DanielLeens commented on issue #12704: URL: https://github.com/apache/seatunnel/issues/12704#issuecomment-6078543030
Classification: A / Connector-V2 Elasticsearch/OpenSearch document identity and data correctness. Thank you for the precise reproduction. Current dev confirms that `KeyExtractor` joins composite key values with `key_delimiter`, so the mapping is not injective when a value contains that delimiter; this can overwrite a different document. A contribution is welcome, but the suggested length-prefix encoding must not silently replace the existing default. `_id` is the persisted upsert identity: changing it for an existing index would make a restarted job create a second set of documents instead of updating the old ones. Please keep legacy encoding as the default and propose an explicit, opt-in collision-safe encoding or migration strategy. Please keep this to one focused PR and include tests for the reported collision, exact legacy IDs, a single-key control, temporal/string values, and both Elasticsearch and OpenSearch paths if their shared serializer is exercised. Update EN/ZH connector documentation with the compatibility and migration behavior. No competing PR is open in the current audit. -- 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]
