[GitHub] [hudi] jsbali commented on a change in pull request #3946: [HUDI-2711] Fallback to fulltable scan for IncrementalRelation if underlying files have been cleared or moved by cleaner

[GitBox <gi...@apache.org>]

jsbali commented on a change in pull request #3946:
URL: https://github.com/apache/hudi/pull/3946#discussion_r790282213



##########
File path: hudi-spark-datasource/hudi-spark/src/main/scala/org/apache/hudi/DefaultSource.scala
##########
@@ -201,7 +201,7 @@ class DefaultSource extends RelationProvider
       HadoopFsRelation(
         fileIndex,
         fileIndex.partitionSchema,
-        fileIndex.dataSchema,
+        fileIndex.schema,

Review comment:
       @danny0405 @nsivabalan 
   Ok from my understanding 
   
   partiton in data implies dataSchema = schema
   for partition not in schema dataSchema is still equal to schema. 
   
   With example say partA is the partition column and is written along with data. so dataSchema is {colA, partA, colB} and schema in commit file will be same {colA, partA, colB}
   Now for partition data not written to data file. dataSchema is {colA, colB} and so is commit file schema {colA, colB}
   Am i missing something




-- 
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: commits-unsubscribe@hudi.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org