RussellSpitzer commented on code in PR #12982:
URL: https://github.com/apache/iceberg/pull/12982#discussion_r2075912711
##########
format/spec.md:
##########
@@ -1680,6 +1680,23 @@ Row-level delete changes:
* These position delete files must be merged into the DV for a data file
when one is created
* Position delete files that contain deletes for more than one data file
need to be kept in table metadata until all deletes are replaced by DVs
+Row lineage changes:
+
+* Writers must set the table's `next-row-id` and use the existing
`next-row-id` as the `first-row-id` to create a new snapshot
+ * When a table is upgraded to v3, `next_row_id` should be initialized to 0
+ * It is recommended to increment `next-row-id` by the total
`added_rows_count` and `existing_rows_count` of all manifests assigned a
`first_row_id`
+* Writers must assign a `first_row_id` to new data manifests when writing a
manifest list
+ * It is recommended to increment `first_row_id` by a manifest's
`added_rows_count` and `existing_rows_count`
+* Readers must assign a `first_row_id` for any data file that does not have an
assigned value written in a manifest
+ * Readers must increment `first_row_id` by the data file's `record_count`
+* When writing an existing data file into a new manifest, its `first_row_id`
must be written into the manifest
+* When a data file has a non-null `first_row_id`, readers must:
Review Comment:
"assigned or non-null" ?
--
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: issues-unsubscr...@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org
For additional commands, e-mail: issues-h...@iceberg.apache.org
