Issue527 load stac metadata

openeo/metadata.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import logging
+import pystac
 import warnings
 from typing import Any, Callable, List, NamedTuple, Optional, Tuple, Union
 
@@ -522,3 +523,61 @@ def _repr_html_(self):
     def __str__(self) -> str:
         bands = self.band_names if self.has_band_dimension() else "no bands dimension"
         return f"CollectionMetadata({self.extent} - {bands} - {self.dimension_names()})"
+
+
+def metadata_from_stac(url: str) -> CubeMetadata:
+    """
+    Reads the band metadata a static STAC catalog or a STAC API Collection and returns it as a :py:class:`CubeMetadata`
+
+    :param url: The URL to a static STAC catalog (STAC Item, STAC Collection, or STAC Catalog) or a specific STAC API Collection
+    :return: A :py:class:`CubeMetadata` containing the DataCube band metadata from the url.
+    """
+
+    def get_band_names(asst: pystac.Asset) -> List[Band]:
+        return [Band(eo_band["name"]) for eo_band in asst.extra_fields["eo:bands"]]
+
+    def is_band_asset(asset: pystac.Asset) -> bool:
+        return "eo:bands" in asset.extra_fields
+
+    stac_object = pystac.read_file(href=url)
+
+    band_names = []
+    collection = None
+
+    if isinstance(stac_object, pystac.Item):
+        item = stac_object
+        if "eo:bands" in item.properties:
+            eo_bands_location = item.properties
+        elif item.get_collection() is not None:
+            collection = item.get_collection()
+            eo_bands_location = item.get_collection().summaries.lists
+        else:
+            eo_bands_location = {}
+        band_names = [Band(b["name"]) for b in eo_bands_location.get("eo:bands", [])]
+
+    elif isinstance(stac_object, pystac.Collection):
+        collection = stac_object
+        band_names = [Band(b["name"]) for b in collection.summaries.lists.get("eo:bands", [])]
+
+        # Summaries is not a required field in a STAC collection, so also check the assets
+        for itm in collection.get_items():
+            band_assets = {
+                asset_id: asset
+                for asset_id, asset in dict(sorted(itm.get_assets().items())).items()
+                if is_band_asset(asset)
+            }
+
+            for asset in band_assets.values():
+                asset_band_names = get_band_names(asset)
+                for asset_band_name in asset_band_names:
+                    if asset_band_name not in band_names:
+                        band_names.append(asset_band_name)
+
+    else:
+        assert isinstance(stac_object, pystac.Catalog)
+        catalog = stac_object
+        band_names = [Band(b["name"]) for b in catalog.extra_fields.get("summaries", {}).get("eo:bands", [])]
+
+    band_dimension = BandDimension(name="bands", bands=band_names)
+    metadata = CubeMetadata(dimensions=[band_dimension])
+    return metadata

openeo/rest/connection.py

@@ -27,7 +27,14 @@
 from openeo.internal.jupyter import VisualDict, VisualList
 from openeo.internal.processes.builder import ProcessBuilderBase
 from openeo.internal.warnings import deprecated, legacy_alias
-from openeo.metadata import Band, BandDimension, CollectionMetadata, SpatialDimension, TemporalDimension
+from openeo.metadata import (
+    Band,
+    BandDimension,
+    CollectionMetadata,
+    SpatialDimension,
+    TemporalDimension,
+    metadata_from_stac,
+)
 from openeo.rest import (
     CapabilitiesException,
     OpenEoApiError,
@@ -1361,6 +1368,10 @@ def load_stac(
                 prop: build_child_callback(pred, parent_parameters=["value"]) for prop, pred in properties.items()
             }
         cube = self.datacube_from_process(process_id="load_stac", **arguments)
+        try:
+            cube.metadata = metadata_from_stac(url)
+        except Exception:
+            _log.warning("Python client could not read band metadata from URL.")
         return cube
 
     def load_ml_model(self, id: Union[str, BatchJob]) -> MlModel:

setup.py

@@ -30,6 +30,7 @@
     "flake8>=5.0.0",
     "time_machine",
     "pyproj>=3.2.0",  # Pyproj is an optional, best-effort runtime dependency
+    "pystac"
 ]
 
 docs_require = [

tests/test_metadata.py

@@ -2,6 +2,7 @@
 
 from typing import List
 
+import json
 import pytest
 
 from openeo.metadata import (
@@ -14,6 +15,7 @@
     MetadataException,
     SpatialDimension,
     TemporalDimension,
+    metadata_from_stac,
 )
 
 
@@ -782,3 +784,50 @@ def filter_bbox(self, bbox):
     assert isinstance(new, MyCubeMetadata)
     assert orig.bbox is None
     assert new.bbox == (1, 2, 3, 4)
+
+
+collection_json = {
+    "type": "Collection",
+    "id": "test-collection",
+    "stac_version": "1.0.0",
+    "description": "Test collection",
+    "links": [],
+    "title": "Test Collection",
+    "extent": {
+        "spatial": {"bbox": [[-180.0, -90.0, 180.0, 90.0]]},
+        "temporal": {"interval": [["2020-01-01T00:00:00Z", "2020-01-10T00:00:00Z"]]},
+    },
+    "license": "proprietary",
+    "summaries": {"eo:bands": [{"name": "B01"}, {"name": "B02"}]},
+}
+
+catalog_json = {
+    "type": "Catalog",
+    "id": "test-catalog",
+    "stac_version": "1.0.0",
+    "description": "Test Catalog",
+    "links": [],
+}
+
+item_json = {
+    "type": "Feature",
+    "stac_version": "1.0.0",
+    "id": "test-item",
+    "properties": {"datetime": "2020-05-22T00:00:00Z", "eo:bands": [{"name": "SCL"}, {"name": "B08"}]},
+    "geometry": {"coordinates": [[[0, 0], [0, 1], [1, 1], [1, 0], [0, 0]]], "type": "Polygon"},
+    "links": [],
+    "assets": {},
+    "bbox": [0, 1, 0, 1],
+    "stac_extensions": [],
+}
+
+
+@pytest.mark.parametrize(
+    "test_stac, expected", [(collection_json, ["B01", "B02"]), (catalog_json, []), (item_json, ["SCL", "B08"])]
+)
+def test_metadata_from_stac(tmp_path, test_stac, expected):
+
+    path = tmp_path / "stac.json"
+    path.write_text(json.dumps(test_stac))
+    metadata = metadata_from_stac(path)
+    assert metadata.band_names == expected