wombatu-kun commented on code in PR #20398:
URL: https://github.com/apache/hudi/pull/20398#discussion_r4344377062
##########
hudi-common/src/main/java/org/apache/hudi/common/schema/internal/convert/InternalSchemaConverter.java:
##########
@@ -63,19 +63,27 @@ public class InternalSchemaConverter {
static final int BLOB_DATA_FIELD_ID = -11;
static final int BLOB_REFERENCE_FIELD_ID = -12;
+ // The four fields of the out-of-line reference record nested in a Blob are
tagged with sentinel ids as
+ // well. Ids of table columns are allocated from 0 upwards, so a
non-negative id here would shadow a table
Review Comment:
Since the only ids still shared between columns are the negative sentinels,
would a duplicate check limited to ids >= 0 in
`InternalSchemaBuilder.buildIdToField` (or a test asserting it) give #19833 the
guard it asks for, without rejecting tables that have several BLOB or VARIANT
columns? If you would rather keep this PR to the id change, a follow-up issue
for the guard would keep it tracked once #19833 closes.
--
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]
