Merge pull request #144 from nansencenter/upgrade_earthdata_normalizer

Make Earthdata CMR normalizer more generic
nansencenter · Apr 17, 2024 · ad109d6 · ad109d6
2 parents aa68a1d + a469ab4
commit ad109d6
Show file tree

Hide file tree

Showing 3 changed files with 54 additions and 12 deletions.
diff --git a/metanorm/normalizers/geospaas/earthdata_cmr.py b/metanorm/normalizers/geospaas/earthdata_cmr.py
@@ -25,21 +25,27 @@ def get_entry_title(self, raw_metadata):
 
     @utils.raises((KeyError, IndexError))
     def get_entry_id(self, raw_metadata):
-        return raw_metadata['umm']['DataGranule']['Identifiers'][0]['Identifier'].rstrip('.nc')
+        try:
+            return raw_metadata['umm']['DataGranule']['Identifiers'][0]['Identifier'].rstrip('.nc')
+        except KeyError:
+            return raw_metadata['umm']['GranuleUR']
 
     @utils.raises((KeyError, IndexError))
     def get_summary(self, raw_metadata):
         summary_fields = {}
         description = ''
         umm = raw_metadata['umm']
 
-        for platform in umm['Platforms']:
-            description += (
-                f"Platform={platform['ShortName']}, " +
-                ', '.join(f"Instrument={i['ShortName']}" for i in platform['Instruments']))
+        try:
+            for platform in umm['Platforms']:
+                description += (
+                    f"Platform={platform['ShortName']}, " +
+                    ', '.join(f"Instrument={i['ShortName']}" for i in platform['Instruments']) +
+                    ', ')
+        except KeyError:
+            pass
+        description += f"Start date={umm['TemporalExtent']['RangeDateTime']['BeginningDateTime']}"
 
-        description += (
-            f", Start date={umm['TemporalExtent']['RangeDateTime']['BeginningDateTime']}")
         summary_fields[utils.SUMMARY_FIELDS['description']] = description
 
         processing_level_match = re.match(
@@ -66,7 +72,10 @@ def get_platform(self, raw_metadata):
         """Only get the first platform from the raw metadata, because
         GeoSPaaS does not support more than one platform per dataset
         """
-        return utils.get_gcmd_platform(raw_metadata['umm']['Platforms'][0]['ShortName'])
+        try:
+            return utils.get_gcmd_platform(raw_metadata['umm']['Platforms'][0]['ShortName'])
+        except KeyError:
+            return utils.get_gcmd_platform(utils.UNKNOWN)
 
     @utils.raises((KeyError, IndexError))
     def get_instrument(self, raw_metadata):

diff --git a/metanorm/utils.py b/metanorm/utils.py
@@ -80,6 +80,7 @@ def export_subclasses(package__all__, package_name, package_dir, base_class):
     # providers
     'ESA/EO': ('ESA',),
     'OB.DAAC': ('OB_DAAC',),
+    'NASA/JPL/PODAAC': ('POCLOUD',),
     'C-SAR': ('SAR-C', 'SAR-C SAR'),
     'EUMETSAT/OSISAF': ('EUMETSAT OSI SAF',),
     'NSIDC': ('NSIDC_ECS',),

diff --git a/tests/normalizers/test_earthdata_cmr.py b/tests/normalizers/test_earthdata_cmr.py
@@ -1,6 +1,7 @@
 """Tests for the ACDD metadata normalizer"""
 import unittest
 import unittest.mock as mock
+from collections import OrderedDict
 from datetime import datetime
 
 from dateutil.tz import tzutc
@@ -38,6 +39,11 @@ def test_entry_id(self):
         }
         self.assertEqual(self.normalizer.get_entry_id(attributes), 'V2020245000600.L2_SNPP_OC')
 
+    def test_entry_id_from_granuleUR(self):
+        """Test getting the ID from the GranuleUR field"""
+        attributes = {'umm': {'GranuleUR': 'foo'}}
+        self.assertEqual(self.normalizer.get_entry_id(attributes), 'foo')
+
     def test_entry_id_missing_attribute(self):
         """A MetadataNormalizationError must be raised if the raw
         attribute is missing
@@ -109,6 +115,27 @@ def test_summary(self):
             'Description: Platform=SUOMI-NPP, ' +
             'Instrument=VIIRS, Start date=2020-09-01T00:06:00Z')
 
+    def test_summary_no_platform(self):
+        """Test getting a summary when no platform info is available
+        """
+        attributes = {
+            "umm": {
+                "TemporalExtent": {
+                    "RangeDateTime": {
+                        "BeginningDateTime": "2020-09-01T00:06:00Z",
+                        "EndingDateTime": "2020-09-01T00:11:59Z"
+                    }
+                },
+                "CollectionReference": {
+                    "ShortName": "VIIRSN_L2_OC",
+                    "Version": "2018"
+                }
+            }
+        }
+        self.assertEqual(
+            self.normalizer.get_summary(attributes),
+            'Description: Start date=2020-09-01T00:06:00Z;Processing level: 2')
+
     def test_summary_missing_attribute(self):
         """A MetadataNormalizationError must be raised if the raw
         attribute is missing
@@ -189,10 +216,15 @@ def test_platform_missing_attribute(self):
         """A MetadataNormalizationError must be raised if the raw
         attribute is missing
         """
-        with self.assertRaises(MetadataNormalizationError):
-            self.normalizer.get_platform({})
-        with self.assertRaises(MetadataNormalizationError):
-            self.normalizer.get_platform({'umm': {'foo': 'bar'}})
+        unknown_platform = OrderedDict([
+            ('Category', 'Unknown'),
+            ('Series_Entity', 'Unknown'),
+            ('Short_Name', 'Unknown'),
+            ('Long_Name', 'Unknown')
+        ])
+        self.assertDictEqual(self.normalizer.get_platform({}), unknown_platform)
+        self.assertDictEqual(self.normalizer.get_platform({'umm': {'foo': 'bar'}}),
+                             unknown_platform)
 
     def test_instrument(self):
         """Test getting the instrument"""