[
https://issues.apache.org/jira/browse/SPARK-40658?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17614565#comment-17614565
]
Sandish Kumar HN commented on SPARK-40658:
------------------------------------------
[~mposdev21] these are the changes I see between proto2 vs proto3
# The latest Proto3 also supports optional fields, the difference is optional
fields which have has_foo() methods, and "singular" fields, which do not. I
don't see any different treatment needed to handle this.
# In contrast to proto3, Proto2 allows custom default values and required
fields.
# Enums: Proto3's default value is the enum 0 index value. Proto2 uses the
first syntactic entry in the enum declaration as the default value if it is not
specified otherwise.
# Proto2 does not validate that inbound and outbound bytes are encoded in
UTF-8. During parsing, all string fields in Proto3 are appropriately UTF-8
encoded.
# Proto2 and proto3 are wire compatible, they will have the same binary
representation.
should we have an optional option setting something like
PROTO_VERSION_SUPPORT=V3 or V2 or ANY? the default can be ANY.
> Protobuf v2 & v3 support
> ------------------------
>
> Key: SPARK-40658
> URL: https://issues.apache.org/jira/browse/SPARK-40658
> Project: Spark
> Issue Type: Improvement
> Components: Structured Streaming
> Affects Versions: 3.3.0
> Reporter: Raghu Angadi
> Priority: Major
>
> We want to ensure Protobuf functions support both Protobuf version 2 and
> version 3 schemas (e.g. descriptor file or compiled classes with v2 and v3).
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
