API reference

The modules are layered: the transform algebra at the bottom, chunk resolution and the wire format built on top of it.

The transform algebra

• zarr_indexing.domain — IndexDomain, a rectangular region of integer coordinates with an explicit (possibly non-zero) origin
• zarr_indexing.output_map — ConstantMap, DimensionMap, and ArrayMap: three representations of a set of integer coordinates, one per storage dimension
• zarr_indexing.transform — IndexTransform, which pairs a domain with output maps, plus the indexing ([...], .oindex, .vindex), intersect, and translate operations, and selection_to_transform
• zarr_indexing.composition — compose, which chains two transforms into one

Chunk resolution

• zarr_indexing.chunk_resolution — iter_chunk_transforms (transform + chunk grid → per-chunk transforms) and sub_transform_to_selections (the bridge back to the selection tuples the current codec pipeline expects)
• zarr_indexing.grid — DimensionGridLike, the Protocol describing the narrow chunk-grid surface chunk resolution consumes, so that nothing here imports zarr

The ndsel wire format (see the guide)

• zarr_indexing.messages — parse_ndsel / normalize_ndsel, the pure JSON→JSON message layer, and NdselError
• zarr_indexing.json — lowering between canonical ndsel bodies and in-memory transforms

Errors

• zarr_indexing.errors — the canonical index-error types, which zarr.errors re-exports by identity

Every name listed in zarr_indexing.__all__ is re-exported at the top level, so from zarr_indexing import IndexTransform and from zarr_indexing.transform import IndexTransform are equivalent.