Merge pull request #73 from megies/time_dependent_station_combos

add possibility to specify complex, time-dependent station combinations

Author: megies

obspyck/example.cfg

@@ -57,6 +57,31 @@ S = blue
 [station_combinations]
 EXAMPLE1 = II.PFO.00.LH?,GE.APE..LH?,GE.PSZ..LH?,CI.BBR..LH*,IV.PRMA..LH?,G.RER..LH*
 EXAMPLE2 = 7A.W01..LH?,CI.BBR..LH*
+# for complex, time-dependent station combinations use an exclamation mark
+# followed by a corresponding section title where the station combination is
+# defined
+EXAMPLE3 = !station_combination_BW
+
+[station_combination_BW]
+# this is an example section for a complex, time-dependent station combination.
+# this can be useful e.g. in case of instrumentation changes that go along with
+# a change in channel naming or when stations get added or removed in the
+# field. each entry consists of a stream identifier as the key and zero, one or
+# two timestamps to denote for which times the stream should be fetched.
+# All times interpreted as UTC.
+#
+# if value is empty, always fetch stream, regardless of requested time
+GR.FUR..HH? =
+# otherwise, value can be..
+#   - two timestamps separated by a comma, to fetch stream only for times in
+#     between these two timestamps
+BW.RTBE..HH? = 2017-07-01,2017-10-01
+#   - a timestamp preceded by a "larger than" sign, to fetch stream only for times after the
+#     given timestamp (including it, so this is actually evaluated as >=)
+BW.RMOA..HH? = >2017-07-04T15:00:00
+#   - a comma followed by a timestamp, to fetch stream only for times before the
+#     given timestamp (including it, so this is actually evaluated as <=)
+BW.RMOA..EH? = <2017-07-04T11:40:00
 
 [seed_id_lookup]
 # wildcards ("?", "*") and regex (e.g. "[EH]*") are only allowed in channel!

obspyck/util.py

@@ -233,6 +233,47 @@ def check_keybinding_conflicts(keys):
             raise Exception(err)
 
 
+def resolve_complex_station_combination(config, section_title, time):
+    """
+    Resolve complex time-dependent station combination from config file
+    """
+    seed_ids = set()
+    for key, value in config.items(section_title):
+        # no value: always use that stream
+        if not value:
+            seed_ids.add(key)
+        # two comma separated time stamps: only use stream if requested time is
+        # in between given timestamps
+        elif ',' in value:
+            start, end = map(UTCDateTime, value.split(','))
+            if end < start:
+                msg = ("Invalid configuration in section [{}], stream '{}'. "
+                       "End time larger than start time for used time "
+                       "span.").format(section_title, key)
+                raise ValueError(msg)
+            if start <= time <= end:
+                seed_ids.add(key)
+        # timestamp preceded by smaller-than sign: only use stream if requested
+        # time is smaller than specified timestamp
+        elif value[0] == '<':
+            end = UTCDateTime(value[1:])
+            if time <= end:
+                seed_ids.add(key)
+        # timestamp preceded by larger-than sign: only use stream if requested
+        # time is after the specified timestamp
+        elif value[0] == '>':
+            start = UTCDateTime(value[1:])
+            if time >= start:
+                seed_ids.add(key)
+        else:
+            msg = ("Invalid configuration in section [{}], stream '{}'. "
+                   "Value can be empty, two comma-separated time stamps or "
+                   "one timestamp preceded by either '>' or '<'.").format(
+                        section_title, key)
+            raise ValueError(msg)
+    return seed_ids
+
+
 def fetch_waveforms_with_metadata(options, args, config):
     """
     Sets up obspy clients and fetches waveforms and metadata according to
@@ -269,7 +310,14 @@ def fetch_waveforms_with_metadata(options, args, config):
     seed_ids_to_fetch = set()
     for station_combination_key in options.station_combinations:
         seed_ids = config.get("station_combinations", station_combination_key)
-        for seed_id in seed_ids.split(","):
+        # starts with an exclamation mark: lookup seed ids from that section
+        if seed_ids[0] == '!':
+            seed_ids = resolve_complex_station_combination(
+                config, seed_ids[1:], t1)
+        # otherwise a list of comma-separated stream labels
+        else:
+            seed_ids = seed_ids.split(",")
+        for seed_id in seed_ids:
             seed_ids_to_fetch.add(seed_id)
     for seed_id in options.seed_ids:
         seed_ids_to_fetch.add(seed_id)