fix(RecordAssembler): optimize field creation using Accessor #3031
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Confluent Schema Registry currently suppresses Avro’s default-value validation (see (confluentinc/schema-registry#1687)), which means even schemas such as
{"name":"f_v2","type":"string","default":null}slip through unchecked (confluentinc/schema-registry#2213). If we rebuilt those fields via new Schema.Field(...), Avro would revalidate and reject them, diverging from what the registry already accepted. (AutoMQ/automq-labs#100 (comment) & #3006)By constructing fields with Accessor.createField(...), we reuse the existing schema object (so logical types stay intact) while bypassing the costly validation step; this keeps the hot path efficient and, more importantly, aligns our behavior with the registry’s current tolerance. If we ever need the original field’s custom props or aliases, we can still copy them after the accessor call.