Some adjustments to simplify ingestion of data

geospaas/base_viewer/apps.py

@@ -2,4 +2,4 @@
 
 
 class BaseViewerConfig(AppConfig):
-    name = 'base_viewer'
+    name = 'geospaas.base_viewer'

geospaas/nansat_ingestor/management/commands/ingest_thredds_crawl.py

@@ -38,7 +38,7 @@ def crawl_and_ingest(url, **options):
             # Create Dataset from OPeNDAP url - this is necessary to get all metadata
             gds, cr = NansatDataset.objects.get_or_create(url, uri_service_name=name,
                                                           uri_service_type=service)
-        except (IOError, AttributeError) as e:
+        except (IOError, AttributeError, ValueError) as e:
             # warnings.warn(e.message)
             continue
         if cr:

geospaas/nansat_ingestor/managers.py

@@ -1,9 +1,10 @@
 import json
+import logging
 import uuid
-import warnings
 
 import pythesint as pti
 from django.contrib.gis.geos import WKTReader
+from django.db import connection
 from django.db import models
 
 from geospaas.catalog.managers import FILE_SERVICE_NAME, LOCAL_FILE_SERVICE
@@ -56,23 +57,37 @@ def get_or_create(self,
         uris = DatasetURI.objects.filter(uri=uri, **uri_filter_args)
         if len(uris) > 0:
             return uris[0].dataset, False
+        connection.close()
 
         # Open file with Nansat
         n = Nansat(nansat_filename(uri), **kwargs)
 
         # get metadata from Nansat and get objects from vocabularies
         n_metadata = n.get_metadata()
 
-        entry_id = n_metadata.get('entry_id', None)
+        entry_id = n_metadata.get('entry_id', n_metadata.get('id', None))
         # set compulsory metadata (source)
-        platform, _ = Platform.objects.get_or_create(
-            json.loads(n_metadata['platform']))
-        instrument, _ = Instrument.objects.get_or_create(
-            json.loads(n_metadata['instrument']))
+        pp = n_metadata['platform']
+        try:
+            pp_dict = json.loads(pp)
+        except json.JSONDecodeError:
+            pp_entry = [elem.strip() for elem in pp.split('>')]
+            pp_dict = pti.get_gcmd_platform(pp_entry[-1].split('(')[1][0:-1])
+        platform, _ = Platform.objects.get_or_create(pp_dict)
+        connection.close()
+        ii = n_metadata['instrument']
+        try:
+            ii_dict = json.loads(ii)
+        except json.JSONDecodeError:
+            ii_entry = [elem.strip() for elem in ii.split('>')]
+            ii_dict = pti.get_gcmd_instrument(ii_entry[-1].split('(')[1][0:-1])
+        instrument, _ = Instrument.objects.get_or_create(ii_dict)
+        connection.close()
         specs = n_metadata.get('specs', '')
         source, _ = Source.objects.get_or_create(platform=platform,
                                                  instrument=instrument,
                                                  specs=specs)
+        connection.close()
 
         default_char_fields = {
             # Adding NERSC_ in front of the id violates the string representation of the uuid
@@ -88,9 +103,10 @@ def get_or_create(self,
             existing_ds = Dataset.objects.get(entry_id=entry_id)
         except Dataset.DoesNotExist:
             existing_ds = None
+        connection.close()
         for name in default_char_fields:
             if name not in n_metadata:
-                warnings.warn('%s is not provided in Nansat metadata!' % name)
+                logging.debug('%s is not provided in Nansat metadata!' % name)
                 # prevent overwriting of existing values by defaults
                 if existing_ds:
                     options[name] = existing_ds.__getattribute__(name)
@@ -103,7 +119,7 @@ def get_or_create(self,
             'gcmd_location': {'model': Location,
                               'value': pti.get_gcmd_location('SEA SURFACE')},
             'data_center': {'model': DataCenter,
-                            'value': pti.get_gcmd_provider('NERSC')},
+                            'value': pti.get_gcmd_provider('NO/MET')},
             'ISO_topic_category': {'model': ISOTopicCategory,
                                    'value': pti.get_iso19115_topic_category('Oceans')},
         }
@@ -113,12 +129,12 @@ def get_or_create(self,
             value = default_foreign_keys[name]['value']
             model = default_foreign_keys[name]['model']
             if name not in n_metadata:
-                warnings.warn('%s is not provided in Nansat metadata!' % name)
+                logging.debug('%s is not provided in Nansat metadata!' % name)
             else:
                 try:
                     value = json.loads(n_metadata[name])
                 except:
-                    warnings.warn('%s value of %s  metadata provided in Nansat is wrong!' %
+                    logging.debug('%s value of %s  metadata provided in Nansat is wrong!' %
                                   (n_metadata[name], name))
             if existing_ds:
                 options[name] = existing_ds.__getattribute__(name)
@@ -130,6 +146,7 @@ def get_or_create(self,
             n.reproject_gcps()
         geolocation = GeographicLocation.objects.get_or_create(
             geometry=WKTReader().read(n.get_border_wkt(nPoints=n_points)))[0]
+        connection.close()
 
         # create dataset
         # - the get_or_create method should use get_or_create here as well,
@@ -145,6 +162,7 @@ def get_or_create(self,
             'entry_title': options["entry_title"],
             'summary': options["summary"]}
         )
+        connection.close()
 
         # create parameter
         all_band_meta = n.bands()
@@ -162,12 +180,14 @@ def get_or_create(self,
                 params = params.filter(units=units)
             if params.count() >= 1:
                 ds.parameters.add(params[0])
+        connection.close()
 
         # create dataset URI
         DatasetURI.objects.get_or_create(
             name=uri_service_name,
             service=uri_service_type,
             uri=uri,
             dataset=ds)
+        connection.close()
 
         return ds, created

geospaas/utils/utils.py

@@ -25,9 +25,15 @@ def module_path(module, root):
     return media_path
 
 
-def path(module, filename, root):
+def path(module, filename, root, date=None):
     mp = module_path(module, root)
 
+    if date is not None:
+        for xx in [date.strftime('%Y'), date.strftime('%m'), date.strftime('%d')]:
+            mp = os.path.join(mp, xx)
+            if not os.path.exists(mp):
+                os.mkdir(mp)
+
     # Get the path of media files created from <filename>
     basename = os.path.split(filename)[-1].split('.')[0]
     dataset_path = os.path.join(mp, basename)
@@ -39,8 +45,8 @@ def path(module, filename, root):
 def media_path(module, filename):
     return path(module, filename, settings.MEDIA_ROOT)
 
-def product_path(module, filename):
-    return path(module, filename, settings.PRODUCTS_ROOT)
+def product_path(module, filename, date=None):
+    return path(module, filename, settings.PRODUCTS_ROOT, date=date)
 
 def validate_uri(uri):
     """ Validation of URI and its existence

geospaas/vocabularies/management/commands/update_vocabularies.py

@@ -35,15 +35,16 @@ def handle(self, *args, **options):
         models = [
             Parameter,
             DataCenter,
-            HorizontalDataResolution,
+            #HorizontalDataResolution,
             Instrument,
             ISOTopicCategory,
             Location,
             Platform,
-            Project,
+            #Project,
             ScienceKeyword,
-            TemporalDataResolution,
-            VerticalDataResolution]
+            #TemporalDataResolution,
+            #VerticalDataResolution,
+            ]
 
         for model in models:
             model.objects.create_from_vocabularies(**options)

geospaas/vocabularies/managers.py

@@ -52,14 +52,20 @@ def create_instances(self, pti_list):
         """
         num = 0
         for entry in pti_list:
-            pp, created = self.get_or_create(entry)
+            if bool(entry):
+                pp, created = self.get_or_create(entry)
+            else:
+                created = False
             if created: num+=1
         print("Successfully added %d new entries" % num)
 
     def get_or_create(self, entry, *args, **kwargs):
         """ Get or create database instance from input pythesint entry """
-
-        params = {key : entry[self.mapping[key]] for key in self.mapping}
+        try:
+            params = {key : entry[self.mapping[key]] for key in self.mapping}
+        except:
+            import ipdb
+            ipdb.set_trace()
         return super(VocabularyManager, self).get_or_create(**params)
 
     def create_from_vocabularies(self, force=False, **kwargs):
@@ -114,11 +120,21 @@ def get_by_natural_key(self, standard_name):
 class PlatformManager(VocabularyManager):
     get_list = pti.get_gcmd_platform_list
     update = pti.update_gcmd_platform
-    mapping = dict(category='Category',
-                    series_entity='Series_Entity',
+    # mapping = dict(category='Category',
+    #                 series_entity='Series_Entity',
+    #                 short_name='Short_Name',
+    #                 long_name='Long_Name')
+    # New mapping to adapt to changes in GCMD
+    mapping = dict(category='Basis',
+                    series_entity='Category',
                     short_name='Short_Name',
                     long_name='Long_Name')
-
+    # Correct mapping needs update of tables..:
+    # mapping = dict(basis='Basis',
+    #                category='Category',
+    #                sub_category='Sub_Category',
+    #                short_name='Short_Name',
+    #                long_name='Long_Name')
 
 
 class InstrumentManager(VocabularyManager):