modify grouping algorithm to use day in US/Central

Author: ank1m

batcher/tempo_filename_parser.py

@@ -29,6 +29,8 @@
 import logging
 import re
 from argparse import ArgumentParser
+from datetime import datetime
+from dateutil import tz
 
 default_logger = logging.getLogger(__name__)
 
@@ -44,6 +46,33 @@
     r"(?P<granule_id>G[0-9]{2}).*\.nc"
 )
 
+def get_day_in_us_central(day_in_granule: str, time_in_granule: str, assume_tz=tz.UTC) -> str:
+    """
+    Convert a datetime to US Central time (US/Central) and return
+    a timezone-aware datetime.
+
+    Parameters
+    ----------
+    day_in_granule: str
+        The day from granule filename
+    time_in_granule: str
+        The time from granule filename
+    assume_tz : timezon, optional (default: UTC)
+        this is the timezone in which `dt` should be interpreted
+        before converting to Central.
+
+    Returns
+    -------
+    str
+        The day for datetime converted to US/Central
+    """
+
+    dt = datetime.strptime(day_in_granule+time_in_granule, "%Y%m%d%H%M%S")
+    dt = dt.replace(tzinfo = assume_tz)
+
+    dt_central = dt.astimezone(tz.gettz('US/Central'))
+    return dt_central.strftime("%Y%m%d")
+
 
 def get_batch_indices(filenames: list, logger: logging.Logger = default_logger) -> list[int]:
     """
@@ -60,7 +89,8 @@ def get_batch_indices(filenames: list, logger: logging.Logger = default_logger)
         matches = tempo_granule_filename_pattern.match(name)
         if matches:
             match_dict = matches.groupdict()
-            day_and_scans.append((match_dict["day_in_granule"], match_dict["daily_scan_id"]))
+            day_in_central= get_day_in_us_central(match_dict["day_in_granule"], match_dict["time_in_granule"])
+            day_and_scans.append((day_in_central, match_dict["daily_scan_id"]))
 
     # Unique day-scans are determined (while keeping the same order). Each will be its own batch.
     unique_day_scans: list[tuple[str, str]] = sorted(set(day_and_scans), key=day_and_scans.index)