Fix minor bugs on querying Hudi MOR tables

[7c00]

This PR fixed two minor bugs on querying Hudi MOR tables.

• Make HudiRealtimeSplitConverter work with MOR splits without log files
• Fix HadoopExtendedFileSystem to support MOR reader constructor

It is extracted from #17463.

Test plan - Unit tests.

== NO RELEASE NOTE ==

[7c00]

@codope @pratyakshsharma Hi, would you like to take a review on this bugfix?

[codope]

Good find @7c00 !
Would it be possible to add a test for this scenario? Possibly in TestCustomSplitConversionUtils

[7c00]

Good find @7c00 ! Would it be possible to add a test for this scenario? Possibly in TestCustomSplitConversionUtils

Sure. I will do that.

[7c00]

rerun the test

[7c00]

@codope I haved added the tests. Could you take a second review?

[codope]

Thanks @7c00 for quick revert. One small question below.

[codope]

Why do we need this API?

[7c00]

If we dont fix this method, MOR tables could not be queried. I added the intergration tests in #17463.
Following is part of the error messages:

Caused by: java.lang.UnsupportedOperationException: Not implemented by the HadoopExtendedFileSystem FileSystem implementation
at org.apache.hadoop.fs.FileSystem.getScheme(FileSystem.java:219)
at org.apache.hudi.common.fs.FSUtils.isGCSFileSystem(FSUtils.java:592)
at org.apache.hudi.common.table.log.HoodieLogFileReader.getFSDataInputStream(HoodieLogFileReader.java:119)
at org.apache.hudi.common.table.log.HoodieLogFileReader.(HoodieLogFileReader.java:95)
at org.apache.hudi.common.table.log.HoodieLogFileReader.(HoodieLogFileReader.java:86)
at org.apache.hudi.common.table.log.HoodieLogFormat.newReader(HoodieLogFormat.java:282)
at org.apache.hudi.common.table.log.LogReaderUtils.readSchemaFromLogFileInReverse(LogReaderUtils.java:49)
at org.apache.hudi.common.table.log.LogReaderUtils.readLatestSchemaFromLogFiles(LogReaderUtils.java:77)
at org.apache.hudi.hadoop.realtime.AbstractRealtimeRecordReader.init(AbstractRealtimeRecordReader.java:85)
at org.apache.hudi.hadoop.realtime.AbstractRealtimeRecordReader.(AbstractRealtimeRecordReader.java:67)
at org.apache.hudi.hadoop.realtime.RealtimeCompactedRecordReader.(RealtimeCompactedRecordReader.java:62)
at org.apache.hudi.hadoop.realtime.HoodieRealtimeRecordReader.constructRecordReader(HoodieRealtimeRecordReader.java:70)
at org.apache.hudi.hadoop.realtime.HoodieRealtimeRecordReader.(HoodieRealtimeRecordReader.java:47)
at org.apache.hudi.hadoop.realtime.HoodieParquetRealtimeInputFormat.getRecordReader(HoodieParquetRealtimeInputFormat.java:323)
at com.facebook.presto.hive.HiveUtil.createRecordReader(HiveUtil.java:272)
at com.facebook.presto.hive.GenericHiveRecordCursorProvider.lambda$createRecordCursor$0(GenericHiveRecordCursorProvider.java:74)
at com.facebook.presto.hive.authentication.NoHdfsAuthentication.doAs(NoHdfsAuthentication.java:23)
at com.facebook.presto.hive.HdfsEnvironment.doAs(HdfsEnvironment.java:81)
at com.facebook.presto.hive.GenericHiveRecordCursorProvider.createRecordCursor(GenericHiveRecordCursorProvider.java:73)
at com.facebook.presto.hive.HivePageSourceProvider.getPageSourceFromCursorProvider(HivePageSourceProvider.java:571)

You can check the full error log at https://github.com/prestodb/presto/runs/5566744421?check_suite_focus=true.

[7c00]

BTW, please review the PR at #17463

[codope]

Got it.

[codope]

@arunthirupathi Could you please take a quick pass? Would be good to have it in the upcoming release.

[arunthirupathi]

Can you please edit the description and title to reflect what bug it fixes ?

[7c00]

@arunthirupathi I have updated the description.