Skip to content

Fix serialization of large integers in JSON (#5038)#5042

Merged
tustvold merged 1 commit intoapache:masterfrom
tustvold:do-not-sign-extend-json-serialize
Nov 6, 2023
Merged

Fix serialization of large integers in JSON (#5038)#5042
tustvold merged 1 commit intoapache:masterfrom
tustvold:do-not-sign-extend-json-serialize

Conversation

@tustvold
Copy link
Contributor

@tustvold tustvold commented Nov 6, 2023

Which issue does this PR close?

Closes #5038

Rationale for this change

When converting i32 to i64 by default Rust will sign extend, we don't want this behaviour

What changes are included in this PR?

Are there any user-facing changes?

@github-actions github-actions bot added the arrow Changes to the arrow crate label Nov 6, 2023
@alamb alamb changed the title Fix serialization of large integers (#5038) Fix serialization of large integers in JSON (#5038) Nov 6, 2023
TapeElement::I64(high) => match self.get(idx + 1) {
TapeElement::I32(low) => {
let val = (high as i64) << 32 | low as i64;
let val = (high as i64) << 32 | (low as u32) as i64;
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can you confirm this code is only used for debugging? it wasn't clear to me that this is covered by a test, but if it is only used in

self.serialize(&mut out, idx);
I don't think that is necessary

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's only used in error messages

Copy link
Contributor

@alamb alamb left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

👨‍🍳 👌

I also double checked and it seems like this pattern to use the unsigned variant is done for all other uses in arrow:

Screenshot 2023-11-06 at 8 56 11 AM

@tustvold tustvold merged commit 91acfb0 into apache:master Nov 6, 2023
@alamb alamb mentioned this pull request Nov 7, 2023
6 tasks
alamb pushed a commit to alamb/arrow-rs that referenced this pull request Nov 8, 2023
alamb added a commit that referenced this pull request Nov 9, 2023
…0.0_maintenance` (#5059)

* Fix serialization of large integers (#5038) (#5042)

* fmt

---------

Co-authored-by: Raphael Taylor-Davies <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

arrow Changes to the arrow crate

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Regression when serializing large json numbers

2 participants