Refactor checks, update plugins and configs, improve docs

Author: Ayoubnac1

README.md

@@ -13,14 +13,20 @@ $ pip install cc-plugin-wcrp
 ```
 See the [**IOOS/compliance-checker**](https://github.com/ioos/compliance-checker#installation) for additional Installation notes.
 
-And then install Esgvoc and universe to get the Controlled Vocabulary :
+And then install Esgvoc and the other projects to get the Controlled Vocabulary :
 
 ```shell
-$ esgvoc config set universe:branch=esgvoc_dev
 $ esgvoc config add cordex-cmip6
+$ esgvoc config add cmip7
 $ esgvoc install
 ```
+You should verify with `esgvoc config show` that all projects are on the `esgvoc` branch.
 
+If not, run:
+```bash
+esgvoc config set project:branch=esgvoc 
+esgvoc install
+```
 ## Usage
 
 ```shell

checks/attribute_checks/check_attribute_suite.py

@@ -62,6 +62,7 @@ def check_attribute_suite(
     severity: int,
     value_type: Optional[str] = None,
     is_required: bool = True,
+    na_value: Optional[Any] = None,
     pattern: Optional[str] = None,
     constant: Any = None,
     enum: Optional[Iterable[Any]] = None,
@@ -95,6 +96,11 @@ def check_attribute_suite(
 
       The ONLY allowed combination is:
         cv_source_collection + cv_source_collection_key
+
+    na_value behavior:
+      - if the attribute is missing -> normal ATTR001 logic
+      - if the attribute exists and its value equals na_value -> stop after ATTR001
+      - otherwise continue with normal validation
     """
 
     results = []
@@ -123,8 +129,6 @@ def check_attribute_suite(
     existence_ctx = TestCtx(severity, f"[ATTR001] {label} existence")
     try:
         attr_value = obj.getncattr(nc_key)
-        existence_ctx.add_pass()
-        results.append(existence_ctx.to_result())
     except AttributeError:
         if is_required:
             existence_ctx.add_failure(
@@ -133,6 +137,16 @@ def check_attribute_suite(
             results.append(existence_ctx.to_result())
         return results  # stop here if missing
 
+    # -------------------------------------------------------------------------
+    # Short-circuit on sentinel / not-applicable value
+    # -------------------------------------------------------------------------
+    if na_value is not None:
+        if str(attr_value).strip().lower() == str(na_value).strip().lower():
+            return results
+
+    existence_ctx.add_pass()
+    results.append(existence_ctx.to_result())
+
     # -------------------------------------------------------------------------
     # ATTR002 - Type check
     # -------------------------------------------------------------------------
@@ -147,6 +161,8 @@ def check_attribute_suite(
             "str": str,
             "int": (int, np.integer),
             "float": (float, np.floating),
+            "double": np.float64,
+            "simple": np.float32,
             "bool": (bool, np.bool_),
             "str_array": list,
         }
@@ -310,22 +326,48 @@ def check_attribute_suite(
         invalid: list[Any] = []
 
         try:
-            # If a key is provided, do the direct get_term_in_collection (as requested).
-            # This checks that the collection contains that term_id key.
             if cv_source_collection_key:
-                term = voc.get_term_in_collection(
+                terms = voc.get_all_terms_in_collection(
                     project_id=project_name,
                     collection_id=cv_source_collection,
-                    term_id=cv_source_collection_key,
                 )
-                if not term:
+
+                if not terms:
                     ctx.add_failure(
-                        f"CV collection '{cv_source_collection}' has no term key '{cv_source_collection_key}'."
+                        f"CV collection '{cv_source_collection}' is empty or could not be retrieved."
                     )
                     results.append(ctx.to_result())
                     return results
 
-            # Then validate the actual attribute value against the collection
+                for val in values:
+                    found = False
+                    val_norm = " ".join(str(val).strip().lower().split())
+
+                    for term in terms:
+                        candidate = getattr(term, str(cv_source_collection_key), None)
+                        if candidate is None:
+                            continue
+
+                        candidate_norm = " ".join(str(candidate).strip().lower().split())
+
+                        if val_norm == candidate_norm or val_norm in candidate_norm:
+                            found = True
+                            break
+
+                    if not found:
+                        invalid.append(val)
+
+                if invalid:
+                    ctx.add_failure(
+                        f"Value(s) {invalid} not found in field '{cv_source_collection_key}' "
+                        f"of any term in CV collection '{cv_source_collection}'."
+                    )
+                else:
+                    ctx.add_pass()
+
+                results.append(ctx.to_result())
+                return results
+
             for val in values:
                 if not voc.valid_term_in_collection(
                     value=val,
@@ -348,4 +390,4 @@ def check_attribute_suite(
         return results
 
     # If no ATTR004 rule configured, we simply return existence/type/utf8 results.
-    return results
+    return results

checks/consistency_checks/check_attributes_match_filename.py

@@ -4,7 +4,7 @@
 from compliance_checker.base import TestCtx
 from checks.utils import _parse_filename_components
 
-# CMIP6: time_range exists in filename, not a GA => parse but don't compare
+# CMIP6
 _FILENAME_KEYS_CMIP6_PARSE = [
     "variable_id",
     "table_id",
@@ -23,7 +23,7 @@
     "grid_label",
 ]
 
-# CMIP7: time_range exists in filename, not a GA => parse but don't compare
+# CMIP7
 _FILENAME_KEYS_CMIP7_PARSE = [
     "variable_id",
     "branding_suffix",
@@ -77,13 +77,6 @@ def _parse_cmip7_filename(filename: str):
 
 
 def _unwrap_facets(maybe_tuple):
-    """
-    In this repo, checks.utils._parse_filename_components may return:
-      - dict
-      - None
-      - (dict, extra)  <-- this is what broke CMIP6
-    We only need the dict.
-    """
     if isinstance(maybe_tuple, tuple) and len(maybe_tuple) > 0 and isinstance(maybe_tuple[0], dict):
         return maybe_tuple[0]
     return maybe_tuple
@@ -96,7 +89,6 @@ def check_filename_vs_global_attrs(ds, severity, filename_template_keys=None):
     Important:
     - time_range is NOT a global attribute in CMIP6 or CMIP7.
       It's parsed from filename but not compared here.
-      (Your VAR009 handles time-range vs data.)
     """
     fixed_check_id = "ATTR005"
     description = f"[{fixed_check_id}] Consistency: Filename vs Global Attributes"
@@ -109,7 +101,7 @@ def check_filename_vs_global_attrs(ds, severity, filename_template_keys=None):
 
     filename = os.path.basename(filepath)
 
-    # ---------------- CMIP7 branch ----------------
+    # ---------------- CMIP7----------------
     if _is_cmip7(ds):
         facets = _parse_cmip7_filename(filename)
         if facets is None:
@@ -119,7 +111,7 @@ def check_filename_vs_global_attrs(ds, severity, filename_template_keys=None):
             return [ctx.to_result()]
         compare_keys = _FILENAME_KEYS_CMIP7_COMPARE
 
-    # ---------------- CMIP6 (default) branch ----------------
+    # ---------------- CMIP6 ----------------
     else:
         parse_keys = filename_template_keys or _FILENAME_KEYS_CMIP6_PARSE
         facets = _parse_filename_components(filename, parse_keys)

checks/consistency_checks/check_drs_consistency.py

@@ -6,11 +6,7 @@
 from checks.utils import _get_drs_facets
 
 def _unwrap_facets(x):
-    """
-    Some implementations of checks.utils._get_drs_facets return tuples like:
-      (facets_dict, extra)
 
-    """
     if isinstance(x, tuple) and len(x) > 0 and isinstance(x[0], dict):
         return x[0]
     return x
@@ -24,13 +20,13 @@ def _unwrap_facets(x):
     "variant_label", "grid_label", "time_range",
 ]
 
-# CMIP7 directoryStructureDD (CMIP7 doc v1.0)
+# CMIP7 directoryStructure
 _dir_template_keys_cmip7 = [
     "drs_specs", "mip_era", "activity_id", "institution_id", "source_id",
     "experiment_id", "variant_label", "region", "frequency", "variable_id",
     "branding_suffix", "grid_label", "directory_date",
 ]
-# CMIP7 fileNameDD
+# CMIP7 fileName
 _filename_template_keys_cmip7 = [
     "variable_id", "branding_suffix", "frequency", "region", "grid_label",
     "source_id", "experiment_id", "variant_label", "time_range",
@@ -92,7 +88,7 @@ def check_attributes_match_directory_structure(
         return [ctx.to_result()]
 
     if _is_cmip7(ds, project_id):
-        # drs_specs is required to anchor the DRS in the filesystem
+        
         try:
             drs_specs = str(ds.getncattr("drs_specs"))
         except Exception:

checks/consistency_checks/check_drs_filename_cv.py

@@ -13,7 +13,7 @@
 
 def _normalize_project_id(project_id: str) -> str:
     if isinstance(project_id, str) and project_id.lower() == "cmip7":
-        return "CMIP7"
+        return "cmip7"
     if isinstance(project_id, str) and project_id.lower() == "cmip6":
         return "cmip6"
     return project_id
@@ -76,7 +76,6 @@ def check_drs_directory(ds, severity, project_id="cmip6"):
         ctx.add_failure("File path could not be determined.")
         return [ctx.to_result()]
 
-    # If dataset declares drs_specs, prefer parsing from that directory name.
     drs_specs = None
     try:
         drs_specs = ds.getncattr("drs_specs")