# -*- coding: utf-8 -*-
"""
-InfoEx <-> NRCS Auto Wx implementation
+InfoEx <-> NRCS/MesoWest Auto Wx implementation
Alexander Vasarab
Wylark Mountaineering LLC
-Version 1.0.0
+Version 2.0.0
-This program fetches data from an NRCS SNOTEL site and pushes it to
-InfoEx using the new automated weather system implementation.
+This program fetches data from either an NRCS SNOTEL site or MesoWest
+weather station and pushes it to InfoEx using the new automated weather
+system implementation.
It is designed to be run hourly, and it asks for the last three hours
of data of each desired type, and selects the most recent one. This
import csv
import datetime
import logging
+import os
+import sys
import time
from collections import OrderedDict
from ftplib import FTP
from optparse import OptionParser
+import requests
+
import zeep
import zeep.cache
import zeep.transports
+__version__ = '2.0.0'
+
log = logging.getLogger(__name__)
-log.setLevel(logging.DEBUG)
+log.setLevel(logging.NOTSET)
try:
from systemd.journal import JournalHandler
log.addHandler(JournalHandler())
except:
- # fallback to syslog
- import logging.handlers
- log.addHandler(logging.handlers.SysLogHandler())
-
-parser = OptionParser()
-parser.add_option("--config", dest="config", metavar="FILE", help="location of config file")
+ ## fallback to syslog
+ #import logging.handlers
+ #log.addHandler(logging.handlers.SysLogHandler())
+ # fallback to stdout
+ handler = logging.StreamHandler(sys.stdout)
+ log.addHandler(handler)
+
+parser = OptionParser(version=__version__)
+
+parser.add_option("--config",
+ dest="config",
+ metavar="FILE",
+ help="location of config file")
+
+parser.add_option("--log-level",
+ dest="log_level",
+ default=None,
+ help="set the log level (debug, info, warning)")
+
+parser.add_option("--dry-run",
+ action="store_true",
+ dest="dry_run",
+ default=False,
+ help="fetch data but don't upload to InfoEx")
(options, args) = parser.parse_args()
config = configparser.ConfigParser(allow_no_value=False)
+
+if not options.config:
+ parser.print_help()
+ print("\nPlease specify a configuration file via --config.")
+ sys.exit(1)
+
config.read(options.config)
-log.debug('STARTING UP')
+# ugly, but passable
+if options.log_level in [None, 'debug', 'info', 'warning']:
+ if options.log_level == 'debug':
+ log.setLevel(logging.DEBUG)
+ elif options.log_level == 'info':
+ log.setLevel(logging.INFO)
+ elif options.log_level == 'warning':
+ log.setLevel(logging.WARNING)
+ else:
+ log.setLevel(logging.NOTSET)
+else:
+ parser.print_help()
+ print("\nPlease select an appropriate log level or remove the switch (--log-level).")
+ sys.exit(1)
-wsdl = 'https://www.wcc.nrcs.usda.gov/awdbWebService/services?WSDL'
+log.debug('STARTING UP')
try:
infoex = {
- 'host': config['ftp']['host'],
- 'uuid': config['ftp']['uuid'],
- 'api_key': config['ftp']['api_key'],
- 'location_uuid': config['wxsite']['location_uuid'],
+ 'host': config['infoex']['host'],
+ 'uuid': config['infoex']['uuid'],
+ 'api_key': config['infoex']['api_key'],
+ 'csv_filename': config['infoex']['csv_filename'],
+ 'location_uuid': config['infoex']['location_uuid'],
'wx_data': {}, # placeholder key, values to come later
- 'csv_filename': config['wxsite']['csv_filename']
}
- station_triplet = config['wxsite']['station_triplet']
+ data = dict()
+ data['provider'] = config['station']['type']
+
+ if data['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:
+ desired_data = config['station']['desired_data'].split(',')
+ except:
+ # desired_data malformed or missing, setting default
+ desired_data = [
+ 'TOBS', # AIR TEMPERATURE OBSERVED (degF)
+ 'SNWD', # SNOW DEPTH (in)
+ 'PREC' # PRECIPITATION ACCUMULATION (in)
+ ]
+
+ # 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 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:
+ desired_data = config['station']['desired_data']
+ except:
+ # desired_data malformed or missing, setting default
+ desired_data = 'air_temp,snow_depth'
+
+ # 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=' + desired_data
- try:
- desired_data = config['wxsite']['desired_data'].split(',')
- except:
- # desired_data malformed or missing, setting default
- desired_data = [
- 'TOBS', # AIR TEMPERATURE OBSERVED (degF)
- 'SNWD', # SNOW DEPTH (in)
- 'PREC' # PRECIPITATION ACCUMULATION (in)
- ]
except KeyError as e:
log.critical("%s not defined in %s" % (e, options.config))
exit(1)
#
# 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
-# are sanely indexing beginning at 0.
+# sanely index beginning at 0.
fmap = {} ; final_data = [None] * 29
fmap['Location UUID'] = 0 ; final_data[0] = infoex['location_uuid']
fmap['obDate'] = 1 ; final_data[1] = None
fmap['hstAuto'] = 27 ; final_data[27] = None
fmap['hstAutoUnit'] = 28 ; final_data[28] = 'in'
-# one final mapping, the NRCS fields that this program supports to
+# one final mapping, the NRCS/MesoWest fields that this program supports to
# their InfoEx counterpart
iemap = {}
-iemap['PREC'] = 'precipitationGauge'
-iemap['TOBS'] = 'tempPres'
-iemap['SNWD'] = 'hS'
+
+if data['provider'] == 'nrcs':
+ iemap['PREC'] = 'precipitationGauge'
+ iemap['TOBS'] = 'tempPres'
+ iemap['SNWD'] = 'hS'
+ iemap['PRES'] = 'baro'
+ iemap['RHUM'] = 'rH'
+ iemap['WSPD'] = 'windSpeedNum'
+ iemap['WDIR'] = 'windDirectionNum'
+ # unsupported by NRCS:
+ # windGustSpeedNum
+elif data['provider'] == 'mesowest':
+ iemap['precip_accum'] = 'precipitationGauge'
+ iemap['air_temp'] = 'tempPres'
+ iemap['snow_depth'] = 'hS'
+ iemap['pressure'] = 'baro'
+ iemap['relative_humidity'] = 'rH'
+ iemap['wind_speed'] = 'windSpeedNum'
+ iemap['wind_direction'] = 'windDirectionNum'
+ iemap['wind_gust'] = 'windGustSpeedNum'
+
+# 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'
# floor time to nearest hour
dt = datetime.datetime.now()
microseconds=dt.microsecond)
begin_date = end_date - datetime.timedelta(hours=3)
-transport = zeep.transports.Transport(cache=zeep.cache.SqliteCache())
-client = zeep.Client(wsdl=wsdl, transport=transport)
-time_all_elements = time.time()
-
+# get the data
log.debug("Getting %s data from %s to %s" % (str(desired_data),
str(begin_date), str(end_date)))
-for elementCd in desired_data:
- time_element = time.time()
-
- # get the last three hours of data for this elementCd
- tmp = client.service.getHourlyData(
- stationTriplets=[station_triplet],
- elementCd=elementCd,
- ordinal=1,
- beginDate=begin_date,
- endDate=end_date)
-
- log.info("Time to get elementCd '%s': %.3f sec" % (elementCd,
- time.time() - time_element))
+time_all_elements = time.time()
- values = tmp[0]['values']
+# 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 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)
- # 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
+ try:
+ json = req.json()
+ except ValueError:
+ log.error("Bad JSON in MesoWest response")
+ sys.exit(1)
-log.info("Time to get all elementCds : %.3f sec" % (time.time() -
+ 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 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
+
+log.info("Time to get all data : %.3f sec" % (time.time() -
time_all_elements))
log.debug("infoex[wx_data]: %s", str(infoex['wx_data']))
# any possible elementCds we may want are any other data
# type than float.
#
- # Another possibility is to query the API with
+ # Another possibility is to query the API with
# getStationElements and temporarily store the
# storedUnitCd. But that's pretty network-intensive and
# may not even be worth it if there's only e.g. one or two
writer.writerow(final_data)
f.close()
-#with open(infoex['csv_filename'], 'rb') as f:
-# log.debug("uploading FTP file '%s'" % (infoex['host']))
-# ftp = FTP(infoex['host'], infoex['uuid'], infoex['api_key'])
-# ftp.storlines('STOR ' + infoex['csv_filename'], f)
-# ftp.close()
-# f.close()
+if not options.dry_run:
+ with open(infoex['csv_filename'], 'rb') as f:
+ log.debug("uploading FTP file '%s'" % (infoex['host']))
+ ftp = FTP(infoex['host'], infoex['uuid'], infoex['api_key'])
+ ftp.storlines('STOR ' + infoex['csv_filename'], f)
+ ftp.close()
+ f.close()
+ os.remove(infoex['csv_filename'])
log.debug('DONE')
wylark / infoex-autowx.git / blobdiff