitHub Bot commented on PARQUET-1719:
-----------------------------------------
dssysolyatin commented on a change in pull request #707:
URL: https://github.com/apache/parquet-mr/pull/707#discussion_r794344312
##########
File path:
parquet-hadoop/src/main/java/org/apache/parquet/hadoop/ParquetReader.java
##########
@@ -104,7 +104,7 @@ public ParquetReader(Configuration conf, Path file,
ReadSupport<T> readSupport,
this(conf, file, readSupport, FilterCompat.get(unboundRecordFilter));
}
- private ParquetReader(Configuration conf,
+ public ParquetReader(Configuration conf,
Review comment:
Yes, you are right. I need another method. Finally, I can get rid of the
hack :)
```
private val constructor = {
val ctor = classTag.runtimeClass.getDeclaredConstructor(
classOf[java.util.List[InputFile]],
classOf[ParquetReadOptions],
classOf[ReadSupport[Row]]
)
ctor.setAccessible(true)
ctor
}
```
--
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]
> Make ParquetReader(List<InputFile>, ParquetReadOptions, ReadSupport<T>)
> constructor public
> ------------------------------------------------------------------------------------------
>
> Key: PARQUET-1719
> URL: https://issues.apache.org/jira/browse/PARQUET-1719
> Project: Parquet
> Issue Type: Improvement
> Components: parquet-mr
> Reporter: Dmitry Sysolyatin
> Priority: Major
> Labels: pull-request-available
> Fix For: 1.11.0
>
>
> My application uses s3 storage to store parquet files. The problem is that
> there is not ability to get rid from `new Configuration(true);` which load
> configuration from files:
> `core-default.xml` and `core-site.xml` every time when ParquetReader instance
> is created.
> I suggest just make ParquetReader(List<InputFile>, ParquetReadOptions,
> ReadSupport<T>) constructor public then there will be ability to do custom
> builder.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
