Skip to content

[Parquet][Python] pq.read_table should allow passing MemoryPool #54637

Closed

Description

@pitrou

Describe the enhancement requested

pq.read_table currently doesn't allow passing a memory pool argument, even though reading a Parquet file can be very allocation-heavy.

Component(s)

Python, Parquet

Activity

  1. changed the title [-][Python] pq.read_table should allow passing MemoryPool[/-] [+][Parquet][Python] pq.read_table should allow passing MemoryPool[/+] on Oct 19, 2026
  2. omkar-334 commented on Oct 19, 2026

    @omkar-334
    Contributor

    @pitrou can i work on this? Id like to give this a shot

  3. akshat-lakhera commented on Oct 19, 2026

    @akshat-lakhera

    take

    I'd like to work on this. My plan is to add a memory_pool argument to pq.read_table (and ParquetDataset / ParquetFile.read where it applies) and pass it down to the Parquet reader, with a test using a proxy memory pool to check that the allocations go through it. I'll open a PR once I have it working.

  4. pitrou commented on Oct 19, 2026

    @pitrou
    MemberAuthor

    @omkar-334 Feel free to work on this, of course.

    @akshat-lakhera You came slightly later, sorry. Also, please avoid posting AI-generated messages.

  5. akshat-lakhera commented on Oct 19, 2026

    @akshat-lakhera

    Okay @pitrou, will browse some other issues. And will not AI generated messages from now on ^^

  6. added 6 commits that reference this issue on Oct 23, 2026
  7. added a commit that references this issue on Oct 28, 2026
  8. added this to the 27.0.0 milestone on Oct 28, 2026
  9. AlenkaF commented on Oct 28, 2026

    @AlenkaF
    Member

    Issue resolved by pull request 54919
    #54919

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

Type

No type

Projects

No projects

    Milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions