You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Commit 8aef319
Browse filesBrowse the repository at this point in the historyBrowse files
refactor(metadata): drop numpy and tuple widening from chunk grid metadata
The chunk grid metadata classes take typed Python values again: a
regular chunk shape of `int`s, rectilinear dimensions of `int` or
`list` in stored documents. `declares_chunk_edges` and the numpy and
tuple handling in `expand_rle`, `_validate_chunk_shapes` and
`RectilinearChunkGridMetadata.from_dict` are removed with their tests;
numpy input belongs to the `chunks=` normalizer. The regular grid still
rejects edge lists (zarr-developersgh-4374), now naming the offending type instead of
echoing the value.
Assisted-by: ClaudeCode:claude-opus-5-5
Co-Authored-By: Claude Opus 5.5 <[email protected]>
0 commit comments