Repository navigation
[HSTACK] feat: add option to modify the low-level Delta schema to make some fields nullable ADDENDUM - #25
Merged
Conversation
|
ACTION NEEDED delta-rs follows the Conventional Commits specification for release automation. The PR title and description are used as the merge commit message. Please update your PR title and description to match the specification. |
VLCDaniel
force-pushed
the
next-nullable-filter-fix
branch
from
October 19, 2026 13:48
362411c to
96d3b92
Compare
…e some fields nullable ADDENDUM
VLCDaniel
force-pushed
the
next-nullable-filter-fix
branch
from
October 19, 2026 14:39
96d3b92 to
a863c8e
Compare
VLCDaniel
changed the title
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Description
aep_dataset_next queries with a partition filter (e.g. WHERE ACP_DATE >= …) fail with:
Found unmasked nulls for non-nullable StructArray field "rowId"Why:
The parquet read schema already gets the
DELTA_FIELD_PATHS_TO_MAKE_NULLABLEoverrides. But when a file needs a kernel transform, for example to add partition columns,DeltaScanStreamevaluates it againstscan.logical_schema(). That schema still has the original strict Delta nullability. The kernel rebuilds the structs with that schema, and Arrow rejects the nulls. Without a partition filter, no transform runs, which is why those queries work.Fix:
In
DeltaScanExec::execute,build the transform output type from the logical schema with the existingrewrite_schema_with_nullable_fields applied(kernel → Arrow → rewrite → kernel). Column-mapped partitioned tables (table_with_partitioning_mapping, table_with_column_mapping) read correctly through this path.Related Issue(s)
Documentation