rahil-c opened a new issue, #19700: URL: https://github.com/apache/hudi/issues/19700
`DocumentParser` and `EmbeddingProvider` were added for unstructured file ingestion and are both implementable by users: `hoodie.streamer.source.unstructured.parser.class` and `hoodie.streamer.transformer.embedding.provider.class` exist specifically to point at a custom implementation. Neither carries a stability annotation, while the surrounding extension points in the same modules do (`Source` and `Transformer` are both `@PublicAPIClass(maturity = STABLE)`). Their shape is still moving: neither signature carries the object's URI, size or content type, so a parser cannot route on content type or read a range rather than a whole stream, and the embedding interface accepts only `List<String>` so it cannot reach image or audio inputs. Declaring them EVOLVING now, before a release includes them, records that they are expected to change and keeps those changes from reading as breaking ones later. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
