Akash3121 commented on code in PR #10158:
URL: https://github.com/apache/paimon/pull/10158#discussion_r4098327458
##########
paimon-python/pypaimon/table/file_store_table.py:
##########
@@ -392,7 +392,7 @@ def path_factory(self) -> 'FileStorePathFactory':
legacy_partition_name=self.options.options.get(CoreOptions.PARTITION_GENERATE_LEGACY_NAME),
file_suffix_include_compression=False,
file_compression=file_compression,
- data_file_path_directory=None,
Review Comment:
This updates Python writes to place files under the configured directory,
but native split planning still resolves them at the old bucket-root location.
The required `Python / Rust Plan` job demonstrates the mismatch: the new test
writes the Parquet file under `data/...`, then the native-planned read fails
with `NotFound / No such file or directory` . Please either add
`data-file.path-directory` support to the native planner/runtime used here or
make `_try_native_plan` fall back to Python planning whenever this option is
configured. The regression test should also run with native planning enabled so
this path cannot silently diverge again.
##########
paimon-python/pypaimon/read/scanner/data_evolution_split_generator.py:
##########
@@ -139,7 +139,8 @@ def _build_split_from_pack_for_data_evolution(
self.table.table_path,
file_entries[0].partition,
file_entries[0].bucket,
- self.default_part_value
+ self.default_part_value,
+ self.table.options.data_file_path_directory()
Review Comment:
This new required method is missing from the existing lightweight `_Options`
implementations used by the data-evolution scanner tests. As a result, the
focused suite now fails six existing cases with `AttributeError: '_Options'
object has no attribute 'data_file_path_directory'`, and the same failures
appear across the Python 3.10 - 3.13 CI jobs. Please update all affected option
doubles (and the equivalent calls in the other split generators as needed), or
retrieve this optional value through a shared compatibility helper that
defaults to `None`.
--
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]
