bharos opened a new pull request, #13635: URL: https://github.com/apache/gravitino/pull/13635
### What changes were proposed in this pull request? Seven documentation corrections, each verified against an authoritative source in this repository. **Privilege scopes** — `CREATE_TOPIC`, `CREATE_FILESET` and `REGISTER_FUNCTION` were listed as grantable on the object they create: | File | Privilege | Was | Now | |---|---|---|---| | `docs/topics.md` | `CREATE_TOPIC` | Metalake, catalog, schema, or topic | Metalake, catalog, or schema | | `docs/filesets.md` | `CREATE_FILESET` | Metalake, catalog, schema, or fileset | Metalake, catalog, or schema | | `docs/functions.md` | `REGISTER_FUNCTION` | Metalake, catalog, schema, or function | Metalake, catalog, or schema | `docs/security/access-control.md` limits all three to Metalake, Catalog and Schema, and `docs/tables-and-views.md` already documents `CREATE_TABLE` and `CREATE_VIEW` as "Metalake, catalog, or schema". The other rows in these tables (`PRODUCE_TOPIC`, `CONSUME_TOPIC`, `READ_FILESET`, `WRITE_FILESET`, `EXECUTE_FUNCTION`, `MODIFY_FUNCTION`) correctly include the leaf object and are unchanged. **Kafka property name** — the broker default is `num.partitions`, not `num.partition`. Corrected in `docs/topics.md` and `docs/kafka-catalog.md`. **Spark implementation language** — `docs/functions.md` said a definition can carry "a Python version for Spark". `docs/spark-connector/spark-connector-udf.md` states that only Java implementations with `RuntimeType.SPARK` are supported and that Python implementations cannot yet be invoked. **Broken link** — the CLI Reference entry in `docs/table-maintenance-service/optimizer-extension-guide.md` pointed at `optimizer-configuration.md`, the same target as the Configuration entry directly above it. ### Why are the changes needed? Fix: #13634 Each of these leads a reader to something that does not work: a grant the server rejects, a Kafka property that does not exist, a Spark UDF language that cannot be invoked, and a link to the wrong page. ### Does this PR introduce any user-facing change? No code change. Documentation only. ### How was this patch tested? No tests; documentation only. Each correction was checked against the authoritative source in-repo: - privilege scopes against the table in `docs/security/access-control.md` - the Spark claim against `docs/spark-connector/spark-connector-udf.md` - the link target exists at `docs/table-maintenance-service/optimizer-cli-reference.md` - a repository-wide grep confirms no remaining `num.partition` (singular) and no other `CREATE_*`/`REGISTER_*` row listing the leaf object ### Note These were found while publishing the 1.3.1 documentation to gravitino-site, where an automated review flagged them. Issue #13634 lists several further items that are intentionally not in this PR: two pre-existing broken `../dev/charts` links that date to 1.0.0, and two OpenAPI questions that need a maintainer decision (a `schema` referencing `components/responses`, which occurs about 30 times across the spec files, and the version string in `open-api/idp/openapi.yaml`). One reported item, the Java example in `trino-connector/udf-support.md`, was investigated and is correct as written; the apparent double wrapping is the varargs array-builder idiom. -- 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]
