Fix :latest checker version resolution for compliance-checker >= 6.0.0

README.md

@@ -88,7 +88,7 @@ $ esgqa [-h] [-P <parallel_processes>] [-o <OUTPUT_DIR>] [-t <TEST>] [-O OPTION]
   - `-h, --help`: show this help message and exit
   - `-P, --parallel_processes`: Specify the maximum number of parallel processes. Default: 0 (= number of cores).
   - `-o, --output_dir OUTPUT_DIR`: Directory to store QA results. Needs to be non-existing or empty or from previous QA run. If not specified, will store results in `./cc-qa-check-results/YYYYMMDD-HHmm_<hash>`.
-  - `-t, --test TEST`: The test to run (eg. `'wcrp_cmip6:latest'`, `'wcrp_cordex_cmip6:latest'` or `'cf:<version>'`, can be specified multiple times, eg.: `'-t wcrp_cmip6:latest -t cf:1.7'`) - default: running latest CF checks `'cf:latest'`. If the version is omitted, `latest` will be used.
+  - `-t, --test TEST`: The test to run (eg. `'wcrp_cmip6:latest'`, `'wcrp_cordex_cmip6:latest'` or `'cf:<version>'`, can be specified multiple times, eg.: `'-t wcrp_cmip6:latest -t cf:1.7'`) - default: running latest CF checks. If the version is omitted, `latest` will be used (`'cf'` and `'cf:latest'` are equivalent).
   - `-O, --option OPTION`: Additional options to be passed to the checkers. Format: `'<checker>:<option_name>[:<option_value>]'`. Multiple invocations possible.
   - `-i, --info INFO`:  Information used to tag the QA results, eg. the simulation id to identify the checked run. Suggested is the original experiment-id you gave the run.
   - `-r, --resume`: Specify to continue a previous QC run. Requires the `<output_dir>` argument to be set.

esgf_qa/run_qa.py

@@ -140,23 +140,57 @@ def get_checker_release_versions(checkers, checker_options={}):
     check_suite = CheckSuite(options=checker_options)
     check_suite.load_all_available_checkers()
     for checker in checkers:
-        if checker.split(":")[0] not in checker_release_versions:
-            if checker.split(":")[0] in checker_dict:
-                checker_release_versions[checker.split(":")[0]] = (
-                    check_suite.checkers.get(
-                        checker, "unknown version"
-                    )._cc_spec_version
-                )
-            elif checker.split(":")[0] in checker_dict_ext:
-                checker_release_versions[checker.split(":")[0]] = version
+        checker_name = checker.split(":")[0]
+        if checker_name not in checker_release_versions:
+            if checker_name in checker_dict_ext and checker_name not in checker_dict:
+                # Internal esgf-qa checker (cons, cont, comp) - use esgf-qa version
+                checker_release_versions[checker_name] = version
             else:
-                checker_release_versions[checker.split(":")[0]] = (
-                    check_suite.checkers.get(
-                        checker, "unknown version"
-                    )._cc_spec_version
+                # compliance-checker plugin: look up _cc_spec_version.
+                # CC >= 6.0.0 removed :latest, so fall back to the highest
+                # explicitly versioned key when the requested key is missing.
+                checker_obj = check_suite.checkers.get(checker)
+                if checker_obj is None:
+                    prefix = checker_name + ":"
+                    candidates = [
+                        k for k in check_suite.checkers if k.startswith(prefix)
+                    ]
+                    if candidates:
+                        resolved_key = max(
+                            candidates,
+                            key=lambda k: pversion.parse(k.split(":")[1]),
+                        )
+                        checker_obj = check_suite.checkers.get(resolved_key)
+                checker_release_versions[checker_name] = (
+                    checker_obj._cc_spec_version
+                    if checker_obj is not None
+                    else "unknown"
                 )
 
 
+def normalize_checker_specs(checkers_versions):
+    """
+    Normalize checker specifications for compliance-checker.
+
+    Parameters
+    ----------
+    checkers_versions : dict
+        Mapping of checker name to requested version string.
+
+    Returns
+    -------
+    list
+        Sorted checker specs where explicit versions are kept as
+        '<checker>:<version>' and 'latest' maps to unversioned '<checker>'.
+    """
+    return sorted(
+        [
+            checker if checker_version == "latest" else f"{checker}:{checker_version}"
+            for checker, checker_version in checkers_versions.items()
+        ]
+    )
+
+
 def run_compliance_checker(file_path, checkers, checker_options={}):
     """
     Run the compliance checker on a file with the specified checkers and options.
@@ -194,7 +228,7 @@ def run_compliance_checker(file_path, checkers, checker_options={}):
     if include_checks:
         results = {}
         for checker in checkers:
-            if include_checks and "cc6:latest" in checker or "mip:latest" in checker:
+            if include_checks and checker.split(":")[0] in ("cc6", "mip"):
                 results.update(
                     check_suite.run_all(ds, [checker], include_checks, skip_checks=[])
                 )
@@ -274,7 +308,10 @@ def process_file(
         and os.path.isfile(result_file)
         and (
             os.path.isfile(consistency_file)
-            or not any(cn.startswith("cc6") or cn.startswith("mip") for cn in checkers)
+            or not any(
+                cn.split(":", 1)[0] in checker_supporting_consistency_checks
+                for cn in checkers
+            )
         )
     ):
         with open(result_file) as file:
@@ -737,9 +774,11 @@ def main():
                 "ERROR: Cannot run both 'cc6' and 'mip' checkers at the same time."
             )
 
-    # Combine checkers and versions
-    #  (checker_options are hardcoded)
-    checkers = sorted([f"{c}:{v}" for c, v in checkers_versions.items()])
+    # Normalize checker specifications for compliance-checker:
+    # - explicit versions are forwarded as '<checker>:<version>'
+    # - omitted versions and ':latest' are both forwarded as '<checker>'
+    #   so compliance-checker selects the highest installed version.
+    checkers = normalize_checker_specs(checkers_versions)
 
     # Does parent_dir exist?
     if parent_dir is None:
@@ -760,13 +799,17 @@ def main():
     with open(os.path.join(result_dir, ".resume_info"), "w") as f:
         json.dump(resume_info, f, sort_keys=True, indent=4)
 
-    # If only cf checker is selected, run cc6 time checks only
+    # If none of the selected checkers support consistency checks,
+    # add mip time checks so consistency output can be generated.
     if (
-        not any(cn.startswith("cc6") or cn.startswith("mip") for cn in checkers)
+        not any(
+            cn.split(":", 1)[0] in checker_supporting_consistency_checks
+            for cn in checkers
+        )
         and include_consistency_checks
     ):
         time_checks_only = True
-        checkers.append("mip:latest")
+        checkers.append("mip")
         checkers.sort()
     else:
         time_checks_only = False
@@ -1026,7 +1069,7 @@ def main():
                 )
                 del result
 
-    # Skip continuity and consistency checks if no cc6/mip checks were run
+    # Skip continuity and consistency checks if no appropriate checkers were run
     #   (and thus no consistency output file was created)
     if any(
         ch.split(":", 1)[0] in checker_supporting_consistency_checks for ch in checkers
@@ -1120,7 +1163,7 @@ def main():
     print()
     print("#" * 50)
     print(
-        f"# QA Part {'3' if 'cc6:latest' in checkers or 'mip:latest' in checkers else '2'} - Summarizing and clustering the results"
+        f"# QA Part {'3' if any(cn.split(':')[0] in checker_supporting_consistency_checks for cn in checkers) else '2'} - Summarizing and clustering the results"
     )
     print("#" * 50)
     print()

pyproject.toml

@@ -29,15 +29,15 @@ classifiers = [
 dependencies = [
   "cftime",
   "cf_xarray",
-  "compliance-checker>=5.3.0",
+  "compliance-checker>=6.0.0",
   "dask",
   "netCDF4",
   "packaging",
   "pandas",
   "textual",
   "xarray",
-  "cc-plugin-cc6>=0.4.0",
-  "cc-plugin-wcrp"
+  "cc-plugin-cc6>=0.4.3",
+  "cc-plugin-wcrp>=2.1.0"
 ]
 dynamic = [
   "version"

tests/test_run_qa.py

@@ -13,6 +13,7 @@
     get_checker_release_versions,
     get_default_result_dir,
     get_dsid,
+    normalize_checker_specs,
     parse_options,
     track_checked_datasets,
 )
@@ -179,3 +180,14 @@ def test_parse_options():
         },
     )
     assert _verify_options_dict(opt_dict) is True
+
+
+def test_latest_and_omitted_versions_are_equivalent_in_internal_specs():
+    checkers_versions = {"cf": "latest", "cc6": "latest", "wcrp_cmip6": "1.7"}
+    checkers = normalize_checker_specs(checkers_versions)
+
+    assert "cf" in checkers
+    assert "cc6" in checkers
+    assert "wcrp_cmip6:1.7" in checkers
+    assert "cf:latest" not in checkers
+    assert "cc6:latest" not in checkers