Skip to content

Conversation

keewis
Copy link
Contributor

@keewis keewis commented Aug 30, 2025

closes #3416

Basically, we're changing the default for endian when constructing BytesCodec using from_dict.

(I also renamed test_endian.py to test_bytes.py as mentioned in #3416 (comment))

TODO:

  • Add unit tests and/or doctests in docstrings
  • Add docstrings and API docs for any new/modified user-facing classes and functions
  • New/modified features documented in docs/user-guide/*.rst
  • Changes documented as a new file in changes/
  • GitHub Actions have all passed
  • Test coverage is 100% (Codecov passes)

@github-actions github-actions bot added the needs release notes Automatically applied to PRs which haven't added release notes label Aug 30, 2025
@github-actions github-actions bot removed the needs release notes Automatically applied to PRs which haven't added release notes label Aug 30, 2025
@keewis
Copy link
Contributor Author

keewis commented Aug 30, 2025

looks like some tests actually relied on this behavior, i.e. choosing the default endian if none was specified for multi-byte dtypes (even though this technically violates the spec).

Since this might be potentially user facing (I'm not certain, but at least virtualizarr uses ArrayV3Metadata directly) someone might be relying on this behavior?

What we could do is start a deprecation cycle by checking for this in ArrayV3Metadata.from_dict and warning, then adding the system default. I'll leave the decision to you.

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.

BytesCodec's endian does not roundtrip
2 participants