geotiff: stamp _xrspatial_geotiff_contract=1 on every read (#1984)

xrspatial/geotiff/_attrs.py

@@ -52,6 +52,13 @@
 _TIFF_SHORT = 3
 
 
+# Contract version emitted on every read; bumped when the attrs contract
+# changes. Downstream code reads ``attrs['_xrspatial_geotiff_contract']``
+# to learn which attrs-contract revision produced the array. See issue
+# #1984 and ``docs/source/user_guide/attrs_contract.rst``.
+_ATTRS_CONTRACT_VERSION = 1
+
+
 # String identifiers (used in xrspatial attrs) -> TIFF ResolutionUnit tag ids.
 _RESOLUTION_UNIT_IDS = {'none': 1, 'inch': 2, 'centimeter': 3}
 
@@ -100,7 +107,19 @@ def _populate_attrs_from_geo_info(attrs: dict, geo_info, *, window=None) -> None
     the outer window through this helper so the resulting DataArray
     advertises the windowed transform. The GPU path does not currently
     expose a windowed read, so it passes ``window=None``.
+
+    ``attrs['_xrspatial_geotiff_contract']`` is stamped unconditionally
+    as the first step. Any pre-existing value on the passed-in dict is
+    overwritten with the current ``_ATTRS_CONTRACT_VERSION``; callers
+    pass freshly built dicts, so this is the intended behaviour.
     """
+    # Stamp the contract version first so every read path that funnels
+    # through this helper carries the marker. The VRT backends build
+    # their attrs dict directly and stamp the version there (see
+    # ``_backends/vrt.py``); keep both sites in sync via the constant
+    # rather than the bare literal.
+    attrs['_xrspatial_geotiff_contract'] = _ATTRS_CONTRACT_VERSION
+
     if geo_info.crs_epsg is not None:
         attrs['crs'] = geo_info.crs_epsg
     if geo_info.crs_wkt is not None:

xrspatial/geotiff/_backends/vrt.py

@@ -13,6 +13,7 @@
 import numpy as np
 import xarray as xr
 
+from .._attrs import _ATTRS_CONTRACT_VERSION
 from .._coords import (
     coords_from_pixel_geometry as _coords_from_pixel_geometry,
     transform_tuple_from_pixel_geometry as _transform_tuple_from_pixel_geometry,
@@ -193,7 +194,10 @@ def read_vrt(source: str, *,
     else:
         coords = {}
 
-    attrs = {}
+    # VRT builds its attrs dict inline rather than going through
+    # ``_populate_attrs_from_geo_info``; stamp the contract version here
+    # so both code paths emit the same marker.
+    attrs = {'_xrspatial_geotiff_contract': _ATTRS_CONTRACT_VERSION}
     if vrt.crs_wkt:
         epsg = _wkt_to_epsg(vrt.crs_wkt)
         if epsg is not None:
@@ -562,7 +566,10 @@ def _read_vrt_chunked(source, *, window, band, name, chunks, gpu, dtype,
     # eager reads share the same x/y arrays.
     gt = vrt.geo_transform
     coords = {}
-    attrs = {}
+    # Mirrors the eager VRT branch: this code path bypasses
+    # ``_populate_attrs_from_geo_info``, so the contract version is
+    # stamped inline using the shared constant to stay in lockstep.
+    attrs = {'_xrspatial_geotiff_contract': _ATTRS_CONTRACT_VERSION}
     if gt is not None:
         origin_x, res_x, _, origin_y, _, res_y = gt
         coord_window = (win_r0, win_c0, win_r0 + full_h, win_c0 + full_w)

xrspatial/geotiff/tests/test_attrs_contract_version_1984.py

@@ -0,0 +1,141 @@
+"""Contract-version marker tests for issue #1984.
+
+PR 3 of the 7-PR plan attached to issue #1984 stamps every DataArray
+returned by an xrspatial geotiff read path with
+``attrs['_xrspatial_geotiff_contract'] = 1``. Downstream code reads
+this marker to learn which attrs-contract revision produced the array.
+
+The stamp must appear on every backend:
+
+* eager numpy (``open_geotiff``)
+* dask + numpy (``open_geotiff(chunks=...)`` / ``read_geotiff_dask``)
+* cupy / GPU (``open_geotiff(gpu=True)`` / ``read_geotiff_gpu``)
+* dask + cupy (``open_geotiff(gpu=True, chunks=...)``)
+* VRT eager (``read_vrt``)
+* VRT dask chunked (``read_vrt(chunks=...)``)
+
+The fixture style mirrors ``test_attrs_parity_1548.py``: build a small
+on-disk TIFF (and a small VRT pointing at one) inside ``tmp_path``,
+open it through each backend, and assert on the resulting attrs.
+"""
+from __future__ import annotations
+
+import importlib.util
+import os
+
+import numpy as np
+import pytest
+
+from xrspatial.geotiff import open_geotiff, read_vrt
+from xrspatial.geotiff._attrs import _ATTRS_CONTRACT_VERSION
+
+tifffile = pytest.importorskip("tifffile")
+
+
+_CONTRACT_KEY = '_xrspatial_geotiff_contract'
+
+
+def _gpu_available() -> bool:
+    if importlib.util.find_spec("cupy") is None:
+        return False
+    try:
+        import cupy
+        return bool(cupy.cuda.is_available())
+    except Exception:
+        return False
+
+
+_HAS_GPU = _gpu_available()
+_gpu_only = pytest.mark.skipif(not _HAS_GPU, reason="cupy + CUDA required")
+
+
+def _write_small_tiff(path):
+    """Write a small tiled float32 TIFF used by every read-path assertion."""
+    arr = np.arange(64 * 64, dtype=np.float32).reshape(64, 64)
+    tifffile.imwrite(
+        path, arr, photometric='minisblack', planarconfig='contig',
+        tile=(32, 32), compression='deflate', metadata=None,
+    )
+    return arr
+
+
+def _write_minimal_vrt(vrt_path, source_name, *, height, width):
+    """Write a VRT that references ``source_name`` as a single-band source."""
+    vrt_path.write_text(
+        f'<VRTDataset rasterXSize="{width}" rasterYSize="{height}">\n'
+        '  <VRTRasterBand dataType="Float32" band="1">\n'
+        '    <SimpleSource>\n'
+        f'      <SourceFilename relativeToVRT="1">{source_name}'
+        '</SourceFilename>\n'
+        '      <SourceBand>1</SourceBand>\n'
+        f'      <SrcRect xOff="0" yOff="0" xSize="{width}" ySize="{height}"/>\n'
+        f'      <DstRect xOff="0" yOff="0" xSize="{width}" ySize="{height}"/>\n'
+        '    </SimpleSource>\n'
+        '  </VRTRasterBand>\n'
+        '</VRTDataset>\n'
+    )
+
+
+def test_attrs_contract_version_constant_is_one():
+    """Pin the integer value so a careless bump shows up here first."""
+    assert _ATTRS_CONTRACT_VERSION == 1
+
+
+def test_eager_numpy_stamps_contract_version(tmp_path):
+    path = str(tmp_path / "contract_v1_eager.tif")
+    _write_small_tiff(path)
+
+    da = open_geotiff(path)
+
+    assert da.attrs[_CONTRACT_KEY] == 1
+
+
+def test_dask_numpy_stamps_contract_version(tmp_path):
+    path = str(tmp_path / "contract_v1_dask.tif")
+    _write_small_tiff(path)
+
+    da = open_geotiff(path, chunks=32)
+
+    assert da.attrs[_CONTRACT_KEY] == 1
+
+
+@_gpu_only
+def test_gpu_stamps_contract_version(tmp_path):
+    path = str(tmp_path / "contract_v1_gpu.tif")
+    _write_small_tiff(path)
+
+    da = open_geotiff(path, gpu=True)
+
+    assert da.attrs[_CONTRACT_KEY] == 1
+
+
+@_gpu_only
+def test_dask_gpu_stamps_contract_version(tmp_path):
+    path = str(tmp_path / "contract_v1_dask_gpu.tif")
+    _write_small_tiff(path)
+
+    da = open_geotiff(path, gpu=True, chunks=32)
+
+    assert da.attrs[_CONTRACT_KEY] == 1
+
+
+def test_vrt_eager_stamps_contract_version(tmp_path):
+    src = tmp_path / "contract_v1_vrt_source.tif"
+    _write_small_tiff(str(src))
+    vrt = tmp_path / "contract_v1_vrt_eager.vrt"
+    _write_minimal_vrt(vrt, os.path.basename(src), height=64, width=64)
+
+    da = read_vrt(str(vrt))
+
+    assert da.attrs[_CONTRACT_KEY] == 1
+
+
+def test_vrt_chunked_stamps_contract_version(tmp_path):
+    src = tmp_path / "contract_v1_vrt_chunked_source.tif"
+    _write_small_tiff(str(src))
+    vrt = tmp_path / "contract_v1_vrt_chunked.vrt"
+    _write_minimal_vrt(vrt, os.path.basename(src), height=64, width=64)
+
+    da = read_vrt(str(vrt), chunks=32)
+
+    assert da.attrs[_CONTRACT_KEY] == 1