imbajin opened a new pull request, #804: URL: https://github.com/apache/hugegraph-toolchain/pull/804
## Purpose of the PR The ordinary Loader distribution cannot start existing GRAPH or Parquet sources: its selected protobuf runtime lacks the PD discovery classes, and Parquet classes are inherited only through Spark's provided dependencies. The ordinary package now carries the dependencies these existing readers require. Reported during [master integration testing](https://github.com/hugegraph/hugegraph-toolchain/pull/47#issuecomment-6064536763), and reproduced through the original GraphSource and InputReader entry points.  ## Main Changes - Use a compatible protobuf 3.25.9 patch runtime while retaining the existing PD/Hadoop generated-message interfaces. - Include existing Parquet 1.12.2 modules and the required current MapReduce core, retaining the existing native codec versions and avoiding YARN/logging dependency changes. - Exercise PD/Hadoop binary messages and real compressed Parquet files; run the Parquet reader with ordinary distribution libraries in CI. - Update the distribution license, notices, and dependency inventory. - Fetch CI MySQL from Docker Official Images on ECR, pinned to the exact digest of the previously successful fixture; service settings and assertions stay identical. ## Verifying these changes - [ ] Trivial rework / code cleanup without any test coverage. (No Need) - [ ] Already covered by existing tests. - [x] Need tests and can be verified as follows: - Clean native packaging, SDK provenance validation, targeted tests, Checkstyle and actionlint pass. - Original distributions fail at the reader entry points. The candidate ordinary classpath reads a real SNAPPY Parquet file with Unicode, numeric fields, and EOF correctly. - GraphSource now passes the protobuf class-loading boundary; the isolated closed-endpoint probe subsequently fails discovery. Full GRAPH ingestion and HDFS validation remain pending. - Three independent readers reviewed the final diff. This does not extend the documented Java 17 HBase bulkload support boundary. ## Does this PR potentially affect the following parts? - [ ] Nope - [x] Dependencies (add/update license info) - [ ] Modify configurations - [ ] The public API - [x] Other affects: ordinary Loader runtime packaging and its CI regression ## Documentation Status - [ ] `Doc - TODO` - [x] `Doc - Done`: distribution licenses and dependency notices are updated. - [ ] `Doc - No Need` Organization core CI passes on the current branch. The final combined distribution also builds and passes the ordinary package runtime linkage checks against a fixed Server SDK. Full GRAPH import acceptance remains in progress. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
