Merge branch 'timezone-support' into develop
[infoex-autowx.git] / infoex-autowx.py
index a16b7e61e16947bf72a783c745ea81dbf59059c6..15adf0832a24cd757fd5b989edc729020ee8eee8 100755 (executable)
@@ -31,7 +31,9 @@ import sys
 import time
 
 from ftplib import FTP
-from optparse import OptionParser
+from argparse import ArgumentParser
+
+import pytz
 
 import requests
 
@@ -39,30 +41,34 @@ import zeep
 import zeep.cache
 import zeep.transports
 
-__version__ = '2.0.1'
+__version__ = '3.0.2'
 
 LOG = logging.getLogger(__name__)
 LOG.setLevel(logging.NOTSET)
 
 def get_parser():
     """Return OptionParser for this program"""
-    parser = OptionParser(version=__version__)
+    parser = ArgumentParser()
+
+    parser.add_argument("--version",
+                        action="version",
+                        version=__version__)
 
-    parser.add_option("--config",
-        dest="config",
-        metavar="FILE",
-        help="location of config file")
+    parser.add_argument("--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_argument("--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")
+    parser.add_argument("--dry-run",
+                        action="store_true",
+                        dest="dry_run",
+                        default=False,
+                        help="fetch data but don't upload to InfoEx")
 
     return parser
 
@@ -81,7 +87,7 @@ def setup_config(config):
         station = dict()
         station['provider'] = config['station']['type']
 
-        if station['provider'] not in ['nrcs', 'mesowest']:
+        if station['provider'] not in ['nrcs', 'mesowest', 'python']:
             print("Please specify either nrcs or mesowest as the station type.")
             sys.exit(1)
 
@@ -102,24 +108,39 @@ def setup_config(config):
             station['desired_data'] = config['station']['desired_data']
 
             # construct full API URL (sans start/end time, added later)
-            station['source'] = station['source'] + '?token=' + config['station']['token'] + '&within=60&units=' + station['units'] + '&stid=' + station['station_id'] + '&vars=' + station['desired_data']
+            station['source'] = station['source'] + '?token=' + \
+                                config['station']['token'] + \
+                                '&within=60&units=' + station['units'] + \
+                                '&stid=' + station['station_id'] + \
+                                '&vars=' + station['desired_data']
+
+        if station['provider'] == 'python':
+            station['path'] = config['station']['path']
 
-    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))
-        exit(1)
+        tz = 'America/Los_Angeles'
+
+        if 'tz' in config['station']:
+            tz = config['station']['tz']
+
+        try:
+            station['tz'] = pytz.timezone(tz)
+        except pytz.exceptions.UnknownTimeZoneError:
+            LOG.critical("%s is not a valid timezone", tz)
+            sys.exit(1)
+
+    except KeyError as err:
+        LOG.critical("%s not defined in configuration file", err)
+        sys.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))
-        exit(1)
+                if not config[key][subkey]:
+                    raise ValueError
+    except ValueError:
+        LOG.critical("Config value '%s.%s' is empty", key, subkey)
+        sys.exit(1)
 
     return (infoex, station)
 
@@ -128,7 +149,7 @@ def setup_logging(log_level):
     try:
         from systemd.journal import JournalHandler
         LOG.addHandler(JournalHandler())
-    except:
+    except ImportError:
         ## fallback to syslog
         #import logging.handlers
         #LOG.addHandler(logging.handlers.SysLogHandler())
@@ -154,7 +175,7 @@ def setup_logging(log_level):
 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)
 
@@ -179,14 +200,23 @@ def main():
     iemap = setup_infoex_counterparts_mapping(station['provider'])
 
     # override units if user selected metric
-    if station['units'] == 'metric':
-        final_data = switch_units_to_metric(final_data, fmap)
+    try:
+        if station['units'] == 'metric':
+            final_data = switch_units_to_metric(final_data, fmap)
+    except KeyError:
+        if station['provider'] != 'python':
+            LOG.error("Please specify the units in the configuration "
+                      "file")
+            sys.exit(1)
 
-    (begin_date, end_date) = setup_time_values()
+    (begin_date, end_date) = setup_time_values(station)
 
-    # get the data
-    LOG.debug("Getting %s data from %s to %s" % (str(station['desired_data']),
-        str(begin_date), str(end_date)))
+    if station['provider'] == 'python':
+        LOG.debug("Getting custom data from external Python program")
+    else:
+        LOG.debug("Getting %s data from %s to %s (%s)",
+                  str(station['desired_data']),
+                  str(begin_date), str(end_date), end_date.tzinfo.zone)
 
     time_all_elements = time.time()
 
@@ -195,24 +225,74 @@ def main():
         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)
+                                              station)
+    elif station['provider'] == 'python':
+        try:
+            import importlib.util
+
+            spec = importlib.util.spec_from_file_location('custom_wx',
+                                                          station['path'])
+            mod = importlib.util.module_from_spec(spec)
+            spec.loader.exec_module(mod)
+            mod.LOG = LOG
+
+            try:
+                infoex['wx_data'] = mod.get_custom_data()
+
+                if infoex['wx_data'] is None:
+                    infoex['wx_data'] = []
+            except Exception as exc:
+                LOG.error("Python program for custom Wx data failed in "
+                          "execution: %s", str(exc))
+                sys.exit(1)
+
+            LOG.info("Successfully executed external Python program")
+        except ImportError:
+            LOG.error("Please upgrade to Python 3.3 or later")
+            sys.exit(1)
+        except FileNotFoundError:
+            LOG.error("Specified Python program for custom Wx data "
+                      "was not found")
+            sys.exit(1)
+        except Exception as exc:
+            LOG.error("A problem was encountered when attempting to "
+                      "load your custom Wx program: %s", str(exc))
+            sys.exit(1)
 
-    LOG.info("Time taken 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']))
 
+    # timezone massaging
+    final_end_date = end_date.astimezone(station['tz'])
+
     # Now we only need to add in what we want to change thanks to that
     # abomination of a variable declaration earlier
     final_data[fmap['Location UUID']] = infoex['location_uuid']
-    final_data[fmap['obDate']] = end_date.strftime('%m/%d/%Y')
-    final_data[fmap['obTime']] = end_date.strftime('%H:%M')
+    final_data[fmap['obDate']] = final_end_date.strftime('%m/%d/%Y')
+    final_data[fmap['obTime']] = final_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, relative humidity, 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', 'RHUM',
+                            'relative_humidity', '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.
@@ -222,17 +302,18 @@ def main():
         #           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;
+    if infoex['wx_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
 
-    if not options.dry_run:
-        upload_csv(infoex['csv_filename'], infoex)
+        if not options.dry_run:
+            upload_csv(infoex['csv_filename'], infoex)
 
     LOG.debug('DONE')
     return 0
@@ -243,16 +324,17 @@ 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
@@ -296,6 +378,8 @@ def setup_infoex_counterparts_mapping(provider):
     if provider == 'nrcs':
         iemap['PREC'] = 'precipitationGauge'
         iemap['TOBS'] = 'tempPres'
+        iemap['TMAX'] = 'tempMaxHour'
+        iemap['TMIN'] = 'tempMinHour'
         iemap['SNWD'] = 'hS'
         iemap['PRES'] = 'baro'
         iemap['RHUM'] = 'rH'
@@ -306,12 +390,26 @@ def setup_infoex_counterparts_mapping(provider):
     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'
         iemap['wind_speed'] = 'windSpeedNum'
         iemap['wind_direction'] = 'windDirectionNum'
         iemap['wind_gust'] = 'windGustSpeedNum'
+    elif provider == 'python':
+        # we expect Python programs to use the InfoEx data type names
+        iemap['precipitationGauge'] = 'precipitationGauge'
+        iemap['tempPres'] = 'tempPres'
+        iemap['tempMaxHour'] = 'tempMaxHour'
+        iemap['tempMinHour'] = 'tempMinHour'
+        iemap['hS'] = 'hS'
+        iemap['baro'] = 'baro'
+        iemap['rH'] = 'rH'
+        iemap['windSpeedNum'] = 'windSpeedNum'
+        iemap['windDirectionNum'] = 'windDirectionNum'
+        iemap['windGustSpeedNum'] = 'windGustSpeedNum'
 
     return iemap
 
@@ -322,19 +420,23 @@ def get_nrcs_data(begin, end, station):
     client = zeep.Client(wsdl=station['source'], transport=transport)
     remote_data = {}
 
-    for elementCd in station['desired_data']:
+    # massage begin/end date format
+    begin_date_str = begin.strftime('%Y-%m-%d %H:%M:00')
+    end_date_str = end.strftime('%Y-%m-%d %H:%M:00')
+
+    for element_cd in station['desired_data']:
         time_element = time.time()
 
-        # get the last three hours of data for this elementCd
+        # get the last three hours of data for this elementCd/element_cd
         tmp = client.service.getHourlyData(
-                stationTriplets=[station['station_id']],
-                elementCd=elementCd,
-                ordinal=1,
-                beginDate=begin,
-                endDate=end)
+            stationTriplets=[station['station_id']],
+            elementCd=element_cd,
+            ordinal=1,
+            beginDate=begin_date_str,
+            endDate=end_date_str)
 
-        LOG.info("Time to get elementCd '%s': %.3f sec" % (elementCd,
-            time.time() - time_element))
+        LOG.info("Time to get NRCS elementCd '%s': %.3f sec", element_cd,
+                 time.time() - time_element)
 
         values = tmp[0]['values']
 
@@ -345,9 +447,9 @@ def get_nrcs_data(begin, end, station):
         #       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[elementCd] = ordered[0]['value']
+            remote_data[element_cd] = ordered[0]['value']
         else:
-            remote_data[elementCd] = None
+            remote_data[element_cd] = None
 
     return remote_data
 
@@ -377,7 +479,7 @@ def get_mesowest_data(begin, end, station):
 
     pos = len(observations['date_time']) - 1
 
-    for elementCd in station['desired_data'].split(','):
+    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
         #
@@ -392,14 +494,21 @@ def get_mesowest_data(begin, end, station):
         #       irregularities
 
         # we may not have the data at all
-        key_name = elementCd + '_set_1'
+        key_name = element_cd + '_set_1'
+
         if key_name in observations:
             if observations[key_name][pos]:
-                remote_data[elementCd] = observations[key_name][pos]
+                remote_data[element_cd] = observations[key_name][pos]
+
+                # mesowest by default provides wind_speed in m/s, but
+                # we specify 'english' units in the request; either way,
+                # we want mph
+                if element_cd in ('wind_speed', 'wind_gust'):
+                    remote_data[element_cd] = kn_to_mph(remote_data[element_cd])
             else:
-                remote_data[elementCd] = None
+                remote_data[element_cd] = None
         else:
-            remote_data[elementCd] = None
+            remote_data[element_cd] = None
 
     return remote_data
 
@@ -423,19 +532,19 @@ def switch_units_to_metric(data_map, mapping):
 # 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)
@@ -444,15 +553,31 @@ def upload_csv(path_to_file, infoex_data):
     os.remove(path_to_file)
 
 # other miscellaneous routines
-def setup_time_values():
+def setup_time_values(station):
     """establish time bounds of data request(s)"""
+
+    # default timezone to UTC (for MesoWest)
+    tz = pytz.utc
+
+    # but for NRCS, use the config-specified timezone
+    if station['provider'] == 'nrcs':
+        tz = station['tz']
+
     # floor time to nearest hour
-    dt = datetime.datetime.now()
-    end_date = dt - datetime.timedelta(minutes=dt.minute % 60,
-                                       seconds=dt.second,
-                                       microseconds=dt.microsecond)
+    date_time = datetime.datetime.now(tz=tz)
+    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)
 
+def ms_to_mph(ms):
+    """convert meters per second to miles per hour"""
+    return ms * 2.236936
+
+def kn_to_mph(kn):
+    """convert knots to miles per hour"""
+    return kn * 1.150779
+
 if __name__ == "__main__":
     sys.exit(main())