jackylee-ch opened a new pull request, #10158:
URL: https://github.com/apache/paimon/pull/10158

   ### Purpose
   
   PyPaimon ignored the `data-file.path-directory` table option. 
`FileStoreTable` built its `FileStorePathFactory` with 
`data_file_path_directory=None`, and on the read side 
`DataFileMeta.set_file_path` built `<table>/<partition>/bucket-N` without the 
directory. A table created with `data-file.path-directory` set therefore wrote 
and read its data files at a location that disagrees with Java/Spark/Flink, 
which place them under `<table>/<dir>/<partition>/bucket-N` — cross-engine 
reads hit `FileNotFound` in both directions.
   
   `options.data_file_path_directory()` is now passed into `path_factory()`, 
and threaded through `set_file_path` (and its three split-generator call 
sites), inserting the directory right after the table root to match Java 
`CoreOptions.dataFilePathDirectory()` and `FileStorePathFactory`. The option 
was already in `IMMUTABLE_OPTIONS`; this adds the missing `ConfigOption` and 
accessor. The unset default keeps files at the bucket root, so existing tables 
are unaffected.
   
   ### Tests
   
   `data_file_path_directory_test` writes a table with the option set and 
asserts the data files land under the directory and read back, plus a 
default-layout case.
   
   Written with Claude Code; verification is mine.
   


-- 
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]

Reply via email to