[
https://issues.apache.org/jira/browse/DRILL-5797?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16456529#comment-16456529
]
Oleksandr Kalinin commented on DRILL-5797:
------------------------------------------
Just for a record, further debugging shows how complex column sneaks into
ReadState:
(1) `ParquetRecordReader.setup()` triggers ParquetSchema
buildSchema/loadParquetSchema for column mapping
(2) `ParquetSchema.loadParquetSchema()` is using `ParqueSchema.fieldSelected()`
for column matching
(3) fieldSelected() takes MaterializedField as an argument and uses it's
getName() method for column name comparison. For column B.A it returns A.
(4) As result of that, column B.A of the file gets positively matched to column
A and gets added to selectedColumnMetadata in the ParquetSchema which is then
passed to ReadState
> Use more often the new parquet reader
> -------------------------------------
>
> Key: DRILL-5797
> URL: https://issues.apache.org/jira/browse/DRILL-5797
> Project: Apache Drill
> Issue Type: Improvement
> Components: Storage - Parquet
> Reporter: Damien Profeta
> Assignee: Oleksandr Kalinin
> Priority: Major
> Fix For: 1.14.0
>
>
> The choice of using the regular parquet reader of the optimized one is based
> of what type of columns is in the file. But the columns that are read by the
> query doesn't matter. We can increase a little bit the cases where the
> optimized reader is used by checking is the projected column are simple or
> not.
> This is an optimization waiting for the fast parquet reader to handle complex
> structure.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
