[
https://issues.apache.org/jira/browse/HUDI-7081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17911290#comment-17911290
]
Y Ethan Guo edited comment on HUDI-7081 at 1/9/25 1:55 AM:
-----------------------------------------------------------
You can write a table with enum type field, apply updates to generate log files
with enum fields, and see if the table can be read. Test both avro and parquet
log block. Test log file on top of parquet base file generated by both create
handle and merge handle (compaction baed on fg reader).
was (Author: JIRAUSER280684):
You can write a table with enum type field, apply updates to generate log files
with enum fields, and see if the table can be read. Test both avro and parquet
log block. Test log file on top of parquet base file generated by both create
handle and merge handle (compaction baed on fg reader)
> Hoodie FilegroupReader cannot read Enums from MOR avro log blocks
> -----------------------------------------------------------------
>
> Key: HUDI-7081
> URL: https://issues.apache.org/jira/browse/HUDI-7081
> Project: Apache Hudi
> Issue Type: Sub-task
> Reporter: Jonathan Vexler
> Assignee: Lin Liu
> Priority: Blocker
> Fix For: 1.0.1
>
>
> {code:java}
> buildFileGroupIterator(
> Option.empty[PartitionedFile => Iterator[InternalRow]],
> partitionValues,
> Option.empty[HoodieBaseFile],
> getLogFilesFromSlice(fileSlice),
> requiredSchemaWithMandatory,
> outputSchema,
> partitionSchema,
> broadcastedHadoopConf.value.value,
> -1,
> -1,
> shouldUseRecordPosition
> ) {code}
> This only uses StructType schemas and converts to avro. So it thinks enums
> are strings
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
