Merge pull request #48 from bcdev/forman-rule_updates

Rule updates

Author: forman

CHANGES.md

@@ -2,19 +2,32 @@
 
 ## Version 0.5.0 (in development)
 
-### Incompatible API changes:
+### Changes
 
-- Renamed nodes and node properties for consistency and clarity:
-  - renamed `DataArrayNode` into `VariableNode`
-  - renamed `DataArrayNode.data_array` into `VariableNode.array`
+- Rule `no-empty-chunks` has taken off the `"recommended"` settings 
+  as there is no easy/efficient way to tell whether a dataset has 
+  been written using `write_emtpy_chunks` option or not.
+  The rule message itself has been fixed. (#45) 
+
+- Adjusted messages of rules `var-units` and `time-coordinate` 
+  to be consistent with messages of other rules.
+
+- Core rule `dataset-title-attr` has been moved into `xcube` plugin
+  and renamed to `xcube/dataset-title` because the core rule `var-descr` 
+  covers checking for dataset titles.
+
+### Incompatible API changes
 
 - Changed general use of term _verify_ into _validate_: 
   - prefixed `RuleOp` methods by `validate_` for clarity.
   - renamed `XRLint.verify_datasets()` into `validate_files()`
   - renamed `Lint.verify_dataset()` into `validate()`
 
-- Various changes for improved clarity and consistency 
-  regarding configuration management:
+- Renamed nodes and node properties for clarity and consistency:
+  - renamed `DataArrayNode` into `VariableNode`
+  - renamed `DataArrayNode.data_array` into `VariableNode.array`
+
+- Various changes for improved clarity regarding configuration management:
   - introduced type aliases `ConfigLike` and `ConfigObjectLike`.
   - renamed `Config` into `ConfigObject` 
   - renamed `ConfigList.configs` into `config_objects` 
@@ -25,6 +38,10 @@
   - added class method `from_config()` to `ConfigList`.
   - removed function `xrlint.config.merge_configs` as it was no longer used.
 
+### Other changes
+
+- Added more tests so we finally reached 100% coverage.
+
 ## Version 0.4.1 (from 2025-01-31)
 
 ### Changes

docs/rule-ref.md

@@ -32,12 +32,6 @@ Dimensions of data variables should have corresponding coordinates.
 
 Contained in:  `all`-:material-lightning-bolt: `recommended`-:material-lightning-bolt:
 
-### :material-lightbulb: `dataset-title-attr`
-
-Datasets should be given a non-empty title.
-
-Contained in:  `all`-:material-lightning-bolt: `recommended`-:material-alert:
-
 ### :material-bug: `grid-mappings`
 
 Grid mappings, if any, shall have valid grid mapping coordinate variables.
@@ -69,7 +63,7 @@ Contained in:  `all`-:material-lightning-bolt: `recommended`-:material-alert:
 Empty chunks should not be encoded and written. The rule currently applies to Zarr format only.
 [:material-information-variant:](https://docs.xarray.dev/en/stable/generated/xarray.Dataset.to_zarr.html#xarray-dataset-to-zarr)
 
-Contained in:  `all`-:material-lightning-bolt: `recommended`-:material-alert:
+Contained in:  `all`-:material-lightning-bolt: `recommended`-:material-circle-off-outline:
 
 ### :material-bug: `time-coordinate`
 
@@ -122,6 +116,13 @@ Spatial data variables should encode xcube color mappings in their metadata.
 
 Contained in:  `all`-:material-lightning-bolt: `recommended`-:material-alert:
 
+### :material-bug: `dataset-title`
+
+Datasets should be given a non-empty title.
+[:material-information-variant:](https://xcube.readthedocs.io/en/latest/cubespec.html#metadata)
+
+Contained in:  `all`-:material-lightning-bolt: `recommended`-:material-lightning-bolt:
+
 ### :material-lightbulb: `grid-mapping-naming`
 
 Grid mapping variables should be called 'spatial_ref' or 'crs' for compatibility with rioxarray and other packages.

docs/todo.md

@@ -24,7 +24,6 @@
 
 ## Nice to have
 
-- add some more tests so we reach 100% coverage
 - support `autofix` feature
 - support `md` (markdown) output format
 - support formatter op args/kwargs and apply validation schema

notebooks/mkdataset.py

@@ -1,16 +1,28 @@
 import numpy as np
 import xarray as xr
 
-nx = 2
-ny = 3
-nt = 4
+nx = 720
+ny = 360
+nt = 5
+
+ncx = 180
+ncy = 180
+nct = 1
 
 
 def make_dataset() -> xr.Dataset:
     """Create a dataset that passes xrlint core rules."""
 
     return xr.Dataset(
-        attrs=dict(title="SST-Climatology Subset"),
+        attrs=dict(
+            Conventions="CF-1.10",
+            title="SST-Climatology Subset",
+            history="2025-01-31 17:31:00 - created;",
+            institution="BC",
+            source="SST CCI L4",
+            references="https://climate.esa.int/en/projects/sea-surface-temperature/",
+            comment="Demo dataset",
+        ),
         coords={
             "x": xr.DataArray(
                 np.linspace(-180, 180, nx),
@@ -53,15 +65,25 @@ def make_dataset() -> xr.Dataset:
             "sst": xr.DataArray(
                 np.random.random((nt, ny, nx)),
                 dims=["time", "y", "x"],
-                attrs={"units": "kelvin", "grid_mapping": "spatial_ref"},
+                attrs={
+                    "standard_name": "sea_surface_temperature",
+                    "long_name": "sea surface temperature",
+                    "units": "kelvin",
+                    "grid_mapping": "spatial_ref",
+                },
             ),
             "sst_anomaly": xr.DataArray(
                 np.random.random((nt, ny, nx)),
                 dims=["time", "y", "x"],
-                attrs={"units": "kelvin", "grid_mapping": "spatial_ref"},
+                attrs={
+                    "standard_name": "sea_surface_temperature_anomaly",
+                    "long_name": "sea surface temperature anomaly",
+                    "units": "kelvin",
+                    "grid_mapping": "spatial_ref",
+                },
             ),
         },
-    )
+    ).chunk(time=nct, y=ncy, x=ncx)
 
 
 def make_dataset_with_issues() -> xr.Dataset: