kotwal-itpro commented on PR #8859: URL: https://github.com/apache/hop/pull/8859#issuecomment-6117021626
Thanks @bamaer and @mattcasters. Fixed in e3483bf14e, rebased on main: **Old names (@bamaer).** The navigation, the transform and metadata indexes and every cross-link you listed now say "Lake table input", "Lake table output" and "Lakehouse catalog", including the option labels in the merge and maintenance pages, the metadata index description and the prose in `lakehouse.adoc`. The entries in `transforms.adoc` and `nav.adoc` are re-sorted under L, and the catalog's nav entry is moved to its place among the metadata types. The only remaining "Spark Catalog" is the intentional "called *Spark Catalog* in earlier versions" on the catalog page. **Dates and timestamps (@mattcasters).** You're right on both; those sentences were written before the type mapping moved to the JVM zone and I didn't update them. - Input: dates and timestamps without a zone are read as local dates and date-times in the Hop JVM's zone, as Parquet Input does; `timestamptz` values are instants. - Output: Date and Timestamp become `timestamp` without zone, holding the wall-clock time in the Hop JVM's zone, as Parquet Output does; into an existing `timestamptz` column they're written as instants. - BigNumber: the field's length (at most 38) and precision are used only when the length is greater than 0; otherwise `decimal(38, 10)`. **Concurrent PATH commits.** Agreed: with the `CREATE_NEW` fallback every local file system is safe. The page now says any `file://` location allows concurrent writers (hard link, or exclusive create where links aren't supported), and asks for one writer, or a REST catalog, only on HDFS, object stores and other non-local VFS locations, where the commit checks and then writes. **Marketplace id.** After rebasing, `hop-tech-lakehouse` resolves: your #8854 added it to `optional-plugins.yaml` (with the note to install it together with the Spark engine), so I kept your entry as is. I added an assertion to `OptionalPluginCatalogTest` that the catalog lists it, and both transform pages now say they're part of the Lakehouse plugin and that the Spark engine plugin is needed too on Spark. **0004 validation.** The description now names the 0004 extract. Tests: `OptionalPluginCatalogTest` passes, `apache-rat:check` is clean. The integration tests (all 4 iceberg scenarios) passed before this change; this commit only touches the 0004 description. -- 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]
