[do not merge] zarrs bindings

.github/workflows/engine.yml

@@ -0,0 +1,49 @@
+name: Array engine
+
+on:
+  push:
+    branches: [ main ]
+    paths:
+      - 'src/zarr/**'
+      - 'tests/engine/**'
+      - 'tests/zarrista/**'
+      - 'pyproject.toml'
+      - '.github/workflows/engine.yml'
+  pull_request:
+    branches: [ main ]
+    paths:
+      - 'src/zarr/**'
+      - 'tests/engine/**'
+      - 'tests/zarrista/**'
+      - 'pyproject.toml'
+      - '.github/workflows/engine.yml'
+  workflow_dispatch:
+
+permissions: {}
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  test:
+    runs-on: ubuntu-latest
+    permissions:
+      contents: read
+    steps:
+      - uses: actions/checkout@df4cb1c069e1874edd31b4311f1884172cec0e10 # v6.0.3
+        with:
+          fetch-depth: 0  # hatch-vcs needs tags to compute zarr's version
+          persist-credentials: false
+      - name: Install Rust
+        uses: dtolnay/rust-toolchain@29eef336d9b2848a0b548edc03f92a220660cdb8 # stable
+      - name: Install uv
+        uses: astral-sh/setup-uv@fac544c07dec837d0ccb6301d7b5580bf5edae39 # v8.2.0
+        with:
+          python-version: '3.12'
+      - name: Sync zarrista group
+        run: uv sync --group zarrista
+      - name: Run engine tests
+        # the ubuntu runner image ships a Rust toolchain; the maturin build
+        # backend for zarrista is fetched by uv on demand
+        run: uv run --group zarrista pytest tests/engine tests/zarrista -v

changes/4181.feature.md

@@ -0,0 +1,22 @@
+`Array` and `AsyncArray` now route data I/O through pluggable *array engines*
+(`zarr.abc.engine.ArrayEngine` / `AsyncArrayEngine`). The default engine
+preserves existing behavior on every store and format. Pass
+`engine="zarrista"` (requires the `zarrista` package) to serve Zarr v3 arrays
+on local-filesystem, obstore, or icechunk storage through the Rust `zarrs`
+implementation. Call `zarr.list_engines()` to discover the available engine
+names.
+
+As part of this rewiring, writing a single field of a structured dtype (e.g.
+`z["x"] = ...`) now preserves the values of the array's other fields; it
+previously clobbered them.
+
+Behavior note: strided and fancy (orthogonal/coordinate) selections are now
+served internally via the contiguous bounding box that covers the selected
+indices, with the requested elements picked out of that box afterwards. For
+sparse selections this can transfer more data from storage than before, and
+with the `array.read_missing_chunks` config set to `False`, a missing chunk
+that lies within the bounding box will raise even if no selected element
+falls inside that chunk. Relatedly, a partial (non-full-box) write now performs
+an internal read of the bounding box before writing it back, so with
+`array.read_missing_chunks` set to `False` such a write raises
+`ChunkNotFoundError` when it covers an uninitialized chunk.

docs/api/zarr/abc/engine.md

@@ -0,0 +1,5 @@
+---
+title: engine
+---
+
+::: zarr.abc.engine

docs/api/zarr/zarrista.md

@@ -0,0 +1,5 @@
+---
+title: zarrista
+---
+
+::: zarr.zarrista