def syncvesk(): sensors = {} service = SensorObservationService(SOSURL, version=SOSVERSION) get_obs = service.get_operation_by_name('GetObservation') describe_sensor = service.get_operation_by_name('DescribeSensor') sensor_ids = describe_sensor.parameters['procedure']['values'] for sensor_id in sensor_ids: response = service.describe_sensor( outputFormat='http://www.opengis.net/sensorML/1.0.1', procedure=sensor_id) tr = etree.fromstring(response) element = tr.find('.//' + nspath_eval('sml:SensorML', namespaces)) ds = SensorML(element) name = ds.members[0].name sensors[sensor_id] = name offerings = [off.id for off in service.offerings] _observed_properties = list( set(op for off in service.offerings for op in off.observed_properties)) observed_properties = SOS_PARAMS.keys() end_time = datetime.now() # - timedelta(hours=120*10) start_time = end_time - timedelta(hours=12) temporalFilter = "om:phenomenonTime,{}/{}".format(start_time.isoformat(), end_time.isoformat()) response_format = 'application/json' _namespaces = "xmlns(om,http://www.opengis.net/om/2.0)" _response = service.get_observation(offerings=offerings, responseFormat=response_format, observedProperties=observed_properties, namespaces=_namespaces, temporalFilter=temporalFilter) response = json.loads(_response) for obs in response['observations']: location = obs['featureOfInterest']['name']['value'] _parameter = obs['observableProperty'] parameter = SOS_PARAMS.get(_parameter) timestamp = obs['phenomenonTime'] network = 'CNR' value = float(obs['result']['value']) rjson = { 'location': location, 'parameter': parameter, "timestamp": timestamp, 'network': network, "value": value } _write([rjson]) pass pass
class IoosSweSos(Collector): def __init__(self, url, xml=None): super(IoosSweSos,self).__init__() self.server = Sos(url, xml=xml) def metadata(self, **kwargs): callback = kwargs.get("feature_name_callback", None) or str kwargs['outputFormat'] = 'text/xml;subtype="sensorML/1.0.1"' responses = [] if self.features is not None: for feature in self.features: kwargs['procedure'] = callback(feature) responses.append(SensorML(self.server.describe_sensor(**kwargs))) return responses def setup_params(self, **kwargs): params = kwargs # TODO: BBOX needs to be implemented as an OGC Filter. # This page has a good example: http://opendap.co-ops.nos.noaa.gov/ioos-dif-sos/index.jsp # Click on "Collection" in the POST example to see what it should look like. # Requires changing from GET to POST... if self.bbox is not None: print "BBOX requests for IOOS SWE SOS services are not yet implemented" if self.start_time is not None: params["eventTime"] = self.start_time.strftime('%Y-%m-%dT%H:%M:%SZ') if self.end_time is not None: params["eventTime"] += "/%s" % self.end_time.strftime('%Y-%m-%dT%H:%M:%SZ') if self.variables is None or len(self.variables) < 1: raise ValueError("You must set a filter for at least one variable (observedProperty)") else: ops = ",".join(self.variables) if isinstance(ops, basestring): ops = [ops] params["observedProperties"] = ops return params def collect(self, **kwargs): kwargs["responseFormat"] = 'text/xml;subtype="om/1.0.0/profiles/ioos_sos/1.0"' return IoosGetObservation(self.raw(**kwargs)).feature def raw(self, **kwargs): params = self.setup_params(**kwargs) return self.server.get_observation(**params)
class CoopsSos(Collector): def __init__(self, **kwargs): super(CoopsSos,self).__init__() url = 'http://opendap.co-ops.nos.noaa.gov/ioos-dif-sos/SOS' self.server = Sos(url) def get_metadata(self, **kwargs): response = self.server.describe_sensor(**kwargs) return IoosDescribeSensor(response) def get_describe_sensor_output_formats(self): return self.server.get_operation_by_name('DescribeSensor').parameters['outputFormat']['values'] #def get_data(self, **kwargs): # response = self.get_raw_data(**kwargs) # return IoosDif(response) def get_raw_data(self, **kwargs): res = self.server.get_observation(**kwargs) if res[41:56] == "ExceptionReport": res = -1 return res
class NdbcSos(Collector): def __init__(self, **kwargs): super(NdbcSos,self).__init__() if kwargs.get('test') is True: url = 'http://sdftest.ndbc.noaa.gov/sos/server.php' else: url = 'http://sdf.ndbc.noaa.gov/sos/server.php' self.server = Sos(url) def get_metadata(self, **kwargs): response = self.server.describe_sensor(**kwargs) return IoosDescribeSensor(response) def get_describe_sensor_output_formats(self): return self.server.get_operation_by_name('DescribeSensor').parameters['outputFormat']['values'] def get_data(self, **kwargs): response = self.get_raw_data(**kwargs) return IoosSwe(response) def get_raw_data(self, **kwargs): return self.server.get_observation(**kwargs)
class CoopsSos(Collector): def __init__(self, **kwargs): super(CoopsSos, self).__init__() url = 'http://opendap.co-ops.nos.noaa.gov/ioos-dif-sos/SOS' self.server = Sos(url) def get_metadata(self, **kwargs): response = self.server.describe_sensor(**kwargs) return IoosDescribeSensor(response) def get_describe_sensor_output_formats(self): return self.server.get_operation_by_name( 'DescribeSensor').parameters['outputFormat']['values'] #def get_data(self, **kwargs): # response = self.get_raw_data(**kwargs) # return IoosDif(response) def get_raw_data(self, **kwargs): res = self.server.get_observation(**kwargs) if res[41:56] == "ExceptionReport": res = -1 return res
class NdbcSos(Collector): def __init__(self, **kwargs): super(NdbcSos, self).__init__() if kwargs.get('test') is True: url = 'http://sdftest.ndbc.noaa.gov/sos/server.php' else: url = 'http://sdf.ndbc.noaa.gov/sos/server.php' self.server = Sos(url) def get_metadata(self, **kwargs): response = self.server.describe_sensor(**kwargs) return IoosDescribeSensor(response) def get_describe_sensor_output_formats(self): return self.server.get_operation_by_name( 'DescribeSensor').parameters['outputFormat']['values'] def get_data(self, **kwargs): response = self.get_raw_data(**kwargs) return IoosSwe(response) def get_raw_data(self, **kwargs): return self.server.get_observation(**kwargs)
class IoosSweSos(Collector): def __init__(self, url, xml=None): super(IoosSweSos, self).__init__() self.server = Sos(url, xml=xml) def metadata(self, output_format=None, feature_name_callback=None, **kwargs): """ Gets SensorML objects for all procedures in your filtered features. You should override the default output_format for servers that do not respond properly. """ callback = feature_name_callback or str if output_format is None: output_format = 'text/xml; subtype="sensorML/1.0.1/profiles/ioos_sos/1.0"' responses = [] if self.features is not None: for feature in self.features: ds_kwargs = kwargs.copy() ds_kwargs.update({'outputFormat': output_format, 'procedure' : callback(feature)}) responses.append(SensorML(self.server.describe_sensor(**ds_kwargs))) return responses def setup_params(self, **kwargs): params = kwargs if self.bbox is not None: params["featureOfInterest"] = "BBOX:%s,%s,%s,%s" % (self.bbox[0], self.bbox[1], self.bbox[2], self.bbox[3]) if self.start_time is not None: params["eventTime"] = self.start_time.strftime('%Y-%m-%dT%H:%M:%SZ') if self.end_time is not None: params["eventTime"] += "/%s" % self.end_time.strftime('%Y-%m-%dT%H:%M:%SZ') if self.variables is None or len(self.variables) < 1: raise ValueError("You must set a filter for at least one variable (observedProperty)") else: ops = ",".join(self.variables) if isinstance(ops, basestring): ops = [ops] params["observedProperties"] = ops return params def collect(self, **kwargs): # there is an unfortunate difference in how 52N and ncSOS handle the response format. # 52N expects subtype, ncSOS expects schema. # consult the observed properties and getcaps to figure out which should be used if none passed if 'responseFormat' not in kwargs: # iterate offerings and see if we need to change to subtype off_dict = {off.name : off for off in self.server.offerings} response_format = None for offering in kwargs.get('offerings', []): if offering not in off_dict: continue ioos_formats = [rf for rf in off_dict[offering].response_formats if 'ioos_sos/1.0' in rf] if not len(ioos_formats): raise StandardError("No ioos_sos/1.0 response format found for offering %s" % offering) if response_format != ioos_formats[0]: response_format = ioos_formats[0] kwargs["responseFormat"] = response_format return IoosGetObservation(self.raw(**kwargs)).observations def raw(self, **kwargs): params = self.setup_params(**kwargs) return self.server.get_observation(**params)
class IoosSweSos(Collector): def __init__(self, url, xml=None): super(IoosSweSos, self).__init__() self.server = Sos(url, xml=xml) def metadata(self, output_format=None, feature_name_callback=None, **kwargs): """ Gets SensorML objects for all procedures in your filtered features. You should override the default output_format for servers that do not respond properly. """ callback = feature_name_callback or str if output_format is None: output_format = 'text/xml; subtype="sensorML/1.0.1/profiles/ioos_sos/1.0"' responses = [] if self.features is not None: for feature in self.features: ds_kwargs = kwargs.copy() ds_kwargs.update({ 'outputFormat': output_format, 'procedure': callback(feature) }) responses.append( SensorML(self.server.describe_sensor(**ds_kwargs))) return responses def metadata_plus_exceptions(self, output_format=None, feature_name_callback=None, **kwargs): """ Gets SensorML objects for all procedures in your filtered features. Return two dictionaries for service responses keyed by 'feature': responses: values are SOS DescribeSensor response text response_failures: values are exception text content furnished from ServiceException, ExceptionReport You should override the default output_format for servers that do not respond properly. """ callback = feature_name_callback or str if output_format is None: output_format = 'text/xml; subtype="sensorML/1.0.1/profiles/ioos_sos/1.0"' responses = {} response_failures = {} if self.features is not None: for feature in self.features: ds_kwargs = kwargs.copy() ds_kwargs.update({ 'outputFormat': output_format, 'procedure': callback(feature) }) try: responses[feature] = (SensorML( self.server.describe_sensor(**ds_kwargs))) except (ServiceException, ExceptionReport) as e: response_failures[feature] = str(e) return (responses, response_failures) def setup_params(self, **kwargs): params = kwargs if self.bbox is not None: params["featureOfInterest"] = "BBOX:%s,%s,%s,%s" % ( self.bbox[0], self.bbox[1], self.bbox[2], self.bbox[3]) if self.start_time is not None: params["eventTime"] = self.start_time.strftime( '%Y-%m-%dT%H:%M:%SZ') if self.end_time is not None: params["eventTime"] += "/%s" % self.end_time.strftime( '%Y-%m-%dT%H:%M:%SZ') if self.variables is None or len(self.variables) < 1: raise ValueError( "You must set a filter for at least one variable (observedProperty)" ) else: ops = ",".join(self.variables) if isinstance(ops, string_types): ops = [ops] params["observedProperties"] = ops return params def collect(self, **kwargs): # there is an unfortunate difference in how 52N and ncSOS handle the response format. # 52N expects subtype, ncSOS expects schema. # consult the observed properties and getcaps to figure out which should be used if none passed if 'responseFormat' not in kwargs: # iterate offerings and see if we need to change to subtype off_dict = {off.name: off for off in self.server.offerings} response_format = None for offering in kwargs.get('offerings', []): if offering not in off_dict: continue ioos_formats = [ rf for rf in off_dict[offering].response_formats if 'ioos_sos/1.0' in rf ] if not len(ioos_formats): raise Exception( "No ioos_sos/1.0 response format found for offering {}" .format(offering)) if response_format != ioos_formats[0]: response_format = ioos_formats[0] kwargs["responseFormat"] = response_format return IoosGetObservation(self.raw(**kwargs)).observations def raw(self, **kwargs): params = self.setup_params(**kwargs) return self.server.get_observation(**params)
class SosHarvest(Harvester): def __init__(self, service): Harvester.__init__(self, service) def harvest(self): self.sos = SensorObservationService(self.service.get('url')) for offering in self.sos.offerings: # TODO: We assume an offering should only have one procedure here # which will be the case in sos 2.0, but may not be the case right now # on some non IOOS supported servers. uid = offering.procedures[0] sp_uid = uid.split(":") # List storing the stations that have already been processed in this SOS server. # This is kept and checked later to avoid servers that have the same stations in many offerings. processed = [] # temnplate: urn:ioos:type:authority:id # sample: ioos:station:wmo:21414 if sp_uid[2] == "station": # Station Offering if not uid in processed: self.process_station(uid) processed.append(uid) elif sp_uid[2] == "network": # Network Offering network_ds = IoosDescribeSensor(self.sos.describe_sensor(outputFormat='text/xml;subtype="sensorML/1.0.1/profiles/ioos_sos/1.0"', procedure=uid)) # Iterate over stations in the network and process them individually for proc in network_ds.procedures: if proc is not None and proc.split(":")[2] == "station": if not proc in processed: self.process_station(proc) processed.append(proc) def process_station(self, uid): """ Makes a DescribeSensor request based on a 'uid' parameter being a station procedure """ GML_NS = "http://www.opengis.net/gml" XLINK_NS = "http://www.w3.org/1999/xlink" with app.app_context(): metadata_value = etree.fromstring(self.sos.describe_sensor(outputFormat='text/xml;subtype="sensorML/1.0.1/profiles/ioos_sos/1.0"', procedure=uid)) station_ds = IoosDescribeSensor(metadata_value) unique_id = station_ds.id if unique_id is None: app.logger.warn("Could not get a 'stationID' from the SensorML identifiers. Looking for a definition of 'http://mmisw.org/ont/ioos/definition/stationID'") return dataset = db.Dataset.find_one( { 'uid' : unicode(unique_id) } ) if dataset is None: dataset = db.Dataset() dataset.uid = unicode(unique_id) # Find service reference in Dataset.services and remove (to replace it) tmp = dataset.services[:] for d in tmp: if d['service_id'] == self.service.get('_id'): dataset.services.remove(d) # Parsing messages messages = [] # NAME name = unicode_or_none(station_ds.shortName) if name is None: messages.append(u"Could not get a 'shortName' from the SensorML identifiers. Looking for a definition of 'http://mmisw.org/ont/ioos/definition/shortName'") # DESCRIPTION description = unicode_or_none(station_ds.longName) if description is None: messages.append(u"Could not get a 'longName' from the SensorML identifiers. Looking for a definition of 'http://mmisw.org/ont/ioos/definition/longName'") # PLATFORM TYPE asset_type = unicode_or_none(station_ds.platformType) if asset_type is None: messages.append(u"Could not get a 'platformType' from the SensorML identifiers. Looking for a definition of 'http://mmisw.org/ont/ioos/definition/platformType'") # LOCATION is in GML gj = None loc = station_ds.location if loc is not None and loc.tag == "{%s}Point" % GML_NS: pos_element = loc.find("{%s}pos" % GML_NS) # strip out points positions = map(float, testXMLValue(pos_element).split(" ")) crs = Crs(testXMLAttribute(pos_element, "srsName")) if crs.axisorder == "yx": gj = json.loads(geojson.dumps(geojson.Point([positions[1], positions[0]]))) else: gj = json.loads(geojson.dumps(geojson.Point([positions[0], positions[1]]))) else: messages.append(u"Found an unrecognized child of the sml:location element and did not attempt to process it: %s" % etree.tostring(loc).strip()) service = { # Reset service 'name' : name, 'description' : description, 'service_type' : self.service.get('service_type'), 'service_id' : ObjectId(self.service.get('_id')), 'data_provider' : self.service.get('data_provider'), 'metadata_type' : u'sensorml', 'metadata_value' : unicode(etree.tostring(metadata_value)).strip(), 'messages' : map(unicode, messages), 'keywords' : map(unicode, sorted(station_ds.keywords)), 'variables' : map(unicode, sorted(station_ds.variables)), 'asset_type' : asset_type, 'geojson' : gj, 'updated' : datetime.utcnow() } dataset.services.append(service) dataset.updated = datetime.utcnow() dataset.save() return "Harvested"
def get_stations_df(self, sos_url, station_urns_sel=None): """ Returns a Pandas Dataframe """ # oFrmts: IOOS SOS OutputFormat strings (first is compliant to the IOOS SOS spec, # second is to accommodate NDBC). More info here: # http://ioos.github.io/sos-guidelines/doc/wsdd/sos_wsdd_github_notoc/#describesensor-request:638e0b263020c13a76a55332bd966dbe oFrmts = [ 'text/xml; subtype="sensorML/1.0.1/profiles/ioos_sos/1.0"', 'text/xml;subtype="sensorML/1.0.1"' ] params = { 'service': 'SOS', 'request': 'GetCapabilities', 'acceptVersions': '1.0.0' } sos_url_params = sos_url + '?' + urlencode(params) # sos_url_params_quoted = quote(sos_url_params,"/=:") # sos_url_params_unquoted = unquote(sos_url_params) try: sosgc = SensorObservationService(sos_url_params) except (ConnectionError, ReadTimeout) as e: self.log.write( u"\nError: unable to connect to SOS service: {url} due to HTTP connection error." .format(url=sos_url_params)) self.log.write( u"\nHTTP connection error: {err}.".format(err=str(e))) sys.exit( "\nError: unable to connect to SOS service: {url}. \nUnderlying HTTP connection error: {err}" .format(url=sos_url_params, err=str(e))) # vars to store returns from sos_collector.metadata_plus_exceptions function: sml_recs = {} sml_errors = {} describe_sensor_url = {} # leverage Pyoos Collector to query for all available stations and obtain SensorML # (if station subset not passed in --stations param) if station_urns_sel is not None: station_urns = station_urns_sel else: sos_collector = IoosSweSos(sos_url) station_urns = [ urn.name for urn in sos_collector.server.offerings if 'network' not in urn.name.split(':') ] sos_collector.features = station_urns # write out stations in SOS that will be handled: if self.verbose: self.log.write(u"\nStations to process for SOS: {sos}".format( sos=sos_url_params)) print("Stations to process for SOS: {sos}".format( sos=sos_url_params)) for feature in sos_collector.features: self.log.write(u"\n - {feature}".format(feature=feature)) print(" - {feature}".format(feature=feature)) # iterate over possible oFrmts expected of the various SOS services (IOOS SOS 1.0, NDBC): # for fmt in reversed(oFrmts): for fmt in oFrmts: try: sml_recs, sml_errors = sos_collector.metadata_plus_exceptions( output_format=fmt, timeout=200) # if no valid SensorML docs returned, try next oFrmt: if not sml_recs: continue else: # assign correct DescribeSensor url (use sos_collector.feature rather than sml_recs.keys() to # create DescribeSensor URLs for failures to record in logs): for station in sos_collector.features: describe_sensor_url[ station] = self.generate_describe_sensor_url( sosgc, procedure=station, oFrmt=fmt) # report on errors returned from metadata_plus_exceptions: if sml_errors: if self.verbose: for station, msg in iteritems(sml_errors): self.log.write( u"\nSOS DescribeSensor error returned for: {station}, skipping. Error msg: {msg}" .format(station=station, msg=msg)) print( "SOS DescribeSensor error returned for: {station}, skipping. Error msg: {msg}" .format(station=station, msg=msg)) else: self.log.write( u"\nSuccess, no errors returned from DescribeSensor requests in service: {sos}" .format(sos=sos_url_params)) print( "Success, no errors returned from DescribeSensor requests in service: {sos}" .format(sos=sos_url_params)) break # ServiceException shouldn't be thrown by metadata_plus_exceptions function, but handle regardless by attempting next oFrmt: except ServiceException as e: continue station_recs = [] failures = [] # generate Pandas DataFrame by populating 'station_recs' list by parsing SensorML strings: for station_idx, station_urn in enumerate(station_urns): if station_urns_sel is not None: # iterate oFrmts items for describe_sensor request (first is IOOS SOS spec-compliant, second is for NDBC SOS): for fmt in oFrmts: try: describe_sensor_url[ station_urn] = self.generate_describe_sensor_url( sosgc, procedure=station_urn, oFrmt=fmt) sml_str = sosgc.describe_sensor(procedure=station_urn, outputFormat=fmt, timeout=200) break except ServiceException as e: sml_errors[station_urn] = str(e) continue sml = SensorML(sml_str) else: # process valid SensorML responses, quietly pass on invalid stations (add to failures list for verbose reporting): try: sml = sml_recs[station_urn] except KeyError: self.log.write( u"\n\nStation: {station} failed (no SensorML in sml_recs dict). URL: {ds}" .format(station=station_urn, ds=describe_sensor_url[station_urn].replace( "&", "&"))) print( "Station: {station} failed (no SensorML in sml_recs dict). URL: {ds}" .format(station=station_urn, ds=describe_sensor_url[station_urn].replace( "&", "&"))) failures.append(station_urn) continue if self.sos_type.lower() == 'ndbc': # later: add an error check sosgc_station_offering = sosgc.contents[ 'station-' + station_urn.split(':')[-1]] else: sosgc_station_offering = None try: ds = IoosDescribeSensor(sml._root) except AttributeError: self.log.write( u"\nInvalid SensorML passed to IoosDescribeSensor. Check DescribeSensor request for : {station}, URL: " .format(station=station, ds=describe_sensor_url[station_urn].replace( "&", "&"))) print( "Invalid SensorML passed to IoosDescribeSensor. Check DescribeSensor request for : {station}, URL: " .format(station=station, ds=describe_sensor_url[station_urn].replace( "&", "&"))) station = OrderedDict() # debug: if self.verbose: self.log.write(u"\n\nProcessing station: {station}".format( station=station_urn)) print("Processing station: {station}".format( station=station_urn)) self.log.write("\n" + etree.tostring(sml._root).decode('utf-8')) # assign 'pos' to GML point location (accommodate 'gml:coordinates' as used by NDBC if gml:Point not found): try: pos = testXMLValue(ds.system.location.find(self.nsp('gml:Point/gml:pos'))) \ if testXMLValue(ds.system.location.find(self.nsp('gml:Point/gml:pos'))) is not None \ else testXMLValue(ds.system.location.find(self.nsp('gml:Point/gml:coordinates'))) station['lon'] = float(pos.split()[1]) station['lat'] = float(pos.split()[0]) except AttributeError as e: station['lon'] = None station['lat'] = None system_el = sml._root.findall(self.nsp('sml:member'))[0].find( self.nsp('sml:System')) # Parse the DocumentList into a dict storing documents by index value 'name' (may cause index duplication # errors but there is not enough information in SensorML for alternatives) # Assume that member corresponds to xlink:arcrole="urn:ogc:def:role:webPage" documents = system_el.findall( self.nsp('sml:documentation/sml:DocumentList/sml:member')) documents_dct = {} for d in documents: document = Documentation(d) name = testXMLAttribute(d, "name") # url = document.documents[0].url documents_dct[name] = document # obtain list of contacts (accommodate 'sml:contact' element repetition used by NDBC insead of ContactList): contacts = system_el.findall(self.nsp('sml:contact/sml:ContactList/sml:member')) \ if system_el.findall(self.nsp('sml:contact/sml:ContactList/sml:member')) \ else system_el.findall(self.nsp('sml:contact')) contacts_dct = {} for c in contacts: contact = Contact(c) role = contact.role.split('/')[-1] contacts_dct[role] = contact # verify a 'publisher' Contact exists (template expects one): if "publisher" not in contacts_dct.keys(): self.log.write( u"\n\nStation: {station} skipped. No \'http://mmisw.org/ont/ioos/definition/publisher\' Contact role defined in SensorML as required. Roles defined: [{roles}]" .format(station=station_urn, roles=", ".join(contacts_dct.keys()))) print( "Station: {station} skipped. No \'http://mmisw.org/ont/ioos/definition/publisher\' Contact role defined in SensorML as required. Roles defined: [{roles}]" .format(station=station_urn, roles=", ".join(contacts_dct.keys()))) failures.append(station_urn) continue sweQuants = system_el.findall( self.nsp('sml:outputs/sml:OutputList/sml:output/swe:Quantity')) quant_lst = [ sweQuant.attrib['definition'] for sweQuant in sweQuants ] parameter_lst = [sweQuant.split('/')[-1] for sweQuant in quant_lst] # attempt to read beginPosition, if available, otherwise use current date # bc ISO requires date value in output location in template: beginPosition = testXMLValue( system_el.find( self.nsp( 'sml:validTime/gml:TimePeriod/gml:beginPosition'))) try: begin_service_date = parser.parse(beginPosition) except (AttributeError, TypeError) as e: begin_service_date = datetime.now(pytz.utc) station['station_urn'] = station_urn station['sos_url'] = sos_url_params station['describesensor_url'] = describe_sensor_url[station_urn] station['shortName'] = ds.shortName station['longName'] = ds.longName if self.sos_type.lower() == 'ndbc': station['wmoID'] = station_urn.split(':')[-1] else: station['wmoID'] = ds.get_ioos_def('wmoID', 'identifier', ont) station['serverName'] = self.server_name # Some capabilities-level metadata: station['title'] = sosgc.identification.title station['abstract'] = sosgc.identification.abstract station['keywords'] = sosgc.identification.keywords station['begin_service_date'] = begin_service_date # Beware that a station can have >1 classifier of the same type # This code does not accommodate that possibility station['platformType'] = ds.platformType station['parentNetwork'] = ds.get_ioos_def('parentNetwork', 'classifier', ont) station['sponsor'] = ds.get_ioos_def('sponsor', 'classifier', ont) # store some nested dictionaries in 'station' for appropriate SensorML sources: station['contacts_dct'] = contacts_dct station['documents_dct'] = documents_dct if self.sos_type.lower( ) == 'ndbc' and sosgc_station_offering is not None: station['starting'] = sosgc_station_offering.begin_position station['ending'] = sosgc_station_offering.end_position else: station['starting'] = ds.starting station['ending'] = ds.ending if self.sos_type.lower( ) == 'ndbc' and sosgc_station_offering is not None: station[ 'variable_uris'] = sosgc_station_offering.observed_properties station['variables'] = [ var.split('/')[-1] for var in sosgc_station_offering.observed_properties ] station['parameter_uris'] = ','.join(station['variable_uris']) station['parameters'] = ','.join(station['variables']) else: station['variable_uris'] = ds.variables station['variables'] = [ var.split('/')[-1] for var in ds.variables ] station['parameter_uris'] = ','.join(quant_lst) station['parameters'] = ','.join(parameter_lst) if self.verbose: for var in station['variable_uris']: self.log.write(u"\nvariable: {var}".format(var=var)) print("variable: {var}".format(var=var)) # print(sosgc.contents) # for id, offering in sosgc.contents.iteritems(): # print("sosgc.contents: {item}".format(item=id)) # parse 'responseFormat' vals from SensorML: # response_formats = sosgc.contents[station_urn].response_formats response_formats = [] for id, sosgc.content in sosgc.contents.items(): if sosgc.content.name == station_urn: response_formats = sosgc.content.response_formats # response_formats = [ sosgc.content.response_formats for id, sosgc.content in sosgc.contents.items() # if sosgc.content.name == station_urn ] # match responseFormats from SensorML (response_formats) against those passed in --response_formats parameter to # populate 'download_formats' list, that is then used to generate GetObservation requests for the template: # (default --response_formats values are: 'application/json,application/zip; subtype=x-netcdf' ) download_formats = [ response_format for response_format in response_formats if response_format in self.response_formats ] station['response_formats'] = response_formats station['download_formats'] = download_formats if self.verbose: for format in response_formats: self.log.write( u"\nresponseFormat: {format}".format(format=format)) print("responseFormat: {format}".format(format=format)) for format in download_formats: self.log.write( u"\ndownloadFormats: {format}".format(format=format)) print("downloadFormats: {format}".format(format=format)) # calculate event_time using self.getobs_req_hours: event_time_formatstr = "{begin:%Y-%m-%dT%H:%M:%S}{utc_code}/{end:%Y-%m-%dT%H:%M:%S}{utc_code}" utc_code = 'Z' if self.sos_type.lower() == 'ndbc' else '' if station['starting'] is not None and station[ 'ending'] is not None: event_time = event_time_formatstr.format( begin=station['ending'] - timedelta(hours=self.getobs_req_hours), end=station['ending'], utc_code=utc_code) if self.verbose: self.log.write( u"\nUsing starting/ending times from SensorML for eventTime" ) print( "Using starting/ending times from SensorML for eventTime" ) self.log.write( u"\nobservationTimeRange: starting: {start}, ending: {end}" .format(start=station['starting'], end=station['ending'])) print( "observationTimeRange: starting: {start}, ending: {end}" .format(start=station['starting'], end=station['ending'])) else: now = datetime.now(pytz.utc) then = now - timedelta(hours=self.getobs_req_hours) event_time = event_time_formatstr.format(begin=then, end=now, utc_code=utc_code) if self.verbose: self.log.write( u"\nNo 'observationTimeRange' present in SensorML. Using present time for eventTime: then: {then:%Y-%m-%dT%H:%M:%S%z}, now: {now:%Y-%m-%dT%H:%M:%S%z}" .format(then=then, now=now)) print( "No 'observationTimeRange' present in SensorML. Using present time for eventTime: then: {then:%Y-%m-%dT%H:%M:%S%z}, now: {now:%Y-%m-%dT%H:%M:%S%z}" .format(then=then, now=now)) if self.verbose: self.log.write(u"\neventTime: {time}".format(time=event_time)) print("eventTime: {time}".format(time=event_time)) # create a dict to store parameters for valid example GetObservation requests for station: getobs_req_dct = {} # populate a parameters dictionary for download links for each 'observedProperty' type # and secondly for each 'responseFormat' per observedProperty: getobs_params_base = { 'service': 'SOS', 'request': 'GetObservation', 'version': '1.0.0', 'offering': station_urn, 'eventTime': event_time } for variable in station['variable_uris']: getobs_params = getobs_params_base.copy() getobs_params['observedProperty'] = variable variable = variable.split('/')[-1] for format in download_formats: getobs_params['responseFormat'] = format getobs_request_url_encoded = sos_url + '?' + urlencode( getobs_params) getobs_request_url = unquote(getobs_request_url_encoded) getobs_req_dct[variable + '-' + format] = { 'variable': variable, 'url': getobs_request_url, 'format_type': self.RESPONSE_FORMAT_TYPE_MAP.get(format, format), 'format_name': self.RESPONSE_FORMAT_NAME_MAP.get(format, format) } if self.verbose: self.log.write( u"\ngetobs_request_url (var: {variable}): {getobs_request_url}" .format(variable=variable.split("/")[-1], getobs_request_url=getobs_request_url)) print( "getobs_request_url (var: {variable}): {getobs_request_url}" .format(variable=variable.split("/")[-1], getobs_request_url=getobs_request_url)) # ToDo: finish adding the 'getobs_req_dct' to the output template station['getobs_req_dct'] = getobs_req_dct station_recs.append(station) # extra debug for failed stations in verbose mode: if self.verbose: self.log.write( u"\n\n\nSOS DescribeSensor request errors recap. Failed requests:" ) print("SOS DescribeSensor request errors recap. Failed requests:") for station_fail, msg in iteritems(sml_errors): self.log.write( u"\n{station} - {msg}. DescribeSensor URL: {ds}".format( station=station_fail, msg=msg, ds=describe_sensor_url[station_fail].replace( "&", "&"))) print("{station} - {msg}. DescribeSensor URL: {ds}".format( station=station_fail, msg=msg, ds=describe_sensor_url[station_fail].replace("&", "&"))) if failures: self.log.write( u"\nStations in 'failures' list (should match DescribeSensor errors):" ) print( "Stations in 'failures' list (should match DescribeSensor errors):" ) for station_fail in failures: self.log.write(u"\n{station}".format(station=station_fail)) print("{station}".format(station=station_fail)) if station_recs: stations_df = pd.DataFrame.from_records(station_recs, columns=station.keys()) stations_df.index = stations_df['station_urn'] return stations_df else: return None
class SosHarvest(Harvester): def __init__(self, service): Harvester.__init__(self, service) def harvest(self): self.sos = SensorObservationService(self.service.get('url')) for offering in self.sos.offerings: # TODO: We assume an offering should only have one procedure here # which will be the case in sos 2.0, but may not be the case right now # on some non IOOS supported servers. uid = offering.procedures[0] sp_uid = uid.split(":") # List storing the stations that have already been processed in this SOS server. # This is kept and checked later to avoid servers that have the same stations in many offerings. processed = [] # temnplate: urn:ioos:type:authority:id # sample: ioos:station:wmo:21414 if sp_uid[2] == "station": # Station Offering if not uid in processed: self.process_station(uid) processed.append(uid) elif sp_uid[2] == "network": # Network Offering network_ds = IoosDescribeSensor( self.sos.describe_sensor( outputFormat= 'text/xml;subtype="sensorML/1.0.1/profiles/ioos_sos/1.0"', procedure=uid)) # Iterate over stations in the network and process them individually for proc in network_ds.procedures: if proc is not None and proc.split(":")[2] == "station": if not proc in processed: self.process_station(proc) processed.append(proc) def process_station(self, uid): """ Makes a DescribeSensor request based on a 'uid' parameter being a station procedure """ GML_NS = "http://www.opengis.net/gml" XLINK_NS = "http://www.w3.org/1999/xlink" with app.app_context(): metadata_value = etree.fromstring( self.sos.describe_sensor( outputFormat= 'text/xml;subtype="sensorML/1.0.1/profiles/ioos_sos/1.0"', procedure=uid)) station_ds = IoosDescribeSensor(metadata_value) unique_id = station_ds.id if unique_id is None: app.logger.warn( "Could not get a 'stationID' from the SensorML identifiers. Looking for a definition of 'http://mmisw.org/ont/ioos/definition/stationID'" ) return dataset = db.Dataset.find_one({'uid': unicode(unique_id)}) if dataset is None: dataset = db.Dataset() dataset.uid = unicode(unique_id) # Find service reference in Dataset.services and remove (to replace it) tmp = dataset.services[:] for d in tmp: if d['service_id'] == self.service.get('_id'): dataset.services.remove(d) # Parsing messages messages = [] # NAME name = unicode_or_none(station_ds.shortName) if name is None: messages.append( u"Could not get a 'shortName' from the SensorML identifiers. Looking for a definition of 'http://mmisw.org/ont/ioos/definition/shortName'" ) # DESCRIPTION description = unicode_or_none(station_ds.longName) if description is None: messages.append( u"Could not get a 'longName' from the SensorML identifiers. Looking for a definition of 'http://mmisw.org/ont/ioos/definition/longName'" ) # PLATFORM TYPE asset_type = unicode_or_none(station_ds.platformType) if asset_type is None: messages.append( u"Could not get a 'platformType' from the SensorML identifiers. Looking for a definition of 'http://mmisw.org/ont/ioos/definition/platformType'" ) # LOCATION is in GML gj = None loc = station_ds.location if loc is not None and loc.tag == "{%s}Point" % GML_NS: pos_element = loc.find("{%s}pos" % GML_NS) # strip out points positions = map(float, testXMLValue(pos_element).split(" ")) crs = Crs(testXMLAttribute(pos_element, "srsName")) if crs.axisorder == "yx": gj = json.loads( geojson.dumps( geojson.Point([positions[1], positions[0]]))) else: gj = json.loads( geojson.dumps( geojson.Point([positions[0], positions[1]]))) else: messages.append( u"Found an unrecognized child of the sml:location element and did not attempt to process it: %s" % etree.tostring(loc).strip()) service = { # Reset service 'name': name, 'description': description, 'service_type': self.service.get('service_type'), 'service_id': ObjectId(self.service.get('_id')), 'data_provider': self.service.get('data_provider'), 'metadata_type': u'sensorml', 'metadata_value': unicode(etree.tostring(metadata_value)).strip(), 'messages': map(unicode, messages), 'keywords': map(unicode, sorted(station_ds.keywords)), 'variables': map(unicode, sorted(station_ds.variables)), 'asset_type': asset_type, 'geojson': gj, 'updated': datetime.utcnow() } dataset.services.append(service) dataset.updated = datetime.utcnow() dataset.save() return "Harvested"
def get_stations_df(self, sos_url, station_urns_sel=None): """ Returns a GeoDataFrame """ # LATER: ADD ERROR TEST/CATCH AFTER EACH WEB REQUEST oFrmt = 'text/xml; subtype="sensorML/1.0.1/profiles/ioos_sos/1.0"' params = { 'service': 'SOS', 'request': 'GetCapabilities', 'acceptVersions': '1.0.0' } sos_url_params = sos_url + '?' + urlencode(params) sos_url_params_esc = sos_url_params.replace("&", "&") # sos_url_params_quoted = quote(sos_url_params,"/=:") # sos_url_params_unquoted = unquote(sos_url_params) try: sosgc = SensorObservationService(sos_url_params) except (ConnectionError, ReadTimeout) as e: self.log.write( u"Error: unable to connect to SOS service: {url} due to HTTP connection error.\n" .format(url=sos_url_params)) self.log.write( u"HTTP connection error: {err}.\n".format(err=e.message)) sys.exit( "Error: unable to connect to SOS service: {url}. \nUnderlying HTTP connection error: {err}" .format(url=sos_url_params, err=e.message)) if station_urns_sel is not None: station_urns = station_urns_sel else: sos_collector = IoosSweSos(sos_url) station_urns = [ urn.name for urn in sos_collector.server.offerings if 'network' not in urn.name.split(':') ] sos_collector.features = station_urns sml_lst = sos_collector.metadata(timeout=200) station_recs = [] for station_idx, station_urn in enumerate(station_urns): if station_urns_sel is not None: sml_str = sosgc.describe_sensor(procedure=station_urn, outputFormat=oFrmt, timeout=200) sml = SensorML(sml_str) else: sml = sml_lst[station_idx] # debug: # if self.verbose: # self.log.write(unicode(etree.tostring(sml._root))) ds = IoosDescribeSensor(sml._root) pos = testXMLValue( ds.system.location.find(self.nsp('gml:Point/gml:pos'))) system_el = sml._root.findall(self.nsp('sml:member'))[0].find( self.nsp('sml:System')) # Parse the DocumentList into a dict storing documents by index value 'name' (may cause index duplication # errors but there is not enough information in SensorML for alternatives) # Assume that member corresponds to xlink:arcrole="urn:ogc:def:role:webPage" documents = system_el.findall( self.nsp('sml:documentation/sml:DocumentList/sml:member')) documents_dct = {} for d in documents: document = Documentation(d) name = testXMLAttribute(d, "name") # url = document.documents[0].url documents_dct[name] = document contacts = system_el.findall( self.nsp('sml:contact/sml:ContactList/sml:member')) contacts_dct = {} for c in contacts: contact = Contact(c) role = contact.role.split('/')[-1] contacts_dct[role] = contact sweQuants = system_el.findall( self.nsp('sml:outputs/sml:OutputList/sml:output/swe:Quantity')) quant_lst = [ sweQuant.attrib['definition'] for sweQuant in sweQuants ] parameter_lst = [sweQuant.split('/')[-1] for sweQuant in quant_lst] # attempt to read beginPosition, if available: beginPosition = testXMLValue( system_el.find( self.nsp( 'sml:validTime/gml:TimePeriod/gml:beginPosition'))) try: begin_service_date = parser.parse(beginPosition) except AttributeError as e: begin_service_date = None station = OrderedDict() station['station_urn'] = station_urn station['sos_url'] = sos_url_params_esc station['lon'] = float(pos.split()[1]) station['lat'] = float(pos.split()[0]) station['shortName'] = ds.shortName station['longName'] = ds.longName station['wmoID'] = ds.get_ioos_def('wmoID', 'identifier', ont) station['serverName'] = self.server_name # Some capabilities-level metadata: station['title'] = sosgc.identification.title station['abstract'] = sosgc.identification.abstract station['keywords'] = sosgc.identification.keywords station['begin_service_date'] = begin_service_date # Beware that a station can have >1 classifier of the same type # This code does not accommodate that possibility station['platformType'] = ds.platformType station['parentNetwork'] = ds.get_ioos_def('parentNetwork', 'classifier', ont) station['sponsor'] = ds.get_ioos_def('sponsor', 'classifier', ont) # store some nested dictionaries in 'station' for appopriate SensorML sources: station['contacts_dct'] = contacts_dct station['documents_dct'] = documents_dct # MW: the 'operator_' and 'publisher_' attributes can be removed bc they are not used # in the template code currently in favor of 'contacts_dct' # station['operatorSector'] = ds.get_ioos_def('operatorSector', 'classifier', ont) # station['operator_org'] = contacts_dct['operator'].organization # station['operator_country'] = contacts_dct['operator'].country # station['operator_url'] = contacts_dct['operator'].url # station['operator_email'] = contacts_dct['operator'].email # station['publisher'] = ds.get_ioos_def('publisher', 'classifier', ont) # station['publisher_org'] = contacts_dct['publisher'].organization # station['publisher_url'] = contacts_dct['publisher'].url # station_dct['publisher_email'] = contacts_dct['publisher'].electronicMailAddress station['starting'] = ds.starting station['ending'] = ds.ending # station['starting_isostr'] = datetime.isoformat(ds.starting) # station['ending_isostr'] = datetime.isoformat(ds.ending) station['parameter_uris'] = ','.join(quant_lst) station['parameters'] = ','.join(parameter_lst) station['variables'] = [var.split('/')[-1] for var in ds.variables] # debug: if self.verbose: self.log.write(u"\nProcessing station: {station}\n".format( station=station_urn)) print("\nProcessing station: {station}".format( station=station_urn)) for var in ds.variables: self.log.write(u"variable: {var}\n".format(var=var)) print("variable: {var}".format(var=var)) # print(sosgc.contents) # for id, offering in sosgc.contents.iteritems(): # print("sosgc.contents: {item}".format(item=id)) # parse 'responseFormat' values and populate list: # response_formats = sosgc.contents[station_urn].response_formats response_formats = [] for id, sosgc.content in sosgc.contents.items(): if sosgc.content.name == station_urn: response_formats = sosgc.content.response_formats # response_formats = [ sosgc.content.response_formats for id, sosgc.content in sosgc.contents.items() if sosgc.content.name == station_urn ] # subset responseFormats (response_formats) for download links matching those passed in --response_formats parameter # (or 'application/json,application/zip; subtype=x-netcdf' by default): download_formats = [ response_format for response_format in response_formats if response_format in self.response_formats ] station['response_formats'] = response_formats station['download_formats'] = download_formats if self.verbose: for format in response_formats: self.log.write( u"responseFormat: {format}\n".format(format=format)) print("responseFormat: {format}".format(format=format)) for format in download_formats: self.log.write( u"downloadFormats: {format}\n".format(format=format)) print("downloadFormats: {format}".format(format=format)) # calculate event_time using self.getobs_req_hours: if ds.starting is not None and ds.ending is not None: event_time = "{begin:%Y-%m-%dT%H:%M:%S}/{end:%Y-%m-%dT%H:%M:%S}\n".format( begin=ds.ending - timedelta(hours=self.getobs_req_hours), end=ds.ending) else: now = datetime.now(pytz.utc) then = now - timedelta(hours=self.getobs_req_hours) event_time = "{begin:%Y-%m-%dT%H:%M:%S}/{end:%Y-%m-%dT%H:%M:%S}\n".format( begin=then, end=now) if self.verbose: self.log.write( u"then: {then:%Y-%m-%dT%H:%M:%S%z}, now: {now:%Y-%m-%dT%H:%M:%S%z}\n" .format(then=then, now=now)) print( "then: {then:%Y-%m-%dT%H:%M:%S%z}, now: {now:%Y-%m-%dT%H:%M:%S%z}" .format(then=then, now=now)) if self.verbose: self.log.write(u"eventTime: {time}\n".format(time=event_time)) print("eventTime: {time}".format(time=event_time)) # create a dict to store parameters for valid example GetObservation requests for station: getobs_req_dct = {} # populate a parameters dictionary for download links for each 'observedProperty' type and secondly for each 'responseFormat' per observedProperty: getobs_params_base = { 'service': 'SOS', 'request': 'GetObservation', 'version': '1.0.0', 'offering': station_urn, 'eventTime': event_time } for variable in ds.variables: getobs_params = getobs_params_base.copy() getobs_params['observedProperty'] = variable variable = variable.split('/')[-1] for format in download_formats: getobs_params['responseFormat'] = format getobs_request_url_encoded = sos_url + '?' + urlencode( getobs_params) getobs_request_url = unquote(getobs_request_url_encoded) getobs_request_url_esc = getobs_request_url.replace( "&", "&") getobs_req_dct[variable + '-' + format] = { 'variable': variable, 'url': getobs_request_url_esc, 'format_type': self.RESPONSE_FORMAT_TYPE_MAP[format], 'format_name': self.RESPONSE_FORMAT_NAME_MAP[format] } if self.verbose: self.log.write( u"getobs_request_url (var: {variable}): {getobs_request_url}\ngetobs_request_url_esc (var: {variable}): {getobs_request_url_esc}\n" .format( variable=variable.split("/")[-1], getobs_request_url=getobs_request_url, getobs_request_url_esc=getobs_request_url_esc)) print( "getobs_request_url (var: {variable}): {getobs_request_url}\ngetobs_request_url_esc (var: {variable}): {getobs_request_url_esc}" .format( variable=variable.split("/")[-1], getobs_request_url=getobs_request_url, getobs_request_url_esc=getobs_request_url_esc)) # ToDo: finish adding the 'getobs_req_dct' to the output template station['getobs_req_dct'] = getobs_req_dct station_recs.append(station) stations_df = pd.DataFrame.from_records(station_recs, columns=station.keys()) stations_df.index = stations_df['station_urn'] return stations_df