import time
from ftplib import FTP
-from optparse import OptionParser
+from argparse import ArgumentParser
import requests
import zeep.cache
import zeep.transports
-__version__ = '2.0.0'
+__version__ = '2.2.0'
LOG = logging.getLogger(__name__)
LOG.setLevel(logging.NOTSET)
def get_parser():
"""Return OptionParser for this program"""
- parser = OptionParser(version=__version__)
+ parser = ArgumentParser()
- parser.add_option("--config",
- dest="config",
- metavar="FILE",
- help="location of config file")
+ parser.add_argument("--version",
+ action="version",
+ version=__version__)
- parser.add_option("--log-level",
- dest="log_level",
- default=None,
- help="set the log level (debug, info, warning)")
+ parser.add_argument("--config",
+ dest="config",
+ metavar="FILE",
+ help="location of config file")
- parser.add_option("--dry-run",
- action="store_true",
- dest="dry_run",
- default=False,
- help="fetch data but don't upload to InfoEx")
+ parser.add_argument("--log-level",
+ dest="log_level",
+ default=None,
+ help="set the log level (debug, info, warning)")
+
+ parser.add_argument("--dry-run",
+ action="store_true",
+ dest="dry_run",
+ default=False,
+ help="fetch data but don't upload to InfoEx")
return parser
'wx_data': {}, # placeholder key, values to come later
}
- data = dict()
- data['provider'] = config['station']['type']
+ station = dict()
+ station['provider'] = config['station']['type']
- if data['provider'] not in ['nrcs', 'mesowest']:
+ if station['provider'] not in ['nrcs', 'mesowest']:
print("Please specify either nrcs or mesowest as the station type.")
sys.exit(1)
- if data['provider'] == 'nrcs':
- data['source'] = 'https://www.wcc.nrcs.usda.gov/awdbWebService/services?WSDL'
- data['station_id'] = config['station']['station_id']
-
- try:
- data['desired_data'] = config['station']['desired_data'].split(',')
- except:
- # desired_data malformed or missing, setting default
- data['desired_data'] = [
- 'TOBS', # AIR TEMPERATURE OBSERVED (degF)
- 'SNWD', # SNOW DEPTH (in)
- 'PREC' # PRECIPITATION ACCUMULATION (in)
- ]
+ if station['provider'] == 'nrcs':
+ station['source'] = 'https://www.wcc.nrcs.usda.gov/awdbWebService/services?WSDL'
+ station['station_id'] = config['station']['station_id']
+ station['desired_data'] = config['station']['desired_data'].split(',')
# XXX: For NRCS, we're manually overriding units for now! Once
# unit conversion is supported for NRCS, REMOVE THIS!
- if 'units' not in data:
- data['units'] = 'imperial'
+ if 'units' not in station:
+ station['units'] = 'imperial'
- if data['provider'] == 'mesowest':
- data['source'] = 'https://api.synopticdata.com/v2/stations/timeseries'
- data['station_id'] = config['station']['station_id']
- data['units'] = config['station']['units']
-
- try:
- data['desired_data'] = config['station']['desired_data']
- except:
- # desired_data malformed or missing, setting default
- data['desired_data'] = 'air_temp,snow_depth'
+ if station['provider'] == 'mesowest':
+ station['source'] = 'https://api.synopticdata.com/v2/stations/timeseries'
+ station['station_id'] = config['station']['station_id']
+ station['units'] = config['station']['units']
+ station['desired_data'] = config['station']['desired_data']
# construct full API URL (sans start/end time, added later)
- data['source'] = data['source'] + '?token=' + config['station']['token'] + '&within=60&units=' + data['units'] + '&stid=' + data['station_id'] + '&vars=' + data['desired_data']
-
- except KeyError as e:
- LOG.critical("%s not defined in %s" % (e, options.config))
- exit(1)
- except Exception as exc:
- LOG.critical("Exception occurred in config parsing: '%s'" % (exc))
+ station['source'] = station['source'] + '?token=' + \
+ config['station']['token'] + \
+ '&within=60&units=' + station['units'] + \
+ '&stid=' + station['station_id'] + \
+ '&vars=' + station['desired_data']
+
+ except KeyError as err:
+ LOG.critical("%s not defined in configuration file", err)
exit(1)
# all sections/values present in config file, final sanity check
try:
for key in config.sections():
for subkey in config[key]:
- if not len(config[key][subkey]):
- raise ValueError;
- except ValueError as exc:
- LOG.critical("Config value '%s.%s' is empty" % (key, subkey))
+ if not config[key][subkey]:
+ raise ValueError
+ except ValueError:
+ LOG.critical("Config value '%s.%s' is empty", key, subkey)
exit(1)
- return (infoex, data)
+ return (infoex, station)
def setup_logging(log_level):
"""Setup our logging infrastructure"""
try:
from systemd.journal import JournalHandler
LOG.addHandler(JournalHandler())
- except:
+ except ImportError:
## fallback to syslog
#import logging.handlers
#LOG.addHandler(logging.handlers.SysLogHandler())
def main():
"""Main routine: sort through args, decide what to do, then do it"""
parser = get_parser()
- (options, args) = parser.parse_args()
+ options = parser.parse_args()
config = configparser.ConfigParser(allow_no_value=False)
print("\nPlease select an appropriate log level or remove the switch (--log-level).")
sys.exit(1)
- (infoex, data) = setup_config(config)
+ (infoex, station) = setup_config(config)
LOG.debug('Config parsed, starting up')
# create mappings
(fmap, final_data) = setup_infoex_fields_mapping(infoex['location_uuid'])
- iemap = setup_infoex_counterparts_mapping(data['provider'])
+ iemap = setup_infoex_counterparts_mapping(station['provider'])
# override units if user selected metric
- #
- # NOTE: to update this, use the fmap<->final_data mapping laid out above
- #
- # NOTE: this only 'works' with MesoWest for now, as the MesoWest API
- # itself handles the unit conversion; in the future, we will also
- # support NRCS unit conversion, but this must be done by this
- # program.
- if data['units'] == 'metric':
- final_data[fmap['tempPresUnit']] = 'C'
- final_data[fmap['hsUnit']] = 'm'
- final_data[fmap['windSpeedUnit']] = 'm/s'
- final_data[fmap['windGustSpeedNumUnit']] = 'm/s'
+ if station['units'] == 'metric':
+ final_data = switch_units_to_metric(final_data, fmap)
- # floor time to nearest hour
- dt = datetime.datetime.now()
- end_date = dt - datetime.timedelta(minutes=dt.minute % 60,
- seconds=dt.second,
- microseconds=dt.microsecond)
- begin_date = end_date - datetime.timedelta(hours=3)
+ (begin_date, end_date) = setup_time_values()
# get the data
- LOG.debug("Getting %s data from %s to %s" % (str(data['desired_data']),
- str(begin_date), str(end_date)))
+ LOG.debug("Getting %s data from %s to %s", str(station['desired_data']),
+ str(begin_date), str(end_date))
time_all_elements = time.time()
- # NRCS-specific code
- if data['provider'] == 'nrcs':
- transport = zeep.transports.Transport(cache=zeep.cache.SqliteCache())
- client = zeep.Client(wsdl=data['source'], transport=transport)
-
- for elementCd in data['desired_data']:
- time_element = time.time()
-
- # get the last three hours of data for this elementCd
- tmp = client.service.getHourlyData(
- stationTriplets=[data['station_id']],
- elementCd=elementCd,
- ordinal=1,
- beginDate=begin_date,
- endDate=end_date)
-
- LOG.info("Time to get elementCd '%s': %.3f sec" % (elementCd,
- time.time() - time_element))
-
- values = tmp[0]['values']
-
- # sort and isolate the most recent
- #
- # NOTE: we do this because sometimes there are gaps in hourly data
- # in NRCS; yes, we may end up with slightly inaccurate data,
- # so perhaps this decision will be re-evaluated in the future
- if values:
- ordered = sorted(values, key=lambda t: t['dateTime'], reverse=True)
- infoex['wx_data'][elementCd] = ordered[0]['value']
- else:
- infoex['wx_data'][elementCd] = None
-
- # MesoWest-specific code
- elif data['provider'] == 'mesowest':
- # massage begin/end date format
- begin_date_str = begin_date.strftime('%Y%m%d%H%M')
- end_date_str = end_date.strftime('%Y%m%d%H%M')
-
- # construct final, completed API URL
- api_req_url = data['source'] + '&start=' + begin_date_str + '&end=' + end_date_str
- req = requests.get(api_req_url)
-
- try:
- json = req.json()
- except ValueError:
- LOG.error("Bad JSON in MesoWest response")
- sys.exit(1)
-
- try:
- observations = json['STATION'][0]['OBSERVATIONS']
- except ValueError:
- LOG.error("Bad JSON in MesoWest response")
- sys.exit(1)
-
- pos = len(observations['date_time']) - 1
-
- for elementCd in data['desired_data'].split(','):
- # sort and isolate the most recent, see note above in NRCS for how and
- # why this is done
- #
- # NOTE: Unlike in the NRCS case, the MesoWest API response contains all
- # data (whereas with NRCS, we have to make a separate request for
- # each element we want). This is nice for network efficiency but
- # it means we have to handle this part differently for each.
- #
- # NOTE: Also unlike NRCS, MesoWest provides more granular data; NRCS
- # provides hourly data, but MesoWest can often provide data every
- # 10 minutes -- though this provides more opportunity for
- # irregularities
-
- # we may not have the data at all
- key_name = elementCd + '_set_1'
- if key_name in observations:
- if observations[key_name][pos]:
- infoex['wx_data'][elementCd] = observations[key_name][pos]
- else:
- infoex['wx_data'][elementCd] = None
- else:
- infoex['wx_data'][elementCd] = None
+ # get the data
+ if station['provider'] == 'nrcs':
+ infoex['wx_data'] = get_nrcs_data(begin_date, end_date, station)
+ elif station['provider'] == 'mesowest':
+ infoex['wx_data'] = get_mesowest_data(begin_date, end_date,
+ station)
- LOG.info("Time to get all data : %.3f sec" % (time.time() -
- time_all_elements))
+ LOG.info("Time taken to get all data : %.3f sec", time.time() -
+ time_all_elements)
LOG.debug("infoex[wx_data]: %s", str(infoex['wx_data']))
final_data[fmap['obDate']] = end_date.strftime('%m/%d/%Y')
final_data[fmap['obTime']] = end_date.strftime('%H:%M')
- for elementCd in infoex['wx_data']:
- if elementCd not in iemap:
- LOG.warning("BAD KEY wx_data['%s']" % (elementCd))
+ for element_cd in infoex['wx_data']:
+ if element_cd not in iemap:
+ LOG.warning("BAD KEY wx_data['%s']", element_cd)
continue
+ # Massage precision of certain values to fit InfoEx's
+ # expectations
+ #
+ # 0 decimal places: wind speed, wind direction, wind gust, snow depth
+ # 1 decimal place: air temp, baro
+ # Avoid transforming None values
+ if infoex['wx_data'][element_cd] is None:
+ continue
+ elif element_cd in ['wind_speed', 'WSPD', 'wind_direction',
+ 'WDIR', 'wind_gust', 'SNWD', 'snow_depth']:
+ infoex['wx_data'][element_cd] = round(infoex['wx_data'][element_cd])
+ elif element_cd in ['TOBS', 'air_temp', 'PRES', 'pressure']:
+ infoex['wx_data'][element_cd] = round(infoex['wx_data'][element_cd], 1)
+
# CONSIDER: Casting every value to Float() -- need to investigate if
# any possible elementCds we may want are any other data
# type than float.
# storedUnitCd. But that's pretty network-intensive and
# may not even be worth it if there's only e.g. one or two
# exceptions to any otherwise uniformly Float value set.
- final_data[fmap[iemap[elementCd]]] = infoex['wx_data'][elementCd]
+ final_data[fmap[iemap[element_cd]]] = infoex['wx_data'][element_cd]
- LOG.debug("final_data: %s" % (str(final_data)))
+ LOG.debug("final_data: %s", str(final_data))
if not write_local_csv(infoex['csv_filename'], final_data):
LOG.warning('Could not write local CSV file: %s',
infoex['csv_filename'])
- return 1;
+ return 1
if not options.dry_run:
upload_csv(infoex['csv_filename'], infoex)
LOG.debug('DONE')
return 0
-# Data structure operations
+# data structure operations
def setup_infoex_fields_mapping(location_uuid):
"""
Create a mapping of InfoEx fields to the local data's indexing scheme.
INFOEX FIELDS
-
+
This won't earn style points in Python, but here we establish a couple
of helpful mappings variables. The reason this is helpful is that the
end result is simply an ordered set, the CSV file. But we still may
want to manipulate the values arbitrarily before writing that file.
-
+
Also note that the current Auto Wx InfoEx documentation shows these
keys in a graphical table with the "index" beginning at 1, but here we
sanely index beginning at 0.
"""
+ # pylint: disable=too-many-statements,multiple-statements,bad-whitespace
fmap = {} ; final_data = [None] * 29
fmap['Location UUID'] = 0 ; final_data[0] = location_uuid
fmap['obDate'] = 1 ; final_data[1] = None
if provider == 'nrcs':
iemap['PREC'] = 'precipitationGauge'
iemap['TOBS'] = 'tempPres'
+ iemap['TMAX'] = 'tempMaxHour'
+ iemap['TMIN'] = 'tempMinHour'
iemap['SNWD'] = 'hS'
iemap['PRES'] = 'baro'
iemap['RHUM'] = 'rH'
elif provider == 'mesowest':
iemap['precip_accum'] = 'precipitationGauge'
iemap['air_temp'] = 'tempPres'
+ iemap['air_temp_high_24_hour'] = 'tempMaxHour'
+ iemap['air_temp_low_24_hour'] = 'tempMinHour'
iemap['snow_depth'] = 'hS'
iemap['pressure'] = 'baro'
iemap['relative_humidity'] = 'rH'
return iemap
+# provider-specific operations
+def get_nrcs_data(begin, end, station):
+ """get the data we're after from the NRCS WSDL"""
+ transport = zeep.transports.Transport(cache=zeep.cache.SqliteCache())
+ client = zeep.Client(wsdl=station['source'], transport=transport)
+ remote_data = {}
+
+ for element_cd in station['desired_data']:
+ time_element = time.time()
+
+ # get the last three hours of data for this elementCd/element_cd
+ tmp = client.service.getHourlyData(
+ stationTriplets=[station['station_id']],
+ elementCd=element_cd,
+ ordinal=1,
+ beginDate=begin,
+ endDate=end)
+
+ LOG.info("Time to get NRCS elementCd '%s': %.3f sec", element_cd,
+ time.time() - time_element)
+
+ values = tmp[0]['values']
+
+ # sort and isolate the most recent
+ #
+ # NOTE: we do this because sometimes there are gaps in hourly data
+ # in NRCS; yes, we may end up with slightly inaccurate data,
+ # so perhaps this decision will be re-evaluated in the future
+ if values:
+ ordered = sorted(values, key=lambda t: t['dateTime'], reverse=True)
+ remote_data[element_cd] = ordered[0]['value']
+ else:
+ remote_data[element_cd] = None
+
+ return remote_data
+
+def get_mesowest_data(begin, end, station):
+ """get the data we're after from the MesoWest/Synoptic API"""
+ remote_data = {}
+
+ # massage begin/end date format
+ begin_date_str = begin.strftime('%Y%m%d%H%M')
+ end_date_str = end.strftime('%Y%m%d%H%M')
+
+ # construct final, completed API URL
+ api_req_url = station['source'] + '&start=' + begin_date_str + '&end=' + end_date_str
+ req = requests.get(api_req_url)
+
+ try:
+ json = req.json()
+ except ValueError:
+ LOG.error("Bad JSON in MesoWest response")
+ sys.exit(1)
+
+ try:
+ observations = json['STATION'][0]['OBSERVATIONS']
+ except ValueError:
+ LOG.error("Bad JSON in MesoWest response")
+ sys.exit(1)
+
+ pos = len(observations['date_time']) - 1
+
+ for element_cd in station['desired_data'].split(','):
+ # sort and isolate the most recent, see note above in NRCS for how and
+ # why this is done
+ #
+ # NOTE: Unlike in the NRCS case, the MesoWest API response contains all
+ # data (whereas with NRCS, we have to make a separate request for
+ # each element we want). This is nice for network efficiency but
+ # it means we have to handle this part differently for each.
+ #
+ # NOTE: Also unlike NRCS, MesoWest provides more granular data; NRCS
+ # provides hourly data, but MesoWest can often provide data every
+ # 10 minutes -- though this provides more opportunity for
+ # irregularities
+
+ # we may not have the data at all
+ key_name = element_cd + '_set_1'
+ if key_name in observations:
+ if observations[key_name][pos]:
+ remote_data[element_cd] = observations[key_name][pos]
+ else:
+ remote_data[element_cd] = None
+ else:
+ remote_data[element_cd] = None
+
+ return remote_data
+
+def switch_units_to_metric(data_map, mapping):
+ """replace units with metric counterparts"""
+
+ # NOTE: to update this, use the fmap<->final_data mapping laid out
+ # in setup_infoex_fields_mapping ()
+ #
+ # NOTE: this only 'works' with MesoWest for now, as the MesoWest API
+ # itself handles the unit conversion; in the future, we will also
+ # support NRCS unit conversion, but this must be done by this
+ # program.
+ data_map[mapping['tempPresUnit']] = 'C'
+ data_map[mapping['hsUnit']] = 'm'
+ data_map[mapping['windSpeedUnit']] = 'm/s'
+ data_map[mapping['windGustSpeedNumUnit']] = 'm/s'
+
+ return data_map
+
# CSV operations
def write_local_csv(path_to_file, data):
"""Write the specified CSV file to disk"""
- with open(path_to_file, 'w') as f:
+ with open(path_to_file, 'w') as file_object:
# The requirement is that empty values are represented in the CSV
# file as "", csv.QUOTE_NONNUMERIC achieves that
- LOG.debug("writing CSV file '%s'" % (path_to_file))
- writer = csv.writer(f, quoting=csv.QUOTE_NONNUMERIC)
+ LOG.debug("writing CSV file '%s'", path_to_file)
+ writer = csv.writer(file_object, quoting=csv.QUOTE_NONNUMERIC)
writer.writerow(data)
- f.close()
+ file_object.close()
return True
def upload_csv(path_to_file, infoex_data):
"""Upload the specified CSV file to InfoEx FTP and remove the file"""
with open(path_to_file, 'rb') as file_object:
- LOG.debug("uploading FTP file '%s'" % (infoex_data['host']))
+ LOG.debug("uploading FTP file '%s'", infoex_data['host'])
ftp = FTP(infoex_data['host'], infoex_data['uuid'],
infoex_data['api_key'])
ftp.storlines('STOR ' + path_to_file, file_object)
file_object.close()
os.remove(path_to_file)
+# other miscellaneous routines
+def setup_time_values():
+ """establish time bounds of data request(s)"""
+ # floor time to nearest hour
+ date_time = datetime.datetime.now()
+ end_date = date_time - datetime.timedelta(minutes=date_time.minute % 60,
+ seconds=date_time.second,
+ microseconds=date_time.microsecond)
+ begin_date = end_date - datetime.timedelta(hours=3)
+ return (begin_date, end_date)
+
if __name__ == "__main__":
sys.exit(main())
wylark / infoex-autowx.git / blobdiff