[
https://issues.apache.org/jira/browse/FLINK-25509?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17708769#comment-17708769
]
Dong Lin commented on FLINK-25509:
----------------------------------
[~martijnvisser] I agree with the discussion mentioned above that we should aim
to let connectors use public API. It will be great for someone to update flink
core so that flink-connector-pulsar use only public API from Flink.
I guess my points are:
- It is not a requirement that externalized connectors should only use public
API (Please let me know if there is consensus on this requirement). We will
have to stick with the existing practice (where connectors use internal API)
until someone volunteers to improve this with new FLIPs.
- The issue here is a bit different from FLINK-31324 in the sense that there is
no good way for flink-connector-pulsar to compile with `SplitFetcherManager`
from both Flink 1.17 and Flink 1.18-snapshot. I would be happy to create a PR
if we can have a good fix for this issue.
- I am not sure it is a good practice to require one code to compile with two
versions. This is typically done with multiple branches if needed.
In this particular case, I am not sure if it is problematic to add an internal
abstract method to `SplitFetcherManager`. I would be happy to create a PR for
flink-connector-pulsar if the maintainer of this connector can create dedicated
branch for each target flink version.
It is probably up to the maintainer to decide how to address this properly.
> FLIP-208: Add RecordEvaluator to dynamically stop source based on
> de-serialized records
> ---------------------------------------------------------------------------------------
>
> Key: FLINK-25509
> URL: https://issues.apache.org/jira/browse/FLINK-25509
> Project: Flink
> Issue Type: New Feature
> Components: Connectors / Common, Connectors / Kafka
> Reporter: Dong Lin
> Assignee: Hang Ruan
> Priority: Major
> Labels: pull-request-available
>
> This feature is needed to migrate applications which uses
> KafkaDeserializationSchema::isEndOfStream() from using FlinkKafkaConsumer to
> using KafkaSource.
> Please checkout
> https://cwiki.apache.org/confluence/display/FLINK/FLIP-208%3A+Add+RecordEvaluator+to+dynamically+stop+source+based+on+de-serialized+records
> for the motivation and the proposed changes.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
