Skip to content

ARROW-3919: [Python] Support 64 bit indices for pyarrow.serialize and pyarrow.deserialize#3062

Closed
pcmoritz wants to merge 9 commits intoapache:masterfrom
pcmoritz:serialize-indices
Closed

ARROW-3919: [Python] Support 64 bit indices for pyarrow.serialize and pyarrow.deserialize#3062
pcmoritz wants to merge 9 commits intoapache:masterfrom
pcmoritz:serialize-indices

Conversation

@pcmoritz
Copy link
Copy Markdown
Contributor

No description provided.

@wesm
Copy link
Copy Markdown
Member

wesm commented Dec 14, 2018

@pcmoritz can you rebase? would you like me to review this?

@pcmoritz
Copy link
Copy Markdown
Contributor Author

@wesm This is not done yet, but do you want to take a brief look if that's the direction we want to go in (make ListArray offsets int64_t)?

@pcmoritz
Copy link
Copy Markdown
Contributor Author

pcmoritz commented Dec 23, 2018

After looking into this a bit more, the alternative route would be to use ChunkedArrays and the new ChunkedBuilders etc. Which wouldn't blow up everybody's ListArray offsets by 2x.

We would need a ChunkedUnionArrayBuilder to implement this I think.

@wesm
Copy link
Copy Markdown
Member

wesm commented Dec 24, 2018

@pcmoritz this isn't consistent with the columnar format specification (list offsets are 32 bits). We'll have to find another way

@wesm
Copy link
Copy Markdown
Member

wesm commented May 16, 2019

Closing per discussions on mailing list

@wesm wesm closed this May 16, 2019
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants