Support transcoding for Flatten in Flink Runner
- Dominant language
- Java
- Stars
- 8.7k
- Forks
- 4.7k
- Avg merge
- 1d 20h
- Merged PRs (30d)
- 196
Description
The Beam documentation mentions that Flatten does not require coders of all inputs to be of the same type, as long as the result type is the same. However, the current implementation in the Flink Runner requires all coders to match.
Imported from Jira [BEAM-6523](https://issues.apache.org/jira/browse/BEAM-6523). Original Jira may contain additional context.
Reported by: mxm.
This issue has child subcomponents which were not migrated over. See the original Jira for more information.
Contributor guide
Research direction
Start with the Beam documentation statement about Flatten coders and inspect the Flink Runner implementation that currently requires matching coders. Read the original Jira issue for the missing child subcomponents and additional context. Done means Flatten accepts inputs with different coders when their result type is the same.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- java
- Domain
- data-engineering, distributed-systems
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100