pavibhai commented on a change in pull request #634:
URL: https://github.com/apache/orc/pull/634#discussion_r564295072
##########
File path: java/core/src/java/org/apache/orc/impl/RecordReaderImpl.java
##########
@@ -104,17 +104,10 @@
*/
static int findColumns(SchemaEvolution evolution,
Review comment:
The behavior change we did to this API is as follows:
* If the column is present is missing in readSchema then throw an exception
of missing column
* If the column is missing in file but present in readSchema then we see
this as missing column in file handled via NullReader
Wanted to confirm that the ask is to not throw an exception when Hive/Spark
use a column name that is missing in the readSchema. I might be missing the
scenario where this is the case, I was under the impression that column
references should always be valid within the readSchema.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
