Review changes: clarify comments.

lib/iris/tests/unit/fileformats/nc_load_rules/actions/test__grid_mappings.py

@@ -607,6 +607,10 @@ def test_mapping_undefined(self):
     #     TODO: it's not clear why this needs to behave differently from case
     #     (3.) : possibly, these two should be made consistent.
     #
+    # TODO: *all* these 'mismatch' cases should probably generate warnings,
+    # except for plain-latlon coords with no grid-mapping.
+    # At present, we _only_ warn when an expected grid-mapping is absent.
+    #
 
     def test_mapping__mismatch__latlon_coords_rotated_system(self):
         # Rules Triggered:
@@ -694,7 +698,6 @@ def test_mapping__mismatch__rotated_coords_nonll_system(self):
         #     006 : fc_build_coordinate_(rotated_longitude)(rotated no-cs : discarded projected cs)
         # Notes:
         #     * coords built : rotated-lat + lon, with no coord-system (see above)
-        # TODO: should this change ??
         result = self.run_testcase(
             mapping_type_name=hh.CF_GRID_MAPPING_ALBERS,
             xco_name="grid_longitude",

lib/iris/tests/unit/fileformats/nc_load_rules/actions/test__time_coords.py

@@ -317,14 +317,17 @@ def test_dim_nonmonotonic(self):
         self.check_result(result, "aux")
 
     def test_dim_fails_typeident(self):
-        # The coord variable is identified as a CFDimensionCoordinate by cf.py,
-        # but having the wrong units causes it to fail the 'is_time' or
-        # 'is_period' test, so the 'provides_coord' rule fails to trigger.
-        # So it is built as a 'miscellaneous' dim-coord.
+        # Provide a coord variable, identified as a CFDimensionCoordinate by
+        # cf.py, but with the "wrong" units for a time or period coord.
+        # This causes it to fail both 'is_time' and 'is_period' tests and so,
+        # within the 'action_provides_coordinate' routine, does not trigger as
+        # a 'provides_coord_(time[[_period]])' rule, but instead as a
+        # 'default_coordinate_(provide-phase)'.
+        # As a result, it is built as a 'miscellaneous' dim-coord.
         # N.B. this makes *no* practical difference, because a 'misc' dim
-        # coord is still a dim coord (albeit one with bad units).
-        # ( N.B.#2 Not quite the same for lat/lon coords, where coord-specific
-        #   'build' rules always use a fixed standard-name ).
+        # coord is still a dim coord (albeit one with incorrect units).
+        # N.B.#2 that is different from lat/lon coords, where the coord-specific
+        # 'build' rules have the extra effect of setting a fixed standard-name.
         #
         # Rules Triggered:
         #     001 : fc_default