AdityRajGupta opened a new pull request, #55346: URL: https://github.com/apache/arrow/pull/55346
### Rationale for this change This closes #39530. The Dataset Parquet reader already lets you pass CacheOptions, but ParquetFile and IpcReadOptions did not, so there was no way to control the pre-buffer cache when reading through those paths. ### What changes are included in this PR? - ParquetFile now accepts cache_options and passes it through to the reader. - ParquetReader.open() now accepts cache_options and sets it on the Arrow reader properties. - IpcReadOptions now exposes pre_buffer_cache_options, so you can read it back and set it, and it round-trips when the options are wrapped. - The C++ field was already there, this only wires it up on the Python side. No C++ behaviour changes. ### Are these changes tested? Yes. <img width="808" height="427" alt="Screenshot 2026-10-31 at 2 33 30 AM" src="https://github.com/user-attachments/assets/3905cdd6-d006-4c41-a4d2-856715c5f19b"; /> ### Are there any user-facing changes? Yes ### Was AI used for this PR? Yes. AI helped draft the initial patch. I reviewed the code, fixed the build issues, and ran the tests on my Mac before submitting. **PR code and description written by:** - [X] Human - [X] AI **Reviewed before submission by:** - [X] Human - [ ] AI - [ ] Not reviewed -- 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]
